diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,153969 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9901902086307742,
+  "eval_steps": 50,
+  "global_step": 21500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 91.2317079978569,
+      "learning_rate": 9.999999952897649e-06,
+      "loss": 2.5261,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.9650615453720093,
+      "eval_runtime": 67.7003,
+      "eval_samples_per_second": 50.369,
+      "eval_steps_per_second": 1.58,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.331649195398498,
+      "learning_rate": 9.999999811590591e-06,
+      "loss": 1.1407,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.543867288474374,
+      "learning_rate": 9.999999576078835e-06,
+      "loss": 1.0416,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.17894097332178,
+      "learning_rate": 9.99999924636238e-06,
+      "loss": 0.7622,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.63758312569935,
+      "learning_rate": 9.999998822441238e-06,
+      "loss": 1.0577,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.457697094247186,
+      "learning_rate": 9.999998304315414e-06,
+      "loss": 0.973,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.791014052293002,
+      "learning_rate": 9.999997691984922e-06,
+      "loss": 0.6403,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.995467634316878,
+      "learning_rate": 9.999996985449774e-06,
+      "loss": 0.838,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.21493305932694,
+      "learning_rate": 9.999996184709983e-06,
+      "loss": 0.8609,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1725481833999947,
+      "learning_rate": 9.999995289765565e-06,
+      "loss": 0.6387,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6092753076851096,
+      "learning_rate": 9.999994300616544e-06,
+      "loss": 0.7093,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9878872977198165,
+      "learning_rate": 9.999993217262936e-06,
+      "loss": 0.5897,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8176581955051367,
+      "learning_rate": 9.999992039704765e-06,
+      "loss": 0.5978,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0398165821603516,
+      "learning_rate": 9.999990767942057e-06,
+      "loss": 0.7011,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.515634979309735,
+      "learning_rate": 9.999989401974836e-06,
+      "loss": 0.8217,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3719897778405232,
+      "learning_rate": 9.999987941803132e-06,
+      "loss": 0.7266,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.985590542895897,
+      "learning_rate": 9.999986387426977e-06,
+      "loss": 0.8182,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5171981139136212,
+      "learning_rate": 9.9999847388464e-06,
+      "loss": 0.7318,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.480558787521623,
+      "learning_rate": 9.999982996061438e-06,
+      "loss": 0.7144,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.774616718381461,
+      "learning_rate": 9.999981159072126e-06,
+      "loss": 0.693,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.923464945416381,
+      "learning_rate": 9.999979227878504e-06,
+      "loss": 0.8238,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.774427783887792,
+      "learning_rate": 9.999977202480613e-06,
+      "loss": 0.6055,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.090070110304742,
+      "learning_rate": 9.999975082878493e-06,
+      "loss": 0.6345,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.091581403872198,
+      "learning_rate": 9.999972869072191e-06,
+      "loss": 0.6819,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7416049794393156,
+      "learning_rate": 9.999970561061751e-06,
+      "loss": 0.7758,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1665530955141428,
+      "learning_rate": 9.999968158847224e-06,
+      "loss": 0.8913,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.519046754876911,
+      "learning_rate": 9.999965662428657e-06,
+      "loss": 0.7466,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9836349806807285,
+      "learning_rate": 9.999963071806103e-06,
+      "loss": 0.7477,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9969643440586786,
+      "learning_rate": 9.99996038697962e-06,
+      "loss": 0.6901,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.04142213903052,
+      "learning_rate": 9.99995760794926e-06,
+      "loss": 0.8904,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2361416516476074,
+      "learning_rate": 9.999954734715082e-06,
+      "loss": 0.5603,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6603872395803445,
+      "learning_rate": 9.999951767277147e-06,
+      "loss": 0.7427,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.931779616649879,
+      "learning_rate": 9.999948705635517e-06,
+      "loss": 0.8328,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2798612487356342,
+      "learning_rate": 9.999945549790255e-06,
+      "loss": 0.5872,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4695373047705305,
+      "learning_rate": 9.999942299741428e-06,
+      "loss": 0.7048,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.785453050662037,
+      "learning_rate": 9.999938955489105e-06,
+      "loss": 0.5642,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.885633650687876,
+      "learning_rate": 9.999935517033354e-06,
+      "loss": 0.7462,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.081775709818994,
+      "learning_rate": 9.99993198437425e-06,
+      "loss": 0.6886,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.543804481498728,
+      "learning_rate": 9.999928357511864e-06,
+      "loss": 0.6223,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6464552796135785,
+      "learning_rate": 9.999924636446272e-06,
+      "loss": 0.7327,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9925364014992732,
+      "learning_rate": 9.999920821177551e-06,
+      "loss": 0.6477,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.905730262546426,
+      "learning_rate": 9.999916911705786e-06,
+      "loss": 0.7177,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7568771991593537,
+      "learning_rate": 9.999912908031054e-06,
+      "loss": 0.8184,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.613905559475549,
+      "learning_rate": 9.99990881015344e-06,
+      "loss": 0.6542,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6905851868480157,
+      "learning_rate": 9.99990461807303e-06,
+      "loss": 0.6467,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9299670197085925,
+      "learning_rate": 9.999900331789911e-06,
+      "loss": 0.6561,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8666171629336366,
+      "learning_rate": 9.999895951304175e-06,
+      "loss": 0.5585,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3089631808211846,
+      "learning_rate": 9.999891476615915e-06,
+      "loss": 0.9326,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3333689241397972,
+      "learning_rate": 9.999886907725217e-06,
+      "loss": 0.7397,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0575394196436543,
+      "learning_rate": 9.999882244632183e-06,
+      "loss": 0.6429,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 0.7258575558662415,
+      "eval_runtime": 67.9482,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7036712099732707,
+      "learning_rate": 9.99987748733691e-06,
+      "loss": 0.7338,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.422303009876831,
+      "learning_rate": 9.999872635839497e-06,
+      "loss": 0.6586,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5763423530171083,
+      "learning_rate": 9.999867690140045e-06,
+      "loss": 0.6194,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9861161105926146,
+      "learning_rate": 9.999862650238657e-06,
+      "loss": 0.7162,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4356205516479172,
+      "learning_rate": 9.99985751613544e-06,
+      "loss": 0.6255,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5320195518851683,
+      "learning_rate": 9.9998522878305e-06,
+      "loss": 0.4924,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.101689023700977,
+      "learning_rate": 9.999846965323949e-06,
+      "loss": 0.929,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.447250620803791,
+      "learning_rate": 9.999841548615895e-06,
+      "loss": 0.7167,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1528131465389118,
+      "learning_rate": 9.999836037706455e-06,
+      "loss": 0.7189,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8304124045436256,
+      "learning_rate": 9.999830432595741e-06,
+      "loss": 0.5994,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8623516909164626,
+      "learning_rate": 9.999824733283872e-06,
+      "loss": 0.7584,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1762439621889547,
+      "learning_rate": 9.999818939770969e-06,
+      "loss": 0.7562,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8498415496473966,
+      "learning_rate": 9.999813052057148e-06,
+      "loss": 0.6799,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.267633127609889,
+      "learning_rate": 9.999807070142539e-06,
+      "loss": 0.6287,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7874487438779543,
+      "learning_rate": 9.999800994027264e-06,
+      "loss": 0.7902,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.266303715453486,
+      "learning_rate": 9.999794823711447e-06,
+      "loss": 0.7149,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7014543628637284,
+      "learning_rate": 9.999788559195223e-06,
+      "loss": 0.671,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1914782663470658,
+      "learning_rate": 9.999782200478719e-06,
+      "loss": 0.6059,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.11531033065683,
+      "learning_rate": 9.99977574756207e-06,
+      "loss": 0.4488,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.097972658321014,
+      "learning_rate": 9.999769200445411e-06,
+      "loss": 0.7792,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.548915607342437,
+      "learning_rate": 9.999762559128878e-06,
+      "loss": 0.5693,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.034595344064918,
+      "learning_rate": 9.99975582361261e-06,
+      "loss": 0.5088,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8291326420865768,
+      "learning_rate": 9.999748993896752e-06,
+      "loss": 0.8646,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.035993814924278,
+      "learning_rate": 9.999742069981442e-06,
+      "loss": 0.8785,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.909803390912722,
+      "learning_rate": 9.999735051866825e-06,
+      "loss": 0.8106,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2039100465772132,
+      "learning_rate": 9.99972793955305e-06,
+      "loss": 0.748,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.514939740559863,
+      "learning_rate": 9.999720733040268e-06,
+      "loss": 0.6235,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0921557413705294,
+      "learning_rate": 9.999713432328624e-06,
+      "loss": 0.6933,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.289189334935671,
+      "learning_rate": 9.999706037418277e-06,
+      "loss": 0.7695,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9388943715460314,
+      "learning_rate": 9.999698548309377e-06,
+      "loss": 0.5782,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.230328363189254,
+      "learning_rate": 9.999690965002083e-06,
+      "loss": 0.6189,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.745571620296296,
+      "learning_rate": 9.999683287496554e-06,
+      "loss": 0.5153,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.438532661670713,
+      "learning_rate": 9.999675515792951e-06,
+      "loss": 0.6971,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5727929049053917,
+      "learning_rate": 9.999667649891434e-06,
+      "loss": 0.6791,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.628949302087166,
+      "learning_rate": 9.999659689792171e-06,
+      "loss": 0.7954,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9657539249286904,
+      "learning_rate": 9.999651635495328e-06,
+      "loss": 0.6986,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.368107055313853,
+      "learning_rate": 9.999643487001072e-06,
+      "loss": 0.6588,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8275258173750584,
+      "learning_rate": 9.999635244309573e-06,
+      "loss": 0.6956,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0726986206671456,
+      "learning_rate": 9.999626907421007e-06,
+      "loss": 0.6796,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8914311428142105,
+      "learning_rate": 9.999618476335545e-06,
+      "loss": 0.6476,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.058763231665245,
+      "learning_rate": 9.999609951053367e-06,
+      "loss": 0.65,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3065094962456962,
+      "learning_rate": 9.999601331574648e-06,
+      "loss": 0.6323,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2441671715219993,
+      "learning_rate": 9.99959261789957e-06,
+      "loss": 0.6363,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9214784792652146,
+      "learning_rate": 9.999583810028316e-06,
+      "loss": 0.7964,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.554270443955648,
+      "learning_rate": 9.99957490796107e-06,
+      "loss": 0.6789,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7689028474982087,
+      "learning_rate": 9.999565911698017e-06,
+      "loss": 0.7596,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.790865994734635,
+      "learning_rate": 9.999556821239345e-06,
+      "loss": 0.7756,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.257739897234312,
+      "learning_rate": 9.999547636585248e-06,
+      "loss": 0.4942,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.072018894137989,
+      "learning_rate": 9.999538357735916e-06,
+      "loss": 0.4705,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6800180226778365,
+      "learning_rate": 9.999528984691542e-06,
+      "loss": 0.7502,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 0.6971283555030823,
+      "eval_runtime": 67.9531,
+      "eval_samples_per_second": 50.182,
+      "eval_steps_per_second": 1.575,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.575350589435018,
+      "learning_rate": 9.999519517452326e-06,
+      "loss": 0.6752,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.841342093367452,
+      "learning_rate": 9.999509956018462e-06,
+      "loss": 0.6307,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6761695915769934,
+      "learning_rate": 9.99950030039015e-06,
+      "loss": 0.7969,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6268755274960913,
+      "learning_rate": 9.999490550567596e-06,
+      "loss": 0.7408,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7248640035496954,
+      "learning_rate": 9.999480706551003e-06,
+      "loss": 0.7043,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.562281593248224,
+      "learning_rate": 9.999470768340574e-06,
+      "loss": 0.6657,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.327871325963673,
+      "learning_rate": 9.99946073593652e-06,
+      "loss": 0.6474,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.628950486915566,
+      "learning_rate": 9.99945060933905e-06,
+      "loss": 0.5531,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2000986741805595,
+      "learning_rate": 9.999440388548377e-06,
+      "loss": 0.5802,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.396045738284757,
+      "learning_rate": 9.999430073564713e-06,
+      "loss": 0.6037,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0459944018349794,
+      "learning_rate": 9.999419664388276e-06,
+      "loss": 0.7088,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.42951975828646,
+      "learning_rate": 9.999409161019284e-06,
+      "loss": 0.8023,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6557240418966876,
+      "learning_rate": 9.999398563457954e-06,
+      "loss": 0.801,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5238162061868437,
+      "learning_rate": 9.99938787170451e-06,
+      "loss": 0.7138,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.417731109911248,
+      "learning_rate": 9.999377085759177e-06,
+      "loss": 0.5846,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.296358101578327,
+      "learning_rate": 9.999366205622178e-06,
+      "loss": 0.7774,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3447637397161554,
+      "learning_rate": 9.999355231293742e-06,
+      "loss": 0.6276,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0201988506373043,
+      "learning_rate": 9.9993441627741e-06,
+      "loss": 0.6662,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.294610613864954,
+      "learning_rate": 9.999333000063483e-06,
+      "loss": 0.7746,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6267033750540203,
+      "learning_rate": 9.999321743162122e-06,
+      "loss": 0.6615,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.678206229890648,
+      "learning_rate": 9.999310392070257e-06,
+      "loss": 0.7769,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.694614240856692,
+      "learning_rate": 9.999298946788123e-06,
+      "loss": 0.7418,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6391165397289433,
+      "learning_rate": 9.99928740731596e-06,
+      "loss": 0.7326,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0681936584547067,
+      "learning_rate": 9.99927577365401e-06,
+      "loss": 0.7066,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.069465831496503,
+      "learning_rate": 9.999264045802515e-06,
+      "loss": 0.5696,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3728463778398887,
+      "learning_rate": 9.999252223761723e-06,
+      "loss": 0.6958,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3693612510013233,
+      "learning_rate": 9.99924030753188e-06,
+      "loss": 0.7008,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1977317619093664,
+      "learning_rate": 9.999228297113234e-06,
+      "loss": 0.6089,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6704067102329487,
+      "learning_rate": 9.999216192506041e-06,
+      "loss": 0.6139,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.441131791798755,
+      "learning_rate": 9.999203993710552e-06,
+      "loss": 0.5886,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.546455753366936,
+      "learning_rate": 9.99919170072702e-06,
+      "loss": 0.7339,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2050747715666974,
+      "learning_rate": 9.999179313555703e-06,
+      "loss": 0.6199,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3831681466615158,
+      "learning_rate": 9.999166832196865e-06,
+      "loss": 0.702,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9356436111696167,
+      "learning_rate": 9.999154256650761e-06,
+      "loss": 0.5726,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.275878301140902,
+      "learning_rate": 9.999141586917658e-06,
+      "loss": 0.6841,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.772105918311177,
+      "learning_rate": 9.99912882299782e-06,
+      "loss": 0.715,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.511770401924075,
+      "learning_rate": 9.999115964891519e-06,
+      "loss": 0.5828,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4272493698746063,
+      "learning_rate": 9.999103012599016e-06,
+      "loss": 0.8003,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5114537472446163,
+      "learning_rate": 9.999089966120586e-06,
+      "loss": 0.6765,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.476381600171169,
+      "learning_rate": 9.9990768254565e-06,
+      "loss": 0.669,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.529080001498705,
+      "learning_rate": 9.999063590607039e-06,
+      "loss": 0.6482,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.185350677654188,
+      "learning_rate": 9.999050261572474e-06,
+      "loss": 0.6419,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.142428716173918,
+      "learning_rate": 9.999036838353086e-06,
+      "loss": 0.7764,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.046489790708567,
+      "learning_rate": 9.999023320949159e-06,
+      "loss": 0.5762,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6392169440869075,
+      "learning_rate": 9.99900970936097e-06,
+      "loss": 0.5703,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1461857271878055,
+      "learning_rate": 9.998996003588809e-06,
+      "loss": 0.7786,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7820859864028327,
+      "learning_rate": 9.99898220363296e-06,
+      "loss": 0.6336,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2563575445405055,
+      "learning_rate": 9.998968309493714e-06,
+      "loss": 0.8062,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3498589246112225,
+      "learning_rate": 9.99895432117136e-06,
+      "loss": 0.6952,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3531522414359745,
+      "learning_rate": 9.998940238666193e-06,
+      "loss": 0.6313,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.687222957611084,
+      "eval_runtime": 67.9717,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.004423020277841,
+      "learning_rate": 9.998926061978505e-06,
+      "loss": 0.7835,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.359162640225442,
+      "learning_rate": 9.998911791108595e-06,
+      "loss": 0.6543,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3633178504509007,
+      "learning_rate": 9.998897426056761e-06,
+      "loss": 0.6499,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.655355899007356,
+      "learning_rate": 9.998882966823304e-06,
+      "loss": 0.7964,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.26408109968416,
+      "learning_rate": 9.998868413408525e-06,
+      "loss": 0.447,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.18891044968983,
+      "learning_rate": 9.998853765812731e-06,
+      "loss": 0.7008,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.379843816375166,
+      "learning_rate": 9.99883902403623e-06,
+      "loss": 0.5726,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3254978177034507,
+      "learning_rate": 9.998824188079326e-06,
+      "loss": 0.6889,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.536353671673955,
+      "learning_rate": 9.998809257942332e-06,
+      "loss": 0.6678,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.892674687465232,
+      "learning_rate": 9.998794233625561e-06,
+      "loss": 0.7425,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0903174888206744,
+      "learning_rate": 9.998779115129325e-06,
+      "loss": 0.4777,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.677903190775237,
+      "learning_rate": 9.998763902453946e-06,
+      "loss": 0.8479,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2075660673918334,
+      "learning_rate": 9.998748595599737e-06,
+      "loss": 0.5691,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9935653119941408,
+      "learning_rate": 9.99873319456702e-06,
+      "loss": 0.4662,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.615520827621649,
+      "learning_rate": 9.998717699356119e-06,
+      "loss": 0.7137,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2321729394781293,
+      "learning_rate": 9.998702109967357e-06,
+      "loss": 0.5814,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.032138024646017,
+      "learning_rate": 9.998686426401061e-06,
+      "loss": 0.5195,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.408627337045337,
+      "learning_rate": 9.998670648657558e-06,
+      "loss": 0.7917,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.43962423087857,
+      "learning_rate": 9.998654776737179e-06,
+      "loss": 0.78,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8685592184153594,
+      "learning_rate": 9.998638810640258e-06,
+      "loss": 0.7358,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.508771088898249,
+      "learning_rate": 9.998622750367126e-06,
+      "loss": 0.656,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4885617641857105,
+      "learning_rate": 9.998606595918122e-06,
+      "loss": 0.6491,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1144440103594917,
+      "learning_rate": 9.998590347293582e-06,
+      "loss": 0.5744,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2680644521330406,
+      "learning_rate": 9.998574004493849e-06,
+      "loss": 0.8524,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.919576218247978,
+      "learning_rate": 9.998557567519261e-06,
+      "loss": 0.829,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.355727716500502,
+      "learning_rate": 9.998541036370166e-06,
+      "loss": 0.8046,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.739339260861417,
+      "learning_rate": 9.998524411046908e-06,
+      "loss": 0.6471,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.048199080819445,
+      "learning_rate": 9.998507691549836e-06,
+      "loss": 0.4536,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3588829199612547,
+      "learning_rate": 9.9984908778793e-06,
+      "loss": 0.6379,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1541454734269214,
+      "learning_rate": 9.998473970035652e-06,
+      "loss": 0.6708,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0651383731085735,
+      "learning_rate": 9.998456968019243e-06,
+      "loss": 0.8363,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5233030600316204,
+      "learning_rate": 9.998439871830434e-06,
+      "loss": 0.6662,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.832759005855989,
+      "learning_rate": 9.998422681469578e-06,
+      "loss": 0.5829,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.597225999392599,
+      "learning_rate": 9.99840539693704e-06,
+      "loss": 0.7322,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5539237640253503,
+      "learning_rate": 9.998388018233177e-06,
+      "loss": 0.6611,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3458278586933994,
+      "learning_rate": 9.998370545358357e-06,
+      "loss": 0.6485,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.502545364783036,
+      "learning_rate": 9.998352978312942e-06,
+      "loss": 0.8195,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.876625564686255,
+      "learning_rate": 9.998335317097302e-06,
+      "loss": 0.5551,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.673487056933717,
+      "learning_rate": 9.998317561711805e-06,
+      "loss": 0.6708,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7496955045847655,
+      "learning_rate": 9.998299712156824e-06,
+      "loss": 0.5245,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2133375183664494,
+      "learning_rate": 9.998281768432735e-06,
+      "loss": 0.5387,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.448157683736773,
+      "learning_rate": 9.998263730539909e-06,
+      "loss": 0.6667,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2949687848926676,
+      "learning_rate": 9.998245598478727e-06,
+      "loss": 0.7105,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.519769499163072,
+      "learning_rate": 9.998227372249566e-06,
+      "loss": 0.6873,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3465121198143075,
+      "learning_rate": 9.99820905185281e-06,
+      "loss": 0.5826,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5200566898185364,
+      "learning_rate": 9.99819063728884e-06,
+      "loss": 0.6114,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.23980325243879,
+      "learning_rate": 9.998172128558045e-06,
+      "loss": 0.6355,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.318247934547554,
+      "learning_rate": 9.998153525660809e-06,
+      "loss": 0.6068,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8864512292928377,
+      "learning_rate": 9.998134828597524e-06,
+      "loss": 0.7911,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2345256007874217,
+      "learning_rate": 9.998116037368578e-06,
+      "loss": 0.6523,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.6727938652038574,
+      "eval_runtime": 67.9269,
+      "eval_samples_per_second": 50.201,
+      "eval_steps_per_second": 1.575,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.356757823081826,
+      "learning_rate": 9.998097151974369e-06,
+      "loss": 0.5949,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.261471407900797,
+      "learning_rate": 9.99807817241529e-06,
+      "loss": 0.7439,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.752081592551988,
+      "learning_rate": 9.998059098691735e-06,
+      "loss": 0.7165,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.456864199489977,
+      "learning_rate": 9.998039930804111e-06,
+      "loss": 0.6404,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2761962036626717,
+      "learning_rate": 9.998020668752813e-06,
+      "loss": 0.5973,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.269866072074436,
+      "learning_rate": 9.998001312538247e-06,
+      "loss": 0.5639,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1478842123586848,
+      "learning_rate": 9.997981862160815e-06,
+      "loss": 0.5423,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3753050945832284,
+      "learning_rate": 9.997962317620927e-06,
+      "loss": 0.5764,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3520337242025873,
+      "learning_rate": 9.997942678918992e-06,
+      "loss": 0.5904,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.902617796521297,
+      "learning_rate": 9.997922946055424e-06,
+      "loss": 0.7054,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4750659270397555,
+      "learning_rate": 9.997903119030629e-06,
+      "loss": 0.782,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.818933541248018,
+      "learning_rate": 9.997883197845026e-06,
+      "loss": 0.8233,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1384740625075107,
+      "learning_rate": 9.997863182499034e-06,
+      "loss": 0.5959,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8173446771131332,
+      "learning_rate": 9.997843072993068e-06,
+      "loss": 0.4901,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.227938341049738,
+      "learning_rate": 9.997822869327554e-06,
+      "loss": 0.5821,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7427949126182836,
+      "learning_rate": 9.99780257150291e-06,
+      "loss": 0.6455,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.503655935959697,
+      "learning_rate": 9.997782179519563e-06,
+      "loss": 0.6212,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1543592740164277,
+      "learning_rate": 9.99776169337794e-06,
+      "loss": 0.5911,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3373635109453863,
+      "learning_rate": 9.997741113078468e-06,
+      "loss": 0.5901,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.153069402026206,
+      "learning_rate": 9.99772043862158e-06,
+      "loss": 0.7071,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4543152103225485,
+      "learning_rate": 9.99769967000771e-06,
+      "loss": 0.737,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.970850170428933,
+      "learning_rate": 9.99767880723729e-06,
+      "loss": 0.7919,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4409568018924763,
+      "learning_rate": 9.997657850310758e-06,
+      "loss": 0.6179,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6406038818629667,
+      "learning_rate": 9.997636799228553e-06,
+      "loss": 0.8172,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1756843707560556,
+      "learning_rate": 9.997615653991115e-06,
+      "loss": 0.711,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0122165470075495,
+      "learning_rate": 9.997594414598886e-06,
+      "loss": 0.6245,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.865003625357823,
+      "learning_rate": 9.99757308105231e-06,
+      "loss": 0.4413,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4542458963060794,
+      "learning_rate": 9.997551653351838e-06,
+      "loss": 0.7263,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.486508261249356,
+      "learning_rate": 9.997530131497915e-06,
+      "loss": 0.7708,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6778856455268003,
+      "learning_rate": 9.997508515490991e-06,
+      "loss": 0.702,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7765760037540295,
+      "learning_rate": 9.997486805331522e-06,
+      "loss": 0.6317,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5704444888255797,
+      "learning_rate": 9.997465001019958e-06,
+      "loss": 0.7261,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4739363041911777,
+      "learning_rate": 9.997443102556758e-06,
+      "loss": 0.6378,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.464063214290251,
+      "learning_rate": 9.99742110994238e-06,
+      "loss": 0.5406,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1855246584785144,
+      "learning_rate": 9.997399023177285e-06,
+      "loss": 0.6281,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7859352018877166,
+      "learning_rate": 9.997376842261937e-06,
+      "loss": 0.6219,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3622723124990364,
+      "learning_rate": 9.997354567196795e-06,
+      "loss": 0.6285,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3522991072812873,
+      "learning_rate": 9.99733219798233e-06,
+      "loss": 0.7292,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4471472649086032,
+      "learning_rate": 9.997309734619008e-06,
+      "loss": 0.5214,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6144687334566976,
+      "learning_rate": 9.9972871771073e-06,
+      "loss": 0.8389,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.372269291297909,
+      "learning_rate": 9.997264525447681e-06,
+      "loss": 0.6112,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.191396452790055,
+      "learning_rate": 9.99724177964062e-06,
+      "loss": 0.6432,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.188347748793101,
+      "learning_rate": 9.997218939686596e-06,
+      "loss": 0.6194,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8953489888660557,
+      "learning_rate": 9.997196005586088e-06,
+      "loss": 0.4603,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7512750591643003,
+      "learning_rate": 9.997172977339572e-06,
+      "loss": 0.6058,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.246630368798266,
+      "learning_rate": 9.997149854947537e-06,
+      "loss": 0.5535,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.201651057731387,
+      "learning_rate": 9.997126638410461e-06,
+      "loss": 0.5703,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.069565714284901,
+      "learning_rate": 9.99710332772883e-06,
+      "loss": 0.606,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.705983235953863,
+      "learning_rate": 9.997079922903137e-06,
+      "loss": 0.5881,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4273534090576896,
+      "learning_rate": 9.997056423933869e-06,
+      "loss": 0.7336,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.6658697128295898,
+      "eval_runtime": 67.9102,
+      "eval_samples_per_second": 50.213,
+      "eval_steps_per_second": 1.576,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0168920069001905,
+      "learning_rate": 9.997032830821518e-06,
+      "loss": 0.6153,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1541187729239417,
+      "learning_rate": 9.997009143566578e-06,
+      "loss": 0.5278,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.601474143427012,
+      "learning_rate": 9.996985362169541e-06,
+      "loss": 0.5919,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.058763039657333,
+      "learning_rate": 9.996961486630912e-06,
+      "loss": 0.9491,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.886703169738353,
+      "learning_rate": 9.996937516951188e-06,
+      "loss": 0.8351,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6230206659803437,
+      "learning_rate": 9.99691345313087e-06,
+      "loss": 0.6241,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7200151566319013,
+      "learning_rate": 9.996889295170461e-06,
+      "loss": 0.9142,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.238678971449497,
+      "learning_rate": 9.996865043070467e-06,
+      "loss": 0.6616,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5849427305027812,
+      "learning_rate": 9.996840696831398e-06,
+      "loss": 0.5484,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.369711930313695,
+      "learning_rate": 9.996816256453761e-06,
+      "loss": 0.6164,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3403513066493358,
+      "learning_rate": 9.99679172193807e-06,
+      "loss": 0.7873,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4657682138481216,
+      "learning_rate": 9.996767093284837e-06,
+      "loss": 1.0274,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4020258247398387,
+      "learning_rate": 9.996742370494576e-06,
+      "loss": 0.7386,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7801153928242854,
+      "learning_rate": 9.996717553567809e-06,
+      "loss": 0.6403,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.45078042389275,
+      "learning_rate": 9.996692642505052e-06,
+      "loss": 0.6442,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0584175123326336,
+      "learning_rate": 9.996667637306829e-06,
+      "loss": 0.7179,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3800970099346985,
+      "learning_rate": 9.996642537973659e-06,
+      "loss": 0.6733,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7440032518568835,
+      "learning_rate": 9.996617344506072e-06,
+      "loss": 0.7417,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0980338360135042,
+      "learning_rate": 9.996592056904593e-06,
+      "loss": 0.6229,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1145059369826127,
+      "learning_rate": 9.996566675169754e-06,
+      "loss": 0.4595,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5388344841496395,
+      "learning_rate": 9.996541199302084e-06,
+      "loss": 0.7124,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.341775936390086,
+      "learning_rate": 9.996515629302116e-06,
+      "loss": 0.6194,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2196897154339776,
+      "learning_rate": 9.996489965170387e-06,
+      "loss": 0.6029,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.40352923113206,
+      "learning_rate": 9.996464206907433e-06,
+      "loss": 0.5579,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.421033598062775,
+      "learning_rate": 9.996438354513794e-06,
+      "loss": 0.6579,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6597494588333386,
+      "learning_rate": 9.996412407990012e-06,
+      "loss": 0.659,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4875443147352256,
+      "learning_rate": 9.996386367336627e-06,
+      "loss": 0.6954,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1985228725281982,
+      "learning_rate": 9.996360232554187e-06,
+      "loss": 0.6382,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.315922999117987,
+      "learning_rate": 9.99633400364324e-06,
+      "loss": 0.4805,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1911715525183393,
+      "learning_rate": 9.996307680604331e-06,
+      "loss": 0.6065,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.199110260542043,
+      "learning_rate": 9.996281263438015e-06,
+      "loss": 0.6743,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9844122372964668,
+      "learning_rate": 9.996254752144842e-06,
+      "loss": 0.5026,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1034104632701345,
+      "learning_rate": 9.996228146725369e-06,
+      "loss": 0.5101,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.366204038242735,
+      "learning_rate": 9.996201447180154e-06,
+      "loss": 0.655,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6955100779540375,
+      "learning_rate": 9.996174653509752e-06,
+      "loss": 0.7038,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3279497807997984,
+      "learning_rate": 9.996147765714726e-06,
+      "loss": 0.4808,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7546371962419047,
+      "learning_rate": 9.996120783795641e-06,
+      "loss": 0.7778,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3502659032449165,
+      "learning_rate": 9.996093707753061e-06,
+      "loss": 0.5523,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.504104457110341,
+      "learning_rate": 9.99606653758755e-06,
+      "loss": 0.7541,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.76768380939897,
+      "learning_rate": 9.996039273299679e-06,
+      "loss": 0.7306,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3426838030554293,
+      "learning_rate": 9.99601191489002e-06,
+      "loss": 0.5857,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.769690518792717,
+      "learning_rate": 9.995984462359141e-06,
+      "loss": 0.8651,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9144829021380507,
+      "learning_rate": 9.995956915707623e-06,
+      "loss": 0.3772,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5540851108792753,
+      "learning_rate": 9.995929274936037e-06,
+      "loss": 0.6885,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.54684770932808,
+      "learning_rate": 9.995901540044966e-06,
+      "loss": 0.5998,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.287523963310316,
+      "learning_rate": 9.995873711034987e-06,
+      "loss": 0.6769,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.046146510228686,
+      "learning_rate": 9.995845787906685e-06,
+      "loss": 0.5121,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2202492852500595,
+      "learning_rate": 9.995817770660645e-06,
+      "loss": 0.5798,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9984338308888947,
+      "learning_rate": 9.995789659297453e-06,
+      "loss": 0.4919,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9010908273191554,
+      "learning_rate": 9.995761453817693e-06,
+      "loss": 0.5345,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.6606611609458923,
+      "eval_runtime": 67.8752,
+      "eval_samples_per_second": 50.239,
+      "eval_steps_per_second": 1.576,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.064251791472141,
+      "learning_rate": 9.995733154221964e-06,
+      "loss": 0.6122,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0163995434025614,
+      "learning_rate": 9.995704760510852e-06,
+      "loss": 0.5325,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3933089180867064,
+      "learning_rate": 9.995676272684952e-06,
+      "loss": 0.5975,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7451860439942704,
+      "learning_rate": 9.995647690744863e-06,
+      "loss": 0.7493,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9506123573099745,
+      "learning_rate": 9.995619014691182e-06,
+      "loss": 0.7501,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1607447540988147,
+      "learning_rate": 9.99559024452451e-06,
+      "loss": 0.6309,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3466753662829545,
+      "learning_rate": 9.995561380245446e-06,
+      "loss": 0.5157,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.320384602285467,
+      "learning_rate": 9.995532421854599e-06,
+      "loss": 0.6349,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.351615011782294,
+      "learning_rate": 9.99550336935257e-06,
+      "loss": 0.7423,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2384749080766966,
+      "learning_rate": 9.995474222739973e-06,
+      "loss": 0.6422,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.477898815157077,
+      "learning_rate": 9.995444982017417e-06,
+      "loss": 0.7586,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.293258797093615,
+      "learning_rate": 9.995415647185508e-06,
+      "loss": 0.6625,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3597649263107847,
+      "learning_rate": 9.995386218244869e-06,
+      "loss": 0.6294,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.021144217897856,
+      "learning_rate": 9.995356695196109e-06,
+      "loss": 0.6289,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.017116024192549,
+      "learning_rate": 9.995327078039849e-06,
+      "loss": 0.5877,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9410926249314682,
+      "learning_rate": 9.995297366776709e-06,
+      "loss": 0.5386,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.638696124304496,
+      "learning_rate": 9.99526756140731e-06,
+      "loss": 0.5684,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.053392918459048,
+      "learning_rate": 9.995237661932277e-06,
+      "loss": 0.7477,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.182188177226204,
+      "learning_rate": 9.995207668352238e-06,
+      "loss": 0.6444,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.552084416204882,
+      "learning_rate": 9.995177580667817e-06,
+      "loss": 0.6363,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3324785126689322,
+      "learning_rate": 9.995147398879645e-06,
+      "loss": 0.682,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8548126093762227,
+      "learning_rate": 9.995117122988354e-06,
+      "loss": 0.5142,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.806957067827579,
+      "learning_rate": 9.995086752994579e-06,
+      "loss": 0.6773,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0895838357197967,
+      "learning_rate": 9.995056288898954e-06,
+      "loss": 0.4338,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.080314524269538,
+      "learning_rate": 9.995025730702118e-06,
+      "loss": 0.6867,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5464664054810373,
+      "learning_rate": 9.994995078404712e-06,
+      "loss": 0.7357,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0174578110164014,
+      "learning_rate": 9.994964332007372e-06,
+      "loss": 0.5722,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.44282969025536,
+      "learning_rate": 9.994933491510748e-06,
+      "loss": 0.7519,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4216677942436333,
+      "learning_rate": 9.994902556915483e-06,
+      "loss": 0.6534,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.16019936500543,
+      "learning_rate": 9.994871528222227e-06,
+      "loss": 0.4995,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.290227320759237,
+      "learning_rate": 9.994840405431623e-06,
+      "loss": 0.7273,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.443008156258964,
+      "learning_rate": 9.994809188544329e-06,
+      "loss": 0.6201,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.01874816401395,
+      "learning_rate": 9.994777877560999e-06,
+      "loss": 0.4943,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.538086753933511,
+      "learning_rate": 9.994746472482283e-06,
+      "loss": 0.7401,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.496455320217115,
+      "learning_rate": 9.994714973308844e-06,
+      "loss": 0.7147,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.817648395899557,
+      "learning_rate": 9.994683380041335e-06,
+      "loss": 0.5793,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4085574983241074,
+      "learning_rate": 9.994651692680424e-06,
+      "loss": 0.7788,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.044242865208468,
+      "learning_rate": 9.99461991122677e-06,
+      "loss": 0.6608,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8814251841070175,
+      "learning_rate": 9.99458803568104e-06,
+      "loss": 0.5378,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.645345516095354,
+      "learning_rate": 9.994556066043903e-06,
+      "loss": 0.6648,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0808756523593464,
+      "learning_rate": 9.994524002316023e-06,
+      "loss": 0.6328,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3464285372917844,
+      "learning_rate": 9.994491844498078e-06,
+      "loss": 0.7786,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.681989261538372,
+      "learning_rate": 9.994459592590737e-06,
+      "loss": 0.7483,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.391499231316183,
+      "learning_rate": 9.994427246594675e-06,
+      "loss": 0.5306,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8875769683393775,
+      "learning_rate": 9.994394806510572e-06,
+      "loss": 0.7462,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6491012154133475,
+      "learning_rate": 9.994362272339104e-06,
+      "loss": 0.6729,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.094750229674224,
+      "learning_rate": 9.994329644080954e-06,
+      "loss": 0.5168,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2964810821744743,
+      "learning_rate": 9.994296921736804e-06,
+      "loss": 0.7366,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.114107032620141,
+      "learning_rate": 9.994264105307339e-06,
+      "loss": 0.7507,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9235412280024546,
+      "learning_rate": 9.994231194793247e-06,
+      "loss": 0.6463,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.6565842032432556,
+      "eval_runtime": 67.9376,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.20433292725093,
+      "learning_rate": 9.994198190195217e-06,
+      "loss": 0.7415,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2172787442093314,
+      "learning_rate": 9.994165091513937e-06,
+      "loss": 0.6688,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5307405215936534,
+      "learning_rate": 9.994131898750105e-06,
+      "loss": 0.6319,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.238102744212973,
+      "learning_rate": 9.994098611904412e-06,
+      "loss": 0.6626,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.041950671769457,
+      "learning_rate": 9.994065230977556e-06,
+      "loss": 0.6065,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2699410467228276,
+      "learning_rate": 9.994031755970235e-06,
+      "loss": 0.62,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.898894856248726,
+      "learning_rate": 9.993998186883152e-06,
+      "loss": 0.68,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.815171922712754,
+      "learning_rate": 9.993964523717008e-06,
+      "loss": 0.7729,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1919474393403724,
+      "learning_rate": 9.993930766472506e-06,
+      "loss": 0.637,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9987075773222633,
+      "learning_rate": 9.993896915150358e-06,
+      "loss": 0.4754,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1288024724605363,
+      "learning_rate": 9.993862969751267e-06,
+      "loss": 0.5997,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.913084371347352,
+      "learning_rate": 9.993828930275948e-06,
+      "loss": 0.7782,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8554189140711022,
+      "learning_rate": 9.993794796725109e-06,
+      "loss": 0.4198,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.615484903453214,
+      "learning_rate": 9.993760569099468e-06,
+      "loss": 0.7453,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.229538900931914,
+      "learning_rate": 9.993726247399742e-06,
+      "loss": 0.5927,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.69484574915474,
+      "learning_rate": 9.993691831626647e-06,
+      "loss": 0.6828,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.472589452289921,
+      "learning_rate": 9.993657321780904e-06,
+      "loss": 0.5923,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4020096258771715,
+      "learning_rate": 9.993622717863235e-06,
+      "loss": 0.8281,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2527070886583744,
+      "learning_rate": 9.99358801987437e-06,
+      "loss": 0.5903,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2734929361439304,
+      "learning_rate": 9.993553227815027e-06,
+      "loss": 0.5554,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1617016283482946,
+      "learning_rate": 9.99351834168594e-06,
+      "loss": 0.5761,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.363906516727502,
+      "learning_rate": 9.993483361487836e-06,
+      "loss": 0.6069,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.454873245094785,
+      "learning_rate": 9.99344828722145e-06,
+      "loss": 0.5419,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.487327251989431,
+      "learning_rate": 9.993413118887515e-06,
+      "loss": 0.5579,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7334931472912642,
+      "learning_rate": 9.993377856486767e-06,
+      "loss": 0.7702,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8604364886414446,
+      "learning_rate": 9.993342500019945e-06,
+      "loss": 0.5522,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.014530951976946,
+      "learning_rate": 9.993307049487789e-06,
+      "loss": 0.6245,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5376742111548385,
+      "learning_rate": 9.99327150489104e-06,
+      "loss": 0.636,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.257690655792207,
+      "learning_rate": 9.993235866230441e-06,
+      "loss": 0.666,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9636972890915017,
+      "learning_rate": 9.993200133506744e-06,
+      "loss": 0.5718,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.338732026865895,
+      "learning_rate": 9.993164306720689e-06,
+      "loss": 0.7897,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4246181929712027,
+      "learning_rate": 9.993128385873033e-06,
+      "loss": 0.7261,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.059293623824566,
+      "learning_rate": 9.993092370964526e-06,
+      "loss": 0.5417,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6798981984205352,
+      "learning_rate": 9.993056261995917e-06,
+      "loss": 0.6192,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.673750078028112,
+      "learning_rate": 9.99302005896797e-06,
+      "loss": 0.6248,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8042533447052698,
+      "learning_rate": 9.992983761881436e-06,
+      "loss": 0.5674,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.24206087482463,
+      "learning_rate": 9.992947370737079e-06,
+      "loss": 0.4653,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6414901344242483,
+      "learning_rate": 9.99291088553566e-06,
+      "loss": 0.7076,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7948974028530018,
+      "learning_rate": 9.99287430627794e-06,
+      "loss": 0.4468,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.770958834139863,
+      "learning_rate": 9.992837632964689e-06,
+      "loss": 0.496,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3549767262674464,
+      "learning_rate": 9.992800865596672e-06,
+      "loss": 0.7437,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9573595027876842,
+      "learning_rate": 9.992764004174658e-06,
+      "loss": 0.6455,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2384608869473293,
+      "learning_rate": 9.992727048699421e-06,
+      "loss": 0.593,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.623678320595085,
+      "learning_rate": 9.992689999171735e-06,
+      "loss": 0.764,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0042278291926565,
+      "learning_rate": 9.992652855592375e-06,
+      "loss": 0.5179,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.754559099155479,
+      "learning_rate": 9.992615617962115e-06,
+      "loss": 0.6647,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9883700774252795,
+      "learning_rate": 9.992578286281742e-06,
+      "loss": 0.6152,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.293132266439996,
+      "learning_rate": 9.992540860552027e-06,
+      "loss": 0.6397,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1834736580096417,
+      "learning_rate": 9.992503340773763e-06,
+      "loss": 0.6858,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4588714810130092,
+      "learning_rate": 9.992465726947733e-06,
+      "loss": 0.6131,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.6518715620040894,
+      "eval_runtime": 67.8944,
+      "eval_samples_per_second": 50.225,
+      "eval_steps_per_second": 1.576,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7871705352100364,
+      "learning_rate": 9.992428019074722e-06,
+      "loss": 0.4888,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.787782045011712,
+      "learning_rate": 9.992390217155521e-06,
+      "loss": 0.6377,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.868472354316693,
+      "learning_rate": 9.99235232119092e-06,
+      "loss": 0.7609,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.640381127279276,
+      "learning_rate": 9.992314331181716e-06,
+      "loss": 0.7488,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.859385776033848,
+      "learning_rate": 9.9922762471287e-06,
+      "loss": 0.4763,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4241679965214518,
+      "learning_rate": 9.99223806903267e-06,
+      "loss": 0.6265,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4381380286983436,
+      "learning_rate": 9.992199796894429e-06,
+      "loss": 0.5366,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4581473680479866,
+      "learning_rate": 9.992161430714774e-06,
+      "loss": 0.615,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4703742314349437,
+      "learning_rate": 9.99212297049451e-06,
+      "loss": 0.5404,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.010906368284515,
+      "learning_rate": 9.992084416234442e-06,
+      "loss": 0.771,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7468945027982397,
+      "learning_rate": 9.992045767935374e-06,
+      "loss": 0.6422,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2884544492832677,
+      "learning_rate": 9.992007025598121e-06,
+      "loss": 0.6754,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.415734269215375,
+      "learning_rate": 9.99196818922349e-06,
+      "loss": 0.63,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.139791697686822,
+      "learning_rate": 9.991929258812295e-06,
+      "loss": 0.6746,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2254137034651227,
+      "learning_rate": 9.99189023436535e-06,
+      "loss": 0.5945,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.18098020392941,
+      "learning_rate": 9.991851115883476e-06,
+      "loss": 0.5798,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.324781745185156,
+      "learning_rate": 9.991811903367487e-06,
+      "loss": 0.5625,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.50362971124322,
+      "learning_rate": 9.991772596818205e-06,
+      "loss": 0.6038,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0271060712796243,
+      "learning_rate": 9.991733196236454e-06,
+      "loss": 0.6115,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6809111514401773,
+      "learning_rate": 9.99169370162306e-06,
+      "loss": 0.6651,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2159917754220406,
+      "learning_rate": 9.991654112978847e-06,
+      "loss": 0.5243,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.369885901041182,
+      "learning_rate": 9.991614430304644e-06,
+      "loss": 0.5841,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.45750399695914,
+      "learning_rate": 9.991574653601286e-06,
+      "loss": 0.5831,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3733948193719034,
+      "learning_rate": 9.991534782869599e-06,
+      "loss": 0.5716,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9425820245149448,
+      "learning_rate": 9.991494818110424e-06,
+      "loss": 0.5099,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2106701973313485,
+      "learning_rate": 9.991454759324593e-06,
+      "loss": 0.7021,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8695708805064744,
+      "learning_rate": 9.991414606512947e-06,
+      "loss": 0.5961,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5044758512033396,
+      "learning_rate": 9.991374359676325e-06,
+      "loss": 0.7241,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.213165353931713,
+      "learning_rate": 9.99133401881557e-06,
+      "loss": 0.5624,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5035831474376335,
+      "learning_rate": 9.99129358393153e-06,
+      "loss": 0.599,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.217513295905469,
+      "learning_rate": 9.991253055025046e-06,
+      "loss": 0.614,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2090002455951288,
+      "learning_rate": 9.99121243209697e-06,
+      "loss": 0.722,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.232647657542322,
+      "learning_rate": 9.99117171514815e-06,
+      "loss": 0.5704,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4604768131163244,
+      "learning_rate": 9.99113090417944e-06,
+      "loss": 0.7277,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2881706652602505,
+      "learning_rate": 9.991089999191696e-06,
+      "loss": 0.6362,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.294709374860985,
+      "learning_rate": 9.991049000185768e-06,
+      "loss": 0.6269,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.978221829427301,
+      "learning_rate": 9.991007907162522e-06,
+      "loss": 0.494,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9176245847721651,
+      "learning_rate": 9.990966720122816e-06,
+      "loss": 0.4929,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.424945884984071,
+      "learning_rate": 9.990925439067509e-06,
+      "loss": 0.8066,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8801549024699695,
+      "learning_rate": 9.990884063997468e-06,
+      "loss": 0.6312,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5387282965073874,
+      "learning_rate": 9.99084259491356e-06,
+      "loss": 0.7301,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2610715133847665,
+      "learning_rate": 9.990801031816648e-06,
+      "loss": 0.9098,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2835763506933318,
+      "learning_rate": 9.990759374707608e-06,
+      "loss": 0.6768,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.609446250050464,
+      "learning_rate": 9.99071762358731e-06,
+      "loss": 0.6999,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.170615529739962,
+      "learning_rate": 9.990675778456627e-06,
+      "loss": 0.6299,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.352865587581697,
+      "learning_rate": 9.990633839316437e-06,
+      "loss": 0.4827,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.08842167940634,
+      "learning_rate": 9.990591806167617e-06,
+      "loss": 0.7092,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.413287383191211,
+      "learning_rate": 9.990549679011046e-06,
+      "loss": 0.6131,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8112919972202364,
+      "learning_rate": 9.990507457847606e-06,
+      "loss": 0.6787,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0479165632577847,
+      "learning_rate": 9.990465142678181e-06,
+      "loss": 0.5402,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.6522212624549866,
+      "eval_runtime": 67.9013,
+      "eval_samples_per_second": 50.22,
+      "eval_steps_per_second": 1.576,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.08970084748411,
+      "learning_rate": 9.99042273350366e-06,
+      "loss": 0.6214,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2695937917697355,
+      "learning_rate": 9.990380230324927e-06,
+      "loss": 0.6043,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7604564123262714,
+      "learning_rate": 9.990337633142872e-06,
+      "loss": 0.6547,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5867913144039383,
+      "learning_rate": 9.990294941958386e-06,
+      "loss": 0.6254,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.355606048802719,
+      "learning_rate": 9.99025215677237e-06,
+      "loss": 0.7868,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5333612225795554,
+      "learning_rate": 9.99020927758571e-06,
+      "loss": 0.7175,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5697113396473044,
+      "learning_rate": 9.99016630439931e-06,
+      "loss": 0.5684,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.325057635865044,
+      "learning_rate": 9.990123237214064e-06,
+      "loss": 0.6392,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0746840418034367,
+      "learning_rate": 9.99008007603088e-06,
+      "loss": 0.7299,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.203016795097804,
+      "learning_rate": 9.990036820850657e-06,
+      "loss": 0.609,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1892932466977113,
+      "learning_rate": 9.989993471674302e-06,
+      "loss": 0.6394,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2102433955962195,
+      "learning_rate": 9.989950028502722e-06,
+      "loss": 0.634,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4243272848759574,
+      "learning_rate": 9.989906491336828e-06,
+      "loss": 0.7044,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.308774744531891,
+      "learning_rate": 9.989862860177532e-06,
+      "loss": 0.7242,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.144371897130929,
+      "learning_rate": 9.989819135025743e-06,
+      "loss": 0.4837,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1533212473893406,
+      "learning_rate": 9.98977531588238e-06,
+      "loss": 0.6853,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6325108435635496,
+      "learning_rate": 9.989731402748359e-06,
+      "loss": 0.5863,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4482695806507317,
+      "learning_rate": 9.989687395624602e-06,
+      "loss": 0.832,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1575878229781797,
+      "learning_rate": 9.989643294512025e-06,
+      "loss": 0.7076,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0243358848122583,
+      "learning_rate": 9.989599099411557e-06,
+      "loss": 0.552,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.017133817321932,
+      "learning_rate": 9.989554810324119e-06,
+      "loss": 0.5163,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4478343864678087,
+      "learning_rate": 9.98951042725064e-06,
+      "loss": 0.5822,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.238155885983619,
+      "learning_rate": 9.989465950192049e-06,
+      "loss": 0.589,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3125966247704897,
+      "learning_rate": 9.989421379149276e-06,
+      "loss": 0.5935,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5412420785679055,
+      "learning_rate": 9.989376714123255e-06,
+      "loss": 0.5313,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9516377729890153,
+      "learning_rate": 9.98933195511492e-06,
+      "loss": 0.5709,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3294406347049477,
+      "learning_rate": 9.98928710212521e-06,
+      "loss": 0.5777,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5891099622909026,
+      "learning_rate": 9.989242155155064e-06,
+      "loss": 0.7458,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7743888527727993,
+      "learning_rate": 9.989197114205421e-06,
+      "loss": 0.5039,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.957614617191209,
+      "learning_rate": 9.989151979277225e-06,
+      "loss": 0.7737,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4099935517260365,
+      "learning_rate": 9.98910675037142e-06,
+      "loss": 0.6452,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6311128024246613,
+      "learning_rate": 9.989061427488954e-06,
+      "loss": 0.7877,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.931233966076697,
+      "learning_rate": 9.989016010630774e-06,
+      "loss": 0.5602,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3157605235576524,
+      "learning_rate": 9.988970499797833e-06,
+      "loss": 0.5997,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7921174283916523,
+      "learning_rate": 9.988924894991083e-06,
+      "loss": 0.5653,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.352443218046975,
+      "learning_rate": 9.98887919621148e-06,
+      "loss": 0.5512,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2614595330512826,
+      "learning_rate": 9.988833403459977e-06,
+      "loss": 0.7729,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2463950703252378,
+      "learning_rate": 9.988787516737534e-06,
+      "loss": 0.5812,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7170652788636573,
+      "learning_rate": 9.988741536045114e-06,
+      "loss": 0.605,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.91797249615054,
+      "learning_rate": 9.988695461383679e-06,
+      "loss": 0.6096,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.113424398019019,
+      "learning_rate": 9.98864929275419e-06,
+      "loss": 0.6052,
+      "step": 491
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.890196500554823,
+      "learning_rate": 9.988603030157618e-06,
+      "loss": 0.6217,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.391074827851259,
+      "learning_rate": 9.988556673594927e-06,
+      "loss": 0.7879,
+      "step": 493
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2420076753023146,
+      "learning_rate": 9.988510223067094e-06,
+      "loss": 0.6068,
+      "step": 494
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0419437642147145,
+      "learning_rate": 9.988463678575085e-06,
+      "loss": 0.6223,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7144579660563262,
+      "learning_rate": 9.988417040119877e-06,
+      "loss": 0.5577,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3027137776803275,
+      "learning_rate": 9.988370307702447e-06,
+      "loss": 0.6441,
+      "step": 497
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.262008590322551,
+      "learning_rate": 9.988323481323772e-06,
+      "loss": 0.6704,
+      "step": 498
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.497677466840614,
+      "learning_rate": 9.98827656098483e-06,
+      "loss": 0.7411,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3691246967292354,
+      "learning_rate": 9.98822954668661e-06,
+      "loss": 0.554,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.6504992842674255,
+      "eval_runtime": 67.9099,
+      "eval_samples_per_second": 50.214,
+      "eval_steps_per_second": 1.576,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.405472106731708,
+      "learning_rate": 9.98818243843009e-06,
+      "loss": 0.7273,
+      "step": 501
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.363364834792573,
+      "learning_rate": 9.98813523621626e-06,
+      "loss": 0.6136,
+      "step": 502
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3086886630623944,
+      "learning_rate": 9.988087940046106e-06,
+      "loss": 0.5824,
+      "step": 503
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2217792181635634,
+      "learning_rate": 9.988040549920618e-06,
+      "loss": 0.7046,
+      "step": 504
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0484358846582227,
+      "learning_rate": 9.987993065840788e-06,
+      "loss": 0.5394,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.218094274130376,
+      "learning_rate": 9.98794548780761e-06,
+      "loss": 0.5235,
+      "step": 506
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.292906726213136,
+      "learning_rate": 9.987897815822084e-06,
+      "loss": 0.7544,
+      "step": 507
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.063227667622881,
+      "learning_rate": 9.987850049885202e-06,
+      "loss": 0.5513,
+      "step": 508
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.197294059183431,
+      "learning_rate": 9.987802189997969e-06,
+      "loss": 0.7375,
+      "step": 509
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.840809382342874,
+      "learning_rate": 9.987754236161382e-06,
+      "loss": 0.4653,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4451994351231705,
+      "learning_rate": 9.987706188376449e-06,
+      "loss": 0.6944,
+      "step": 511
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.009012028748248,
+      "learning_rate": 9.987658046644176e-06,
+      "loss": 0.6928,
+      "step": 512
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.395342450898293,
+      "learning_rate": 9.987609810965566e-06,
+      "loss": 0.7105,
+      "step": 513
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.646278819095508,
+      "learning_rate": 9.987561481341633e-06,
+      "loss": 0.7141,
+      "step": 514
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.791136863877273,
+      "learning_rate": 9.98751305777339e-06,
+      "loss": 0.6155,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4815215855317168,
+      "learning_rate": 9.987464540261846e-06,
+      "loss": 0.7977,
+      "step": 516
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7021260455306897,
+      "learning_rate": 9.987415928808018e-06,
+      "loss": 0.6825,
+      "step": 517
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1674093720702854,
+      "learning_rate": 9.987367223412927e-06,
+      "loss": 0.5068,
+      "step": 518
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.655871105143718,
+      "learning_rate": 9.987318424077592e-06,
+      "loss": 0.4538,
+      "step": 519
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2952603411371237,
+      "learning_rate": 9.987269530803029e-06,
+      "loss": 0.6336,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3419293935710934,
+      "learning_rate": 9.987220543590268e-06,
+      "loss": 0.6396,
+      "step": 521
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.664964626639138,
+      "learning_rate": 9.987171462440332e-06,
+      "loss": 0.715,
+      "step": 522
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.869836577814254,
+      "learning_rate": 9.987122287354247e-06,
+      "loss": 0.5719,
+      "step": 523
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8665703285720356,
+      "learning_rate": 9.987073018333045e-06,
+      "loss": 0.5977,
+      "step": 524
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.837536381115212,
+      "learning_rate": 9.987023655377757e-06,
+      "loss": 0.7428,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3483352778417945,
+      "learning_rate": 9.986974198489413e-06,
+      "loss": 0.6177,
+      "step": 526
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.073763742233442,
+      "learning_rate": 9.986924647669054e-06,
+      "loss": 0.5026,
+      "step": 527
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0283157781753087,
+      "learning_rate": 9.986875002917713e-06,
+      "loss": 0.4383,
+      "step": 528
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0901008081945567,
+      "learning_rate": 9.986825264236431e-06,
+      "loss": 0.6042,
+      "step": 529
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2117624063931745,
+      "learning_rate": 9.98677543162625e-06,
+      "loss": 0.5858,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9214905251899106,
+      "learning_rate": 9.98672550508821e-06,
+      "loss": 0.5725,
+      "step": 531
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.206783209453518,
+      "learning_rate": 9.98667548462336e-06,
+      "loss": 0.635,
+      "step": 532
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0874479121017218,
+      "learning_rate": 9.986625370232747e-06,
+      "loss": 0.5397,
+      "step": 533
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0610623860451054,
+      "learning_rate": 9.986575161917417e-06,
+      "loss": 0.5609,
+      "step": 534
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9468213634211353,
+      "learning_rate": 9.986524859678423e-06,
+      "loss": 0.612,
+      "step": 535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.285085718028519,
+      "learning_rate": 9.986474463516816e-06,
+      "loss": 0.523,
+      "step": 536
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.362404192033087,
+      "learning_rate": 9.986423973433654e-06,
+      "loss": 0.5504,
+      "step": 537
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.332330278999187,
+      "learning_rate": 9.986373389429993e-06,
+      "loss": 0.6763,
+      "step": 538
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4169590841794375,
+      "learning_rate": 9.986322711506891e-06,
+      "loss": 0.6294,
+      "step": 539
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4734539184287554,
+      "learning_rate": 9.98627193966541e-06,
+      "loss": 0.5215,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9154770017154674,
+      "learning_rate": 9.986221073906615e-06,
+      "loss": 0.5298,
+      "step": 541
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.512371009062796,
+      "learning_rate": 9.986170114231565e-06,
+      "loss": 0.8539,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.557175211314174,
+      "learning_rate": 9.986119060641331e-06,
+      "loss": 0.662,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6472086608924466,
+      "learning_rate": 9.98606791313698e-06,
+      "loss": 0.7439,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6724439549664525,
+      "learning_rate": 9.986016671719585e-06,
+      "loss": 0.8091,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.482965393087905,
+      "learning_rate": 9.985965336390217e-06,
+      "loss": 0.7429,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.29564750961961,
+      "learning_rate": 9.98591390714995e-06,
+      "loss": 0.6056,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.362789601548628,
+      "learning_rate": 9.985862383999864e-06,
+      "loss": 0.5442,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1308413052942914,
+      "learning_rate": 9.985810766941031e-06,
+      "loss": 0.4233,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6926407912306827,
+      "learning_rate": 9.98575905597454e-06,
+      "loss": 0.9473,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6443497538566589,
+      "eval_runtime": 67.9195,
+      "eval_samples_per_second": 50.206,
+      "eval_steps_per_second": 1.575,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.177123640028348,
+      "learning_rate": 9.985707251101468e-06,
+      "loss": 0.6967,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4204477653919056,
+      "learning_rate": 9.985655352322902e-06,
+      "loss": 0.6838,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.15348045350381,
+      "learning_rate": 9.985603359639926e-06,
+      "loss": 0.5793,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.01803264785735,
+      "learning_rate": 9.985551273053631e-06,
+      "loss": 0.5646,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.055852306597278,
+      "learning_rate": 9.985499092565104e-06,
+      "loss": 0.782,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0223157067593025,
+      "learning_rate": 9.985446818175441e-06,
+      "loss": 0.5892,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4012398250276914,
+      "learning_rate": 9.985394449885735e-06,
+      "loss": 0.6915,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.215520883366924,
+      "learning_rate": 9.985341987697081e-06,
+      "loss": 0.6419,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4854926007534286,
+      "learning_rate": 9.985289431610579e-06,
+      "loss": 0.6772,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2382164211263875,
+      "learning_rate": 9.98523678162733e-06,
+      "loss": 0.6586,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3678497987124096,
+      "learning_rate": 9.985184037748433e-06,
+      "loss": 0.5682,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2080399763674206,
+      "learning_rate": 9.985131199974994e-06,
+      "loss": 0.6416,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8507632453723653,
+      "learning_rate": 9.985078268308122e-06,
+      "loss": 0.6951,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.39046546137784,
+      "learning_rate": 9.98502524274892e-06,
+      "loss": 0.7814,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.214525296290768,
+      "learning_rate": 9.9849721232985e-06,
+      "loss": 0.6893,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.024440610111406,
+      "learning_rate": 9.984918909957975e-06,
+      "loss": 0.6113,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7727876294213005,
+      "learning_rate": 9.984865602728456e-06,
+      "loss": 0.6386,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2613020045969288,
+      "learning_rate": 9.984812201611064e-06,
+      "loss": 0.6361,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9421350266490545,
+      "learning_rate": 9.984758706606914e-06,
+      "loss": 0.7881,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3373288448383134,
+      "learning_rate": 9.984705117717125e-06,
+      "loss": 0.6774,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.536373826454876,
+      "learning_rate": 9.98465143494282e-06,
+      "loss": 0.8546,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1079482223727446,
+      "learning_rate": 9.984597658285124e-06,
+      "loss": 0.7098,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3183843235561463,
+      "learning_rate": 9.98454378774516e-06,
+      "loss": 0.4946,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.815683505512489,
+      "learning_rate": 9.98448982332406e-06,
+      "loss": 0.634,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.542176185918228,
+      "learning_rate": 9.984435765022947e-06,
+      "loss": 0.5731,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0672883161923736,
+      "learning_rate": 9.984381612842959e-06,
+      "loss": 0.5914,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.574572114286184,
+      "learning_rate": 9.984327366785228e-06,
+      "loss": 0.7061,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.019511772210478,
+      "learning_rate": 9.984273026850888e-06,
+      "loss": 0.6728,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2840244279427155,
+      "learning_rate": 9.984218593041078e-06,
+      "loss": 0.6894,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.250796249987599,
+      "learning_rate": 9.984164065356935e-06,
+      "loss": 0.6166,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5100855211950015,
+      "learning_rate": 9.984109443799604e-06,
+      "loss": 0.6332,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2279549205956504,
+      "learning_rate": 9.984054728370225e-06,
+      "loss": 0.5364,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.501395638453589,
+      "learning_rate": 9.983999919069949e-06,
+      "loss": 0.537,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8407534750461616,
+      "learning_rate": 9.983945015899917e-06,
+      "loss": 0.5808,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9865787094135134,
+      "learning_rate": 9.983890018861285e-06,
+      "loss": 0.54,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9581764405984876,
+      "learning_rate": 9.983834927955196e-06,
+      "loss": 0.5483,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.536420932596496,
+      "learning_rate": 9.983779743182808e-06,
+      "loss": 0.4861,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.681205272482087,
+      "learning_rate": 9.983724464545279e-06,
+      "loss": 0.7074,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.67272853179374,
+      "learning_rate": 9.983669092043761e-06,
+      "loss": 0.6521,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.806258801628573,
+      "learning_rate": 9.983613625679415e-06,
+      "loss": 0.572,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3027831545997324,
+      "learning_rate": 9.983558065453406e-06,
+      "loss": 0.6075,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7604003017815137,
+      "learning_rate": 9.983502411366891e-06,
+      "loss": 0.7332,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1382327939546517,
+      "learning_rate": 9.983446663421038e-06,
+      "loss": 0.6586,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1121252010491394,
+      "learning_rate": 9.983390821617013e-06,
+      "loss": 0.5807,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1599065273974003,
+      "learning_rate": 9.98333488595599e-06,
+      "loss": 0.5364,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0711439708168173,
+      "learning_rate": 9.983278856439133e-06,
+      "loss": 0.6117,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.175927692922714,
+      "learning_rate": 9.983222733067617e-06,
+      "loss": 0.834,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9033877329450102,
+      "learning_rate": 9.983166515842618e-06,
+      "loss": 0.5974,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4305234536926945,
+      "learning_rate": 9.983110204765314e-06,
+      "loss": 0.6397,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.275195175764484,
+      "learning_rate": 9.98305379983688e-06,
+      "loss": 0.688,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6451931595802307,
+      "eval_runtime": 67.8503,
+      "eval_samples_per_second": 50.258,
+      "eval_steps_per_second": 1.577,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1186536240817913,
+      "learning_rate": 9.982997301058504e-06,
+      "loss": 0.5881,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8249733075323515,
+      "learning_rate": 9.982940708431359e-06,
+      "loss": 0.6337,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2291453241645773,
+      "learning_rate": 9.982884021956636e-06,
+      "loss": 0.679,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.929939737097305,
+      "learning_rate": 9.98282724163552e-06,
+      "loss": 0.651,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1362331869550104,
+      "learning_rate": 9.982770367469203e-06,
+      "loss": 0.4555,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2954426363590144,
+      "learning_rate": 9.98271339945887e-06,
+      "loss": 0.5857,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9609908759918162,
+      "learning_rate": 9.982656337605718e-06,
+      "loss": 0.5723,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0275352082246614,
+      "learning_rate": 9.982599181910938e-06,
+      "loss": 0.4763,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9763890926034156,
+      "learning_rate": 9.98254193237573e-06,
+      "loss": 0.5225,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.295267087667883,
+      "learning_rate": 9.98248458900129e-06,
+      "loss": 0.6268,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6583726631537057,
+      "learning_rate": 9.982427151788817e-06,
+      "loss": 0.6987,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9113150286588616,
+      "learning_rate": 9.98236962073952e-06,
+      "loss": 0.4919,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9896151462810259,
+      "learning_rate": 9.9823119958546e-06,
+      "loss": 0.5295,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.239772365546213,
+      "learning_rate": 9.982254277135257e-06,
+      "loss": 0.539,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1151281607632364,
+      "learning_rate": 9.98219646458271e-06,
+      "loss": 0.7257,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.285332743337256,
+      "learning_rate": 9.98213855819816e-06,
+      "loss": 0.5574,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3209877137933,
+      "learning_rate": 9.982080557982824e-06,
+      "loss": 0.4949,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.992046264384305,
+      "learning_rate": 9.982022463937916e-06,
+      "loss": 0.6145,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0172502276851394,
+      "learning_rate": 9.981964276064651e-06,
+      "loss": 0.6327,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.375229160522024,
+      "learning_rate": 9.98190599436425e-06,
+      "loss": 0.4989,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2857958236711475,
+      "learning_rate": 9.98184761883793e-06,
+      "loss": 0.7562,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.428010015103615,
+      "learning_rate": 9.981789149486912e-06,
+      "loss": 0.5442,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.316841385505971,
+      "learning_rate": 9.981730586312423e-06,
+      "loss": 0.7041,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3365056496974272,
+      "learning_rate": 9.981671929315688e-06,
+      "loss": 0.5545,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.212030857950589,
+      "learning_rate": 9.981613178497935e-06,
+      "loss": 0.7088,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2941873895932505,
+      "learning_rate": 9.981554333860395e-06,
+      "loss": 0.6141,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.285490041503892,
+      "learning_rate": 9.981495395404294e-06,
+      "loss": 0.6293,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5967012632647313,
+      "learning_rate": 9.981436363130876e-06,
+      "loss": 0.8976,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9055165450575688,
+      "learning_rate": 9.981377237041367e-06,
+      "loss": 0.4469,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7080735434106424,
+      "learning_rate": 9.981318017137011e-06,
+      "loss": 0.4687,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1817093097464846,
+      "learning_rate": 9.981258703419047e-06,
+      "loss": 0.6142,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2046353379492554,
+      "learning_rate": 9.981199295888713e-06,
+      "loss": 0.6934,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8185635689357702,
+      "learning_rate": 9.981139794547257e-06,
+      "loss": 0.4543,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7280285135456297,
+      "learning_rate": 9.981080199395922e-06,
+      "loss": 0.4116,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7144905730334408,
+      "learning_rate": 9.981020510435955e-06,
+      "loss": 0.5827,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0488674996430474,
+      "learning_rate": 9.980960727668609e-06,
+      "loss": 0.5822,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6770744410348914,
+      "learning_rate": 9.980900851095132e-06,
+      "loss": 0.7381,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.195537285985906,
+      "learning_rate": 9.980840880716782e-06,
+      "loss": 0.7179,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.301114102184674,
+      "learning_rate": 9.980780816534806e-06,
+      "loss": 0.5168,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.148388030955687,
+      "learning_rate": 9.98072065855047e-06,
+      "loss": 0.666,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1651677041251807,
+      "learning_rate": 9.98066040676503e-06,
+      "loss": 0.6231,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2238992509430373,
+      "learning_rate": 9.980600061179747e-06,
+      "loss": 0.5788,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8499697601129383,
+      "learning_rate": 9.980539621795886e-06,
+      "loss": 0.6121,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6277801603835114,
+      "learning_rate": 9.980479088614708e-06,
+      "loss": 0.6927,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1539046968483495,
+      "learning_rate": 9.980418461637486e-06,
+      "loss": 0.5085,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2989171554721484,
+      "learning_rate": 9.980357740865484e-06,
+      "loss": 0.5266,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.347902835238327,
+      "learning_rate": 9.980296926299977e-06,
+      "loss": 0.7149,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.461024052882349,
+      "learning_rate": 9.980236017942236e-06,
+      "loss": 0.8497,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.468033835717077,
+      "learning_rate": 9.980175015793537e-06,
+      "loss": 0.7545,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.99185325496491,
+      "learning_rate": 9.980113919855158e-06,
+      "loss": 0.8131,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6392638683319092,
+      "eval_runtime": 67.9492,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.207629933455369,
+      "learning_rate": 9.980052730128375e-06,
+      "loss": 0.7467,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.090440510207781,
+      "learning_rate": 9.979991446614472e-06,
+      "loss": 0.5586,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5955629387147274,
+      "learning_rate": 9.979930069314732e-06,
+      "loss": 0.7517,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2062136484888426,
+      "learning_rate": 9.979868598230436e-06,
+      "loss": 0.5059,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7582004325430232,
+      "learning_rate": 9.979807033362874e-06,
+      "loss": 0.4443,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.437207939129225,
+      "learning_rate": 9.979745374713334e-06,
+      "loss": 0.6665,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.981607127798861,
+      "learning_rate": 9.979683622283108e-06,
+      "loss": 0.669,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7307106509146615,
+      "learning_rate": 9.979621776073486e-06,
+      "loss": 0.5008,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.121220175818491,
+      "learning_rate": 9.979559836085768e-06,
+      "loss": 0.6091,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.413798278519164,
+      "learning_rate": 9.979497802321244e-06,
+      "loss": 0.6034,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5266214616769913,
+      "learning_rate": 9.979435674781217e-06,
+      "loss": 0.6833,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.048548654611901,
+      "learning_rate": 9.979373453466987e-06,
+      "loss": 0.507,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4991070412288674,
+      "learning_rate": 9.979311138379855e-06,
+      "loss": 0.6183,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.275967177840959,
+      "learning_rate": 9.979248729521128e-06,
+      "loss": 0.5542,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0316750105877057,
+      "learning_rate": 9.97918622689211e-06,
+      "loss": 0.5691,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1701537608822963,
+      "learning_rate": 9.97912363049411e-06,
+      "loss": 0.6717,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2471499344661225,
+      "learning_rate": 9.979060940328439e-06,
+      "loss": 0.8093,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2495140570891414,
+      "learning_rate": 9.978998156396409e-06,
+      "loss": 0.6124,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1973285605065795,
+      "learning_rate": 9.978935278699335e-06,
+      "loss": 0.5656,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1762172150969916,
+      "learning_rate": 9.978872307238534e-06,
+      "loss": 0.5757,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.04420807236556,
+      "learning_rate": 9.97880924201532e-06,
+      "loss": 0.6423,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.520611755648099,
+      "learning_rate": 9.97874608303102e-06,
+      "loss": 0.7428,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5078601286567723,
+      "learning_rate": 9.978682830286952e-06,
+      "loss": 0.7222,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4727142043449204,
+      "learning_rate": 9.97861948378444e-06,
+      "loss": 0.6576,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2425779800595675,
+      "learning_rate": 9.978556043524811e-06,
+      "loss": 0.663,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1563152418969653,
+      "learning_rate": 9.978492509509392e-06,
+      "loss": 0.5491,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8852420960636949,
+      "learning_rate": 9.978428881739515e-06,
+      "loss": 0.5004,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9819051556333358,
+      "learning_rate": 9.97836516021651e-06,
+      "loss": 0.6308,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9315826919583614,
+      "learning_rate": 9.978301344941713e-06,
+      "loss": 0.5028,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7377277470727808,
+      "learning_rate": 9.978237435916458e-06,
+      "loss": 0.5135,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1605995811108367,
+      "learning_rate": 9.978173433142087e-06,
+      "loss": 0.623,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.226848124048226,
+      "learning_rate": 9.978109336619933e-06,
+      "loss": 0.5847,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0983858565209483,
+      "learning_rate": 9.978045146351343e-06,
+      "loss": 0.6581,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8324796703744708,
+      "learning_rate": 9.977980862337659e-06,
+      "loss": 0.5768,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2637354442864943,
+      "learning_rate": 9.977916484580228e-06,
+      "loss": 0.7897,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3674465504562203,
+      "learning_rate": 9.977852013080395e-06,
+      "loss": 0.6234,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2824333960142207,
+      "learning_rate": 9.977787447839512e-06,
+      "loss": 0.7705,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.164811638872485,
+      "learning_rate": 9.97772278885893e-06,
+      "loss": 0.6,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.057821317606659,
+      "learning_rate": 9.977658036140004e-06,
+      "loss": 0.5209,
+      "step": 689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.919095439236305,
+      "learning_rate": 9.977593189684087e-06,
+      "loss": 0.5353,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.183742593882193,
+      "learning_rate": 9.977528249492537e-06,
+      "loss": 0.5326,
+      "step": 691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0847137263835664,
+      "learning_rate": 9.977463215566715e-06,
+      "loss": 0.4588,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.145687299390185,
+      "learning_rate": 9.977398087907983e-06,
+      "loss": 0.5429,
+      "step": 693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.01251968508159,
+      "learning_rate": 9.9773328665177e-06,
+      "loss": 0.5438,
+      "step": 694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8698992139852204,
+      "learning_rate": 9.977267551397235e-06,
+      "loss": 0.5111,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9755641585770887,
+      "learning_rate": 9.977202142547956e-06,
+      "loss": 0.5951,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7917795019360045,
+      "learning_rate": 9.977136639971232e-06,
+      "loss": 0.5434,
+      "step": 697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.808807271800855,
+      "learning_rate": 9.977071043668433e-06,
+      "loss": 0.6074,
+      "step": 698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0407251290378556,
+      "learning_rate": 9.977005353640931e-06,
+      "loss": 0.5065,
+      "step": 699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.368419092618227,
+      "learning_rate": 9.976939569890102e-06,
+      "loss": 0.5037,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6368488669395447,
+      "eval_runtime": 67.9507,
+      "eval_samples_per_second": 50.183,
+      "eval_steps_per_second": 1.575,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0635159802978618,
+      "learning_rate": 9.976873692417326e-06,
+      "loss": 0.6025,
+      "step": 701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.162998906617707,
+      "learning_rate": 9.976807721223979e-06,
+      "loss": 0.5899,
+      "step": 702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9037301654105396,
+      "learning_rate": 9.976741656311442e-06,
+      "loss": 0.4791,
+      "step": 703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7017199335457494,
+      "learning_rate": 9.9766754976811e-06,
+      "loss": 0.5524,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3321051828751003,
+      "learning_rate": 9.976609245334338e-06,
+      "loss": 0.6006,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2905888122407716,
+      "learning_rate": 9.97654289927254e-06,
+      "loss": 0.6276,
+      "step": 706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0419151166296077,
+      "learning_rate": 9.9764764594971e-06,
+      "loss": 0.6177,
+      "step": 707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1812091086119194,
+      "learning_rate": 9.976409926009403e-06,
+      "loss": 0.6506,
+      "step": 708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.406639299046162,
+      "learning_rate": 9.976343298810846e-06,
+      "loss": 0.5612,
+      "step": 709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.399483288491632,
+      "learning_rate": 9.976276577902822e-06,
+      "loss": 0.6125,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.563069667817887,
+      "learning_rate": 9.976209763286728e-06,
+      "loss": 0.5344,
+      "step": 711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.148235447847774,
+      "learning_rate": 9.976142854963962e-06,
+      "loss": 0.4856,
+      "step": 712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.018633889182592,
+      "learning_rate": 9.976075852935926e-06,
+      "loss": 0.6835,
+      "step": 713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3748897662723607,
+      "learning_rate": 9.976008757204023e-06,
+      "loss": 0.6735,
+      "step": 714
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7132412778987918,
+      "learning_rate": 9.975941567769657e-06,
+      "loss": 0.4247,
+      "step": 715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.279286143406438,
+      "learning_rate": 9.975874284634234e-06,
+      "loss": 0.667,
+      "step": 716
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.186352898297302,
+      "learning_rate": 9.975806907799162e-06,
+      "loss": 0.7223,
+      "step": 717
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.398329005148641,
+      "learning_rate": 9.975739437265853e-06,
+      "loss": 0.782,
+      "step": 718
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5747947141957126,
+      "learning_rate": 9.97567187303572e-06,
+      "loss": 0.65,
+      "step": 719
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0261682949681328,
+      "learning_rate": 9.975604215110174e-06,
+      "loss": 0.5306,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3030820683078868,
+      "learning_rate": 9.975536463490635e-06,
+      "loss": 0.7732,
+      "step": 721
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6070520818430776,
+      "learning_rate": 9.975468618178521e-06,
+      "loss": 0.6264,
+      "step": 722
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3995755404023225,
+      "learning_rate": 9.97540067917525e-06,
+      "loss": 0.5112,
+      "step": 723
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.207359372202964,
+      "learning_rate": 9.975332646482246e-06,
+      "loss": 0.5763,
+      "step": 724
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5758495076797083,
+      "learning_rate": 9.975264520100933e-06,
+      "loss": 0.6109,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7141837057226805,
+      "learning_rate": 9.975196300032736e-06,
+      "loss": 0.4644,
+      "step": 726
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.128515331441306,
+      "learning_rate": 9.975127986279085e-06,
+      "loss": 0.5851,
+      "step": 727
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.036249993810959,
+      "learning_rate": 9.975059578841408e-06,
+      "loss": 0.4479,
+      "step": 728
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9133979087587,
+      "learning_rate": 9.974991077721138e-06,
+      "loss": 0.5359,
+      "step": 729
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.037066248905105,
+      "learning_rate": 9.974922482919711e-06,
+      "loss": 0.5321,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5630233407394916,
+      "learning_rate": 9.974853794438562e-06,
+      "loss": 0.6173,
+      "step": 731
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.194607821384316,
+      "learning_rate": 9.974785012279125e-06,
+      "loss": 0.5854,
+      "step": 732
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8477010162711027,
+      "learning_rate": 9.974716136442847e-06,
+      "loss": 0.6149,
+      "step": 733
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.650463218009841,
+      "learning_rate": 9.974647166931162e-06,
+      "loss": 0.5785,
+      "step": 734
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3982426982424743,
+      "learning_rate": 9.974578103745521e-06,
+      "loss": 0.6599,
+      "step": 735
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8924557869033802,
+      "learning_rate": 9.974508946887367e-06,
+      "loss": 0.5518,
+      "step": 736
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.126589901397435,
+      "learning_rate": 9.974439696358146e-06,
+      "loss": 0.4312,
+      "step": 737
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.657117426162493,
+      "learning_rate": 9.974370352159308e-06,
+      "loss": 0.6909,
+      "step": 738
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3246283639165912,
+      "learning_rate": 9.974300914292309e-06,
+      "loss": 0.7905,
+      "step": 739
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9806519395059001,
+      "learning_rate": 9.974231382758595e-06,
+      "loss": 0.4941,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1364454311214733,
+      "learning_rate": 9.974161757559629e-06,
+      "loss": 0.596,
+      "step": 741
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1879926129321627,
+      "learning_rate": 9.974092038696866e-06,
+      "loss": 0.6568,
+      "step": 742
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5949375420000003,
+      "learning_rate": 9.974022226171763e-06,
+      "loss": 0.7234,
+      "step": 743
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.718435939465915,
+      "learning_rate": 9.973952319985786e-06,
+      "loss": 0.5814,
+      "step": 744
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.145432121633336,
+      "learning_rate": 9.973882320140395e-06,
+      "loss": 0.6605,
+      "step": 745
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9435797580189407,
+      "learning_rate": 9.973812226637057e-06,
+      "loss": 0.7005,
+      "step": 746
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1369700844207897,
+      "learning_rate": 9.973742039477238e-06,
+      "loss": 0.7271,
+      "step": 747
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2325543406869857,
+      "learning_rate": 9.973671758662408e-06,
+      "loss": 0.6176,
+      "step": 748
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.613833315537388,
+      "learning_rate": 9.97360138419404e-06,
+      "loss": 0.6817,
+      "step": 749
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8826692448968196,
+      "learning_rate": 9.973530916073602e-06,
+      "loss": 0.3843,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6310673952102661,
+      "eval_runtime": 67.9126,
+      "eval_samples_per_second": 50.212,
+      "eval_steps_per_second": 1.576,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.108949813892191,
+      "learning_rate": 9.973460354302577e-06,
+      "loss": 0.6581,
+      "step": 751
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.229616271203554,
+      "learning_rate": 9.973389698882436e-06,
+      "loss": 0.5165,
+      "step": 752
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3660743253568555,
+      "learning_rate": 9.973318949814662e-06,
+      "loss": 0.4858,
+      "step": 753
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1097840638443275,
+      "learning_rate": 9.973248107100732e-06,
+      "loss": 0.544,
+      "step": 754
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.212078935799179,
+      "learning_rate": 9.973177170742132e-06,
+      "loss": 0.5255,
+      "step": 755
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2712741756316586,
+      "learning_rate": 9.973106140740346e-06,
+      "loss": 0.4988,
+      "step": 756
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2867069442815082,
+      "learning_rate": 9.973035017096862e-06,
+      "loss": 0.6246,
+      "step": 757
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4955043984371925,
+      "learning_rate": 9.972963799813169e-06,
+      "loss": 0.7721,
+      "step": 758
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5411227429599417,
+      "learning_rate": 9.972892488890755e-06,
+      "loss": 0.613,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8195134513538576,
+      "learning_rate": 9.972821084331116e-06,
+      "loss": 0.3871,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.663578931793509,
+      "learning_rate": 9.972749586135747e-06,
+      "loss": 0.5229,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.101338297866585,
+      "learning_rate": 9.972677994306143e-06,
+      "loss": 0.5124,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3701421629103403,
+      "learning_rate": 9.972606308843801e-06,
+      "loss": 0.7753,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.377595799050022,
+      "learning_rate": 9.972534529750226e-06,
+      "loss": 0.6101,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.776249324780756,
+      "learning_rate": 9.972462657026919e-06,
+      "loss": 0.7263,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.357867619888339,
+      "learning_rate": 9.972390690675384e-06,
+      "loss": 0.4071,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1055758882732545,
+      "learning_rate": 9.972318630697126e-06,
+      "loss": 0.584,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2144345173855755,
+      "learning_rate": 9.972246477093657e-06,
+      "loss": 0.7038,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.102017661980524,
+      "learning_rate": 9.972174229866486e-06,
+      "loss": 0.506,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.538908876116329,
+      "learning_rate": 9.972101889017125e-06,
+      "loss": 0.6609,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9759076788249388,
+      "learning_rate": 9.972029454547088e-06,
+      "loss": 0.5045,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.262686245438969,
+      "learning_rate": 9.971956926457892e-06,
+      "loss": 0.6784,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.281208697862028,
+      "learning_rate": 9.971884304751056e-06,
+      "loss": 0.7235,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.073481733068263,
+      "learning_rate": 9.971811589428098e-06,
+      "loss": 0.63,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8006183764233,
+      "learning_rate": 9.971738780490544e-06,
+      "loss": 0.5299,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4439314354020607,
+      "learning_rate": 9.971665877939916e-06,
+      "loss": 0.5941,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1334776518515755,
+      "learning_rate": 9.97159288177774e-06,
+      "loss": 0.5701,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0577588174238204,
+      "learning_rate": 9.971519792005545e-06,
+      "loss": 0.592,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.132632076925366,
+      "learning_rate": 9.971446608624858e-06,
+      "loss": 0.5098,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7746131783935,
+      "learning_rate": 9.971373331637218e-06,
+      "loss": 0.614,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2439560467775057,
+      "learning_rate": 9.971299961044149e-06,
+      "loss": 0.5394,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.089620523369969,
+      "learning_rate": 9.971226496847195e-06,
+      "loss": 0.5141,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.206594458410942,
+      "learning_rate": 9.971152939047894e-06,
+      "loss": 0.6563,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.329579442026561,
+      "learning_rate": 9.971079287647782e-06,
+      "loss": 0.6839,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7522034537984568,
+      "learning_rate": 9.9710055426484e-06,
+      "loss": 0.6968,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.54243420404521,
+      "learning_rate": 9.970931704051297e-06,
+      "loss": 0.7139,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2779348880126378,
+      "learning_rate": 9.970857771858013e-06,
+      "loss": 0.6274,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2071772248101227,
+      "learning_rate": 9.970783746070101e-06,
+      "loss": 0.5842,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.664430227550503,
+      "learning_rate": 9.970709626689107e-06,
+      "loss": 0.7583,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0215111973989166,
+      "learning_rate": 9.970635413716583e-06,
+      "loss": 0.5661,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0740173213084385,
+      "learning_rate": 9.970561107154085e-06,
+      "loss": 0.4719,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.383281970608379,
+      "learning_rate": 9.970486707003167e-06,
+      "loss": 0.7645,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2181210232983473,
+      "learning_rate": 9.970412213265384e-06,
+      "loss": 0.6779,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5767121269572857,
+      "learning_rate": 9.970337625942301e-06,
+      "loss": 0.7265,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3785201632562987,
+      "learning_rate": 9.970262945035474e-06,
+      "loss": 0.7406,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.147411238906917,
+      "learning_rate": 9.970188170546471e-06,
+      "loss": 0.7284,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.852649856275618,
+      "learning_rate": 9.970113302476853e-06,
+      "loss": 0.538,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6800334629536247,
+      "learning_rate": 9.97003834082819e-06,
+      "loss": 0.4711,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9789799520138143,
+      "learning_rate": 9.969963285602052e-06,
+      "loss": 0.5857,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.78299542863447,
+      "learning_rate": 9.969888136800008e-06,
+      "loss": 0.7206,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6279563307762146,
+      "eval_runtime": 67.9997,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1145992368116526,
+      "learning_rate": 9.969812894423634e-06,
+      "loss": 0.6418,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9459048639682976,
+      "learning_rate": 9.9697375584745e-06,
+      "loss": 0.5376,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9793841326608415,
+      "learning_rate": 9.96966212895419e-06,
+      "loss": 0.5527,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1834859897834895,
+      "learning_rate": 9.969586605864278e-06,
+      "loss": 0.5256,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.481512385042544,
+      "learning_rate": 9.969510989206347e-06,
+      "loss": 0.708,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2764244836070007,
+      "learning_rate": 9.969435278981979e-06,
+      "loss": 0.6281,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4842207461599015,
+      "learning_rate": 9.969359475192759e-06,
+      "loss": 0.7211,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7657718356831424,
+      "learning_rate": 9.969283577840276e-06,
+      "loss": 0.5109,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.238251300436945,
+      "learning_rate": 9.969207586926116e-06,
+      "loss": 0.6274,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3620915519947596,
+      "learning_rate": 9.969131502451871e-06,
+      "loss": 0.6584,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3817858476529308,
+      "learning_rate": 9.969055324419134e-06,
+      "loss": 0.7403,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4932784910350656,
+      "learning_rate": 9.9689790528295e-06,
+      "loss": 0.7788,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9815051353640718,
+      "learning_rate": 9.968902687684567e-06,
+      "loss": 0.5625,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8751284238753625,
+      "learning_rate": 9.96882622898593e-06,
+      "loss": 0.5303,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.27414067900264,
+      "learning_rate": 9.968749676735192e-06,
+      "loss": 0.6259,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.727489081218162,
+      "learning_rate": 9.968673030933953e-06,
+      "loss": 0.6947,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.732106662753845,
+      "learning_rate": 9.968596291583823e-06,
+      "loss": 0.842,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0731341320588177,
+      "learning_rate": 9.968519458686401e-06,
+      "loss": 0.5509,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.111837596861375,
+      "learning_rate": 9.968442532243306e-06,
+      "loss": 0.5649,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4621111757712932,
+      "learning_rate": 9.968365512256138e-06,
+      "loss": 0.6665,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4951830704788587,
+      "learning_rate": 9.968288398726514e-06,
+      "loss": 0.7569,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.170712979200858,
+      "learning_rate": 9.968211191656047e-06,
+      "loss": 0.7921,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0558312831533896,
+      "learning_rate": 9.968133891046355e-06,
+      "loss": 0.7441,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090782460312724,
+      "learning_rate": 9.968056496899054e-06,
+      "loss": 0.5179,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.891226345075196,
+      "learning_rate": 9.967979009215765e-06,
+      "loss": 0.4601,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0156696135510592,
+      "learning_rate": 9.967901427998114e-06,
+      "loss": 0.5159,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3163926791055536,
+      "learning_rate": 9.967823753247719e-06,
+      "loss": 0.7348,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.087606770207695,
+      "learning_rate": 9.967745984966209e-06,
+      "loss": 0.5744,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.934156665669209,
+      "learning_rate": 9.967668123155213e-06,
+      "loss": 0.4751,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.704842523410819,
+      "learning_rate": 9.967590167816358e-06,
+      "loss": 0.6967,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8919657422064255,
+      "learning_rate": 9.96751211895128e-06,
+      "loss": 0.4423,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.178359438202123,
+      "learning_rate": 9.96743397656161e-06,
+      "loss": 0.6551,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5259320535412786,
+      "learning_rate": 9.967355740648983e-06,
+      "loss": 0.7345,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0292290147850576,
+      "learning_rate": 9.96727741121504e-06,
+      "loss": 0.5004,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.156471784510935,
+      "learning_rate": 9.967198988261418e-06,
+      "loss": 0.554,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6117152139556294,
+      "learning_rate": 9.96712047178976e-06,
+      "loss": 0.7543,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1935979579694584,
+      "learning_rate": 9.967041861801712e-06,
+      "loss": 0.7131,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0176104035129727,
+      "learning_rate": 9.966963158298915e-06,
+      "loss": 0.5506,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4114774094430986,
+      "learning_rate": 9.966884361283017e-06,
+      "loss": 0.7172,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1277440469612063,
+      "learning_rate": 9.966805470755672e-06,
+      "loss": 0.4755,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8906100938865775,
+      "learning_rate": 9.966726486718528e-06,
+      "loss": 0.4817,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.406330692439847,
+      "learning_rate": 9.96664740917324e-06,
+      "loss": 0.6599,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4616193252617955,
+      "learning_rate": 9.966568238121463e-06,
+      "loss": 0.6711,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.278262044194951,
+      "learning_rate": 9.966488973564852e-06,
+      "loss": 0.5629,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.585203014257702,
+      "learning_rate": 9.966409615505071e-06,
+      "loss": 0.6616,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3073724951729817,
+      "learning_rate": 9.966330163943778e-06,
+      "loss": 0.7403,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9813990884427044,
+      "learning_rate": 9.966250618882639e-06,
+      "loss": 0.5386,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.200687926201259,
+      "learning_rate": 9.966170980323315e-06,
+      "loss": 0.6677,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0907935434364875,
+      "learning_rate": 9.966091248267476e-06,
+      "loss": 0.5508,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2414032777706177,
+      "learning_rate": 9.96601142271679e-06,
+      "loss": 0.6027,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6308086514472961,
+      "eval_runtime": 67.9969,
+      "eval_samples_per_second": 50.149,
+      "eval_steps_per_second": 1.574,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0377981305251933,
+      "learning_rate": 9.965931503672931e-06,
+      "loss": 0.5107,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.816833492417442,
+      "learning_rate": 9.965851491137568e-06,
+      "loss": 0.7436,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5940100924919562,
+      "learning_rate": 9.965771385112379e-06,
+      "loss": 0.7422,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4224437495016256,
+      "learning_rate": 9.96569118559904e-06,
+      "loss": 0.8086,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8057960056531173,
+      "learning_rate": 9.965610892599228e-06,
+      "loss": 0.5611,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9368584617019102,
+      "learning_rate": 9.965530506114627e-06,
+      "loss": 0.5852,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.288270398188512,
+      "learning_rate": 9.965450026146919e-06,
+      "loss": 0.6523,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.335330161660769,
+      "learning_rate": 9.965369452697787e-06,
+      "loss": 0.6408,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8704481934757975,
+      "learning_rate": 9.965288785768919e-06,
+      "loss": 0.4944,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2659508851419528,
+      "learning_rate": 9.965208025362005e-06,
+      "loss": 0.4909,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.631013602689971,
+      "learning_rate": 9.965127171478733e-06,
+      "loss": 0.5895,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.218483054384295,
+      "learning_rate": 9.965046224120799e-06,
+      "loss": 0.5821,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6019109234168676,
+      "learning_rate": 9.964965183289894e-06,
+      "loss": 0.6335,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8093868176032912,
+      "learning_rate": 9.964884048987718e-06,
+      "loss": 0.6378,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.527457778593169,
+      "learning_rate": 9.964802821215966e-06,
+      "loss": 0.7555,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3482833274679757,
+      "learning_rate": 9.96472149997634e-06,
+      "loss": 0.7974,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2844594255646222,
+      "learning_rate": 9.964640085270543e-06,
+      "loss": 0.6755,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.520281212913232,
+      "learning_rate": 9.964558577100279e-06,
+      "loss": 0.5879,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.094498330249168,
+      "learning_rate": 9.964476975467254e-06,
+      "loss": 0.5329,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8777457939858504,
+      "learning_rate": 9.964395280373176e-06,
+      "loss": 0.5938,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.560232242818329,
+      "learning_rate": 9.964313491819758e-06,
+      "loss": 0.713,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6649061208976588,
+      "learning_rate": 9.964231609808709e-06,
+      "loss": 0.7751,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.318385020829182,
+      "learning_rate": 9.964149634341742e-06,
+      "loss": 0.5726,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6140360696928475,
+      "learning_rate": 9.964067565420577e-06,
+      "loss": 0.6128,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4569090569891427,
+      "learning_rate": 9.96398540304693e-06,
+      "loss": 0.8148,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.162380753009988,
+      "learning_rate": 9.963903147222522e-06,
+      "loss": 0.5746,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3791087764890513,
+      "learning_rate": 9.963820797949073e-06,
+      "loss": 0.6089,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7600235262895394,
+      "learning_rate": 9.963738355228308e-06,
+      "loss": 0.7008,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1066794759389955,
+      "learning_rate": 9.963655819061955e-06,
+      "loss": 0.5411,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4696390319100976,
+      "learning_rate": 9.963573189451738e-06,
+      "loss": 0.6491,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4757148707250285,
+      "learning_rate": 9.963490466399388e-06,
+      "loss": 0.6364,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7046773501977421,
+      "learning_rate": 9.96340764990664e-06,
+      "loss": 0.4118,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1225628585115,
+      "learning_rate": 9.963324739975224e-06,
+      "loss": 0.6737,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.213603711767757,
+      "learning_rate": 9.963241736606875e-06,
+      "loss": 0.7281,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2333315919190904,
+      "learning_rate": 9.963158639803336e-06,
+      "loss": 0.5511,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9124296633191524,
+      "learning_rate": 9.963075449566341e-06,
+      "loss": 0.6199,
+      "step": 886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.133951253338335,
+      "learning_rate": 9.962992165897634e-06,
+      "loss": 0.8055,
+      "step": 887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7103200800254745,
+      "learning_rate": 9.962908788798959e-06,
+      "loss": 0.4117,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5783064938133418,
+      "learning_rate": 9.962825318272058e-06,
+      "loss": 0.7419,
+      "step": 889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3815056965769066,
+      "learning_rate": 9.962741754318683e-06,
+      "loss": 0.671,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.27407417093172,
+      "learning_rate": 9.96265809694058e-06,
+      "loss": 0.6238,
+      "step": 891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0160968467654663,
+      "learning_rate": 9.962574346139503e-06,
+      "loss": 0.56,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8657512366057194,
+      "learning_rate": 9.962490501917202e-06,
+      "loss": 0.575,
+      "step": 893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7243533107844222,
+      "learning_rate": 9.962406564275436e-06,
+      "loss": 0.4732,
+      "step": 894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.648890335826359,
+      "learning_rate": 9.96232253321596e-06,
+      "loss": 0.7891,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9888829200817757,
+      "learning_rate": 9.962238408740533e-06,
+      "loss": 0.5682,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8394941345804725,
+      "learning_rate": 9.962154190850916e-06,
+      "loss": 0.4809,
+      "step": 897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9639702943076944,
+      "learning_rate": 9.962069879548873e-06,
+      "loss": 0.6509,
+      "step": 898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0838405366161674,
+      "learning_rate": 9.961985474836168e-06,
+      "loss": 0.568,
+      "step": 899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5036866904747517,
+      "learning_rate": 9.961900976714571e-06,
+      "loss": 0.7639,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6294571757316589,
+      "eval_runtime": 67.9568,
+      "eval_samples_per_second": 50.179,
+      "eval_steps_per_second": 1.575,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.139648051680403,
+      "learning_rate": 9.961816385185845e-06,
+      "loss": 0.6338,
+      "step": 901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7782139879178342,
+      "learning_rate": 9.961731700251766e-06,
+      "loss": 0.5193,
+      "step": 902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.087614680823697,
+      "learning_rate": 9.961646921914106e-06,
+      "loss": 0.6056,
+      "step": 903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6656617902092663,
+      "learning_rate": 9.961562050174638e-06,
+      "loss": 0.4388,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9751872918028228,
+      "learning_rate": 9.961477085035138e-06,
+      "loss": 0.5451,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1417528127063994,
+      "learning_rate": 9.961392026497388e-06,
+      "loss": 0.7066,
+      "step": 906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.185538405896253,
+      "learning_rate": 9.961306874563168e-06,
+      "loss": 0.7577,
+      "step": 907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4318068993451747,
+      "learning_rate": 9.961221629234257e-06,
+      "loss": 0.6295,
+      "step": 908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3244378622881463,
+      "learning_rate": 9.961136290512445e-06,
+      "loss": 0.6162,
+      "step": 909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.150841136629287,
+      "learning_rate": 9.961050858399513e-06,
+      "loss": 0.5849,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3843968369013226,
+      "learning_rate": 9.960965332897254e-06,
+      "loss": 0.8958,
+      "step": 911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6071747505222116,
+      "learning_rate": 9.960879714007457e-06,
+      "loss": 0.5598,
+      "step": 912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2562083170031215,
+      "learning_rate": 9.960794001731913e-06,
+      "loss": 0.7176,
+      "step": 913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3512410619298505,
+      "learning_rate": 9.960708196072418e-06,
+      "loss": 0.7073,
+      "step": 914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.188737309757373,
+      "learning_rate": 9.960622297030766e-06,
+      "loss": 0.6446,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.849980339636717,
+      "learning_rate": 9.960536304608756e-06,
+      "loss": 0.5798,
+      "step": 916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.136636306565855,
+      "learning_rate": 9.960450218808192e-06,
+      "loss": 0.65,
+      "step": 917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3339713124170363,
+      "learning_rate": 9.960364039630873e-06,
+      "loss": 0.5801,
+      "step": 918
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1820953335730535,
+      "learning_rate": 9.960277767078601e-06,
+      "loss": 0.4984,
+      "step": 919
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.78042841700571,
+      "learning_rate": 9.960191401153185e-06,
+      "loss": 0.49,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.612109326535376,
+      "learning_rate": 9.960104941856433e-06,
+      "loss": 0.4951,
+      "step": 921
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6610498945052639,
+      "learning_rate": 9.960018389190154e-06,
+      "loss": 0.5089,
+      "step": 922
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.038625875848433,
+      "learning_rate": 9.95993174315616e-06,
+      "loss": 0.5739,
+      "step": 923
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.159524271328175,
+      "learning_rate": 9.959845003756265e-06,
+      "loss": 0.7896,
+      "step": 924
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.287739400929301,
+      "learning_rate": 9.959758170992285e-06,
+      "loss": 0.5945,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4819909427191744,
+      "learning_rate": 9.959671244866039e-06,
+      "loss": 0.6637,
+      "step": 926
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.093357496507407,
+      "learning_rate": 9.959584225379345e-06,
+      "loss": 0.6105,
+      "step": 927
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1031281250855294,
+      "learning_rate": 9.959497112534024e-06,
+      "loss": 0.7064,
+      "step": 928
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.346651899418216,
+      "learning_rate": 9.959409906331902e-06,
+      "loss": 0.6091,
+      "step": 929
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.339573888510028,
+      "learning_rate": 9.959322606774804e-06,
+      "loss": 0.5822,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4194602701399877,
+      "learning_rate": 9.959235213864556e-06,
+      "loss": 0.7445,
+      "step": 931
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.156419964484535,
+      "learning_rate": 9.959147727602988e-06,
+      "loss": 0.5302,
+      "step": 932
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0363545210840632,
+      "learning_rate": 9.959060147991933e-06,
+      "loss": 0.7335,
+      "step": 933
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6193988180035626,
+      "learning_rate": 9.958972475033223e-06,
+      "loss": 0.6433,
+      "step": 934
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.482649412591417,
+      "learning_rate": 9.958884708728694e-06,
+      "loss": 0.7197,
+      "step": 935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.118651040813668,
+      "learning_rate": 9.958796849080183e-06,
+      "loss": 0.6124,
+      "step": 936
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.194939910944753,
+      "learning_rate": 9.95870889608953e-06,
+      "loss": 0.6404,
+      "step": 937
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.048353416238605,
+      "learning_rate": 9.958620849758575e-06,
+      "loss": 0.5476,
+      "step": 938
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1433253251043856,
+      "learning_rate": 9.958532710089162e-06,
+      "loss": 0.6612,
+      "step": 939
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.963758492200796,
+      "learning_rate": 9.958444477083135e-06,
+      "loss": 0.6253,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1730731918226103,
+      "learning_rate": 9.958356150742343e-06,
+      "loss": 0.6767,
+      "step": 941
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4068550773367456,
+      "learning_rate": 9.958267731068635e-06,
+      "loss": 0.556,
+      "step": 942
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1690221785677766,
+      "learning_rate": 9.95817921806386e-06,
+      "loss": 0.4987,
+      "step": 943
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1805207040360557,
+      "learning_rate": 9.958090611729873e-06,
+      "loss": 0.7042,
+      "step": 944
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9484823678335592,
+      "learning_rate": 9.958001912068528e-06,
+      "loss": 0.6254,
+      "step": 945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5695019129653063,
+      "learning_rate": 9.957913119081681e-06,
+      "loss": 0.6544,
+      "step": 946
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7301030330419138,
+      "learning_rate": 9.957824232771193e-06,
+      "loss": 0.3714,
+      "step": 947
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0297242955614516,
+      "learning_rate": 9.957735253138924e-06,
+      "loss": 0.6877,
+      "step": 948
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0931713977540913,
+      "learning_rate": 9.957646180186732e-06,
+      "loss": 0.6469,
+      "step": 949
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2900600127507373,
+      "learning_rate": 9.95755701391649e-06,
+      "loss": 0.8297,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6285670399665833,
+      "eval_runtime": 67.9687,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.048787274753643,
+      "learning_rate": 9.957467754330062e-06,
+      "loss": 0.5379,
+      "step": 951
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3780990152820483,
+      "learning_rate": 9.957378401429312e-06,
+      "loss": 0.7173,
+      "step": 952
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1975444648558238,
+      "learning_rate": 9.957288955216116e-06,
+      "loss": 0.6578,
+      "step": 953
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7039136575029559,
+      "learning_rate": 9.957199415692343e-06,
+      "loss": 0.5562,
+      "step": 954
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.53619991433737,
+      "learning_rate": 9.957109782859867e-06,
+      "loss": 0.7014,
+      "step": 955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9750513416887134,
+      "learning_rate": 9.957020056720568e-06,
+      "loss": 0.4224,
+      "step": 956
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.222444922625765,
+      "learning_rate": 9.956930237276324e-06,
+      "loss": 0.5981,
+      "step": 957
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3045595961145064,
+      "learning_rate": 9.95684032452901e-06,
+      "loss": 0.5921,
+      "step": 958
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.445565370897363,
+      "learning_rate": 9.956750318480515e-06,
+      "loss": 0.7007,
+      "step": 959
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.978263308391473,
+      "learning_rate": 9.956660219132719e-06,
+      "loss": 0.5542,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2441434030307197,
+      "learning_rate": 9.95657002648751e-06,
+      "loss": 0.8227,
+      "step": 961
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7141117383062197,
+      "learning_rate": 9.956479740546775e-06,
+      "loss": 0.623,
+      "step": 962
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4589174151964084,
+      "learning_rate": 9.956389361312405e-06,
+      "loss": 0.7594,
+      "step": 963
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5574009167533525,
+      "learning_rate": 9.95629888878629e-06,
+      "loss": 0.3821,
+      "step": 964
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3854264625710586,
+      "learning_rate": 9.956208322970329e-06,
+      "loss": 0.7579,
+      "step": 965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.150059103172721,
+      "learning_rate": 9.956117663866412e-06,
+      "loss": 0.4927,
+      "step": 966
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2954470664128963,
+      "learning_rate": 9.956026911476442e-06,
+      "loss": 0.5568,
+      "step": 967
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.323969846842787,
+      "learning_rate": 9.955936065802314e-06,
+      "loss": 0.6386,
+      "step": 968
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9570420741693249,
+      "learning_rate": 9.955845126845934e-06,
+      "loss": 0.5649,
+      "step": 969
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0712041788320588,
+      "learning_rate": 9.955754094609202e-06,
+      "loss": 0.5759,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.036162931177781,
+      "learning_rate": 9.955662969094027e-06,
+      "loss": 0.6214,
+      "step": 971
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.351008923116886,
+      "learning_rate": 9.955571750302314e-06,
+      "loss": 0.6235,
+      "step": 972
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.187862873081234,
+      "learning_rate": 9.955480438235975e-06,
+      "loss": 0.8439,
+      "step": 973
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.02088841271691,
+      "learning_rate": 9.955389032896921e-06,
+      "loss": 0.5647,
+      "step": 974
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.589067624363356,
+      "learning_rate": 9.955297534287064e-06,
+      "loss": 0.5809,
+      "step": 975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7329843648378696,
+      "learning_rate": 9.95520594240832e-06,
+      "loss": 0.5285,
+      "step": 976
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.262090446738416,
+      "learning_rate": 9.955114257262607e-06,
+      "loss": 0.628,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.013579933645185,
+      "learning_rate": 9.955022478851844e-06,
+      "loss": 0.5342,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1889446497679304,
+      "learning_rate": 9.954930607177952e-06,
+      "loss": 0.623,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.607749909780862,
+      "learning_rate": 9.954838642242856e-06,
+      "loss": 0.6312,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9637086976233142,
+      "learning_rate": 9.954746584048481e-06,
+      "loss": 0.5641,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8129280127669745,
+      "learning_rate": 9.954654432596752e-06,
+      "loss": 0.4714,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8047816572774658,
+      "learning_rate": 9.9545621878896e-06,
+      "loss": 0.5586,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.378979699217867,
+      "learning_rate": 9.954469849928954e-06,
+      "loss": 0.7375,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8349204965902934,
+      "learning_rate": 9.954377418716749e-06,
+      "loss": 0.5687,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1688409144304015,
+      "learning_rate": 9.95428489425492e-06,
+      "loss": 0.5556,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0798863632786513,
+      "learning_rate": 9.954192276545402e-06,
+      "loss": 0.5355,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8594750136425555,
+      "learning_rate": 9.954099565590137e-06,
+      "loss": 0.514,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.096497009992355,
+      "learning_rate": 9.954006761391063e-06,
+      "loss": 0.5619,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0943209045539364,
+      "learning_rate": 9.953913863950125e-06,
+      "loss": 0.5949,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.810080162784594,
+      "learning_rate": 9.953820873269267e-06,
+      "loss": 0.6601,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2580373063682764,
+      "learning_rate": 9.953727789350434e-06,
+      "loss": 0.6141,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9295381672794263,
+      "learning_rate": 9.953634612195577e-06,
+      "loss": 0.4714,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.900891554363039,
+      "learning_rate": 9.953541341806643e-06,
+      "loss": 0.5945,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3487229410615615,
+      "learning_rate": 9.953447978185591e-06,
+      "loss": 0.5641,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.949528783354681,
+      "learning_rate": 9.95335452133437e-06,
+      "loss": 0.696,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.599326439553034,
+      "learning_rate": 9.95326097125494e-06,
+      "loss": 0.8644,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.007601001977275,
+      "learning_rate": 9.953167327949254e-06,
+      "loss": 0.6065,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.192067995419778,
+      "learning_rate": 9.953073591419277e-06,
+      "loss": 0.5689,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8291631871564096,
+      "learning_rate": 9.952979761666972e-06,
+      "loss": 0.4882,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.6287229061126709,
+      "eval_runtime": 67.937,
+      "eval_samples_per_second": 50.194,
+      "eval_steps_per_second": 1.575,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5260299443007117,
+      "learning_rate": 9.9528858386943e-06,
+      "loss": 0.7438,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8054511789352583,
+      "learning_rate": 9.95279182250323e-06,
+      "loss": 0.6094,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.791409292844339,
+      "learning_rate": 9.95269771309573e-06,
+      "loss": 0.5077,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8673331722885285,
+      "learning_rate": 9.952603510473765e-06,
+      "loss": 0.6268,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.434571742368227,
+      "learning_rate": 9.952509214639316e-06,
+      "loss": 0.519,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.375718353878284,
+      "learning_rate": 9.952414825594348e-06,
+      "loss": 0.629,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9264969764064785,
+      "learning_rate": 9.952320343340844e-06,
+      "loss": 0.4974,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1591272662646186,
+      "learning_rate": 9.952225767880777e-06,
+      "loss": 0.5861,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1888790394557334,
+      "learning_rate": 9.95213109921613e-06,
+      "loss": 0.7425,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6380946779600163,
+      "learning_rate": 9.952036337348883e-06,
+      "loss": 0.5876,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.025421841943854,
+      "learning_rate": 9.951941482281022e-06,
+      "loss": 0.519,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1792438522290905,
+      "learning_rate": 9.95184653401453e-06,
+      "loss": 0.6826,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.258142872446322,
+      "learning_rate": 9.951751492551396e-06,
+      "loss": 0.6162,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1984421336275233,
+      "learning_rate": 9.95165635789361e-06,
+      "loss": 0.6367,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.623220727815112,
+      "learning_rate": 9.951561130043162e-06,
+      "loss": 0.6291,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9500155214443915,
+      "learning_rate": 9.951465809002048e-06,
+      "loss": 0.7522,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.153051826662773,
+      "learning_rate": 9.951370394772262e-06,
+      "loss": 0.5618,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.202281617568718,
+      "learning_rate": 9.951274887355801e-06,
+      "loss": 0.7793,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.090408686143678,
+      "learning_rate": 9.951179286754666e-06,
+      "loss": 0.6829,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.232879548758057,
+      "learning_rate": 9.951083592970858e-06,
+      "loss": 0.5299,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6738852169055547,
+      "learning_rate": 9.950987806006379e-06,
+      "loss": 0.7045,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9492206994864434,
+      "learning_rate": 9.950891925863234e-06,
+      "loss": 0.4584,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1089191512582013,
+      "learning_rate": 9.950795952543431e-06,
+      "loss": 0.5616,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.067127242926489,
+      "learning_rate": 9.95069988604898e-06,
+      "loss": 0.537,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5675011284105027,
+      "learning_rate": 9.950603726381892e-06,
+      "loss": 0.5111,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0930265546463085,
+      "learning_rate": 9.950507473544177e-06,
+      "loss": 0.5086,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9603494676635786,
+      "learning_rate": 9.950411127537855e-06,
+      "loss": 0.5095,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4556967157412113,
+      "learning_rate": 9.95031468836494e-06,
+      "loss": 0.7159,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.608753756683961,
+      "learning_rate": 9.950218156027451e-06,
+      "loss": 0.6849,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.546991008582623,
+      "learning_rate": 9.950121530527407e-06,
+      "loss": 0.4137,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3034200719196476,
+      "learning_rate": 9.950024811866835e-06,
+      "loss": 0.6852,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7192315639446,
+      "learning_rate": 9.949928000047757e-06,
+      "loss": 0.7114,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8346513439802037,
+      "learning_rate": 9.949831095072198e-06,
+      "loss": 0.6444,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2627771611611345,
+      "learning_rate": 9.949734096942193e-06,
+      "loss": 0.6287,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5311917332605094,
+      "learning_rate": 9.949637005659764e-06,
+      "loss": 0.755,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0324416732351263,
+      "learning_rate": 9.94953982122695e-06,
+      "loss": 0.5774,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.715109056743983,
+      "learning_rate": 9.949442543645785e-06,
+      "loss": 0.7827,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.083694285920605,
+      "learning_rate": 9.949345172918305e-06,
+      "loss": 0.4616,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.488900404719324,
+      "learning_rate": 9.949247709046545e-06,
+      "loss": 0.5403,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1930418220673515,
+      "learning_rate": 9.949150152032549e-06,
+      "loss": 0.4457,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9411235802521525,
+      "learning_rate": 9.949052501878356e-06,
+      "loss": 0.5491,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3716817561028045,
+      "learning_rate": 9.948954758586014e-06,
+      "loss": 0.813,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9404639752376631,
+      "learning_rate": 9.948856922157567e-06,
+      "loss": 0.6238,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.791743382091998,
+      "learning_rate": 9.948758992595063e-06,
+      "loss": 0.5695,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.017204164407256,
+      "learning_rate": 9.948660969900554e-06,
+      "loss": 0.6294,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2509830416029253,
+      "learning_rate": 9.94856285407609e-06,
+      "loss": 0.7019,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.345596451387822,
+      "learning_rate": 9.948464645123726e-06,
+      "loss": 0.5282,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0739631473585147,
+      "learning_rate": 9.948366343045516e-06,
+      "loss": 0.5116,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2144760275164366,
+      "learning_rate": 9.948267947843523e-06,
+      "loss": 0.6851,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.007579753550482,
+      "learning_rate": 9.948169459519801e-06,
+      "loss": 0.5249,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.6267303824424744,
+      "eval_runtime": 67.9394,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0296622954718555,
+      "learning_rate": 9.948070878076415e-06,
+      "loss": 0.6025,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0924220618325924,
+      "learning_rate": 9.947972203515428e-06,
+      "loss": 0.6161,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7773544089621094,
+      "learning_rate": 9.947873435838904e-06,
+      "loss": 0.4945,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.845739372998572,
+      "learning_rate": 9.947774575048913e-06,
+      "loss": 0.5064,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1711162998164886,
+      "learning_rate": 9.947675621147525e-06,
+      "loss": 0.604,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2029769408847133,
+      "learning_rate": 9.947576574136813e-06,
+      "loss": 0.6393,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9652874788898482,
+      "learning_rate": 9.947477434018843e-06,
+      "loss": 0.6221,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7811592507327663,
+      "learning_rate": 9.947378200795698e-06,
+      "loss": 0.4841,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7967094145003348,
+      "learning_rate": 9.947278874469453e-06,
+      "loss": 0.5257,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.246918709107761,
+      "learning_rate": 9.947179455042186e-06,
+      "loss": 0.6373,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.445397878599672,
+      "learning_rate": 9.947079942515979e-06,
+      "loss": 0.6095,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9520478969577901,
+      "learning_rate": 9.946980336892915e-06,
+      "loss": 0.5168,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3015574783330153,
+      "learning_rate": 9.946880638175082e-06,
+      "loss": 0.685,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.26538404650593,
+      "learning_rate": 9.946780846364563e-06,
+      "loss": 0.7466,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.284132417868166,
+      "learning_rate": 9.946680961463449e-06,
+      "loss": 0.7448,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.219007163001191,
+      "learning_rate": 9.946580983473832e-06,
+      "loss": 0.7161,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.516773036973467,
+      "learning_rate": 9.946480912397805e-06,
+      "loss": 0.595,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8549914613211549,
+      "learning_rate": 9.946380748237459e-06,
+      "loss": 0.475,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9089351454546537,
+      "learning_rate": 9.946280490994896e-06,
+      "loss": 0.4714,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0209247005278326,
+      "learning_rate": 9.946180140672211e-06,
+      "loss": 0.4443,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4980910431930217,
+      "learning_rate": 9.946079697271509e-06,
+      "loss": 0.6192,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6701403932494354,
+      "learning_rate": 9.945979160794889e-06,
+      "loss": 0.3833,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.428042067463832,
+      "learning_rate": 9.945878531244455e-06,
+      "loss": 0.7074,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0480162265201325,
+      "learning_rate": 9.945777808622316e-06,
+      "loss": 0.4613,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.063653639384926,
+      "learning_rate": 9.94567699293058e-06,
+      "loss": 0.441,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.238468658650116,
+      "learning_rate": 9.94557608417136e-06,
+      "loss": 0.8037,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4266687579991038,
+      "learning_rate": 9.945475082346764e-06,
+      "loss": 0.6831,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7308417440550543,
+      "learning_rate": 9.945373987458907e-06,
+      "loss": 0.6924,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.926522871018752,
+      "learning_rate": 9.94527279950991e-06,
+      "loss": 0.5267,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.867532338294738,
+      "learning_rate": 9.945171518501886e-06,
+      "loss": 0.5495,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.129116458990077,
+      "learning_rate": 9.945070144436956e-06,
+      "loss": 0.7099,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0459032208322037,
+      "learning_rate": 9.944968677317245e-06,
+      "loss": 0.5225,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.054530803875191,
+      "learning_rate": 9.944867117144875e-06,
+      "loss": 0.7072,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.06938791146229,
+      "learning_rate": 9.944765463921974e-06,
+      "loss": 0.7115,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9139290597260965,
+      "learning_rate": 9.944663717650667e-06,
+      "loss": 0.6404,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2583839294040864,
+      "learning_rate": 9.944561878333086e-06,
+      "loss": 0.5587,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6787093032503746,
+      "learning_rate": 9.944459945971365e-06,
+      "loss": 0.6228,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.21683789297645,
+      "learning_rate": 9.944357920567633e-06,
+      "loss": 0.724,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.049096504530957,
+      "learning_rate": 9.94425580212403e-06,
+      "loss": 0.7019,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9140833888297024,
+      "learning_rate": 9.944153590642691e-06,
+      "loss": 0.5044,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7592327608971685,
+      "learning_rate": 9.944051286125757e-06,
+      "loss": 0.4858,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4176614071301654,
+      "learning_rate": 9.943948888575371e-06,
+      "loss": 0.6198,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5266957102800633,
+      "learning_rate": 9.943846397993674e-06,
+      "loss": 0.6165,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2864951084964398,
+      "learning_rate": 9.943743814382814e-06,
+      "loss": 0.5511,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.017069778425312,
+      "learning_rate": 9.943641137744936e-06,
+      "loss": 0.6275,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.433932475751596,
+      "learning_rate": 9.94353836808219e-06,
+      "loss": 0.6306,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5583488950113553,
+      "learning_rate": 9.94343550539673e-06,
+      "loss": 0.6862,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.746688940467564,
+      "learning_rate": 9.943332549690707e-06,
+      "loss": 0.6946,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9414710623885516,
+      "learning_rate": 9.943229500966279e-06,
+      "loss": 0.5393,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0171724040242847,
+      "learning_rate": 9.943126359225598e-06,
+      "loss": 0.6273,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.6247506141662598,
+      "eval_runtime": 67.9164,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.239185972430749,
+      "learning_rate": 9.943023124470828e-06,
+      "loss": 0.4896,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9906530070422719,
+      "learning_rate": 9.942919796704128e-06,
+      "loss": 0.6616,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.005470587363348,
+      "learning_rate": 9.942816375927662e-06,
+      "loss": 0.4171,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8567683427523698,
+      "learning_rate": 9.942712862143595e-06,
+      "loss": 0.4819,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2089972573445755,
+      "learning_rate": 9.942609255354092e-06,
+      "loss": 0.5787,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1587408224023212,
+      "learning_rate": 9.942505555561326e-06,
+      "loss": 0.6035,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.323820581866551,
+      "learning_rate": 9.942401762767464e-06,
+      "loss": 0.7876,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2843811608677163,
+      "learning_rate": 9.94229787697468e-06,
+      "loss": 0.6654,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2618710150168053,
+      "learning_rate": 9.942193898185154e-06,
+      "loss": 0.6392,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.968329507507358,
+      "learning_rate": 9.942089826401052e-06,
+      "loss": 0.5269,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1133598927417783,
+      "learning_rate": 9.941985661624558e-06,
+      "loss": 0.5166,
+      "step": 1111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7107680559383813,
+      "learning_rate": 9.941881403857859e-06,
+      "loss": 0.4939,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.997439528279977,
+      "learning_rate": 9.941777053103127e-06,
+      "loss": 0.6673,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.403930988719047,
+      "learning_rate": 9.941672609362553e-06,
+      "loss": 0.5775,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2407361538566293,
+      "learning_rate": 9.94156807263832e-06,
+      "loss": 0.5418,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.928874397115719,
+      "learning_rate": 9.941463442932622e-06,
+      "loss": 0.5548,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4237472300850644,
+      "learning_rate": 9.941358720247642e-06,
+      "loss": 0.7536,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4714435714856813,
+      "learning_rate": 9.941253904585577e-06,
+      "loss": 0.8634,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9152427320064231,
+      "learning_rate": 9.94114899594862e-06,
+      "loss": 0.6104,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2748852044431644,
+      "learning_rate": 9.941043994338968e-06,
+      "loss": 0.6359,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5827578291074964,
+      "learning_rate": 9.940938899758818e-06,
+      "loss": 0.5441,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2200281167824927,
+      "learning_rate": 9.940833712210371e-06,
+      "loss": 0.6882,
+      "step": 1122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0045705853410913,
+      "learning_rate": 9.940728431695828e-06,
+      "loss": 0.6088,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2342441955410544,
+      "learning_rate": 9.940623058217392e-06,
+      "loss": 0.6445,
+      "step": 1124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1410882823005153,
+      "learning_rate": 9.94051759177727e-06,
+      "loss": 0.5933,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0463685583927647,
+      "learning_rate": 9.940412032377674e-06,
+      "loss": 0.7168,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9182619459861021,
+      "learning_rate": 9.940306380020808e-06,
+      "loss": 0.7046,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7083555596834514,
+      "learning_rate": 9.940200634708886e-06,
+      "loss": 0.4962,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0550017520028923,
+      "learning_rate": 9.940094796444121e-06,
+      "loss": 0.5427,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2143189816887543,
+      "learning_rate": 9.939988865228731e-06,
+      "loss": 0.5939,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6166266342775857,
+      "learning_rate": 9.939882841064931e-06,
+      "loss": 0.7737,
+      "step": 1131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2659488740136435,
+      "learning_rate": 9.939776723954942e-06,
+      "loss": 0.6435,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.978186154385361,
+      "learning_rate": 9.939670513900985e-06,
+      "loss": 0.5322,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1726803439788305,
+      "learning_rate": 9.939564210905282e-06,
+      "loss": 0.6512,
+      "step": 1134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1203191410721085,
+      "learning_rate": 9.939457814970061e-06,
+      "loss": 0.578,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2321277252774405,
+      "learning_rate": 9.939351326097547e-06,
+      "loss": 0.7148,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8775037653387314,
+      "learning_rate": 9.93924474428997e-06,
+      "loss": 0.5013,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4574639107526832,
+      "learning_rate": 9.939138069549564e-06,
+      "loss": 0.6119,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.500898953476219,
+      "learning_rate": 9.939031301878559e-06,
+      "loss": 0.5052,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.722607006542359,
+      "learning_rate": 9.938924441279188e-06,
+      "loss": 0.7928,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9841233114038375,
+      "learning_rate": 9.938817487753694e-06,
+      "loss": 0.5486,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.922197189495082,
+      "learning_rate": 9.938710441304312e-06,
+      "loss": 0.5217,
+      "step": 1142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8221533162324066,
+      "learning_rate": 9.938603301933283e-06,
+      "loss": 0.519,
+      "step": 1143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5231087116254427,
+      "learning_rate": 9.938496069642853e-06,
+      "loss": 0.5431,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4750902800023344,
+      "learning_rate": 9.938388744435262e-06,
+      "loss": 0.7054,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9749300012500588,
+      "learning_rate": 9.938281326312762e-06,
+      "loss": 0.5713,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2089356167977616,
+      "learning_rate": 9.938173815277598e-06,
+      "loss": 0.6816,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.197949028110613,
+      "learning_rate": 9.938066211332021e-06,
+      "loss": 0.7038,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6086284335484464,
+      "learning_rate": 9.937958514478285e-06,
+      "loss": 0.6917,
+      "step": 1149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3383052433279707,
+      "learning_rate": 9.937850724718644e-06,
+      "loss": 0.5761,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.6236903071403503,
+      "eval_runtime": 67.9815,
+      "eval_samples_per_second": 50.161,
+      "eval_steps_per_second": 1.574,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.022724397881433,
+      "learning_rate": 9.937742842055356e-06,
+      "loss": 0.6009,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3402000721988063,
+      "learning_rate": 9.937634866490676e-06,
+      "loss": 0.658,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4256705985413043,
+      "learning_rate": 9.937526798026867e-06,
+      "loss": 0.6312,
+      "step": 1153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8744397359693392,
+      "learning_rate": 9.93741863666619e-06,
+      "loss": 0.5499,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.042742412165368,
+      "learning_rate": 9.937310382410913e-06,
+      "loss": 0.5856,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9743869343978122,
+      "learning_rate": 9.937202035263295e-06,
+      "loss": 0.4184,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4356731679896217,
+      "learning_rate": 9.937093595225611e-06,
+      "loss": 0.7002,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0082769516688157,
+      "learning_rate": 9.93698506230013e-06,
+      "loss": 0.6204,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.727009996733485,
+      "learning_rate": 9.93687643648912e-06,
+      "loss": 0.8401,
+      "step": 1159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2715333694279494,
+      "learning_rate": 9.936767717794858e-06,
+      "loss": 0.6471,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.225277480906987,
+      "learning_rate": 9.936658906219621e-06,
+      "loss": 0.6577,
+      "step": 1161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7673950706562713,
+      "learning_rate": 9.936550001765685e-06,
+      "loss": 0.5587,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8427975441028377,
+      "learning_rate": 9.93644100443533e-06,
+      "loss": 0.6659,
+      "step": 1163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7869609309172758,
+      "learning_rate": 9.936331914230837e-06,
+      "loss": 0.5039,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.289388632953859,
+      "learning_rate": 9.936222731154491e-06,
+      "loss": 0.6228,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7829570400497503,
+      "learning_rate": 9.936113455208579e-06,
+      "loss": 0.6485,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.097453063737706,
+      "learning_rate": 9.936004086395387e-06,
+      "loss": 0.6439,
+      "step": 1167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.618474475921416,
+      "learning_rate": 9.935894624717203e-06,
+      "loss": 0.8032,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0659487997240324,
+      "learning_rate": 9.935785070176322e-06,
+      "loss": 0.6265,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.972238383530873,
+      "learning_rate": 9.935675422775036e-06,
+      "loss": 0.6107,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9474583760458168,
+      "learning_rate": 9.93556568251564e-06,
+      "loss": 0.6163,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3605082928963554,
+      "learning_rate": 9.935455849400432e-06,
+      "loss": 0.6136,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6216971710664945,
+      "learning_rate": 9.935345923431708e-06,
+      "loss": 0.4425,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.300126745594844,
+      "learning_rate": 9.935235904611774e-06,
+      "loss": 0.5779,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3101502608239555,
+      "learning_rate": 9.935125792942932e-06,
+      "loss": 0.5618,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5606374396217357,
+      "learning_rate": 9.935015588427486e-06,
+      "loss": 0.56,
+      "step": 1176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.003633389421397,
+      "learning_rate": 9.934905291067742e-06,
+      "loss": 0.7326,
+      "step": 1177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5252824987062845,
+      "learning_rate": 9.93479490086601e-06,
+      "loss": 0.6701,
+      "step": 1178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6433000741223625,
+      "learning_rate": 9.934684417824603e-06,
+      "loss": 0.5405,
+      "step": 1179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0587960379144983,
+      "learning_rate": 9.934573841945831e-06,
+      "loss": 0.5917,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6707087617195597,
+      "learning_rate": 9.93446317323201e-06,
+      "loss": 0.6281,
+      "step": 1181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2876435830503055,
+      "learning_rate": 9.934352411685458e-06,
+      "loss": 0.5865,
+      "step": 1182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.112838394476519,
+      "learning_rate": 9.934241557308491e-06,
+      "loss": 0.5763,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3195668091143022,
+      "learning_rate": 9.934130610103433e-06,
+      "loss": 0.5637,
+      "step": 1184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.499979817849408,
+      "learning_rate": 9.934019570072603e-06,
+      "loss": 0.7916,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.065968957543929,
+      "learning_rate": 9.933908437218328e-06,
+      "loss": 0.6135,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0887191757546475,
+      "learning_rate": 9.933797211542934e-06,
+      "loss": 0.6813,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.34669646036186,
+      "learning_rate": 9.93368589304875e-06,
+      "loss": 0.5915,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.295434275178162,
+      "learning_rate": 9.933574481738105e-06,
+      "loss": 0.6573,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.20493460936106,
+      "learning_rate": 9.933462977613334e-06,
+      "loss": 0.5626,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.240336386431723,
+      "learning_rate": 9.933351380676764e-06,
+      "loss": 0.7053,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.897462557820873,
+      "learning_rate": 9.93323969093074e-06,
+      "loss": 0.5025,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0260885023677786,
+      "learning_rate": 9.933127908377598e-06,
+      "loss": 0.5039,
+      "step": 1193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2964310334731737,
+      "learning_rate": 9.933016033019675e-06,
+      "loss": 0.6329,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4832280999511402,
+      "learning_rate": 9.932904064859315e-06,
+      "loss": 0.7102,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1266540747052347,
+      "learning_rate": 9.932792003898863e-06,
+      "loss": 0.4825,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.516213348796498,
+      "learning_rate": 9.932679850140664e-06,
+      "loss": 0.7296,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0901424399634183,
+      "learning_rate": 9.932567603587062e-06,
+      "loss": 0.4944,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2662096083215544,
+      "learning_rate": 9.932455264240412e-06,
+      "loss": 0.5249,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.148558697095543,
+      "learning_rate": 9.932342832103066e-06,
+      "loss": 0.5558,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6191983222961426,
+      "eval_runtime": 67.949,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0151106372955505,
+      "learning_rate": 9.932230307177374e-06,
+      "loss": 0.6908,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.126503611740913,
+      "learning_rate": 9.932117689465694e-06,
+      "loss": 0.6245,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7534960833760487,
+      "learning_rate": 9.932004978970384e-06,
+      "loss": 0.4368,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.548825964172726,
+      "learning_rate": 9.9318921756938e-06,
+      "loss": 0.6068,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6543833201129368,
+      "learning_rate": 9.931779279638308e-06,
+      "loss": 0.7331,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.055351796635739,
+      "learning_rate": 9.931666290806269e-06,
+      "loss": 0.4925,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.023088790154702,
+      "learning_rate": 9.93155320920005e-06,
+      "loss": 0.5755,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0694773517841036,
+      "learning_rate": 9.931440034822015e-06,
+      "loss": 0.5947,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7632760291519134,
+      "learning_rate": 9.931326767674538e-06,
+      "loss": 0.4271,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.563853608518765,
+      "learning_rate": 9.931213407759985e-06,
+      "loss": 0.5672,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.104101174179593,
+      "learning_rate": 9.931099955080734e-06,
+      "loss": 0.6795,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.099261878818303,
+      "learning_rate": 9.930986409639155e-06,
+      "loss": 0.6842,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.244011362882482,
+      "learning_rate": 9.930872771437629e-06,
+      "loss": 0.8375,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9848614673625375,
+      "learning_rate": 9.930759040478533e-06,
+      "loss": 0.6331,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0115673148486235,
+      "learning_rate": 9.93064521676425e-06,
+      "loss": 0.5427,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5325635673090092,
+      "learning_rate": 9.93053130029716e-06,
+      "loss": 0.6328,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6191850779909283,
+      "learning_rate": 9.930417291079649e-06,
+      "loss": 0.3909,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8322734994081942,
+      "learning_rate": 9.930303189114104e-06,
+      "loss": 0.4416,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.293301751463478,
+      "learning_rate": 9.930188994402915e-06,
+      "loss": 0.5975,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8864726414347588,
+      "learning_rate": 9.93007470694847e-06,
+      "loss": 0.5161,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1014409013911,
+      "learning_rate": 9.929960326753163e-06,
+      "loss": 0.5567,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.118325861896362,
+      "learning_rate": 9.929845853819388e-06,
+      "loss": 0.5835,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6575927289863015,
+      "learning_rate": 9.929731288149542e-06,
+      "loss": 0.3633,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.197563657727865,
+      "learning_rate": 9.929616629746022e-06,
+      "loss": 0.5218,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.104416662540688,
+      "learning_rate": 9.92950187861123e-06,
+      "loss": 0.482,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7483221654145495,
+      "learning_rate": 9.929387034747566e-06,
+      "loss": 0.5024,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8987485871246914,
+      "learning_rate": 9.929272098157436e-06,
+      "loss": 0.4803,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0469143710192528,
+      "learning_rate": 9.929157068843247e-06,
+      "loss": 0.4371,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.199988564376756,
+      "learning_rate": 9.929041946807406e-06,
+      "loss": 0.6511,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2807197311694067,
+      "learning_rate": 9.928926732052321e-06,
+      "loss": 0.5754,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6492246241581903,
+      "learning_rate": 9.928811424580405e-06,
+      "loss": 0.7165,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.168062106156853,
+      "learning_rate": 9.928696024394076e-06,
+      "loss": 0.5548,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.179666758866191,
+      "learning_rate": 9.928580531495745e-06,
+      "loss": 0.5234,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.23315381911021,
+      "learning_rate": 9.928464945887831e-06,
+      "loss": 0.5245,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1181958362427893,
+      "learning_rate": 9.928349267572755e-06,
+      "loss": 0.7092,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6134112839759505,
+      "learning_rate": 9.928233496552936e-06,
+      "loss": 0.4983,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.072025549090961,
+      "learning_rate": 9.928117632830799e-06,
+      "loss": 0.683,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8945368495896027,
+      "learning_rate": 9.92800167640877e-06,
+      "loss": 0.5434,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8014405463907988,
+      "learning_rate": 9.92788562728928e-06,
+      "loss": 0.4937,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2926840972700915,
+      "learning_rate": 9.927769485474749e-06,
+      "loss": 0.7659,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.019067546917606,
+      "learning_rate": 9.927653250967617e-06,
+      "loss": 0.6357,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.025955492766553,
+      "learning_rate": 9.927536923770315e-06,
+      "loss": 0.4961,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5128898072332086,
+      "learning_rate": 9.927420503885276e-06,
+      "loss": 0.5877,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5596281800108858,
+      "learning_rate": 9.927303991314937e-06,
+      "loss": 0.6355,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9961609919573344,
+      "learning_rate": 9.927187386061742e-06,
+      "loss": 0.5843,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8636175408699798,
+      "learning_rate": 9.927070688128128e-06,
+      "loss": 0.4927,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8776047150148427,
+      "learning_rate": 9.92695389751654e-06,
+      "loss": 0.6729,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1479552576633827,
+      "learning_rate": 9.92683701422942e-06,
+      "loss": 0.6854,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.968900760532879,
+      "learning_rate": 9.926720038269219e-06,
+      "loss": 0.632,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.214069343534823,
+      "learning_rate": 9.92660296963838e-06,
+      "loss": 0.5054,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6207420825958252,
+      "eval_runtime": 67.9375,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7086771534247531,
+      "learning_rate": 9.92648580833936e-06,
+      "loss": 0.482,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0282472192472842,
+      "learning_rate": 9.926368554374608e-06,
+      "loss": 0.5785,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.507711432309921,
+      "learning_rate": 9.92625120774658e-06,
+      "loss": 0.6881,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3645726377762792,
+      "learning_rate": 9.926133768457733e-06,
+      "loss": 0.6353,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4680853896324955,
+      "learning_rate": 9.926016236510522e-06,
+      "loss": 0.6939,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.154746517648218,
+      "learning_rate": 9.925898611907411e-06,
+      "loss": 0.5117,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3153016299568425,
+      "learning_rate": 9.925780894650864e-06,
+      "loss": 0.6663,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9510464642548004,
+      "learning_rate": 9.925663084743342e-06,
+      "loss": 0.5088,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8101192285395082,
+      "learning_rate": 9.92554518218731e-06,
+      "loss": 0.5914,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0968583323891727,
+      "learning_rate": 9.92542718698524e-06,
+      "loss": 0.4378,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0592011462351096,
+      "learning_rate": 9.925309099139602e-06,
+      "loss": 0.4247,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.132258792115898,
+      "learning_rate": 9.925190918652865e-06,
+      "loss": 0.6572,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.89090962556597,
+      "learning_rate": 9.925072645527505e-06,
+      "loss": 0.5219,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.558474863457466,
+      "learning_rate": 9.924954279765999e-06,
+      "loss": 0.6067,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.638198524398693,
+      "learning_rate": 9.924835821370821e-06,
+      "loss": 0.7317,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.516728799551422,
+      "learning_rate": 9.924717270344456e-06,
+      "loss": 0.5905,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2889560755202156,
+      "learning_rate": 9.924598626689382e-06,
+      "loss": 0.5798,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0839232354537285,
+      "learning_rate": 9.924479890408084e-06,
+      "loss": 0.6763,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3874844404584685,
+      "learning_rate": 9.924361061503046e-06,
+      "loss": 0.5385,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1689264412582867,
+      "learning_rate": 9.924242139976759e-06,
+      "loss": 0.6216,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6915640451445,
+      "learning_rate": 9.924123125831711e-06,
+      "loss": 0.5205,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8105751514362352,
+      "learning_rate": 9.924004019070394e-06,
+      "loss": 0.5139,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8610051401208638,
+      "learning_rate": 9.923884819695299e-06,
+      "loss": 0.5349,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8565069378718129,
+      "learning_rate": 9.923765527708925e-06,
+      "loss": 0.5799,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.335259836530268,
+      "learning_rate": 9.923646143113764e-06,
+      "loss": 0.6809,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1441785904110366,
+      "learning_rate": 9.923526665912323e-06,
+      "loss": 0.5285,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2219887750418943,
+      "learning_rate": 9.923407096107096e-06,
+      "loss": 0.6441,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1306827129750747,
+      "learning_rate": 9.92328743370059e-06,
+      "loss": 0.6194,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3164406016090373,
+      "learning_rate": 9.923167678695306e-06,
+      "loss": 0.5391,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.267843360126205,
+      "learning_rate": 9.923047831093757e-06,
+      "loss": 0.5111,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8020999191103404,
+      "learning_rate": 9.922927890898447e-06,
+      "loss": 0.4701,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8504760913444163,
+      "learning_rate": 9.922807858111891e-06,
+      "loss": 0.5211,
+      "step": 1282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2491566159736402,
+      "learning_rate": 9.922687732736598e-06,
+      "loss": 0.6979,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.117332029455522,
+      "learning_rate": 9.922567514775087e-06,
+      "loss": 0.6854,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.495555352257943,
+      "learning_rate": 9.922447204229871e-06,
+      "loss": 0.6336,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4606140258002975,
+      "learning_rate": 9.922326801103468e-06,
+      "loss": 0.611,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.272269236584174,
+      "learning_rate": 9.9222063053984e-06,
+      "loss": 0.4938,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8885834422926475,
+      "learning_rate": 9.922085717117192e-06,
+      "loss": 0.6033,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.315735722029618,
+      "learning_rate": 9.921965036262365e-06,
+      "loss": 0.7378,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0569313031739975,
+      "learning_rate": 9.921844262836447e-06,
+      "loss": 0.5031,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.31018289832683,
+      "learning_rate": 9.921723396841964e-06,
+      "loss": 0.6027,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2020165388804105,
+      "learning_rate": 9.92160243828145e-06,
+      "loss": 0.5834,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1192882061790455,
+      "learning_rate": 9.921481387157435e-06,
+      "loss": 0.4803,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7628767156460863,
+      "learning_rate": 9.921360243472452e-06,
+      "loss": 0.5298,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.330811186377573,
+      "learning_rate": 9.921239007229039e-06,
+      "loss": 0.5887,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6050501563554507,
+      "learning_rate": 9.921117678429735e-06,
+      "loss": 0.6761,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9694169991180348,
+      "learning_rate": 9.920996257077077e-06,
+      "loss": 0.6408,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2468195019750197,
+      "learning_rate": 9.92087474317361e-06,
+      "loss": 0.635,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.973875422702964,
+      "learning_rate": 9.920753136721876e-06,
+      "loss": 0.6075,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4221004824084607,
+      "learning_rate": 9.920631437724419e-06,
+      "loss": 0.6425,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.617007851600647,
+      "eval_runtime": 68.0148,
+      "eval_samples_per_second": 50.136,
+      "eval_steps_per_second": 1.573,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9399491500513133,
+      "learning_rate": 9.920509646183789e-06,
+      "loss": 0.5198,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.962904918388579,
+      "learning_rate": 9.920387762102538e-06,
+      "loss": 0.5319,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1405878263697495,
+      "learning_rate": 9.920265785483212e-06,
+      "loss": 0.4978,
+      "step": 1303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7376022913233617,
+      "learning_rate": 9.920143716328368e-06,
+      "loss": 0.5799,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.733809967549822,
+      "learning_rate": 9.920021554640561e-06,
+      "loss": 0.5034,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9100979599765724,
+      "learning_rate": 9.919899300422348e-06,
+      "loss": 0.4281,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.151934188812606,
+      "learning_rate": 9.919776953676288e-06,
+      "loss": 0.6444,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1886457719307915,
+      "learning_rate": 9.919654514404942e-06,
+      "loss": 0.57,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0919090808974596,
+      "learning_rate": 9.919531982610877e-06,
+      "loss": 0.479,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.794023210757667,
+      "learning_rate": 9.919409358296653e-06,
+      "loss": 0.547,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2566557679399826,
+      "learning_rate": 9.919286641464837e-06,
+      "loss": 0.6526,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8488484922005637,
+      "learning_rate": 9.919163832118003e-06,
+      "loss": 0.5577,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.105694894914224,
+      "learning_rate": 9.919040930258717e-06,
+      "loss": 0.6295,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.269484205491601,
+      "learning_rate": 9.918917935889557e-06,
+      "loss": 0.6332,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.916838174215787,
+      "learning_rate": 9.918794849013093e-06,
+      "loss": 0.5852,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0889847636708128,
+      "learning_rate": 9.918671669631901e-06,
+      "loss": 0.6749,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2520376244894327,
+      "learning_rate": 9.918548397748565e-06,
+      "loss": 0.686,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2366552560911415,
+      "learning_rate": 9.918425033365661e-06,
+      "loss": 0.7366,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1893812896359717,
+      "learning_rate": 9.918301576485776e-06,
+      "loss": 0.5262,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1254116095285207,
+      "learning_rate": 9.918178027111491e-06,
+      "loss": 0.8558,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8402808740695058,
+      "learning_rate": 9.918054385245392e-06,
+      "loss": 0.6037,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.959086596861861,
+      "learning_rate": 9.917930650890069e-06,
+      "loss": 0.5954,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1126654232654256,
+      "learning_rate": 9.917806824048114e-06,
+      "loss": 0.6681,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0799231745074613,
+      "learning_rate": 9.917682904722112e-06,
+      "loss": 0.7548,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1597427616641243,
+      "learning_rate": 9.917558892914667e-06,
+      "loss": 0.6949,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9472608942803078,
+      "learning_rate": 9.917434788628368e-06,
+      "loss": 0.4656,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.046186121552074,
+      "learning_rate": 9.917310591865817e-06,
+      "loss": 0.6347,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.185511410542886,
+      "learning_rate": 9.917186302629612e-06,
+      "loss": 0.6845,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1434777051749316,
+      "learning_rate": 9.917061920922354e-06,
+      "loss": 0.4902,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.182336771160904,
+      "learning_rate": 9.91693744674665e-06,
+      "loss": 0.623,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9527217985139653,
+      "learning_rate": 9.916812880105104e-06,
+      "loss": 0.4971,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.104299768106947,
+      "learning_rate": 9.916688221000323e-06,
+      "loss": 0.6047,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.333334175813318,
+      "learning_rate": 9.916563469434917e-06,
+      "loss": 0.6182,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.433704523652343,
+      "learning_rate": 9.916438625411497e-06,
+      "loss": 0.6218,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7143573853969944,
+      "learning_rate": 9.91631368893268e-06,
+      "loss": 0.4228,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.061925647735927,
+      "learning_rate": 9.916188660001078e-06,
+      "loss": 0.6932,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.359939438294256,
+      "learning_rate": 9.91606353861931e-06,
+      "loss": 0.6435,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1075951782856297,
+      "learning_rate": 9.915938324789994e-06,
+      "loss": 0.7501,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.686999608440556,
+      "learning_rate": 9.915813018515752e-06,
+      "loss": 0.5362,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.193913458755748,
+      "learning_rate": 9.915687619799208e-06,
+      "loss": 0.6535,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.184536192146895,
+      "learning_rate": 9.915562128642985e-06,
+      "loss": 0.6614,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.296402374392694,
+      "learning_rate": 9.915436545049713e-06,
+      "loss": 0.6294,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.305520120378623,
+      "learning_rate": 9.915310869022018e-06,
+      "loss": 0.6467,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.233780498994017,
+      "learning_rate": 9.915185100562533e-06,
+      "loss": 0.5468,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8269701382124603,
+      "learning_rate": 9.91505923967389e-06,
+      "loss": 0.5266,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.95045715828709,
+      "learning_rate": 9.914933286358725e-06,
+      "loss": 0.6105,
+      "step": 1346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.350555579972682,
+      "learning_rate": 9.914807240619674e-06,
+      "loss": 0.7849,
+      "step": 1347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.302796175652717,
+      "learning_rate": 9.914681102459375e-06,
+      "loss": 0.7519,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6317476177963006,
+      "learning_rate": 9.914554871880468e-06,
+      "loss": 0.5813,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7842232676765697,
+      "learning_rate": 9.9144285488856e-06,
+      "loss": 0.7698,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6148520708084106,
+      "eval_runtime": 67.9788,
+      "eval_samples_per_second": 50.163,
+      "eval_steps_per_second": 1.574,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8204345027634277,
+      "learning_rate": 9.914302133477408e-06,
+      "loss": 0.5354,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.205934271313879,
+      "learning_rate": 9.914175625658546e-06,
+      "loss": 0.5185,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9135006656886875,
+      "learning_rate": 9.914049025431657e-06,
+      "loss": 0.6673,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.308678483087819,
+      "learning_rate": 9.913922332799394e-06,
+      "loss": 0.7382,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9765882172979163,
+      "learning_rate": 9.913795547764409e-06,
+      "loss": 0.5597,
+      "step": 1355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6485627629539807,
+      "learning_rate": 9.913668670329356e-06,
+      "loss": 0.5869,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.098985176319322,
+      "learning_rate": 9.913541700496889e-06,
+      "loss": 0.6153,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.213123641898974,
+      "learning_rate": 9.913414638269669e-06,
+      "loss": 0.6385,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0078959296436327,
+      "learning_rate": 9.913287483650356e-06,
+      "loss": 0.5942,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.193300152512081,
+      "learning_rate": 9.913160236641608e-06,
+      "loss": 0.5577,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0214122399014083,
+      "learning_rate": 9.913032897246093e-06,
+      "loss": 0.7062,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.809473622921829,
+      "learning_rate": 9.912905465466477e-06,
+      "loss": 0.571,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2563919493705398,
+      "learning_rate": 9.912777941305424e-06,
+      "loss": 0.6162,
+      "step": 1363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9483272884164038,
+      "learning_rate": 9.912650324765606e-06,
+      "loss": 0.6086,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.490576109049697,
+      "learning_rate": 9.912522615849694e-06,
+      "loss": 0.5666,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0815021469417863,
+      "learning_rate": 9.912394814560362e-06,
+      "loss": 0.6192,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9088078621914157,
+      "learning_rate": 9.912266920900285e-06,
+      "loss": 0.5943,
+      "step": 1367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8767027285194446,
+      "learning_rate": 9.912138934872141e-06,
+      "loss": 0.5971,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7530882158544425,
+      "learning_rate": 9.91201085647861e-06,
+      "loss": 0.4129,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9745090528088918,
+      "learning_rate": 9.91188268572237e-06,
+      "loss": 0.4732,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.81465916814097,
+      "learning_rate": 9.911754422606108e-06,
+      "loss": 0.5493,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.883484402516311,
+      "learning_rate": 9.911626067132506e-06,
+      "loss": 0.6445,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.103486927192038,
+      "learning_rate": 9.911497619304253e-06,
+      "loss": 0.5875,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.647002047516207,
+      "learning_rate": 9.911369079124038e-06,
+      "loss": 0.6705,
+      "step": 1374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1838842552002204,
+      "learning_rate": 9.91124044659455e-06,
+      "loss": 0.7811,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9392135174421803,
+      "learning_rate": 9.911111721718484e-06,
+      "loss": 0.5225,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2851555631303606,
+      "learning_rate": 9.910982904498535e-06,
+      "loss": 0.635,
+      "step": 1377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1771800294602275,
+      "learning_rate": 9.910853994937397e-06,
+      "loss": 0.7763,
+      "step": 1378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4988576072481608,
+      "learning_rate": 9.91072499303777e-06,
+      "loss": 0.7875,
+      "step": 1379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0890688398765684,
+      "learning_rate": 9.910595898802357e-06,
+      "loss": 0.6382,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8745891773341767,
+      "learning_rate": 9.910466712233855e-06,
+      "loss": 0.5156,
+      "step": 1381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.15315346988987,
+      "learning_rate": 9.910337433334976e-06,
+      "loss": 0.6466,
+      "step": 1382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9960335091134,
+      "learning_rate": 9.910208062108419e-06,
+      "loss": 0.569,
+      "step": 1383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8424273304033625,
+      "learning_rate": 9.910078598556898e-06,
+      "loss": 0.4201,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.081730134056539,
+      "learning_rate": 9.909949042683119e-06,
+      "loss": 0.7477,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0220618529237684,
+      "learning_rate": 9.909819394489796e-06,
+      "loss": 0.5998,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0594579196881275,
+      "learning_rate": 9.909689653979643e-06,
+      "loss": 0.5155,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6364880220616689,
+      "learning_rate": 9.909559821155376e-06,
+      "loss": 0.3453,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7933819930842563,
+      "learning_rate": 9.909429896019715e-06,
+      "loss": 0.6062,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.045017564883952,
+      "learning_rate": 9.909299878575376e-06,
+      "loss": 0.4457,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.67875278630339,
+      "learning_rate": 9.909169768825082e-06,
+      "loss": 0.5134,
+      "step": 1391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7978882092116202,
+      "learning_rate": 9.909039566771559e-06,
+      "loss": 0.5495,
+      "step": 1392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.000031476188519,
+      "learning_rate": 9.908909272417531e-06,
+      "loss": 0.7352,
+      "step": 1393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.594118385359425,
+      "learning_rate": 9.908778885765727e-06,
+      "loss": 0.5997,
+      "step": 1394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9896444062929484,
+      "learning_rate": 9.908648406818872e-06,
+      "loss": 0.6167,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0634025275764247,
+      "learning_rate": 9.908517835579705e-06,
+      "loss": 0.6956,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.174739281429982,
+      "learning_rate": 9.908387172050953e-06,
+      "loss": 0.4641,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9877383306231868,
+      "learning_rate": 9.908256416235354e-06,
+      "loss": 0.5775,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.158855476555598,
+      "learning_rate": 9.908125568135644e-06,
+      "loss": 0.8926,
+      "step": 1399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7212490089826713,
+      "learning_rate": 9.907994627754565e-06,
+      "loss": 0.7016,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6136987805366516,
+      "eval_runtime": 67.9593,
+      "eval_samples_per_second": 50.177,
+      "eval_steps_per_second": 1.574,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7847454849057363,
+      "learning_rate": 9.907863595094856e-06,
+      "loss": 0.5346,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2145164335236482,
+      "learning_rate": 9.90773247015926e-06,
+      "loss": 0.6648,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9712105523076422,
+      "learning_rate": 9.907601252950522e-06,
+      "loss": 0.6245,
+      "step": 1403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8188083835605608,
+      "learning_rate": 9.90746994347139e-06,
+      "loss": 0.5503,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.093711806774773,
+      "learning_rate": 9.907338541724612e-06,
+      "loss": 0.6456,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6683694347298594,
+      "learning_rate": 9.90720704771294e-06,
+      "loss": 0.6281,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9783618306829127,
+      "learning_rate": 9.907075461439127e-06,
+      "loss": 0.5088,
+      "step": 1407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.402051422248389,
+      "learning_rate": 9.906943782905925e-06,
+      "loss": 0.5914,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.781265187845192,
+      "learning_rate": 9.906812012116092e-06,
+      "loss": 0.5094,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0225102257250485,
+      "learning_rate": 9.906680149072386e-06,
+      "loss": 0.586,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9009287937828294,
+      "learning_rate": 9.90654819377757e-06,
+      "loss": 0.5909,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4379571192341856,
+      "learning_rate": 9.906416146234402e-06,
+      "loss": 0.6158,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.363444142157378,
+      "learning_rate": 9.90628400644565e-06,
+      "loss": 0.6187,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.041404269253575,
+      "learning_rate": 9.906151774414081e-06,
+      "loss": 0.5048,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2589939128934784,
+      "learning_rate": 9.90601945014246e-06,
+      "loss": 0.6378,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2623369471790387,
+      "learning_rate": 9.905887033633558e-06,
+      "loss": 0.6035,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4475619584735755,
+      "learning_rate": 9.905754524890148e-06,
+      "loss": 0.8735,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9222493479068028,
+      "learning_rate": 9.905621923915004e-06,
+      "loss": 0.4024,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7593741277810546,
+      "learning_rate": 9.9054892307109e-06,
+      "loss": 0.6069,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8128984547084785,
+      "learning_rate": 9.905356445280617e-06,
+      "loss": 0.6018,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.938308746202162,
+      "learning_rate": 9.905223567626933e-06,
+      "loss": 0.636,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.437331753142477,
+      "learning_rate": 9.905090597752629e-06,
+      "loss": 0.5906,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3245082927511476,
+      "learning_rate": 9.904957535660488e-06,
+      "loss": 0.7147,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.054355547516497,
+      "learning_rate": 9.904824381353299e-06,
+      "loss": 0.4467,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.10735196589516,
+      "learning_rate": 9.904691134833848e-06,
+      "loss": 0.6846,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7115982549768427,
+      "learning_rate": 9.904557796104923e-06,
+      "loss": 0.7826,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.870549688383597,
+      "learning_rate": 9.904424365169316e-06,
+      "loss": 0.5418,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1149802273815124,
+      "learning_rate": 9.90429084202982e-06,
+      "loss": 0.6437,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5483149143678874,
+      "learning_rate": 9.904157226689234e-06,
+      "loss": 0.4401,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.000476854722676,
+      "learning_rate": 9.904023519150349e-06,
+      "loss": 0.5722,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.076417770168202,
+      "learning_rate": 9.903889719415968e-06,
+      "loss": 0.5299,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.529055553612549,
+      "learning_rate": 9.90375582748889e-06,
+      "loss": 0.4188,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6598249575447777,
+      "learning_rate": 9.90362184337192e-06,
+      "loss": 0.4135,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9525158560320672,
+      "learning_rate": 9.903487767067864e-06,
+      "loss": 0.6599,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.115485446668057,
+      "learning_rate": 9.903353598579524e-06,
+      "loss": 0.5851,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.269608608118388,
+      "learning_rate": 9.903219337909712e-06,
+      "loss": 0.6739,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6805314151402753,
+      "learning_rate": 9.903084985061238e-06,
+      "loss": 0.657,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0478193824477446,
+      "learning_rate": 9.902950540036914e-06,
+      "loss": 0.6412,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.251138693278515,
+      "learning_rate": 9.902816002839558e-06,
+      "loss": 0.6735,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7868043181136657,
+      "learning_rate": 9.902681373471979e-06,
+      "loss": 0.4858,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0524390587558603,
+      "learning_rate": 9.902546651937003e-06,
+      "loss": 0.7377,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0955566616215746,
+      "learning_rate": 9.902411838237447e-06,
+      "loss": 0.587,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4524599477273723,
+      "learning_rate": 9.902276932376134e-06,
+      "loss": 0.6456,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9922200449731728,
+      "learning_rate": 9.90214193435589e-06,
+      "loss": 0.4721,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4151833213232674,
+      "learning_rate": 9.902006844179536e-06,
+      "loss": 0.5934,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9062711782879616,
+      "learning_rate": 9.901871661849901e-06,
+      "loss": 0.5213,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6706356948486427,
+      "learning_rate": 9.90173638736982e-06,
+      "loss": 0.5626,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7539625255799869,
+      "learning_rate": 9.90160102074212e-06,
+      "loss": 0.5176,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8628264248055528,
+      "learning_rate": 9.90146556196964e-06,
+      "loss": 0.5089,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6496414266243815,
+      "learning_rate": 9.90133001105521e-06,
+      "loss": 0.5656,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.6129624247550964,
+      "eval_runtime": 67.9272,
+      "eval_samples_per_second": 50.201,
+      "eval_steps_per_second": 1.575,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6254882599107854,
+      "learning_rate": 9.901194368001668e-06,
+      "loss": 0.5624,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8340272181672763,
+      "learning_rate": 9.901058632811858e-06,
+      "loss": 0.4946,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8238896186485418,
+      "learning_rate": 9.90092280548862e-06,
+      "loss": 0.4676,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1071728883048877,
+      "learning_rate": 9.900786886034793e-06,
+      "loss": 0.6664,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6954052285805128,
+      "learning_rate": 9.900650874453228e-06,
+      "loss": 0.5205,
+      "step": 1455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.276604957995286,
+      "learning_rate": 9.900514770746769e-06,
+      "loss": 0.5684,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.50146659119374,
+      "learning_rate": 9.900378574918269e-06,
+      "loss": 0.7681,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.314329116400808,
+      "learning_rate": 9.900242286970577e-06,
+      "loss": 0.6133,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.278613099230212,
+      "learning_rate": 9.900105906906544e-06,
+      "loss": 0.6885,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.021861545704924,
+      "learning_rate": 9.899969434729027e-06,
+      "loss": 0.5368,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6882963705349647,
+      "learning_rate": 9.899832870440883e-06,
+      "loss": 0.644,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3511270659537202,
+      "learning_rate": 9.899696214044968e-06,
+      "loss": 0.6606,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9548015664392437,
+      "learning_rate": 9.89955946554415e-06,
+      "loss": 0.6304,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.18656443982212,
+      "learning_rate": 9.899422624941287e-06,
+      "loss": 0.6136,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5791621936930005,
+      "learning_rate": 9.89928569223924e-06,
+      "loss": 0.6688,
+      "step": 1465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0691562961554784,
+      "learning_rate": 9.899148667440883e-06,
+      "loss": 0.7633,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1324622012207115,
+      "learning_rate": 9.89901155054908e-06,
+      "loss": 0.5648,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9744060637373109,
+      "learning_rate": 9.898874341566703e-06,
+      "loss": 0.6246,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.194881375033989,
+      "learning_rate": 9.898737040496623e-06,
+      "loss": 0.6244,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6195583180933504,
+      "learning_rate": 9.898599647341717e-06,
+      "loss": 0.8666,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.121172266194213,
+      "learning_rate": 9.898462162104857e-06,
+      "loss": 0.6403,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.097143309791853,
+      "learning_rate": 9.898324584788925e-06,
+      "loss": 0.5662,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.318773626563772,
+      "learning_rate": 9.8981869153968e-06,
+      "loss": 0.6809,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.116049170627387,
+      "learning_rate": 9.898049153931363e-06,
+      "loss": 0.8742,
+      "step": 1474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.231194801601838,
+      "learning_rate": 9.897911300395499e-06,
+      "loss": 0.6673,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3626574506877747,
+      "learning_rate": 9.897773354792093e-06,
+      "loss": 0.5127,
+      "step": 1476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1608348248548555,
+      "learning_rate": 9.897635317124034e-06,
+      "loss": 0.533,
+      "step": 1477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0935740564835923,
+      "learning_rate": 9.89749718739421e-06,
+      "loss": 0.6962,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1143533280426206,
+      "learning_rate": 9.897358965605516e-06,
+      "loss": 0.6493,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.060047633289368,
+      "learning_rate": 9.897220651760843e-06,
+      "loss": 0.633,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4725486905647185,
+      "learning_rate": 9.897082245863085e-06,
+      "loss": 0.6343,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9442928173503178,
+      "learning_rate": 9.896943747915142e-06,
+      "loss": 0.5584,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8747534240036101,
+      "learning_rate": 9.896805157919913e-06,
+      "loss": 0.562,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.011389206230691,
+      "learning_rate": 9.8966664758803e-06,
+      "loss": 0.6255,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9639711596106448,
+      "learning_rate": 9.896527701799202e-06,
+      "loss": 0.5181,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1187700361501536,
+      "learning_rate": 9.896388835679529e-06,
+      "loss": 0.6205,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.774773911737284,
+      "learning_rate": 9.896249877524188e-06,
+      "loss": 0.6493,
+      "step": 1487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1833663186013608,
+      "learning_rate": 9.896110827336084e-06,
+      "loss": 0.6511,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.974825352326313,
+      "learning_rate": 9.895971685118133e-06,
+      "loss": 0.6868,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0804572612287506,
+      "learning_rate": 9.895832450873244e-06,
+      "loss": 0.5532,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.217745323221171,
+      "learning_rate": 9.895693124604332e-06,
+      "loss": 0.5573,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9581899265985232,
+      "learning_rate": 9.895553706314315e-06,
+      "loss": 0.5634,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.468270063831074,
+      "learning_rate": 9.895414196006113e-06,
+      "loss": 0.5462,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8722010802616762,
+      "learning_rate": 9.895274593682644e-06,
+      "loss": 0.5012,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6026651748410012,
+      "learning_rate": 9.895134899346831e-06,
+      "loss": 0.6301,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7853409933864295,
+      "learning_rate": 9.894995113001599e-06,
+      "loss": 0.4846,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1212924034491736,
+      "learning_rate": 9.894855234649876e-06,
+      "loss": 0.5317,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2302652769616835,
+      "learning_rate": 9.894715264294584e-06,
+      "loss": 0.5929,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7496587049400056,
+      "learning_rate": 9.89457520193866e-06,
+      "loss": 0.4726,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.344367942019827,
+      "learning_rate": 9.894435047585034e-06,
+      "loss": 0.6656,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.6119856834411621,
+      "eval_runtime": 67.979,
+      "eval_samples_per_second": 50.163,
+      "eval_steps_per_second": 1.574,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0200354530473534,
+      "learning_rate": 9.89429480123664e-06,
+      "loss": 0.721,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4814510001910444,
+      "learning_rate": 9.894154462896414e-06,
+      "loss": 0.7538,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3734603884941934,
+      "learning_rate": 9.894014032567292e-06,
+      "loss": 0.5816,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.85157903965519,
+      "learning_rate": 9.893873510252217e-06,
+      "loss": 0.5544,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9584640879917874,
+      "learning_rate": 9.893732895954129e-06,
+      "loss": 0.6294,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.185570106302621,
+      "learning_rate": 9.893592189675972e-06,
+      "loss": 0.6055,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7158762890498531,
+      "learning_rate": 9.893451391420691e-06,
+      "loss": 0.5033,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3816599730508736,
+      "learning_rate": 9.893310501191233e-06,
+      "loss": 0.5514,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3328561672884383,
+      "learning_rate": 9.89316951899055e-06,
+      "loss": 0.5994,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3563645764128247,
+      "learning_rate": 9.893028444821592e-06,
+      "loss": 0.6447,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2351602974165985,
+      "learning_rate": 9.89288727868731e-06,
+      "loss": 0.5925,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2513537447554084,
+      "learning_rate": 9.892746020590663e-06,
+      "loss": 0.5501,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.439542938943426,
+      "learning_rate": 9.892604670534605e-06,
+      "loss": 0.7176,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.025870582790434,
+      "learning_rate": 9.892463228522099e-06,
+      "loss": 0.6191,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9170792633840923,
+      "learning_rate": 9.892321694556101e-06,
+      "loss": 0.4828,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0141995863988154,
+      "learning_rate": 9.89218006863958e-06,
+      "loss": 0.4966,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9139117792603029,
+      "learning_rate": 9.892038350775493e-06,
+      "loss": 0.5852,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.099252785668439,
+      "learning_rate": 9.891896540966814e-06,
+      "loss": 0.5811,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.512790928077131,
+      "learning_rate": 9.891754639216507e-06,
+      "loss": 0.6163,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7191513344988072,
+      "learning_rate": 9.891612645527548e-06,
+      "loss": 0.4541,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.295205266927205,
+      "learning_rate": 9.891470559902903e-06,
+      "loss": 0.743,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.171444795253062,
+      "learning_rate": 9.891328382345549e-06,
+      "loss": 0.6002,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1989682314715355,
+      "learning_rate": 9.891186112858464e-06,
+      "loss": 0.6042,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.37851115435087,
+      "learning_rate": 9.891043751444625e-06,
+      "loss": 0.6612,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8579185403147147,
+      "learning_rate": 9.890901298107013e-06,
+      "loss": 0.6318,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7057516668494024,
+      "learning_rate": 9.89075875284861e-06,
+      "loss": 0.5785,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9730790980971702,
+      "learning_rate": 9.890616115672398e-06,
+      "loss": 0.6106,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6722412429625577,
+      "learning_rate": 9.890473386581366e-06,
+      "loss": 0.6413,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.042224512440595,
+      "learning_rate": 9.8903305655785e-06,
+      "loss": 0.6869,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3242504824972787,
+      "learning_rate": 9.89018765266679e-06,
+      "loss": 0.6206,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3226421072106738,
+      "learning_rate": 9.89004464784923e-06,
+      "loss": 0.6345,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2605615588197345,
+      "learning_rate": 9.88990155112881e-06,
+      "loss": 0.624,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.100054305382676,
+      "learning_rate": 9.889758362508529e-06,
+      "loss": 0.5787,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1914079167610176,
+      "learning_rate": 9.889615081991385e-06,
+      "loss": 0.5476,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.219671497841816,
+      "learning_rate": 9.889471709580374e-06,
+      "loss": 0.5992,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.576140578190344,
+      "learning_rate": 9.889328245278497e-06,
+      "loss": 0.8787,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9282978044774879,
+      "learning_rate": 9.889184689088763e-06,
+      "loss": 0.5404,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.023634181322372,
+      "learning_rate": 9.889041041014172e-06,
+      "loss": 0.487,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.235309720550558,
+      "learning_rate": 9.888897301057733e-06,
+      "loss": 0.6883,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0973707924886402,
+      "learning_rate": 9.888753469222455e-06,
+      "loss": 0.6194,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0389567821155943,
+      "learning_rate": 9.88860954551135e-06,
+      "loss": 0.7456,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.602365570018614,
+      "learning_rate": 9.888465529927428e-06,
+      "loss": 0.6719,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0103840076040798,
+      "learning_rate": 9.888321422473709e-06,
+      "loss": 0.5426,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9300931765269165,
+      "learning_rate": 9.888177223153205e-06,
+      "loss": 0.5356,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.482590044698629,
+      "learning_rate": 9.888032931968936e-06,
+      "loss": 0.707,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7692670842646454,
+      "learning_rate": 9.887888548923924e-06,
+      "loss": 0.4705,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.220930749103081,
+      "learning_rate": 9.88774407402119e-06,
+      "loss": 0.7105,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.161587294065335,
+      "learning_rate": 9.88759950726376e-06,
+      "loss": 0.6366,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.890925963780074,
+      "learning_rate": 9.887454848654659e-06,
+      "loss": 0.6178,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.305151876441939,
+      "learning_rate": 9.887310098196914e-06,
+      "loss": 0.6714,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.6125446557998657,
+      "eval_runtime": 67.963,
+      "eval_samples_per_second": 50.174,
+      "eval_steps_per_second": 1.574,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0621771343996493,
+      "learning_rate": 9.887165255893561e-06,
+      "loss": 0.6852,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6215229826142794,
+      "learning_rate": 9.887020321747625e-06,
+      "loss": 0.4445,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.39549426645569,
+      "learning_rate": 9.886875295762147e-06,
+      "loss": 0.7542,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.969988852357434,
+      "learning_rate": 9.886730177940155e-06,
+      "loss": 0.675,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8991220766861614,
+      "learning_rate": 9.886584968284693e-06,
+      "loss": 0.5319,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3999301849543127,
+      "learning_rate": 9.8864396667988e-06,
+      "loss": 0.607,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1252366343430213,
+      "learning_rate": 9.886294273485518e-06,
+      "loss": 0.5687,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.024672320991454,
+      "learning_rate": 9.886148788347889e-06,
+      "loss": 0.6521,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3904080808013544,
+      "learning_rate": 9.886003211388959e-06,
+      "loss": 0.4874,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.141511512389054,
+      "learning_rate": 9.885857542611776e-06,
+      "loss": 0.5748,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.358517446991475,
+      "learning_rate": 9.88571178201939e-06,
+      "loss": 0.665,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5963691775842346,
+      "learning_rate": 9.885565929614851e-06,
+      "loss": 0.5942,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.470560901101373,
+      "learning_rate": 9.885419985401216e-06,
+      "loss": 0.6301,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.221160450270207,
+      "learning_rate": 9.885273949381534e-06,
+      "loss": 0.5226,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5734416807160394,
+      "learning_rate": 9.885127821558867e-06,
+      "loss": 0.593,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.339859349008004,
+      "learning_rate": 9.884981601936274e-06,
+      "loss": 0.5262,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.064353878046537,
+      "learning_rate": 9.884835290516812e-06,
+      "loss": 0.5858,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9707436430240806,
+      "learning_rate": 9.884688887303549e-06,
+      "loss": 0.4084,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1847321807434166,
+      "learning_rate": 9.884542392299547e-06,
+      "loss": 0.708,
+      "step": 1569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.123441229189967,
+      "learning_rate": 9.884395805507873e-06,
+      "loss": 0.6216,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.644838172681213,
+      "learning_rate": 9.884249126931597e-06,
+      "loss": 0.6107,
+      "step": 1571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0642485380352453,
+      "learning_rate": 9.884102356573787e-06,
+      "loss": 0.6021,
+      "step": 1572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2041084353883327,
+      "learning_rate": 9.88395549443752e-06,
+      "loss": 0.6675,
+      "step": 1573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9468888993056492,
+      "learning_rate": 9.883808540525866e-06,
+      "loss": 0.5488,
+      "step": 1574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4297885425555736,
+      "learning_rate": 9.883661494841903e-06,
+      "loss": 0.7486,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2730100097356427,
+      "learning_rate": 9.883514357388709e-06,
+      "loss": 0.6686,
+      "step": 1576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.014352675661412,
+      "learning_rate": 9.883367128169364e-06,
+      "loss": 0.4268,
+      "step": 1577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3952501874357113,
+      "learning_rate": 9.883219807186951e-06,
+      "loss": 0.6415,
+      "step": 1578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8775755308154527,
+      "learning_rate": 9.883072394444557e-06,
+      "loss": 0.545,
+      "step": 1579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4338593083140947,
+      "learning_rate": 9.882924889945261e-06,
+      "loss": 0.5905,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.211550153343158,
+      "learning_rate": 9.882777293692156e-06,
+      "loss": 0.5472,
+      "step": 1581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.20241119839163,
+      "learning_rate": 9.882629605688333e-06,
+      "loss": 0.6272,
+      "step": 1582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1324699033470496,
+      "learning_rate": 9.882481825936879e-06,
+      "loss": 0.6159,
+      "step": 1583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8117820561790845,
+      "learning_rate": 9.88233395444089e-06,
+      "loss": 0.4956,
+      "step": 1584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8601028198501757,
+      "learning_rate": 9.882185991203461e-06,
+      "loss": 0.4848,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2968565511296997,
+      "learning_rate": 9.882037936227691e-06,
+      "loss": 0.6711,
+      "step": 1586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.19019979956196,
+      "learning_rate": 9.88188978951668e-06,
+      "loss": 0.8119,
+      "step": 1587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.248956890886912,
+      "learning_rate": 9.881741551073524e-06,
+      "loss": 0.5208,
+      "step": 1588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9584509209059713,
+      "learning_rate": 9.881593220901334e-06,
+      "loss": 0.5784,
+      "step": 1589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8157691006784216,
+      "learning_rate": 9.88144479900321e-06,
+      "loss": 0.5714,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3596948234794066,
+      "learning_rate": 9.881296285382261e-06,
+      "loss": 0.5976,
+      "step": 1591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.275635136123225,
+      "learning_rate": 9.881147680041596e-06,
+      "loss": 0.5559,
+      "step": 1592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.313978705118664,
+      "learning_rate": 9.880998982984324e-06,
+      "loss": 0.4953,
+      "step": 1593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3933157766918676,
+      "learning_rate": 9.88085019421356e-06,
+      "loss": 0.6361,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.626227715021951,
+      "learning_rate": 9.88070131373242e-06,
+      "loss": 0.5071,
+      "step": 1595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0248930909565432,
+      "learning_rate": 9.880552341544018e-06,
+      "loss": 0.6179,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5754971909818782,
+      "learning_rate": 9.880403277651472e-06,
+      "loss": 0.6972,
+      "step": 1597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1040420586732744,
+      "learning_rate": 9.880254122057907e-06,
+      "loss": 0.4997,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.50235152856785,
+      "learning_rate": 9.88010487476644e-06,
+      "loss": 0.7071,
+      "step": 1599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2682861975561766,
+      "learning_rate": 9.8799555357802e-06,
+      "loss": 0.6097,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.6118566393852234,
+      "eval_runtime": 67.9424,
+      "eval_samples_per_second": 50.19,
+      "eval_steps_per_second": 1.575,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4249090374089373,
+      "learning_rate": 9.879806105102309e-06,
+      "loss": 0.6318,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5128480138278824,
+      "learning_rate": 9.8796565827359e-06,
+      "loss": 0.6751,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.571425759878021,
+      "learning_rate": 9.879506968684098e-06,
+      "loss": 0.6624,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.249586327180744,
+      "learning_rate": 9.879357262950041e-06,
+      "loss": 0.5963,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2789366258006636,
+      "learning_rate": 9.879207465536858e-06,
+      "loss": 0.7093,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2306296909138728,
+      "learning_rate": 9.879057576447687e-06,
+      "loss": 0.6594,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8050034556589765,
+      "learning_rate": 9.878907595685665e-06,
+      "loss": 0.4929,
+      "step": 1607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.68980402946938,
+      "learning_rate": 9.878757523253934e-06,
+      "loss": 0.6771,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.722818763380077,
+      "learning_rate": 9.878607359155634e-06,
+      "loss": 0.5521,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.794419375022026,
+      "learning_rate": 9.878457103393906e-06,
+      "loss": 0.7376,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.465461476775007,
+      "learning_rate": 9.878306755971899e-06,
+      "loss": 0.5927,
+      "step": 1611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9417878529744086,
+      "learning_rate": 9.87815631689276e-06,
+      "loss": 0.5091,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2908787369634247,
+      "learning_rate": 9.878005786159638e-06,
+      "loss": 0.5138,
+      "step": 1613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1272600029942907,
+      "learning_rate": 9.877855163775685e-06,
+      "loss": 0.6326,
+      "step": 1614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9404086454918446,
+      "learning_rate": 9.877704449744052e-06,
+      "loss": 0.5134,
+      "step": 1615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7014444577036323,
+      "learning_rate": 9.877553644067898e-06,
+      "loss": 0.4029,
+      "step": 1616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2289583593107185,
+      "learning_rate": 9.877402746750374e-06,
+      "loss": 0.71,
+      "step": 1617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5687975523408118,
+      "learning_rate": 9.877251757794644e-06,
+      "loss": 0.5882,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9300672096484557,
+      "learning_rate": 9.877100677203867e-06,
+      "loss": 0.5152,
+      "step": 1619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.250778016503785,
+      "learning_rate": 9.876949504981207e-06,
+      "loss": 0.6934,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9794984540546596,
+      "learning_rate": 9.876798241129825e-06,
+      "loss": 0.5864,
+      "step": 1621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3437475998992072,
+      "learning_rate": 9.876646885652895e-06,
+      "loss": 0.7136,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.183643373755056,
+      "learning_rate": 9.876495438553576e-06,
+      "loss": 0.5826,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.807239987189366,
+      "learning_rate": 9.876343899835046e-06,
+      "loss": 0.5542,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.022111407326927,
+      "learning_rate": 9.876192269500475e-06,
+      "loss": 0.5673,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1819781139525034,
+      "learning_rate": 9.876040547553035e-06,
+      "loss": 0.6168,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0871306288958245,
+      "learning_rate": 9.875888733995906e-06,
+      "loss": 0.7131,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.698798603266553,
+      "learning_rate": 9.875736828832264e-06,
+      "loss": 0.5103,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8167715899037866,
+      "learning_rate": 9.875584832065288e-06,
+      "loss": 0.496,
+      "step": 1629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8772836288938393,
+      "learning_rate": 9.875432743698163e-06,
+      "loss": 0.5591,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.795509447408253,
+      "learning_rate": 9.875280563734071e-06,
+      "loss": 0.6631,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9975534660586378,
+      "learning_rate": 9.875128292176197e-06,
+      "loss": 0.4785,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8204200272466187,
+      "learning_rate": 9.87497592902773e-06,
+      "loss": 0.4664,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2131332160356543,
+      "learning_rate": 9.874823474291859e-06,
+      "loss": 0.7233,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2260592568020763,
+      "learning_rate": 9.874670927971777e-06,
+      "loss": 0.5979,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8696520568016644,
+      "learning_rate": 9.874518290070676e-06,
+      "loss": 0.5893,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2244304612073034,
+      "learning_rate": 9.87436556059175e-06,
+      "loss": 0.6904,
+      "step": 1637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9721934873041715,
+      "learning_rate": 9.8742127395382e-06,
+      "loss": 0.5012,
+      "step": 1638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3915313949464805,
+      "learning_rate": 9.874059826913223e-06,
+      "loss": 0.7101,
+      "step": 1639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7606181347859342,
+      "learning_rate": 9.87390682272002e-06,
+      "loss": 0.5452,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.26171464436341,
+      "learning_rate": 9.873753726961795e-06,
+      "loss": 0.7107,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8035059579664272,
+      "learning_rate": 9.873600539641751e-06,
+      "loss": 0.5536,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6979921745309956,
+      "learning_rate": 9.873447260763099e-06,
+      "loss": 0.4882,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.029527245203075,
+      "learning_rate": 9.873293890329041e-06,
+      "loss": 0.5603,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7155048970157385,
+      "learning_rate": 9.873140428342794e-06,
+      "loss": 0.4924,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8996148062904623,
+      "learning_rate": 9.872986874807568e-06,
+      "loss": 0.5607,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0344549509811127,
+      "learning_rate": 9.872833229726579e-06,
+      "loss": 0.6257,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0135494686679745,
+      "learning_rate": 9.872679493103041e-06,
+      "loss": 0.6054,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.494040127738728,
+      "learning_rate": 9.872525664940173e-06,
+      "loss": 0.4187,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2110094863445786,
+      "learning_rate": 9.8723717452412e-06,
+      "loss": 0.6758,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.612139880657196,
+      "eval_runtime": 67.9116,
+      "eval_samples_per_second": 50.212,
+      "eval_steps_per_second": 1.576,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.433694905234314,
+      "learning_rate": 9.872217734009336e-06,
+      "loss": 0.6893,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5932370255705073,
+      "learning_rate": 9.872063631247813e-06,
+      "loss": 0.7577,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8891649747920134,
+      "learning_rate": 9.87190943695985e-06,
+      "loss": 0.6377,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0134708869947646,
+      "learning_rate": 9.87175515114868e-06,
+      "loss": 0.6976,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.011604512621044,
+      "learning_rate": 9.871600773817532e-06,
+      "loss": 0.6226,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.407218517132755,
+      "learning_rate": 9.871446304969636e-06,
+      "loss": 0.7097,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.294189410153498,
+      "learning_rate": 9.871291744608228e-06,
+      "loss": 0.5711,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.008185610547691,
+      "learning_rate": 9.871137092736542e-06,
+      "loss": 0.612,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.046980222216155,
+      "learning_rate": 9.870982349357816e-06,
+      "loss": 0.5723,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.995738950636849,
+      "learning_rate": 9.870827514475287e-06,
+      "loss": 0.4862,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5107671078291727,
+      "learning_rate": 9.870672588092202e-06,
+      "loss": 0.7526,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6283122234836296,
+      "learning_rate": 9.8705175702118e-06,
+      "loss": 0.4543,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0338617038722,
+      "learning_rate": 9.870362460837327e-06,
+      "loss": 0.6122,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1814110509682387,
+      "learning_rate": 9.870207259972031e-06,
+      "loss": 0.6645,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2485448696322834,
+      "learning_rate": 9.870051967619161e-06,
+      "loss": 0.6749,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.471233449132997,
+      "learning_rate": 9.869896583781966e-06,
+      "loss": 0.5368,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7443532017407448,
+      "learning_rate": 9.8697411084637e-06,
+      "loss": 0.4695,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.379187960586249,
+      "learning_rate": 9.869585541667619e-06,
+      "loss": 0.664,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8929067309979626,
+      "learning_rate": 9.86942988339698e-06,
+      "loss": 0.4499,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6860326183604704,
+      "learning_rate": 9.869274133655037e-06,
+      "loss": 0.703,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1046996026897737,
+      "learning_rate": 9.869118292445056e-06,
+      "loss": 0.601,
+      "step": 1671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9524652439577677,
+      "learning_rate": 9.868962359770298e-06,
+      "loss": 0.5112,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.177346099480733,
+      "learning_rate": 9.868806335634026e-06,
+      "loss": 0.6122,
+      "step": 1673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8279219752461304,
+      "learning_rate": 9.868650220039506e-06,
+      "loss": 0.4439,
+      "step": 1674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6646820956357011,
+      "learning_rate": 9.868494012990008e-06,
+      "loss": 0.5398,
+      "step": 1675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.232048745818608,
+      "learning_rate": 9.868337714488801e-06,
+      "loss": 0.5857,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3366623859064246,
+      "learning_rate": 9.868181324539157e-06,
+      "loss": 0.602,
+      "step": 1677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.031697405905677,
+      "learning_rate": 9.868024843144351e-06,
+      "loss": 0.6594,
+      "step": 1678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5364205298690767,
+      "learning_rate": 9.867868270307656e-06,
+      "loss": 0.5987,
+      "step": 1679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5180557373732175,
+      "learning_rate": 9.867711606032354e-06,
+      "loss": 0.6293,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0043136041330287,
+      "learning_rate": 9.867554850321721e-06,
+      "loss": 0.5431,
+      "step": 1681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.777456955555788,
+      "learning_rate": 9.867398003179041e-06,
+      "loss": 0.4374,
+      "step": 1682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9371749323064387,
+      "learning_rate": 9.867241064607597e-06,
+      "loss": 0.4534,
+      "step": 1683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3550116443062175,
+      "learning_rate": 9.867084034610673e-06,
+      "loss": 0.6828,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9580503810802123,
+      "learning_rate": 9.866926913191558e-06,
+      "loss": 0.6502,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.216606206188794,
+      "learning_rate": 9.86676970035354e-06,
+      "loss": 0.4448,
+      "step": 1686
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9931830541923912,
+      "learning_rate": 9.866612396099909e-06,
+      "loss": 0.5885,
+      "step": 1687
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7882518935486162,
+      "learning_rate": 9.866455000433962e-06,
+      "loss": 0.4768,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.27406021596559,
+      "learning_rate": 9.866297513358989e-06,
+      "loss": 0.8609,
+      "step": 1689
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.799143853759185,
+      "learning_rate": 9.866139934878292e-06,
+      "loss": 0.6548,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1545233475624412,
+      "learning_rate": 9.865982264995166e-06,
+      "loss": 0.7694,
+      "step": 1691
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.104608807203573,
+      "learning_rate": 9.865824503712915e-06,
+      "loss": 0.541,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.089214208283419,
+      "learning_rate": 9.865666651034837e-06,
+      "loss": 0.5248,
+      "step": 1693
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.436546724136029,
+      "learning_rate": 9.865508706964243e-06,
+      "loss": 0.4799,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.420841481887485,
+      "learning_rate": 9.865350671504433e-06,
+      "loss": 0.7198,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1098284315313762,
+      "learning_rate": 9.865192544658717e-06,
+      "loss": 0.545,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1410942893715945,
+      "learning_rate": 9.865034326430409e-06,
+      "loss": 0.6029,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.068053068251166,
+      "learning_rate": 9.864876016822817e-06,
+      "loss": 0.6118,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.725096858978011,
+      "learning_rate": 9.864717615839257e-06,
+      "loss": 0.5793,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9559480237926374,
+      "learning_rate": 9.864559123483045e-06,
+      "loss": 0.5358,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.6079229116439819,
+      "eval_runtime": 67.9687,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.650267425991632,
+      "learning_rate": 9.864400539757498e-06,
+      "loss": 0.6959,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.967381510994838,
+      "learning_rate": 9.864241864665939e-06,
+      "loss": 0.5775,
+      "step": 1702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7441141457253215,
+      "learning_rate": 9.864083098211686e-06,
+      "loss": 0.5538,
+      "step": 1703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2414096143307227,
+      "learning_rate": 9.863924240398062e-06,
+      "loss": 0.6392,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1118737465177735,
+      "learning_rate": 9.863765291228396e-06,
+      "loss": 0.5694,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.939360739103044,
+      "learning_rate": 9.863606250706014e-06,
+      "loss": 0.583,
+      "step": 1706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.039109551852669,
+      "learning_rate": 9.863447118834245e-06,
+      "loss": 0.6443,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9908542544052965,
+      "learning_rate": 9.863287895616423e-06,
+      "loss": 0.6346,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1117713963178666,
+      "learning_rate": 9.863128581055876e-06,
+      "loss": 0.5838,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6667361642287493,
+      "learning_rate": 9.862969175155943e-06,
+      "loss": 0.6134,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5616511403755724,
+      "learning_rate": 9.862809677919961e-06,
+      "loss": 0.7452,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2235203458727395,
+      "learning_rate": 9.862650089351269e-06,
+      "loss": 0.6152,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1830974212003027,
+      "learning_rate": 9.862490409453208e-06,
+      "loss": 0.6528,
+      "step": 1713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7776095276893862,
+      "learning_rate": 9.862330638229116e-06,
+      "loss": 0.5121,
+      "step": 1714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.288722000926598,
+      "learning_rate": 9.862170775682344e-06,
+      "loss": 0.6861,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.23009571219253,
+      "learning_rate": 9.862010821816236e-06,
+      "loss": 0.6986,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0938607795950412,
+      "learning_rate": 9.861850776634141e-06,
+      "loss": 0.6204,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.316442601685224,
+      "learning_rate": 9.86169064013941e-06,
+      "loss": 0.6223,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9585100222175846,
+      "learning_rate": 9.861530412335392e-06,
+      "loss": 0.6681,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.038018020835124,
+      "learning_rate": 9.861370093225447e-06,
+      "loss": 0.5919,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7615418195411612,
+      "learning_rate": 9.861209682812926e-06,
+      "loss": 0.5584,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.695401045236409,
+      "learning_rate": 9.861049181101189e-06,
+      "loss": 0.6205,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8708627721793238,
+      "learning_rate": 9.860888588093598e-06,
+      "loss": 0.5182,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8889704606887918,
+      "learning_rate": 9.860727903793512e-06,
+      "loss": 0.5865,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7485936162218774,
+      "learning_rate": 9.860567128204296e-06,
+      "loss": 0.4803,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6162754583301902,
+      "learning_rate": 9.860406261329315e-06,
+      "loss": 0.5276,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6400104831098243,
+      "learning_rate": 9.860245303171937e-06,
+      "loss": 0.8692,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1137895432733704,
+      "learning_rate": 9.860084253735533e-06,
+      "loss": 0.5583,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0426637306108613,
+      "learning_rate": 9.859923113023472e-06,
+      "loss": 0.6085,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.100552342897661,
+      "learning_rate": 9.85976188103913e-06,
+      "loss": 0.6232,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8886037585457824,
+      "learning_rate": 9.85960055778588e-06,
+      "loss": 0.5243,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5066633658786333,
+      "learning_rate": 9.859439143267101e-06,
+      "loss": 0.4081,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3911873240462893,
+      "learning_rate": 9.85927763748617e-06,
+      "loss": 0.6522,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8657239078535126,
+      "learning_rate": 9.859116040446471e-06,
+      "loss": 0.4637,
+      "step": 1734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.080240155698604,
+      "learning_rate": 9.858954352151384e-06,
+      "loss": 0.7531,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3533336535385865,
+      "learning_rate": 9.858792572604297e-06,
+      "loss": 0.7097,
+      "step": 1736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6938035825589701,
+      "learning_rate": 9.858630701808592e-06,
+      "loss": 0.5008,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.259647323284662,
+      "learning_rate": 9.858468739767662e-06,
+      "loss": 0.7669,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.052550085312085,
+      "learning_rate": 9.858306686484896e-06,
+      "loss": 0.5734,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9665765003001106,
+      "learning_rate": 9.858144541963686e-06,
+      "loss": 0.5358,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0793826482142643,
+      "learning_rate": 9.857982306207428e-06,
+      "loss": 0.6028,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8285747962201433,
+      "learning_rate": 9.857819979219518e-06,
+      "loss": 0.4439,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7852360547015513,
+      "learning_rate": 9.857657561003351e-06,
+      "loss": 0.5362,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3376930506699884,
+      "learning_rate": 9.85749505156233e-06,
+      "loss": 0.7538,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.008810300826036,
+      "learning_rate": 9.857332450899858e-06,
+      "loss": 0.5345,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7272547795259492,
+      "learning_rate": 9.857169759019337e-06,
+      "loss": 0.5217,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6128673629551498,
+      "learning_rate": 9.857006975924174e-06,
+      "loss": 0.3447,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6436900910691494,
+      "learning_rate": 9.856844101617776e-06,
+      "loss": 0.4292,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.937487184582922,
+      "learning_rate": 9.856681136103553e-06,
+      "loss": 0.5977,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7190147336932247,
+      "learning_rate": 9.856518079384914e-06,
+      "loss": 0.5605,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.611440896987915,
+      "eval_runtime": 68.029,
+      "eval_samples_per_second": 50.126,
+      "eval_steps_per_second": 1.573,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.899664997121826,
+      "learning_rate": 9.856354931465278e-06,
+      "loss": 0.5906,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1470075547799867,
+      "learning_rate": 9.856191692348056e-06,
+      "loss": 0.5042,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2783092324059195,
+      "learning_rate": 9.856028362036668e-06,
+      "loss": 0.5288,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.474398698976454,
+      "learning_rate": 9.855864940534529e-06,
+      "loss": 0.6342,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.956052730292292,
+      "learning_rate": 9.855701427845065e-06,
+      "loss": 0.5179,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1099548829107606,
+      "learning_rate": 9.855537823971696e-06,
+      "loss": 0.6651,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3071937186386715,
+      "learning_rate": 9.85537412891785e-06,
+      "loss": 0.644,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.047754010327356,
+      "learning_rate": 9.855210342686951e-06,
+      "loss": 0.5689,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7069184740402654,
+      "learning_rate": 9.855046465282427e-06,
+      "loss": 0.5617,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.119240280887836,
+      "learning_rate": 9.854882496707712e-06,
+      "loss": 0.5822,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8754162020132785,
+      "learning_rate": 9.854718436966237e-06,
+      "loss": 0.5819,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9340673657880993,
+      "learning_rate": 9.854554286061435e-06,
+      "loss": 0.5994,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.181548795231045,
+      "learning_rate": 9.854390043996744e-06,
+      "loss": 0.732,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0461322988704915,
+      "learning_rate": 9.854225710775603e-06,
+      "loss": 0.6745,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1733499064892943,
+      "learning_rate": 9.854061286401451e-06,
+      "loss": 0.6165,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.068301801674872,
+      "learning_rate": 9.85389677087773e-06,
+      "loss": 0.5737,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.167409506635329,
+      "learning_rate": 9.853732164207883e-06,
+      "loss": 0.6416,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9752112029366724,
+      "learning_rate": 9.853567466395361e-06,
+      "loss": 0.6086,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1669531018719357,
+      "learning_rate": 9.853402677443605e-06,
+      "loss": 0.5778,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.065064177283796,
+      "learning_rate": 9.85323779735607e-06,
+      "loss": 0.4536,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.954080565770206,
+      "learning_rate": 9.853072826136205e-06,
+      "loss": 0.6959,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0268979063753263,
+      "learning_rate": 9.852907763787463e-06,
+      "loss": 0.6475,
+      "step": 1772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.767964090163848,
+      "learning_rate": 9.852742610313301e-06,
+      "loss": 0.6138,
+      "step": 1773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1846683358030234,
+      "learning_rate": 9.852577365717176e-06,
+      "loss": 0.5802,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9682549548394672,
+      "learning_rate": 9.85241203000255e-06,
+      "loss": 0.5992,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3392455345957197,
+      "learning_rate": 9.852246603172878e-06,
+      "loss": 0.5623,
+      "step": 1776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6605304331040858,
+      "learning_rate": 9.852081085231628e-06,
+      "loss": 0.5111,
+      "step": 1777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9647385998122877,
+      "learning_rate": 9.851915476182264e-06,
+      "loss": 0.6013,
+      "step": 1778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.225742340287655,
+      "learning_rate": 9.851749776028252e-06,
+      "loss": 0.6333,
+      "step": 1779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.762305063781781,
+      "learning_rate": 9.851583984773062e-06,
+      "loss": 0.5465,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6947977348337815,
+      "learning_rate": 9.851418102420164e-06,
+      "loss": 0.5083,
+      "step": 1781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.026995264542644,
+      "learning_rate": 9.851252128973028e-06,
+      "loss": 0.5366,
+      "step": 1782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.993234026661848,
+      "learning_rate": 9.851086064435136e-06,
+      "loss": 0.5688,
+      "step": 1783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4010188157070016,
+      "learning_rate": 9.850919908809958e-06,
+      "loss": 0.5619,
+      "step": 1784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7732529534333115,
+      "learning_rate": 9.850753662100973e-06,
+      "loss": 0.4611,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.960924203390651,
+      "learning_rate": 9.850587324311663e-06,
+      "loss": 0.6357,
+      "step": 1786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0151977678242012,
+      "learning_rate": 9.85042089544551e-06,
+      "loss": 0.641,
+      "step": 1787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.897758198366653,
+      "learning_rate": 9.850254375506e-06,
+      "loss": 0.4763,
+      "step": 1788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5090142808672753,
+      "learning_rate": 9.850087764496612e-06,
+      "loss": 0.4803,
+      "step": 1789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9611089577704726,
+      "learning_rate": 9.84992106242084e-06,
+      "loss": 0.5387,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8877830914479452,
+      "learning_rate": 9.849754269282175e-06,
+      "loss": 0.46,
+      "step": 1791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.303759812110683,
+      "learning_rate": 9.849587385084104e-06,
+      "loss": 0.6377,
+      "step": 1792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6385364114277545,
+      "learning_rate": 9.849420409830124e-06,
+      "loss": 0.451,
+      "step": 1793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.067320452371506,
+      "learning_rate": 9.849253343523728e-06,
+      "loss": 0.6237,
+      "step": 1794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7382837063877334,
+      "learning_rate": 9.849086186168415e-06,
+      "loss": 0.445,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1930657194488252,
+      "learning_rate": 9.848918937767683e-06,
+      "loss": 0.5658,
+      "step": 1796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0912423908288784,
+      "learning_rate": 9.848751598325037e-06,
+      "loss": 0.7696,
+      "step": 1797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9023047605036074,
+      "learning_rate": 9.848584167843975e-06,
+      "loss": 0.6578,
+      "step": 1798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.103384263287123,
+      "learning_rate": 9.848416646328005e-06,
+      "loss": 0.6017,
+      "step": 1799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0348369704117517,
+      "learning_rate": 9.848249033780633e-06,
+      "loss": 0.779,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.6079102158546448,
+      "eval_runtime": 67.9838,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.698210762907348,
+      "learning_rate": 9.848081330205369e-06,
+      "loss": 0.5812,
+      "step": 1801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.231110092069755,
+      "learning_rate": 9.847913535605722e-06,
+      "loss": 0.7925,
+      "step": 1802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.19060754815724,
+      "learning_rate": 9.847745649985208e-06,
+      "loss": 0.7454,
+      "step": 1803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.009322071913332,
+      "learning_rate": 9.847577673347338e-06,
+      "loss": 0.7148,
+      "step": 1804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4214478958472325,
+      "learning_rate": 9.84740960569563e-06,
+      "loss": 0.7445,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8491930701683108,
+      "learning_rate": 9.847241447033601e-06,
+      "loss": 0.589,
+      "step": 1806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0849466691336214,
+      "learning_rate": 9.847073197364772e-06,
+      "loss": 0.5698,
+      "step": 1807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2355482279435637,
+      "learning_rate": 9.84690485669267e-06,
+      "loss": 0.5387,
+      "step": 1808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.187249183305946,
+      "learning_rate": 9.84673642502081e-06,
+      "loss": 0.67,
+      "step": 1809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8496896040585091,
+      "learning_rate": 9.846567902352725e-06,
+      "loss": 0.5834,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.730912238300789,
+      "learning_rate": 9.84639928869194e-06,
+      "loss": 0.4288,
+      "step": 1811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5291226629897565,
+      "learning_rate": 9.846230584041986e-06,
+      "loss": 0.6064,
+      "step": 1812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.243978016825108,
+      "learning_rate": 9.846061788406394e-06,
+      "loss": 0.6914,
+      "step": 1813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.354526057308139,
+      "learning_rate": 9.845892901788697e-06,
+      "loss": 0.718,
+      "step": 1814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9507116616679487,
+      "learning_rate": 9.845723924192434e-06,
+      "loss": 0.5265,
+      "step": 1815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.453504027785381,
+      "learning_rate": 9.845554855621138e-06,
+      "loss": 0.6257,
+      "step": 1816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0347600206909138,
+      "learning_rate": 9.84538569607835e-06,
+      "loss": 0.5164,
+      "step": 1817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8878090096755065,
+      "learning_rate": 9.845216445567612e-06,
+      "loss": 0.4995,
+      "step": 1818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7753801742645707,
+      "learning_rate": 9.845047104092467e-06,
+      "loss": 0.5624,
+      "step": 1819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.135542640766017,
+      "learning_rate": 9.84487767165646e-06,
+      "loss": 0.5551,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.550582100025421,
+      "learning_rate": 9.844708148263137e-06,
+      "loss": 0.4117,
+      "step": 1821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0322303440460385,
+      "learning_rate": 9.844538533916047e-06,
+      "loss": 0.5649,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.954806938358545,
+      "learning_rate": 9.844368828618742e-06,
+      "loss": 0.5754,
+      "step": 1823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.651758274225083,
+      "learning_rate": 9.844199032374774e-06,
+      "loss": 0.7019,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1015960363340125,
+      "learning_rate": 9.844029145187697e-06,
+      "loss": 0.68,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2794226553492116,
+      "learning_rate": 9.843859167061069e-06,
+      "loss": 0.7332,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3855267677793544,
+      "learning_rate": 9.843689097998446e-06,
+      "loss": 0.582,
+      "step": 1827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.813687596114917,
+      "learning_rate": 9.84351893800339e-06,
+      "loss": 0.5808,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0557983546038066,
+      "learning_rate": 9.843348687079463e-06,
+      "loss": 0.6668,
+      "step": 1829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.362786919267076,
+      "learning_rate": 9.84317834523023e-06,
+      "loss": 0.5996,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8815815162494092,
+      "learning_rate": 9.843007912459254e-06,
+      "loss": 0.4645,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6338266862534225,
+      "learning_rate": 9.842837388770108e-06,
+      "loss": 0.6002,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.517891976378241,
+      "learning_rate": 9.842666774166357e-06,
+      "loss": 0.6634,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3186166950309355,
+      "learning_rate": 9.842496068651573e-06,
+      "loss": 0.6454,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.021991052320413,
+      "learning_rate": 9.842325272229334e-06,
+      "loss": 0.6829,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.870575401826906,
+      "learning_rate": 9.84215438490321e-06,
+      "loss": 0.5942,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.560645910806879,
+      "learning_rate": 9.841983406676781e-06,
+      "loss": 0.4252,
+      "step": 1837
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1434793520064024,
+      "learning_rate": 9.841812337553628e-06,
+      "loss": 0.5413,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5734594196126823,
+      "learning_rate": 9.84164117753733e-06,
+      "loss": 0.4387,
+      "step": 1839
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.766409841331108,
+      "learning_rate": 9.84146992663147e-06,
+      "loss": 0.6161,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1054210001632216,
+      "learning_rate": 9.841298584839634e-06,
+      "loss": 0.6206,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2977660820201087,
+      "learning_rate": 9.84112715216541e-06,
+      "loss": 0.7061,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6748712447998462,
+      "learning_rate": 9.840955628612383e-06,
+      "loss": 0.4491,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8004539372141464,
+      "learning_rate": 9.840784014184146e-06,
+      "loss": 0.4334,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7508940954488046,
+      "learning_rate": 9.840612308884293e-06,
+      "loss": 0.4428,
+      "step": 1845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3120553493543774,
+      "learning_rate": 9.840440512716418e-06,
+      "loss": 0.6302,
+      "step": 1846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6606211318204918,
+      "learning_rate": 9.840268625684116e-06,
+      "loss": 0.6444,
+      "step": 1847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7059364086228648,
+      "learning_rate": 9.840096647790986e-06,
+      "loss": 0.5877,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.322439166311427,
+      "learning_rate": 9.83992457904063e-06,
+      "loss": 0.825,
+      "step": 1849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.444562506784104,
+      "learning_rate": 9.839752419436647e-06,
+      "loss": 0.5104,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.6092979311943054,
+      "eval_runtime": 67.996,
+      "eval_samples_per_second": 50.15,
+      "eval_steps_per_second": 1.574,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.911227675984383,
+      "learning_rate": 9.839580168982643e-06,
+      "loss": 0.5519,
+      "step": 1851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9558653478035215,
+      "learning_rate": 9.839407827682224e-06,
+      "loss": 0.5009,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.457622975881183,
+      "learning_rate": 9.839235395538997e-06,
+      "loss": 0.7043,
+      "step": 1853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8065715576233157,
+      "learning_rate": 9.839062872556572e-06,
+      "loss": 0.6197,
+      "step": 1854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.808131163947032,
+      "learning_rate": 9.838890258738561e-06,
+      "loss": 0.4735,
+      "step": 1855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.151317928664454,
+      "learning_rate": 9.838717554088577e-06,
+      "loss": 0.519,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3929041326891864,
+      "learning_rate": 9.838544758610235e-06,
+      "loss": 0.5697,
+      "step": 1857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.092625246039508,
+      "learning_rate": 9.838371872307156e-06,
+      "loss": 0.5285,
+      "step": 1858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0234250456388643,
+      "learning_rate": 9.838198895182955e-06,
+      "loss": 0.5457,
+      "step": 1859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9944367134341214,
+      "learning_rate": 9.838025827241255e-06,
+      "loss": 0.5546,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7674851375176732,
+      "learning_rate": 9.837852668485678e-06,
+      "loss": 0.3848,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1653492858966756,
+      "learning_rate": 9.837679418919851e-06,
+      "loss": 0.6191,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9679464326019689,
+      "learning_rate": 9.837506078547398e-06,
+      "loss": 0.6249,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5114835247912988,
+      "learning_rate": 9.837332647371951e-06,
+      "loss": 0.5502,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0928992900943815,
+      "learning_rate": 9.837159125397138e-06,
+      "loss": 0.6733,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2693786274326038,
+      "learning_rate": 9.836985512626593e-06,
+      "loss": 0.6342,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0746302359053455,
+      "learning_rate": 9.83681180906395e-06,
+      "loss": 0.6275,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9178786498245946,
+      "learning_rate": 9.836638014712844e-06,
+      "loss": 0.5806,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.774301491537367,
+      "learning_rate": 9.836464129576916e-06,
+      "loss": 0.4272,
+      "step": 1869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2295845621758015,
+      "learning_rate": 9.836290153659807e-06,
+      "loss": 0.6233,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6908870655525323,
+      "learning_rate": 9.836116086965153e-06,
+      "loss": 0.432,
+      "step": 1871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8647937873982323,
+      "learning_rate": 9.835941929496604e-06,
+      "loss": 0.5532,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6630235376233933,
+      "learning_rate": 9.835767681257806e-06,
+      "loss": 0.663,
+      "step": 1873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9835832738951849,
+      "learning_rate": 9.835593342252402e-06,
+      "loss": 0.5894,
+      "step": 1874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9538443210987462,
+      "learning_rate": 9.835418912484045e-06,
+      "loss": 0.6783,
+      "step": 1875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.222868713169519,
+      "learning_rate": 9.835244391956386e-06,
+      "loss": 0.4268,
+      "step": 1876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3261485521142213,
+      "learning_rate": 9.83506978067308e-06,
+      "loss": 0.5777,
+      "step": 1877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9890131194961227,
+      "learning_rate": 9.834895078637776e-06,
+      "loss": 0.6899,
+      "step": 1878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.906730074917037,
+      "learning_rate": 9.834720285854139e-06,
+      "loss": 0.4818,
+      "step": 1879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7353820912134093,
+      "learning_rate": 9.834545402325827e-06,
+      "loss": 0.4549,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8575259956586028,
+      "learning_rate": 9.834370428056497e-06,
+      "loss": 0.5192,
+      "step": 1881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3432804319170186,
+      "learning_rate": 9.834195363049815e-06,
+      "loss": 0.7037,
+      "step": 1882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.300647507685586,
+      "learning_rate": 9.834020207309444e-06,
+      "loss": 0.6291,
+      "step": 1883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8925833392191291,
+      "learning_rate": 9.833844960839053e-06,
+      "loss": 0.5644,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.142665543683844,
+      "learning_rate": 9.83366962364231e-06,
+      "loss": 0.6042,
+      "step": 1885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.586679422599711,
+      "learning_rate": 9.833494195722885e-06,
+      "loss": 0.5229,
+      "step": 1886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1144141039891746,
+      "learning_rate": 9.833318677084449e-06,
+      "loss": 0.6804,
+      "step": 1887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.101642285028283,
+      "learning_rate": 9.83314306773068e-06,
+      "loss": 0.5191,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9171510173114277,
+      "learning_rate": 9.83296736766525e-06,
+      "loss": 0.7212,
+      "step": 1889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.004061452135989,
+      "learning_rate": 9.832791576891842e-06,
+      "loss": 0.6927,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5744390336525194,
+      "learning_rate": 9.832615695414133e-06,
+      "loss": 0.5946,
+      "step": 1891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.223606748125307,
+      "learning_rate": 9.832439723235805e-06,
+      "loss": 0.5812,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3208610531987164,
+      "learning_rate": 9.832263660360541e-06,
+      "loss": 0.5302,
+      "step": 1893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.050499050614514,
+      "learning_rate": 9.83208750679203e-06,
+      "loss": 0.6488,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4983267375131377,
+      "learning_rate": 9.831911262533959e-06,
+      "loss": 0.6002,
+      "step": 1895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.831017265112215,
+      "learning_rate": 9.831734927590013e-06,
+      "loss": 0.5069,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.235115350223994,
+      "learning_rate": 9.831558501963889e-06,
+      "loss": 0.5068,
+      "step": 1897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4223669562210866,
+      "learning_rate": 9.831381985659278e-06,
+      "loss": 0.6583,
+      "step": 1898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.025254647326447,
+      "learning_rate": 9.831205378679876e-06,
+      "loss": 0.605,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9181233434572575,
+      "learning_rate": 9.831028681029378e-06,
+      "loss": 0.5449,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.6070182919502258,
+      "eval_runtime": 68.021,
+      "eval_samples_per_second": 50.132,
+      "eval_steps_per_second": 1.573,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.398112517166929,
+      "learning_rate": 9.830851892711485e-06,
+      "loss": 0.7775,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8962184244228517,
+      "learning_rate": 9.830675013729898e-06,
+      "loss": 0.4339,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1338860051358686,
+      "learning_rate": 9.83049804408832e-06,
+      "loss": 0.6937,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.940691757455467,
+      "learning_rate": 9.830320983790455e-06,
+      "loss": 0.5347,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2180605474756154,
+      "learning_rate": 9.83014383284001e-06,
+      "loss": 0.6104,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0856513008709023,
+      "learning_rate": 9.829966591240692e-06,
+      "loss": 0.5576,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2516930929089054,
+      "learning_rate": 9.829789258996214e-06,
+      "loss": 0.6249,
+      "step": 1907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0155079820346766,
+      "learning_rate": 9.829611836110288e-06,
+      "loss": 0.6937,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.699820258113503,
+      "learning_rate": 9.829434322586625e-06,
+      "loss": 0.7624,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.424541390227065,
+      "learning_rate": 9.829256718428947e-06,
+      "loss": 0.6952,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.38639384638379,
+      "learning_rate": 9.829079023640965e-06,
+      "loss": 0.666,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3070059610060656,
+      "learning_rate": 9.828901238226405e-06,
+      "loss": 0.5726,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6559223514962738,
+      "learning_rate": 9.828723362188985e-06,
+      "loss": 0.5498,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.16625867249227,
+      "learning_rate": 9.828545395532432e-06,
+      "loss": 0.5987,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.839447679211662,
+      "learning_rate": 9.828367338260467e-06,
+      "loss": 0.6836,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8597609852129142,
+      "learning_rate": 9.828189190376823e-06,
+      "loss": 0.5725,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7296481958558159,
+      "learning_rate": 9.828010951885223e-06,
+      "loss": 0.374,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2803542630064415,
+      "learning_rate": 9.827832622789406e-06,
+      "loss": 0.4918,
+      "step": 1918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7461248789664607,
+      "learning_rate": 9.827654203093098e-06,
+      "loss": 0.5852,
+      "step": 1919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4540174266101613,
+      "learning_rate": 9.827475692800039e-06,
+      "loss": 0.7522,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1578171364267957,
+      "learning_rate": 9.827297091913964e-06,
+      "loss": 0.6041,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.11781781457771,
+      "learning_rate": 9.827118400438612e-06,
+      "loss": 0.7363,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9219715868094367,
+      "learning_rate": 9.826939618377724e-06,
+      "loss": 0.7124,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2968732464535413,
+      "learning_rate": 9.826760745735042e-06,
+      "loss": 0.648,
+      "step": 1924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9072219526896232,
+      "learning_rate": 9.826581782514312e-06,
+      "loss": 0.5209,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.094738898365208,
+      "learning_rate": 9.82640272871928e-06,
+      "loss": 0.6811,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2654748101473765,
+      "learning_rate": 9.826223584353694e-06,
+      "loss": 0.7503,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1632024174196967,
+      "learning_rate": 9.826044349421303e-06,
+      "loss": 0.5993,
+      "step": 1928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2746628987398227,
+      "learning_rate": 9.825865023925863e-06,
+      "loss": 0.7718,
+      "step": 1929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1365292120647426,
+      "learning_rate": 9.825685607871123e-06,
+      "loss": 0.588,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.051670814729897,
+      "learning_rate": 9.825506101260843e-06,
+      "loss": 0.6175,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.853534815631942,
+      "learning_rate": 9.825326504098779e-06,
+      "loss": 0.5421,
+      "step": 1932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0476808444050487,
+      "learning_rate": 9.825146816388692e-06,
+      "loss": 0.5656,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.851222108410715,
+      "learning_rate": 9.824967038134343e-06,
+      "loss": 0.6042,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9202317428751379,
+      "learning_rate": 9.824787169339493e-06,
+      "loss": 0.5895,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1804557352391853,
+      "learning_rate": 9.824607210007913e-06,
+      "loss": 0.6249,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1406849596965305,
+      "learning_rate": 9.824427160143367e-06,
+      "loss": 0.607,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.009350586525911,
+      "learning_rate": 9.824247019749622e-06,
+      "loss": 0.4978,
+      "step": 1938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7022798991553505,
+      "learning_rate": 9.824066788830452e-06,
+      "loss": 0.4905,
+      "step": 1939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3958377778249615,
+      "learning_rate": 9.82388646738963e-06,
+      "loss": 0.5672,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9049521528513451,
+      "learning_rate": 9.823706055430932e-06,
+      "loss": 0.529,
+      "step": 1941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7305092079641107,
+      "learning_rate": 9.823525552958131e-06,
+      "loss": 0.5185,
+      "step": 1942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6876404709331683,
+      "learning_rate": 9.82334495997501e-06,
+      "loss": 0.5333,
+      "step": 1943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1967702077916624,
+      "learning_rate": 9.823164276485345e-06,
+      "loss": 0.5892,
+      "step": 1944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.339543419419911,
+      "learning_rate": 9.822983502492924e-06,
+      "loss": 0.6986,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7795616581399496,
+      "learning_rate": 9.822802638001527e-06,
+      "loss": 0.5472,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.726096253491475,
+      "learning_rate": 9.822621683014941e-06,
+      "loss": 0.7323,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.552536186210813,
+      "learning_rate": 9.822440637536956e-06,
+      "loss": 0.3994,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0252003125155045,
+      "learning_rate": 9.822259501571361e-06,
+      "loss": 0.5317,
+      "step": 1949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.55159469031004,
+      "learning_rate": 9.822078275121948e-06,
+      "loss": 0.4233,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.6056253910064697,
+      "eval_runtime": 67.9861,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0129208954583175,
+      "learning_rate": 9.82189695819251e-06,
+      "loss": 0.5517,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.316195894450904,
+      "learning_rate": 9.821715550786844e-06,
+      "loss": 0.6213,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3249028923022697,
+      "learning_rate": 9.821534052908748e-06,
+      "loss": 0.7028,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7387279763447132,
+      "learning_rate": 9.821352464562019e-06,
+      "loss": 0.5592,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7460804732461737,
+      "learning_rate": 9.821170785750464e-06,
+      "loss": 0.5307,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.873022683350261,
+      "learning_rate": 9.82098901647788e-06,
+      "loss": 0.5879,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0384213691410347,
+      "learning_rate": 9.820807156748075e-06,
+      "loss": 0.4051,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5748108901804128,
+      "learning_rate": 9.820625206564855e-06,
+      "loss": 0.4376,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.812297021111459,
+      "learning_rate": 9.820443165932029e-06,
+      "loss": 0.6029,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.680959512418545,
+      "learning_rate": 9.820261034853412e-06,
+      "loss": 0.7279,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9812950132358689,
+      "learning_rate": 9.820078813332812e-06,
+      "loss": 0.507,
+      "step": 1961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.42110271801623,
+      "learning_rate": 9.819896501374046e-06,
+      "loss": 0.5746,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7837319117106518,
+      "learning_rate": 9.81971409898093e-06,
+      "loss": 0.484,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.16724564088432,
+      "learning_rate": 9.819531606157283e-06,
+      "loss": 0.6233,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6706737529835505,
+      "learning_rate": 9.819349022906924e-06,
+      "loss": 0.5176,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1526286021785217,
+      "learning_rate": 9.819166349233676e-06,
+      "loss": 0.8141,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.950166818805396,
+      "learning_rate": 9.818983585141362e-06,
+      "loss": 0.7453,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3201074668566233,
+      "learning_rate": 9.81880073063381e-06,
+      "loss": 0.723,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7869906431750118,
+      "learning_rate": 9.818617785714847e-06,
+      "loss": 0.532,
+      "step": 1969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0397250021041784,
+      "learning_rate": 9.818434750388304e-06,
+      "loss": 0.5615,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7438619565142828,
+      "learning_rate": 9.818251624658011e-06,
+      "loss": 0.3897,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.258636509571346,
+      "learning_rate": 9.818068408527805e-06,
+      "loss": 0.5847,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0081814148737704,
+      "learning_rate": 9.817885102001515e-06,
+      "loss": 0.6314,
+      "step": 1973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.034717213723714,
+      "learning_rate": 9.817701705082985e-06,
+      "loss": 0.5687,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.612000357226754,
+      "learning_rate": 9.81751821777605e-06,
+      "loss": 0.8523,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2787746375282505,
+      "learning_rate": 9.817334640084555e-06,
+      "loss": 0.6508,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7084165704157146,
+      "learning_rate": 9.817150972012336e-06,
+      "loss": 0.4983,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.13348882150106,
+      "learning_rate": 9.816967213563248e-06,
+      "loss": 0.6219,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2144928537645083,
+      "learning_rate": 9.816783364741131e-06,
+      "loss": 0.7452,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.126525922258691,
+      "learning_rate": 9.816599425549835e-06,
+      "loss": 0.4868,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2744029459790145,
+      "learning_rate": 9.816415395993211e-06,
+      "loss": 0.5092,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4220477130717268,
+      "learning_rate": 9.816231276075113e-06,
+      "loss": 0.342,
+      "step": 1982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9489067831388727,
+      "learning_rate": 9.816047065799394e-06,
+      "loss": 0.6585,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9537430766413497,
+      "learning_rate": 9.81586276516991e-06,
+      "loss": 0.5117,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.279671077367545,
+      "learning_rate": 9.81567837419052e-06,
+      "loss": 0.6471,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5111857353158036,
+      "learning_rate": 9.815493892865084e-06,
+      "loss": 0.6949,
+      "step": 1986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.214394147090711,
+      "learning_rate": 9.815309321197462e-06,
+      "loss": 0.4923,
+      "step": 1987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.150853729794921,
+      "learning_rate": 9.815124659191522e-06,
+      "loss": 0.4597,
+      "step": 1988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9486848926005818,
+      "learning_rate": 9.814939906851127e-06,
+      "loss": 0.5224,
+      "step": 1989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7079286872804567,
+      "learning_rate": 9.814755064180144e-06,
+      "loss": 0.6415,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7412469147230762,
+      "learning_rate": 9.814570131182444e-06,
+      "loss": 0.6353,
+      "step": 1991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6675067901914518,
+      "learning_rate": 9.814385107861898e-06,
+      "loss": 0.5528,
+      "step": 1992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.316156137603184,
+      "learning_rate": 9.81419999422238e-06,
+      "loss": 0.482,
+      "step": 1993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0823110199149175,
+      "learning_rate": 9.814014790267767e-06,
+      "loss": 0.625,
+      "step": 1994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.27560939238802,
+      "learning_rate": 9.813829496001931e-06,
+      "loss": 0.7612,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2604732750702987,
+      "learning_rate": 9.813644111428754e-06,
+      "loss": 0.678,
+      "step": 1996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5232352399177556,
+      "learning_rate": 9.813458636552116e-06,
+      "loss": 0.5257,
+      "step": 1997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7419152021571043,
+      "learning_rate": 9.813273071375905e-06,
+      "loss": 0.4582,
+      "step": 1998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6134406972880637,
+      "learning_rate": 9.813087415903997e-06,
+      "loss": 0.4188,
+      "step": 1999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.047021706653522,
+      "learning_rate": 9.812901670140284e-06,
+      "loss": 0.6453,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.6049335598945618,
+      "eval_runtime": 67.9952,
+      "eval_samples_per_second": 50.151,
+      "eval_steps_per_second": 1.574,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0542752069134855,
+      "learning_rate": 9.812715834088655e-06,
+      "loss": 0.6126,
+      "step": 2001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1010566735147083,
+      "learning_rate": 9.812529907752998e-06,
+      "loss": 0.625,
+      "step": 2002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9044361673675143,
+      "learning_rate": 9.812343891137204e-06,
+      "loss": 0.5504,
+      "step": 2003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.456798724148524,
+      "learning_rate": 9.812157784245173e-06,
+      "loss": 0.5708,
+      "step": 2004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.122628190783095,
+      "learning_rate": 9.811971587080793e-06,
+      "loss": 0.6398,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9745418448667509,
+      "learning_rate": 9.811785299647967e-06,
+      "loss": 0.5642,
+      "step": 2006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1448480131930605,
+      "learning_rate": 9.811598921950596e-06,
+      "loss": 0.7057,
+      "step": 2007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.895647078376675,
+      "learning_rate": 9.811412453992577e-06,
+      "loss": 0.5118,
+      "step": 2008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.901340236853861,
+      "learning_rate": 9.81122589577782e-06,
+      "loss": 0.6307,
+      "step": 2009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7627498840343145,
+      "learning_rate": 9.811039247310224e-06,
+      "loss": 0.5818,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.061178673526384,
+      "learning_rate": 9.810852508593698e-06,
+      "loss": 0.526,
+      "step": 2011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0728647405473546,
+      "learning_rate": 9.810665679632154e-06,
+      "loss": 0.6274,
+      "step": 2012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2355971013129463,
+      "learning_rate": 9.8104787604295e-06,
+      "loss": 0.6364,
+      "step": 2013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.197369434181052,
+      "learning_rate": 9.810291750989652e-06,
+      "loss": 0.4979,
+      "step": 2014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.151136788352086,
+      "learning_rate": 9.810104651316524e-06,
+      "loss": 0.6091,
+      "step": 2015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7835698897079773,
+      "learning_rate": 9.809917461414031e-06,
+      "loss": 0.459,
+      "step": 2016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.313847418812589,
+      "learning_rate": 9.809730181286092e-06,
+      "loss": 0.659,
+      "step": 2017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9169083190961391,
+      "learning_rate": 9.80954281093663e-06,
+      "loss": 0.6326,
+      "step": 2018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1303025592245337,
+      "learning_rate": 9.809355350369566e-06,
+      "loss": 0.6344,
+      "step": 2019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9344942547832802,
+      "learning_rate": 9.809167799588825e-06,
+      "loss": 0.5153,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8462740687640853,
+      "learning_rate": 9.808980158598331e-06,
+      "loss": 0.3818,
+      "step": 2021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3086293729338307,
+      "learning_rate": 9.808792427402013e-06,
+      "loss": 0.6583,
+      "step": 2022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.922178976459067,
+      "learning_rate": 9.808604606003804e-06,
+      "loss": 0.4243,
+      "step": 2023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.028445916194619,
+      "learning_rate": 9.808416694407634e-06,
+      "loss": 0.548,
+      "step": 2024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1575516177292395,
+      "learning_rate": 9.808228692617435e-06,
+      "loss": 0.6256,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8993262240763191,
+      "learning_rate": 9.808040600637144e-06,
+      "loss": 0.5338,
+      "step": 2026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.199417934087717,
+      "learning_rate": 9.8078524184707e-06,
+      "loss": 0.6717,
+      "step": 2027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.305340688237323,
+      "learning_rate": 9.80766414612204e-06,
+      "loss": 0.5532,
+      "step": 2028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8263547823291593,
+      "learning_rate": 9.80747578359511e-06,
+      "loss": 0.5182,
+      "step": 2029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.936738097160926,
+      "learning_rate": 9.807287330893847e-06,
+      "loss": 0.6134,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9638875871088652,
+      "learning_rate": 9.8070987880222e-06,
+      "loss": 0.5959,
+      "step": 2031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3920701274642013,
+      "learning_rate": 9.806910154984114e-06,
+      "loss": 0.6148,
+      "step": 2032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1365030423772633,
+      "learning_rate": 9.80672143178354e-06,
+      "loss": 0.6335,
+      "step": 2033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.111148154058807,
+      "learning_rate": 9.806532618424427e-06,
+      "loss": 0.6857,
+      "step": 2034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3075238509332587,
+      "learning_rate": 9.80634371491073e-06,
+      "loss": 0.4871,
+      "step": 2035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.49056797693049,
+      "learning_rate": 9.806154721246402e-06,
+      "loss": 0.7631,
+      "step": 2036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.31020587633623,
+      "learning_rate": 9.8059656374354e-06,
+      "loss": 0.6817,
+      "step": 2037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8932261269884862,
+      "learning_rate": 9.805776463481678e-06,
+      "loss": 0.472,
+      "step": 2038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.412982886147579,
+      "learning_rate": 9.805587199389205e-06,
+      "loss": 0.6946,
+      "step": 2039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6967537266614752,
+      "learning_rate": 9.805397845161934e-06,
+      "loss": 0.4503,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.229235115119224,
+      "learning_rate": 9.805208400803837e-06,
+      "loss": 0.6307,
+      "step": 2041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1609524040298913,
+      "learning_rate": 9.805018866318874e-06,
+      "loss": 0.5756,
+      "step": 2042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.066459441362632,
+      "learning_rate": 9.804829241711017e-06,
+      "loss": 0.7448,
+      "step": 2043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8818050669879587,
+      "learning_rate": 9.80463952698423e-06,
+      "loss": 0.5893,
+      "step": 2044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6154737866241742,
+      "learning_rate": 9.804449722142492e-06,
+      "loss": 0.3888,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8679347818613745,
+      "learning_rate": 9.804259827189772e-06,
+      "loss": 0.5462,
+      "step": 2046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4347841760214104,
+      "learning_rate": 9.804069842130044e-06,
+      "loss": 0.5827,
+      "step": 2047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.247066752076186,
+      "learning_rate": 9.803879766967289e-06,
+      "loss": 0.5203,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.118273147792905,
+      "learning_rate": 9.803689601705484e-06,
+      "loss": 0.5579,
+      "step": 2049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.230872933065364,
+      "learning_rate": 9.80349934634861e-06,
+      "loss": 0.6582,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.6042057275772095,
+      "eval_runtime": 68.0257,
+      "eval_samples_per_second": 50.128,
+      "eval_steps_per_second": 1.573,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4338322520721976,
+      "learning_rate": 9.803309000900651e-06,
+      "loss": 0.5994,
+      "step": 2051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.171513442780239,
+      "learning_rate": 9.803118565365588e-06,
+      "loss": 0.6556,
+      "step": 2052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3295104161156037,
+      "learning_rate": 9.802928039747415e-06,
+      "loss": 0.561,
+      "step": 2053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6217018554254667,
+      "learning_rate": 9.802737424050113e-06,
+      "loss": 0.5207,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.343733916938275,
+      "learning_rate": 9.802546718277677e-06,
+      "loss": 0.7535,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6490452914430598,
+      "learning_rate": 9.802355922434097e-06,
+      "loss": 0.5635,
+      "step": 2056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9169686583376038,
+      "learning_rate": 9.802165036523368e-06,
+      "loss": 0.4548,
+      "step": 2057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.782262617742172,
+      "learning_rate": 9.801974060549487e-06,
+      "loss": 0.651,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7410612407778805,
+      "learning_rate": 9.80178299451645e-06,
+      "loss": 0.4405,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0202340751864125,
+      "learning_rate": 9.80159183842826e-06,
+      "loss": 0.5705,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1933165313729575,
+      "learning_rate": 9.801400592288914e-06,
+      "loss": 0.746,
+      "step": 2061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2420174239619413,
+      "learning_rate": 9.801209256102419e-06,
+      "loss": 0.4444,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0938650949433435,
+      "learning_rate": 9.801017829872781e-06,
+      "loss": 0.7085,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.314810980052965,
+      "learning_rate": 9.800826313604005e-06,
+      "loss": 0.5257,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0041784925107566,
+      "learning_rate": 9.800634707300103e-06,
+      "loss": 0.5674,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.348052343608789,
+      "learning_rate": 9.800443010965083e-06,
+      "loss": 0.6629,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.420737863457457,
+      "learning_rate": 9.80025122460296e-06,
+      "loss": 0.6383,
+      "step": 2067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3155918458854217,
+      "learning_rate": 9.80005934821775e-06,
+      "loss": 0.5542,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2292712924348894,
+      "learning_rate": 9.799867381813468e-06,
+      "loss": 0.7248,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2030458988734165,
+      "learning_rate": 9.799675325394131e-06,
+      "loss": 0.549,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1795619207393635,
+      "learning_rate": 9.799483178963764e-06,
+      "loss": 0.6081,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.20267159724542,
+      "learning_rate": 9.799290942526387e-06,
+      "loss": 0.5109,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9955828967487799,
+      "learning_rate": 9.799098616086023e-06,
+      "loss": 0.607,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8295563868639966,
+      "learning_rate": 9.7989061996467e-06,
+      "loss": 0.5315,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.283040356206018,
+      "learning_rate": 9.798713693212447e-06,
+      "loss": 0.6223,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2867648837248584,
+      "learning_rate": 9.79852109678729e-06,
+      "loss": 0.7447,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6708398965968996,
+      "learning_rate": 9.798328410375268e-06,
+      "loss": 0.4643,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8942000045085945,
+      "learning_rate": 9.798135633980408e-06,
+      "loss": 0.4738,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.156850796707967,
+      "learning_rate": 9.797942767606749e-06,
+      "loss": 0.843,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.086454172223535,
+      "learning_rate": 9.797749811258326e-06,
+      "loss": 0.5214,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.256840156182796,
+      "learning_rate": 9.797556764939182e-06,
+      "loss": 0.6502,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9013795642836377,
+      "learning_rate": 9.797363628653357e-06,
+      "loss": 0.4898,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6439230839881445,
+      "learning_rate": 9.797170402404892e-06,
+      "loss": 0.4901,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.401214691748773,
+      "learning_rate": 9.796977086197834e-06,
+      "loss": 0.5792,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0672529536455224,
+      "learning_rate": 9.79678368003623e-06,
+      "loss": 0.5825,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.865378713099675,
+      "learning_rate": 9.796590183924129e-06,
+      "loss": 0.5782,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6147037786252993,
+      "learning_rate": 9.79639659786558e-06,
+      "loss": 0.7837,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.810280375384069,
+      "learning_rate": 9.796202921864639e-06,
+      "loss": 0.6104,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.150510328431689,
+      "learning_rate": 9.796009155925356e-06,
+      "loss": 0.7442,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2603451297523325,
+      "learning_rate": 9.79581530005179e-06,
+      "loss": 0.8325,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1145420385292564,
+      "learning_rate": 9.795621354248e-06,
+      "loss": 0.573,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7581681927098542,
+      "learning_rate": 9.795427318518044e-06,
+      "loss": 0.5242,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7584073702340637,
+      "learning_rate": 9.795233192865986e-06,
+      "loss": 0.5232,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.899900673191256,
+      "learning_rate": 9.795038977295887e-06,
+      "loss": 0.569,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0353766294831876,
+      "learning_rate": 9.794844671811817e-06,
+      "loss": 0.5729,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.287189598715911,
+      "learning_rate": 9.794650276417841e-06,
+      "loss": 0.5831,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6355246546906683,
+      "learning_rate": 9.794455791118028e-06,
+      "loss": 0.5773,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.429918272915482,
+      "learning_rate": 9.794261215916452e-06,
+      "loss": 0.7554,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9661744225726396,
+      "learning_rate": 9.794066550817183e-06,
+      "loss": 0.6207,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.010002493965582,
+      "learning_rate": 9.793871795824298e-06,
+      "loss": 0.5183,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.6029190421104431,
+      "eval_runtime": 68.0076,
+      "eval_samples_per_second": 50.141,
+      "eval_steps_per_second": 1.573,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.352553740708236,
+      "learning_rate": 9.793676950941876e-06,
+      "loss": 0.6406,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.303001686842405,
+      "learning_rate": 9.793482016173994e-06,
+      "loss": 0.7269,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.850731975859487,
+      "learning_rate": 9.793286991524728e-06,
+      "loss": 0.5102,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6786349611820022,
+      "learning_rate": 9.793091876998169e-06,
+      "loss": 0.4861,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5651947234082038,
+      "learning_rate": 9.792896672598399e-06,
+      "loss": 0.489,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0579305706704476,
+      "learning_rate": 9.792701378329502e-06,
+      "loss": 0.5883,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8017557022082427,
+      "learning_rate": 9.792505994195567e-06,
+      "loss": 0.5287,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2491702105345777,
+      "learning_rate": 9.792310520200687e-06,
+      "loss": 0.6714,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1328186280338204,
+      "learning_rate": 9.792114956348953e-06,
+      "loss": 0.627,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.963118830062618,
+      "learning_rate": 9.791919302644458e-06,
+      "loss": 0.5784,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4691438000440797,
+      "learning_rate": 9.791723559091298e-06,
+      "loss": 0.6278,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9733626564598186,
+      "learning_rate": 9.79152772569357e-06,
+      "loss": 0.5445,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2892107601814304,
+      "learning_rate": 9.791331802455374e-06,
+      "loss": 0.5678,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.775414374845999,
+      "learning_rate": 9.791135789380814e-06,
+      "loss": 0.4941,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.07916872702188,
+      "learning_rate": 9.790939686473991e-06,
+      "loss": 0.5791,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9328547427925395,
+      "learning_rate": 9.79074349373901e-06,
+      "loss": 0.4983,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8394336658455315,
+      "learning_rate": 9.79054721117998e-06,
+      "loss": 0.4637,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2476358496227418,
+      "learning_rate": 9.790350838801009e-06,
+      "loss": 0.5833,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.939070241821524,
+      "learning_rate": 9.790154376606207e-06,
+      "loss": 0.5666,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.168454955909568,
+      "learning_rate": 9.789957824599688e-06,
+      "loss": 0.6845,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5370499138738856,
+      "learning_rate": 9.789761182785567e-06,
+      "loss": 0.5457,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8701709654649412,
+      "learning_rate": 9.789564451167961e-06,
+      "loss": 0.4794,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2795726654977,
+      "learning_rate": 9.789367629750986e-06,
+      "loss": 0.6476,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8006478321759283,
+      "learning_rate": 9.789170718538766e-06,
+      "loss": 0.5952,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7738141255655988,
+      "learning_rate": 9.78897371753542e-06,
+      "loss": 0.6546,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9719881581906633,
+      "learning_rate": 9.788776626745073e-06,
+      "loss": 0.4941,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8716414606791512,
+      "learning_rate": 9.788579446171851e-06,
+      "loss": 0.5476,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.295695509615188,
+      "learning_rate": 9.788382175819881e-06,
+      "loss": 0.723,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1770270843100414,
+      "learning_rate": 9.788184815693296e-06,
+      "loss": 0.6965,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5026844176276437,
+      "learning_rate": 9.787987365796224e-06,
+      "loss": 0.4595,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.171351694347793,
+      "learning_rate": 9.7877898261328e-06,
+      "loss": 0.6435,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3011214910246323,
+      "learning_rate": 9.78759219670716e-06,
+      "loss": 0.6748,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9668252813859182,
+      "learning_rate": 9.78739447752344e-06,
+      "loss": 0.4119,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6912928678414711,
+      "learning_rate": 9.78719666858578e-06,
+      "loss": 0.5556,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6848815150121468,
+      "learning_rate": 9.78699876989832e-06,
+      "loss": 0.4969,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8443535070135315,
+      "learning_rate": 9.786800781465203e-06,
+      "loss": 0.551,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0432733447370257,
+      "learning_rate": 9.786602703290574e-06,
+      "loss": 0.5228,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.585246583548223,
+      "learning_rate": 9.786404535378583e-06,
+      "loss": 0.567,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8167529108935783,
+      "learning_rate": 9.786206277733372e-06,
+      "loss": 0.4886,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8061010740547838,
+      "learning_rate": 9.786007930359097e-06,
+      "loss": 0.5436,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.823415631179881,
+      "learning_rate": 9.785809493259906e-06,
+      "loss": 0.521,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.011961787840989,
+      "learning_rate": 9.785610966439957e-06,
+      "loss": 0.5326,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7247370850707056,
+      "learning_rate": 9.785412349903404e-06,
+      "loss": 0.4799,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.072440555754284,
+      "learning_rate": 9.785213643654405e-06,
+      "loss": 0.5202,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.765499580240947,
+      "learning_rate": 9.785014847697121e-06,
+      "loss": 0.4891,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.040525414530265,
+      "learning_rate": 9.784815962035711e-06,
+      "loss": 0.5597,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.118885799551228,
+      "learning_rate": 9.784616986674341e-06,
+      "loss": 0.5756,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0079492202836464,
+      "learning_rate": 9.784417921617177e-06,
+      "loss": 0.6218,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0454494590034282,
+      "learning_rate": 9.784218766868383e-06,
+      "loss": 0.6001,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9993157928253227,
+      "learning_rate": 9.78401952243213e-06,
+      "loss": 0.6057,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.6040220260620117,
+      "eval_runtime": 68.0045,
+      "eval_samples_per_second": 50.144,
+      "eval_steps_per_second": 1.573,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.585410567692139,
+      "learning_rate": 9.783820188312591e-06,
+      "loss": 0.5048,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9809582208948584,
+      "learning_rate": 9.783620764513936e-06,
+      "loss": 0.6454,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.681455320730406,
+      "learning_rate": 9.783421251040341e-06,
+      "loss": 0.6113,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2509272848049133,
+      "learning_rate": 9.783221647895983e-06,
+      "loss": 0.6323,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8738823078806137,
+      "learning_rate": 9.78302195508504e-06,
+      "loss": 0.5996,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1375631587912745,
+      "learning_rate": 9.782822172611692e-06,
+      "loss": 0.6461,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.730183814158284,
+      "learning_rate": 9.782622300480122e-06,
+      "loss": 0.4592,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0273016841733162,
+      "learning_rate": 9.782422338694515e-06,
+      "loss": 0.6473,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.577434108540485,
+      "learning_rate": 9.782222287259055e-06,
+      "loss": 0.6808,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2611264883252047,
+      "learning_rate": 9.782022146177933e-06,
+      "loss": 0.591,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.194061422405238,
+      "learning_rate": 9.781821915455336e-06,
+      "loss": 0.6661,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7201739670426544,
+      "learning_rate": 9.781621595095457e-06,
+      "loss": 0.3963,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9209673667617007,
+      "learning_rate": 9.78142118510249e-06,
+      "loss": 0.5285,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.206256339548605,
+      "learning_rate": 9.781220685480629e-06,
+      "loss": 0.6427,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0391833242418445,
+      "learning_rate": 9.781020096234073e-06,
+      "loss": 0.5275,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8621955862420714,
+      "learning_rate": 9.780819417367018e-06,
+      "loss": 0.5122,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1116607469313884,
+      "learning_rate": 9.78061864888367e-06,
+      "loss": 0.6357,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6011171510267797,
+      "learning_rate": 9.780417790788228e-06,
+      "loss": 0.4212,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0890508266659924,
+      "learning_rate": 9.780216843084898e-06,
+      "loss": 0.6407,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.456374837064346,
+      "learning_rate": 9.780015805777886e-06,
+      "loss": 0.6734,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0080352155589427,
+      "learning_rate": 9.779814678871404e-06,
+      "loss": 0.441,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2202431513934835,
+      "learning_rate": 9.779613462369659e-06,
+      "loss": 0.5741,
+      "step": 2172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.303391412561931,
+      "learning_rate": 9.779412156276863e-06,
+      "loss": 0.5216,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.098228234210067,
+      "learning_rate": 9.779210760597235e-06,
+      "loss": 0.6429,
+      "step": 2174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0441286390337536,
+      "learning_rate": 9.779009275334985e-06,
+      "loss": 0.6051,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.192553607967509,
+      "learning_rate": 9.778807700494334e-06,
+      "loss": 0.455,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7917603131527506,
+      "learning_rate": 9.778606036079502e-06,
+      "loss": 0.5147,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2555705762850398,
+      "learning_rate": 9.77840428209471e-06,
+      "loss": 0.6316,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1521830622677,
+      "learning_rate": 9.778202438544181e-06,
+      "loss": 0.5637,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4141193770373097,
+      "learning_rate": 9.77800050543214e-06,
+      "loss": 0.6621,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.743854841937344,
+      "learning_rate": 9.777798482762818e-06,
+      "loss": 0.5258,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8069292595136326,
+      "learning_rate": 9.777596370540441e-06,
+      "loss": 0.5282,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.20337515727014,
+      "learning_rate": 9.77739416876924e-06,
+      "loss": 0.6626,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.823039401138172,
+      "learning_rate": 9.777191877453452e-06,
+      "loss": 0.633,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.118969946821114,
+      "learning_rate": 9.776989496597305e-06,
+      "loss": 0.5258,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5115103625812965,
+      "learning_rate": 9.776787026205042e-06,
+      "loss": 0.6845,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8218826408414919,
+      "learning_rate": 9.776584466280899e-06,
+      "loss": 0.5276,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7839931838837866,
+      "learning_rate": 9.776381816829114e-06,
+      "loss": 0.5726,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.39568223023029,
+      "learning_rate": 9.776179077853933e-06,
+      "loss": 0.537,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3754779726710678,
+      "learning_rate": 9.7759762493596e-06,
+      "loss": 0.744,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3102979547055607,
+      "learning_rate": 9.77577333135036e-06,
+      "loss": 0.5624,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7380642480138926,
+      "learning_rate": 9.77557032383046e-06,
+      "loss": 0.3686,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7099509748858455,
+      "learning_rate": 9.775367226804152e-06,
+      "loss": 0.7507,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.176903791514755,
+      "learning_rate": 9.775164040275687e-06,
+      "loss": 0.6683,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.988753978029167,
+      "learning_rate": 9.774960764249317e-06,
+      "loss": 0.551,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4809920142866884,
+      "learning_rate": 9.774757398729297e-06,
+      "loss": 0.7361,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.983297860556445,
+      "learning_rate": 9.774553943719889e-06,
+      "loss": 0.7336,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7211273489525243,
+      "learning_rate": 9.774350399225348e-06,
+      "loss": 0.6886,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3574499260206205,
+      "learning_rate": 9.774146765249937e-06,
+      "loss": 0.5804,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.953461457272104,
+      "learning_rate": 9.773943041797919e-06,
+      "loss": 0.5497,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.6035058498382568,
+      "eval_runtime": 67.9873,
+      "eval_samples_per_second": 50.156,
+      "eval_steps_per_second": 1.574,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3673672673424746,
+      "learning_rate": 9.773739228873555e-06,
+      "loss": 0.6584,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.137683556822467,
+      "learning_rate": 9.773535326481117e-06,
+      "loss": 0.8815,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9898501214910609,
+      "learning_rate": 9.773331334624872e-06,
+      "loss": 0.6464,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3660842593905502,
+      "learning_rate": 9.773127253309088e-06,
+      "loss": 0.58,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6925904291332456,
+      "learning_rate": 9.77292308253804e-06,
+      "loss": 0.4682,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5942941877967707,
+      "learning_rate": 9.772718822316e-06,
+      "loss": 0.824,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9692339045293812,
+      "learning_rate": 9.772514472647246e-06,
+      "loss": 0.6267,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8377201991958123,
+      "learning_rate": 9.772310033536057e-06,
+      "loss": 0.5508,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8505854312670798,
+      "learning_rate": 9.77210550498671e-06,
+      "loss": 0.5419,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.102045017697259,
+      "learning_rate": 9.771900887003487e-06,
+      "loss": 0.6361,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.135482751016329,
+      "learning_rate": 9.771696179590672e-06,
+      "loss": 0.656,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.943253770050598,
+      "learning_rate": 9.771491382752552e-06,
+      "loss": 0.619,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.147746095975063,
+      "learning_rate": 9.771286496493413e-06,
+      "loss": 0.6027,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2950403332426816,
+      "learning_rate": 9.771081520817544e-06,
+      "loss": 0.6373,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.167136006481388,
+      "learning_rate": 9.770876455729236e-06,
+      "loss": 0.6982,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2080202282728116,
+      "learning_rate": 9.770671301232782e-06,
+      "loss": 0.5249,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9157308570117686,
+      "learning_rate": 9.770466057332476e-06,
+      "loss": 0.5759,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1013790124855563,
+      "learning_rate": 9.770260724032617e-06,
+      "loss": 0.5381,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7290507272462463,
+      "learning_rate": 9.770055301337502e-06,
+      "loss": 0.7112,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1103709976221143,
+      "learning_rate": 9.76984978925143e-06,
+      "loss": 0.6131,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0261836308096277,
+      "learning_rate": 9.769644187778708e-06,
+      "loss": 0.6752,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8783364376000549,
+      "learning_rate": 9.769438496923633e-06,
+      "loss": 0.54,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9079360276902566,
+      "learning_rate": 9.769232716690518e-06,
+      "loss": 0.4917,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5805831813252902,
+      "learning_rate": 9.769026847083666e-06,
+      "loss": 0.4559,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4161693866491065,
+      "learning_rate": 9.76882088810739e-06,
+      "loss": 0.6791,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8341247378282433,
+      "learning_rate": 9.768614839765999e-06,
+      "loss": 0.5242,
+      "step": 2226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1924937979011343,
+      "learning_rate": 9.768408702063807e-06,
+      "loss": 0.5965,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0282937848122065,
+      "learning_rate": 9.768202475005131e-06,
+      "loss": 0.696,
+      "step": 2228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8580050236295567,
+      "learning_rate": 9.767996158594288e-06,
+      "loss": 0.5458,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.107473202523936,
+      "learning_rate": 9.767789752835595e-06,
+      "loss": 0.5499,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2043289483018484,
+      "learning_rate": 9.767583257733376e-06,
+      "loss": 0.7731,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6733562739015004,
+      "learning_rate": 9.767376673291951e-06,
+      "loss": 0.5963,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.816184395626058,
+      "learning_rate": 9.767169999515646e-06,
+      "loss": 0.4903,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3143959249323687,
+      "learning_rate": 9.766963236408788e-06,
+      "loss": 0.747,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6416850559521488,
+      "learning_rate": 9.766756383975704e-06,
+      "loss": 0.4853,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.716488419215361,
+      "learning_rate": 9.766549442220726e-06,
+      "loss": 0.5048,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7332944214766721,
+      "learning_rate": 9.766342411148187e-06,
+      "loss": 0.6256,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6618653997150545,
+      "learning_rate": 9.766135290762415e-06,
+      "loss": 0.4205,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1717885565970767,
+      "learning_rate": 9.765928081067756e-06,
+      "loss": 0.5244,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8309366011256003,
+      "learning_rate": 9.765720782068539e-06,
+      "loss": 0.5007,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6231554744225365,
+      "learning_rate": 9.765513393769107e-06,
+      "loss": 0.739,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3235694685216703,
+      "learning_rate": 9.765305916173802e-06,
+      "loss": 0.6824,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1050488961624234,
+      "learning_rate": 9.765098349286967e-06,
+      "loss": 0.5986,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.342142419774009,
+      "learning_rate": 9.764890693112947e-06,
+      "loss": 0.6181,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.364604882562847,
+      "learning_rate": 9.76468294765609e-06,
+      "loss": 0.6922,
+      "step": 2245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9895559832518621,
+      "learning_rate": 9.764475112920743e-06,
+      "loss": 0.6126,
+      "step": 2246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.325574591092873,
+      "learning_rate": 9.76426718891126e-06,
+      "loss": 0.5111,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8585909668938754,
+      "learning_rate": 9.764059175631992e-06,
+      "loss": 0.4996,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1679353209022802,
+      "learning_rate": 9.763851073087292e-06,
+      "loss": 0.6836,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8871236331737749,
+      "learning_rate": 9.763642881281519e-06,
+      "loss": 0.6539,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.6020047068595886,
+      "eval_runtime": 68.0348,
+      "eval_samples_per_second": 50.121,
+      "eval_steps_per_second": 1.573,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9459551806761641,
+      "learning_rate": 9.763434600219032e-06,
+      "loss": 0.4183,
+      "step": 2251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9256242433014046,
+      "learning_rate": 9.763226229904187e-06,
+      "loss": 0.6032,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2572422684756597,
+      "learning_rate": 9.76301777034135e-06,
+      "loss": 0.586,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.24326831533724,
+      "learning_rate": 9.762809221534886e-06,
+      "loss": 0.7232,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.89375841480946,
+      "learning_rate": 9.762600583489156e-06,
+      "loss": 0.5359,
+      "step": 2255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.782173662241324,
+      "learning_rate": 9.76239185620853e-06,
+      "loss": 0.4742,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6024223216295297,
+      "learning_rate": 9.76218303969738e-06,
+      "loss": 0.4949,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.813055786674652,
+      "learning_rate": 9.761974133960075e-06,
+      "loss": 0.4821,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.298498804706096,
+      "learning_rate": 9.761765139000988e-06,
+      "loss": 0.6449,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1277001918692875,
+      "learning_rate": 9.761556054824495e-06,
+      "loss": 0.5465,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.299895373950384,
+      "learning_rate": 9.761346881434972e-06,
+      "loss": 0.7638,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9262380676337667,
+      "learning_rate": 9.7611376188368e-06,
+      "loss": 0.5099,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7751750954182808,
+      "learning_rate": 9.760928267034358e-06,
+      "loss": 0.5637,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8483580321046578,
+      "learning_rate": 9.760718826032031e-06,
+      "loss": 0.3857,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0633167721093515,
+      "learning_rate": 9.7605092958342e-06,
+      "loss": 0.6543,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7495529583748914,
+      "learning_rate": 9.760299676445254e-06,
+      "loss": 0.6576,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0525697770642615,
+      "learning_rate": 9.760089967869581e-06,
+      "loss": 0.6387,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9701688764963627,
+      "learning_rate": 9.75988017011157e-06,
+      "loss": 0.4899,
+      "step": 2268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.167808518199565,
+      "learning_rate": 9.759670283175613e-06,
+      "loss": 0.6541,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0091293455455754,
+      "learning_rate": 9.759460307066104e-06,
+      "loss": 0.5819,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1093415424732074,
+      "learning_rate": 9.75925024178744e-06,
+      "loss": 0.6778,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7944096979799957,
+      "learning_rate": 9.75904008734402e-06,
+      "loss": 0.5907,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.68141611300549,
+      "learning_rate": 9.758829843740237e-06,
+      "loss": 0.5086,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9543445118802283,
+      "learning_rate": 9.758619510980498e-06,
+      "loss": 0.6173,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.940108791344798,
+      "learning_rate": 9.758409089069207e-06,
+      "loss": 0.5339,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.245186703617659,
+      "learning_rate": 9.758198578010764e-06,
+      "loss": 0.6365,
+      "step": 2276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.038961805438824,
+      "learning_rate": 9.75798797780958e-06,
+      "loss": 0.5902,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.776360338886455,
+      "learning_rate": 9.75777728847006e-06,
+      "loss": 0.5771,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0118526732702184,
+      "learning_rate": 9.757566509996618e-06,
+      "loss": 0.6698,
+      "step": 2279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7792990219793081,
+      "learning_rate": 9.757355642393666e-06,
+      "loss": 0.4343,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.195952402232668,
+      "learning_rate": 9.757144685665621e-06,
+      "loss": 0.5576,
+      "step": 2281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9075120563926982,
+      "learning_rate": 9.756933639816892e-06,
+      "loss": 0.5001,
+      "step": 2282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.157942845486913,
+      "learning_rate": 9.756722504851904e-06,
+      "loss": 0.6838,
+      "step": 2283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5716590396575043,
+      "learning_rate": 9.756511280775073e-06,
+      "loss": 0.5971,
+      "step": 2284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1439245825243196,
+      "learning_rate": 9.756299967590822e-06,
+      "loss": 0.5466,
+      "step": 2285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8653417431305743,
+      "learning_rate": 9.756088565303574e-06,
+      "loss": 0.6372,
+      "step": 2286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0371610308251062,
+      "learning_rate": 9.755877073917758e-06,
+      "loss": 0.525,
+      "step": 2287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.889352812911372,
+      "learning_rate": 9.755665493437796e-06,
+      "loss": 0.432,
+      "step": 2288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9421656690186329,
+      "learning_rate": 9.755453823868122e-06,
+      "loss": 0.6004,
+      "step": 2289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.589894156366646,
+      "learning_rate": 9.755242065213166e-06,
+      "loss": 0.6686,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9125024765760796,
+      "learning_rate": 9.755030217477359e-06,
+      "loss": 0.5275,
+      "step": 2291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9528033673033667,
+      "learning_rate": 9.754818280665139e-06,
+      "loss": 0.6544,
+      "step": 2292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9109310429344113,
+      "learning_rate": 9.754606254780939e-06,
+      "loss": 0.4719,
+      "step": 2293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3549789382558357,
+      "learning_rate": 9.7543941398292e-06,
+      "loss": 0.6986,
+      "step": 2294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9407120065714036,
+      "learning_rate": 9.754181935814364e-06,
+      "loss": 0.6045,
+      "step": 2295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4091417667193507,
+      "learning_rate": 9.753969642740869e-06,
+      "loss": 0.6641,
+      "step": 2296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9446897381462342,
+      "learning_rate": 9.753757260613163e-06,
+      "loss": 0.6762,
+      "step": 2297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6027848357548165,
+      "learning_rate": 9.75354478943569e-06,
+      "loss": 0.5209,
+      "step": 2298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9783704118403953,
+      "learning_rate": 9.7533322292129e-06,
+      "loss": 0.6324,
+      "step": 2299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2559075485828672,
+      "learning_rate": 9.753119579949243e-06,
+      "loss": 0.7586,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.6023300886154175,
+      "eval_runtime": 68.047,
+      "eval_samples_per_second": 50.112,
+      "eval_steps_per_second": 1.572,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3896229680595424,
+      "learning_rate": 9.752906841649166e-06,
+      "loss": 0.6631,
+      "step": 2301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.107902633874953,
+      "learning_rate": 9.752694014317128e-06,
+      "loss": 0.6291,
+      "step": 2302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.028985315181291,
+      "learning_rate": 9.752481097957581e-06,
+      "loss": 0.6103,
+      "step": 2303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7650745605445697,
+      "learning_rate": 9.752268092574984e-06,
+      "loss": 0.6712,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.095864062096409,
+      "learning_rate": 9.752054998173797e-06,
+      "loss": 0.4745,
+      "step": 2305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1191151773784678,
+      "learning_rate": 9.751841814758479e-06,
+      "loss": 0.7118,
+      "step": 2306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1168990235312433,
+      "learning_rate": 9.751628542333493e-06,
+      "loss": 0.456,
+      "step": 2307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3240299509370494,
+      "learning_rate": 9.751415180903304e-06,
+      "loss": 0.6544,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.301355605679963,
+      "learning_rate": 9.75120173047238e-06,
+      "loss": 0.5658,
+      "step": 2309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1219885338327553,
+      "learning_rate": 9.750988191045187e-06,
+      "loss": 0.5439,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1382488006620637,
+      "learning_rate": 9.750774562626196e-06,
+      "loss": 0.5691,
+      "step": 2311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5846172058975774,
+      "learning_rate": 9.750560845219882e-06,
+      "loss": 0.6893,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0783677849967153,
+      "learning_rate": 9.750347038830713e-06,
+      "loss": 0.5141,
+      "step": 2313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9886004171611134,
+      "learning_rate": 9.750133143463173e-06,
+      "loss": 0.6764,
+      "step": 2314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.381616869091684,
+      "learning_rate": 9.749919159121734e-06,
+      "loss": 0.6215,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9127664643119426,
+      "learning_rate": 9.749705085810876e-06,
+      "loss": 0.4806,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7556200025486584,
+      "learning_rate": 9.749490923535082e-06,
+      "loss": 0.5978,
+      "step": 2317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2808509026559904,
+      "learning_rate": 9.749276672298837e-06,
+      "loss": 0.6717,
+      "step": 2318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5511244160843347,
+      "learning_rate": 9.749062332106623e-06,
+      "loss": 0.4526,
+      "step": 2319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0025900427865824,
+      "learning_rate": 9.748847902962928e-06,
+      "loss": 0.5028,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3889040172848324,
+      "learning_rate": 9.74863338487224e-06,
+      "loss": 0.6957,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1749360037916365,
+      "learning_rate": 9.748418777839053e-06,
+      "loss": 0.5509,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8125054720754297,
+      "learning_rate": 9.748204081867857e-06,
+      "loss": 0.4907,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1912464315681444,
+      "learning_rate": 9.747989296963147e-06,
+      "loss": 0.5567,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0024892705034105,
+      "learning_rate": 9.747774423129421e-06,
+      "loss": 0.6204,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2242459353887702,
+      "learning_rate": 9.747559460371174e-06,
+      "loss": 0.7268,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.727648757416851,
+      "learning_rate": 9.747344408692909e-06,
+      "loss": 0.5067,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.458044386903328,
+      "learning_rate": 9.747129268099128e-06,
+      "loss": 0.6134,
+      "step": 2328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9049840323874683,
+      "learning_rate": 9.746914038594331e-06,
+      "loss": 0.6244,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8302631466873975,
+      "learning_rate": 9.746698720183028e-06,
+      "loss": 0.5194,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.141858721254371,
+      "learning_rate": 9.746483312869724e-06,
+      "loss": 0.6203,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.613724963363752,
+      "learning_rate": 9.74626781665893e-06,
+      "loss": 0.4923,
+      "step": 2332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.382671396597036,
+      "learning_rate": 9.746052231555157e-06,
+      "loss": 0.8577,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9122513107518622,
+      "learning_rate": 9.745836557562918e-06,
+      "loss": 0.6115,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.114399930003481,
+      "learning_rate": 9.745620794686728e-06,
+      "loss": 0.5216,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7085833533993964,
+      "learning_rate": 9.745404942931103e-06,
+      "loss": 0.6887,
+      "step": 2336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9858853446071763,
+      "learning_rate": 9.745189002300563e-06,
+      "loss": 0.7167,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6635220050730086,
+      "learning_rate": 9.744972972799628e-06,
+      "loss": 0.7277,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.440228786731097,
+      "learning_rate": 9.74475685443282e-06,
+      "loss": 0.6904,
+      "step": 2339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6938089438900576,
+      "learning_rate": 9.744540647204664e-06,
+      "loss": 0.3369,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.972091466848916,
+      "learning_rate": 9.744324351119685e-06,
+      "loss": 0.5993,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.056996165835404,
+      "learning_rate": 9.744107966182412e-06,
+      "loss": 0.6052,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.298133033218171,
+      "learning_rate": 9.743891492397376e-06,
+      "loss": 0.6166,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3308428790219886,
+      "learning_rate": 9.743674929769108e-06,
+      "loss": 0.7331,
+      "step": 2344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0499188280664407,
+      "learning_rate": 9.74345827830214e-06,
+      "loss": 0.6076,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4962220944872477,
+      "learning_rate": 9.74324153800101e-06,
+      "loss": 0.3953,
+      "step": 2346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0691155409845106,
+      "learning_rate": 9.743024708870253e-06,
+      "loss": 0.5143,
+      "step": 2347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.43052817872077,
+      "learning_rate": 9.74280779091441e-06,
+      "loss": 0.7245,
+      "step": 2348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2211680887864333,
+      "learning_rate": 9.74259078413802e-06,
+      "loss": 0.7863,
+      "step": 2349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9844014317997276,
+      "learning_rate": 9.742373688545628e-06,
+      "loss": 0.5622,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.5980060696601868,
+      "eval_runtime": 68.0289,
+      "eval_samples_per_second": 50.126,
+      "eval_steps_per_second": 1.573,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.949911936199172,
+      "learning_rate": 9.742156504141778e-06,
+      "loss": 0.8034,
+      "step": 2351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7899486108080238,
+      "learning_rate": 9.741939230931015e-06,
+      "loss": 0.646,
+      "step": 2352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0030234132273304,
+      "learning_rate": 9.74172186891789e-06,
+      "loss": 0.7141,
+      "step": 2353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.74402788006936,
+      "learning_rate": 9.741504418106951e-06,
+      "loss": 0.4576,
+      "step": 2354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7438433775493927,
+      "learning_rate": 9.741286878502753e-06,
+      "loss": 0.5364,
+      "step": 2355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.891003681664117,
+      "learning_rate": 9.74106925010985e-06,
+      "loss": 0.7006,
+      "step": 2356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6532889051607509,
+      "learning_rate": 9.740851532932794e-06,
+      "loss": 0.6662,
+      "step": 2357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8234399141711666,
+      "learning_rate": 9.740633726976147e-06,
+      "loss": 0.5004,
+      "step": 2358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9100433204647422,
+      "learning_rate": 9.740415832244464e-06,
+      "loss": 0.4953,
+      "step": 2359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9828648921051597,
+      "learning_rate": 9.740197848742311e-06,
+      "loss": 0.4925,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0657858906540474,
+      "learning_rate": 9.73997977647425e-06,
+      "loss": 0.5924,
+      "step": 2361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.274960619214605,
+      "learning_rate": 9.739761615444845e-06,
+      "loss": 0.692,
+      "step": 2362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4919388669877678,
+      "learning_rate": 9.739543365658667e-06,
+      "loss": 0.4773,
+      "step": 2363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9979993531562903,
+      "learning_rate": 9.73932502712028e-06,
+      "loss": 0.4794,
+      "step": 2364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9365973463579074,
+      "learning_rate": 9.739106599834255e-06,
+      "loss": 0.7208,
+      "step": 2365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.156246486489306,
+      "learning_rate": 9.738888083805168e-06,
+      "loss": 0.5262,
+      "step": 2366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3217755716093276,
+      "learning_rate": 9.738669479037592e-06,
+      "loss": 0.6056,
+      "step": 2367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8686781380030184,
+      "learning_rate": 9.738450785536104e-06,
+      "loss": 0.542,
+      "step": 2368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0103227280239246,
+      "learning_rate": 9.73823200330528e-06,
+      "loss": 0.3823,
+      "step": 2369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5396051923085055,
+      "learning_rate": 9.7380131323497e-06,
+      "loss": 0.6606,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0302417453721806,
+      "learning_rate": 9.73779417267395e-06,
+      "loss": 0.537,
+      "step": 2371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6126972959014598,
+      "learning_rate": 9.73757512428261e-06,
+      "loss": 0.4059,
+      "step": 2372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.187505241662759,
+      "learning_rate": 9.737355987180269e-06,
+      "loss": 0.5142,
+      "step": 2373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.123487953812243,
+      "learning_rate": 9.73713676137151e-06,
+      "loss": 0.5472,
+      "step": 2374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8017103245688788,
+      "learning_rate": 9.736917446860925e-06,
+      "loss": 0.4823,
+      "step": 2375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1854808413162985,
+      "learning_rate": 9.736698043653104e-06,
+      "loss": 0.5761,
+      "step": 2376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.920713705211076,
+      "learning_rate": 9.736478551752643e-06,
+      "loss": 0.5313,
+      "step": 2377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.985023823071054,
+      "learning_rate": 9.736258971164134e-06,
+      "loss": 0.5147,
+      "step": 2378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7484744096621507,
+      "learning_rate": 9.736039301892174e-06,
+      "loss": 0.4148,
+      "step": 2379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7194403374546434,
+      "learning_rate": 9.735819543941363e-06,
+      "loss": 0.5877,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5699611887605394,
+      "learning_rate": 9.735599697316302e-06,
+      "loss": 0.6297,
+      "step": 2381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6192839980815674,
+      "learning_rate": 9.73537976202159e-06,
+      "loss": 0.5457,
+      "step": 2382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.154469199375218,
+      "learning_rate": 9.735159738061834e-06,
+      "loss": 0.6426,
+      "step": 2383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8455121226050961,
+      "learning_rate": 9.734939625441641e-06,
+      "loss": 0.4908,
+      "step": 2384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.370355492442237,
+      "learning_rate": 9.734719424165615e-06,
+      "loss": 0.7144,
+      "step": 2385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1157191853488166,
+      "learning_rate": 9.73449913423837e-06,
+      "loss": 0.6842,
+      "step": 2386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.240682214876457,
+      "learning_rate": 9.734278755664514e-06,
+      "loss": 0.6507,
+      "step": 2387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5425668535979724,
+      "learning_rate": 9.734058288448663e-06,
+      "loss": 0.4597,
+      "step": 2388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0555084058982573,
+      "learning_rate": 9.733837732595433e-06,
+      "loss": 0.6256,
+      "step": 2389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7433134397143806,
+      "learning_rate": 9.733617088109437e-06,
+      "loss": 0.6003,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9257646973644729,
+      "learning_rate": 9.7333963549953e-06,
+      "loss": 0.4505,
+      "step": 2391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9339727266744275,
+      "learning_rate": 9.733175533257637e-06,
+      "loss": 0.5294,
+      "step": 2392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0015835454314916,
+      "learning_rate": 9.732954622901076e-06,
+      "loss": 0.6982,
+      "step": 2393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.323690844367972,
+      "learning_rate": 9.732733623930237e-06,
+      "loss": 0.5937,
+      "step": 2394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7486389024050633,
+      "learning_rate": 9.73251253634975e-06,
+      "loss": 0.4659,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0588950816402627,
+      "learning_rate": 9.732291360164242e-06,
+      "loss": 0.5705,
+      "step": 2396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.129943848449124,
+      "learning_rate": 9.732070095378342e-06,
+      "loss": 0.5909,
+      "step": 2397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1449993611690235,
+      "learning_rate": 9.731848741996684e-06,
+      "loss": 0.6208,
+      "step": 2398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8794033618898687,
+      "learning_rate": 9.731627300023902e-06,
+      "loss": 0.5158,
+      "step": 2399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.056701470954038,
+      "learning_rate": 9.73140576946463e-06,
+      "loss": 0.4573,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.5988402962684631,
+      "eval_runtime": 68.007,
+      "eval_samples_per_second": 50.142,
+      "eval_steps_per_second": 1.573,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7460517775791915,
+      "learning_rate": 9.731184150323505e-06,
+      "loss": 0.378,
+      "step": 2401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6428561770886034,
+      "learning_rate": 9.730962442605169e-06,
+      "loss": 0.4346,
+      "step": 2402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6843523242749923,
+      "learning_rate": 9.730740646314265e-06,
+      "loss": 0.4801,
+      "step": 2403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0907208659088914,
+      "learning_rate": 9.730518761455428e-06,
+      "loss": 0.5087,
+      "step": 2404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8691759652351518,
+      "learning_rate": 9.730296788033313e-06,
+      "loss": 0.7109,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1083086938231026,
+      "learning_rate": 9.73007472605256e-06,
+      "loss": 0.5686,
+      "step": 2406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.157473518884618,
+      "learning_rate": 9.72985257551782e-06,
+      "loss": 0.6201,
+      "step": 2407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9577124463226343,
+      "learning_rate": 9.729630336433743e-06,
+      "loss": 0.7408,
+      "step": 2408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9501215628947672,
+      "learning_rate": 9.729408008804984e-06,
+      "loss": 0.4379,
+      "step": 2409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.152755808420283,
+      "learning_rate": 9.729185592636194e-06,
+      "loss": 0.6811,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.802869417657196,
+      "learning_rate": 9.72896308793203e-06,
+      "loss": 0.5138,
+      "step": 2411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8525033979435006,
+      "learning_rate": 9.728740494697152e-06,
+      "loss": 0.6089,
+      "step": 2412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.558234281137451,
+      "learning_rate": 9.728517812936219e-06,
+      "loss": 0.6543,
+      "step": 2413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7638434288720406,
+      "learning_rate": 9.72829504265389e-06,
+      "loss": 0.4002,
+      "step": 2414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0672879572331273,
+      "learning_rate": 9.728072183854831e-06,
+      "loss": 0.4093,
+      "step": 2415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1644211813853262,
+      "learning_rate": 9.727849236543706e-06,
+      "loss": 0.5993,
+      "step": 2416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.714292068017349,
+      "learning_rate": 9.727626200725185e-06,
+      "loss": 0.5367,
+      "step": 2417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.843381484429317,
+      "learning_rate": 9.727403076403935e-06,
+      "loss": 0.6839,
+      "step": 2418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.882030990699036,
+      "learning_rate": 9.727179863584627e-06,
+      "loss": 0.6093,
+      "step": 2419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5530998820331905,
+      "learning_rate": 9.726956562271934e-06,
+      "loss": 0.6793,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2276537181526552,
+      "learning_rate": 9.726733172470531e-06,
+      "loss": 0.5414,
+      "step": 2421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3090133853658554,
+      "learning_rate": 9.726509694185095e-06,
+      "loss": 0.6897,
+      "step": 2422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.94755413467111,
+      "learning_rate": 9.726286127420303e-06,
+      "loss": 0.5284,
+      "step": 2423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1810856464954256,
+      "learning_rate": 9.726062472180837e-06,
+      "loss": 0.6221,
+      "step": 2424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6694562214611912,
+      "learning_rate": 9.725838728471377e-06,
+      "loss": 0.543,
+      "step": 2425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.445644731923156,
+      "learning_rate": 9.725614896296608e-06,
+      "loss": 0.6255,
+      "step": 2426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3453518489949423,
+      "learning_rate": 9.725390975661217e-06,
+      "loss": 0.5088,
+      "step": 2427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.420659315344231,
+      "learning_rate": 9.725166966569887e-06,
+      "loss": 0.644,
+      "step": 2428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0404778353845185,
+      "learning_rate": 9.724942869027314e-06,
+      "loss": 0.6009,
+      "step": 2429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.349304976730716,
+      "learning_rate": 9.724718683038186e-06,
+      "loss": 0.6985,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5206623834728883,
+      "learning_rate": 9.724494408607196e-06,
+      "loss": 0.6631,
+      "step": 2431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9643270543614644,
+      "learning_rate": 9.72427004573904e-06,
+      "loss": 0.4224,
+      "step": 2432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0560627782734247,
+      "learning_rate": 9.724045594438412e-06,
+      "loss": 0.5824,
+      "step": 2433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.05650639691887,
+      "learning_rate": 9.723821054710016e-06,
+      "loss": 0.517,
+      "step": 2434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.07994481031927,
+      "learning_rate": 9.723596426558549e-06,
+      "loss": 0.6243,
+      "step": 2435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.825097649514746,
+      "learning_rate": 9.723371709988713e-06,
+      "loss": 0.4959,
+      "step": 2436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2855669905960005,
+      "learning_rate": 9.723146905005215e-06,
+      "loss": 0.5191,
+      "step": 2437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7206256210470063,
+      "learning_rate": 9.722922011612759e-06,
+      "loss": 0.6073,
+      "step": 2438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3728868473562397,
+      "learning_rate": 9.722697029816054e-06,
+      "loss": 0.6283,
+      "step": 2439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6790730014996664,
+      "learning_rate": 9.722471959619809e-06,
+      "loss": 0.5819,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.849059922631394,
+      "learning_rate": 9.722246801028737e-06,
+      "loss": 0.5369,
+      "step": 2441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2239032307655733,
+      "learning_rate": 9.72202155404755e-06,
+      "loss": 0.5992,
+      "step": 2442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8618288495229254,
+      "learning_rate": 9.721796218680964e-06,
+      "loss": 0.4734,
+      "step": 2443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9484187687739798,
+      "learning_rate": 9.721570794933695e-06,
+      "loss": 0.5628,
+      "step": 2444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8969885907870165,
+      "learning_rate": 9.721345282810467e-06,
+      "loss": 0.5582,
+      "step": 2445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8975762265012721,
+      "learning_rate": 9.721119682315994e-06,
+      "loss": 0.5658,
+      "step": 2446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3142762716168797,
+      "learning_rate": 9.720893993455006e-06,
+      "loss": 0.627,
+      "step": 2447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6598104189328753,
+      "learning_rate": 9.720668216232222e-06,
+      "loss": 0.509,
+      "step": 2448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0870347068362745,
+      "learning_rate": 9.720442350652368e-06,
+      "loss": 0.6483,
+      "step": 2449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8314711259222685,
+      "learning_rate": 9.720216396720177e-06,
+      "loss": 0.4697,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.5962224006652832,
+      "eval_runtime": 68.0277,
+      "eval_samples_per_second": 50.127,
+      "eval_steps_per_second": 1.573,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6788755578678562,
+      "learning_rate": 9.719990354440379e-06,
+      "loss": 0.43,
+      "step": 2451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.941256019517972,
+      "learning_rate": 9.719764223817702e-06,
+      "loss": 0.5726,
+      "step": 2452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.261215406704037,
+      "learning_rate": 9.719538004856881e-06,
+      "loss": 0.5694,
+      "step": 2453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6022154756546023,
+      "learning_rate": 9.719311697562654e-06,
+      "loss": 0.4575,
+      "step": 2454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0273771466024573,
+      "learning_rate": 9.719085301939758e-06,
+      "loss": 0.6511,
+      "step": 2455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1390141873887556,
+      "learning_rate": 9.71885881799293e-06,
+      "loss": 0.6118,
+      "step": 2456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.095126662312115,
+      "learning_rate": 9.718632245726916e-06,
+      "loss": 0.4685,
+      "step": 2457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8605353009760097,
+      "learning_rate": 9.718405585146454e-06,
+      "loss": 0.4578,
+      "step": 2458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2780270771374713,
+      "learning_rate": 9.71817883625629e-06,
+      "loss": 0.7207,
+      "step": 2459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9870435469602716,
+      "learning_rate": 9.717951999061174e-06,
+      "loss": 0.5928,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6193219411845523,
+      "learning_rate": 9.717725073565854e-06,
+      "loss": 0.4511,
+      "step": 2461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1741311473499163,
+      "learning_rate": 9.717498059775077e-06,
+      "loss": 0.6747,
+      "step": 2462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.142408278133826,
+      "learning_rate": 9.717270957693598e-06,
+      "loss": 0.7302,
+      "step": 2463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3067256682829678,
+      "learning_rate": 9.717043767326172e-06,
+      "loss": 0.4667,
+      "step": 2464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.336812731860474,
+      "learning_rate": 9.716816488677553e-06,
+      "loss": 0.8161,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9928361491226163,
+      "learning_rate": 9.7165891217525e-06,
+      "loss": 0.7041,
+      "step": 2466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1361969503231433,
+      "learning_rate": 9.716361666555772e-06,
+      "loss": 0.5932,
+      "step": 2467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8958506622867553,
+      "learning_rate": 9.716134123092134e-06,
+      "loss": 0.4997,
+      "step": 2468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8750739102578582,
+      "learning_rate": 9.715906491366345e-06,
+      "loss": 0.425,
+      "step": 2469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.090667705608544,
+      "learning_rate": 9.715678771383172e-06,
+      "loss": 0.5276,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7646806124229992,
+      "learning_rate": 9.715450963147383e-06,
+      "loss": 0.6911,
+      "step": 2471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9554774367074048,
+      "learning_rate": 9.715223066663747e-06,
+      "loss": 0.6277,
+      "step": 2472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0531873933355524,
+      "learning_rate": 9.714995081937033e-06,
+      "loss": 0.6324,
+      "step": 2473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1298246471619415,
+      "learning_rate": 9.714767008972016e-06,
+      "loss": 0.5506,
+      "step": 2474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0907265202675056,
+      "learning_rate": 9.71453884777347e-06,
+      "loss": 0.5241,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.246345516097174,
+      "learning_rate": 9.71431059834617e-06,
+      "loss": 0.6887,
+      "step": 2476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9589685069494256,
+      "learning_rate": 9.714082260694894e-06,
+      "loss": 0.4548,
+      "step": 2477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2251633123453587,
+      "learning_rate": 9.713853834824424e-06,
+      "loss": 0.6159,
+      "step": 2478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7468109202134734,
+      "learning_rate": 9.713625320739542e-06,
+      "loss": 0.3939,
+      "step": 2479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3683857406954174,
+      "learning_rate": 9.713396718445031e-06,
+      "loss": 0.6357,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.541145162182025,
+      "learning_rate": 9.713168027945678e-06,
+      "loss": 0.6969,
+      "step": 2481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.201566282628014,
+      "learning_rate": 9.712939249246267e-06,
+      "loss": 0.6826,
+      "step": 2482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.887071626782733,
+      "learning_rate": 9.71271038235159e-06,
+      "loss": 0.591,
+      "step": 2483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9101841636767516,
+      "learning_rate": 9.71248142726644e-06,
+      "loss": 0.4354,
+      "step": 2484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1659531801335263,
+      "learning_rate": 9.712252383995606e-06,
+      "loss": 0.4956,
+      "step": 2485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7217385345509875,
+      "learning_rate": 9.712023252543885e-06,
+      "loss": 0.4743,
+      "step": 2486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6248917081859149,
+      "learning_rate": 9.711794032916072e-06,
+      "loss": 0.5681,
+      "step": 2487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.997653284073418,
+      "learning_rate": 9.71156472511697e-06,
+      "loss": 0.6112,
+      "step": 2488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2482875025887616,
+      "learning_rate": 9.711335329151374e-06,
+      "loss": 0.6799,
+      "step": 2489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8063418100295532,
+      "learning_rate": 9.71110584502409e-06,
+      "loss": 0.4803,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7047210969786297,
+      "learning_rate": 9.710876272739918e-06,
+      "loss": 0.4984,
+      "step": 2491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.578180160520587,
+      "learning_rate": 9.71064661230367e-06,
+      "loss": 0.5918,
+      "step": 2492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2305265849936395,
+      "learning_rate": 9.710416863720148e-06,
+      "loss": 0.7872,
+      "step": 2493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0552647004201114,
+      "learning_rate": 9.710187026994166e-06,
+      "loss": 0.6321,
+      "step": 2494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9880253131752388,
+      "learning_rate": 9.709957102130533e-06,
+      "loss": 0.7467,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8968864486491475,
+      "learning_rate": 9.709727089134063e-06,
+      "loss": 0.4975,
+      "step": 2496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.119449332879374,
+      "learning_rate": 9.709496988009572e-06,
+      "loss": 0.577,
+      "step": 2497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8979771747463958,
+      "learning_rate": 9.709266798761874e-06,
+      "loss": 0.7404,
+      "step": 2498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8344322959131587,
+      "learning_rate": 9.709036521395791e-06,
+      "loss": 0.4837,
+      "step": 2499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8366044939839956,
+      "learning_rate": 9.708806155916142e-06,
+      "loss": 0.5867,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5953241586685181,
+      "eval_runtime": 68.0013,
+      "eval_samples_per_second": 50.146,
+      "eval_steps_per_second": 1.573,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1972689422807044,
+      "learning_rate": 9.708575702327752e-06,
+      "loss": 0.7765,
+      "step": 2501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.946754768009859,
+      "learning_rate": 9.708345160635444e-06,
+      "loss": 0.5984,
+      "step": 2502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.940165378276273,
+      "learning_rate": 9.708114530844043e-06,
+      "loss": 0.4406,
+      "step": 2503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6281953010168515,
+      "learning_rate": 9.707883812958378e-06,
+      "loss": 0.4249,
+      "step": 2504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.151840270629321,
+      "learning_rate": 9.707653006983278e-06,
+      "loss": 0.4206,
+      "step": 2505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1066890435573646,
+      "learning_rate": 9.707422112923578e-06,
+      "loss": 0.6826,
+      "step": 2506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6741577633792735,
+      "learning_rate": 9.707191130784108e-06,
+      "loss": 0.4619,
+      "step": 2507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6305217267546532,
+      "learning_rate": 9.706960060569704e-06,
+      "loss": 0.5285,
+      "step": 2508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1054353583979117,
+      "learning_rate": 9.706728902285206e-06,
+      "loss": 0.5186,
+      "step": 2509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.646309369749316,
+      "learning_rate": 9.70649765593545e-06,
+      "loss": 0.5358,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9988395446935718,
+      "learning_rate": 9.70626632152528e-06,
+      "loss": 0.5771,
+      "step": 2511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8099674459985158,
+      "learning_rate": 9.706034899059535e-06,
+      "loss": 0.4975,
+      "step": 2512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6727486519657007,
+      "learning_rate": 9.705803388543062e-06,
+      "loss": 0.5143,
+      "step": 2513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7205480848782408,
+      "learning_rate": 9.70557178998071e-06,
+      "loss": 0.4786,
+      "step": 2514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.941598792314158,
+      "learning_rate": 9.705340103377321e-06,
+      "loss": 0.587,
+      "step": 2515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6582136132486904,
+      "learning_rate": 9.705108328737752e-06,
+      "loss": 0.6276,
+      "step": 2516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0552959021188073,
+      "learning_rate": 9.704876466066852e-06,
+      "loss": 0.5672,
+      "step": 2517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1440999256972817,
+      "learning_rate": 9.704644515369474e-06,
+      "loss": 0.6396,
+      "step": 2518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.273657801635733,
+      "learning_rate": 9.704412476650474e-06,
+      "loss": 0.7126,
+      "step": 2519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.516391694945768,
+      "learning_rate": 9.70418034991471e-06,
+      "loss": 0.6176,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.836209182805601,
+      "learning_rate": 9.703948135167044e-06,
+      "loss": 0.5293,
+      "step": 2521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.974474366103548,
+      "learning_rate": 9.703715832412335e-06,
+      "loss": 0.5368,
+      "step": 2522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1946416477267148,
+      "learning_rate": 9.703483441655443e-06,
+      "loss": 0.5813,
+      "step": 2523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.094256906593942,
+      "learning_rate": 9.703250962901238e-06,
+      "loss": 0.5052,
+      "step": 2524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.206549989523539,
+      "learning_rate": 9.703018396154583e-06,
+      "loss": 0.5132,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9331355761028512,
+      "learning_rate": 9.702785741420351e-06,
+      "loss": 0.4886,
+      "step": 2526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8135499974001323,
+      "learning_rate": 9.702552998703408e-06,
+      "loss": 0.4505,
+      "step": 2527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7961295679220044,
+      "learning_rate": 9.70232016800863e-06,
+      "loss": 0.5183,
+      "step": 2528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.784272670247362,
+      "learning_rate": 9.702087249340888e-06,
+      "loss": 0.8269,
+      "step": 2529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.030876369514637,
+      "learning_rate": 9.70185424270506e-06,
+      "loss": 0.5896,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.46955917416195,
+      "learning_rate": 9.701621148106024e-06,
+      "loss": 0.4054,
+      "step": 2531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.85445183029858,
+      "learning_rate": 9.701387965548658e-06,
+      "loss": 0.5815,
+      "step": 2532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1898176250618033,
+      "learning_rate": 9.701154695037844e-06,
+      "loss": 0.5966,
+      "step": 2533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0845424076957144,
+      "learning_rate": 9.700921336578468e-06,
+      "loss": 0.6914,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3248510683476504,
+      "learning_rate": 9.700687890175411e-06,
+      "loss": 0.7652,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9506648526278967,
+      "learning_rate": 9.700454355833565e-06,
+      "loss": 0.5477,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7029263959218575,
+      "learning_rate": 9.700220733557812e-06,
+      "loss": 0.4776,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.430175747928661,
+      "learning_rate": 9.699987023353049e-06,
+      "loss": 0.5106,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7765543003864777,
+      "learning_rate": 9.699753225224166e-06,
+      "loss": 0.4258,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1311312632289914,
+      "learning_rate": 9.69951933917606e-06,
+      "loss": 0.6407,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.115338398847289,
+      "learning_rate": 9.699285365213621e-06,
+      "loss": 0.5982,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6218019164593245,
+      "learning_rate": 9.699051303341754e-06,
+      "loss": 0.6054,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1484264473618926,
+      "learning_rate": 9.698817153565355e-06,
+      "loss": 0.5127,
+      "step": 2543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7807540168874125,
+      "learning_rate": 9.698582915889328e-06,
+      "loss": 0.5113,
+      "step": 2544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8046187705725178,
+      "learning_rate": 9.698348590318574e-06,
+      "loss": 0.5863,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1518609042727648,
+      "learning_rate": 9.698114176858e-06,
+      "loss": 0.7063,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9111413991318968,
+      "learning_rate": 9.697879675512514e-06,
+      "loss": 0.5399,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.381100956439649,
+      "learning_rate": 9.697645086287026e-06,
+      "loss": 0.6258,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7897176840096387,
+      "learning_rate": 9.697410409186443e-06,
+      "loss": 0.5861,
+      "step": 2549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.348624781747092,
+      "learning_rate": 9.697175644215681e-06,
+      "loss": 0.6911,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5978184342384338,
+      "eval_runtime": 67.949,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6844217926123155,
+      "learning_rate": 9.696940791379653e-06,
+      "loss": 0.4435,
+      "step": 2551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8566972286999797,
+      "learning_rate": 9.696705850683277e-06,
+      "loss": 0.639,
+      "step": 2552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.791022448893091,
+      "learning_rate": 9.696470822131472e-06,
+      "loss": 0.5666,
+      "step": 2553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8689820328191096,
+      "learning_rate": 9.696235705729154e-06,
+      "loss": 0.7753,
+      "step": 2554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6829847273424783,
+      "learning_rate": 9.69600050148125e-06,
+      "loss": 0.5102,
+      "step": 2555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2447861380841454,
+      "learning_rate": 9.695765209392681e-06,
+      "loss": 0.7214,
+      "step": 2556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7182683377247132,
+      "learning_rate": 9.695529829468374e-06,
+      "loss": 0.5288,
+      "step": 2557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.517132377895972,
+      "learning_rate": 9.695294361713255e-06,
+      "loss": 0.7327,
+      "step": 2558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1862098185911556,
+      "learning_rate": 9.695058806132255e-06,
+      "loss": 0.5379,
+      "step": 2559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.303953838291029,
+      "learning_rate": 9.694823162730307e-06,
+      "loss": 0.7179,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.507053115155635,
+      "learning_rate": 9.694587431512337e-06,
+      "loss": 0.4787,
+      "step": 2561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7365852070064454,
+      "learning_rate": 9.694351612483287e-06,
+      "loss": 0.5591,
+      "step": 2562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3071676431226527,
+      "learning_rate": 9.69411570564809e-06,
+      "loss": 0.6122,
+      "step": 2563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2395185411035072,
+      "learning_rate": 9.693879711011687e-06,
+      "loss": 0.5813,
+      "step": 2564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.887990060649477,
+      "learning_rate": 9.693643628579016e-06,
+      "loss": 0.4901,
+      "step": 2565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.015546306292932,
+      "learning_rate": 9.693407458355022e-06,
+      "loss": 0.5635,
+      "step": 2566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5990077018876572,
+      "learning_rate": 9.693171200344647e-06,
+      "loss": 0.4485,
+      "step": 2567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9589142688513077,
+      "learning_rate": 9.692934854552836e-06,
+      "loss": 0.5941,
+      "step": 2568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0000902353977064,
+      "learning_rate": 9.69269842098454e-06,
+      "loss": 0.5931,
+      "step": 2569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.353397463746833,
+      "learning_rate": 9.692461899644705e-06,
+      "loss": 0.5454,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8635248868177277,
+      "learning_rate": 9.692225290538285e-06,
+      "loss": 0.6951,
+      "step": 2571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9032773432518275,
+      "learning_rate": 9.691988593670233e-06,
+      "loss": 0.5618,
+      "step": 2572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.703493841095575,
+      "learning_rate": 9.691751809045502e-06,
+      "loss": 0.507,
+      "step": 2573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9612066487918551,
+      "learning_rate": 9.691514936669052e-06,
+      "loss": 0.5854,
+      "step": 2574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.82820714855996,
+      "learning_rate": 9.69127797654584e-06,
+      "loss": 0.5898,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7078221964675797,
+      "learning_rate": 9.691040928680825e-06,
+      "loss": 0.5262,
+      "step": 2576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6666349012700061,
+      "learning_rate": 9.690803793078975e-06,
+      "loss": 0.464,
+      "step": 2577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.801492483954776,
+      "learning_rate": 9.690566569745249e-06,
+      "loss": 0.751,
+      "step": 2578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4438735397344982,
+      "learning_rate": 9.690329258684611e-06,
+      "loss": 0.3942,
+      "step": 2579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9730588933053805,
+      "learning_rate": 9.690091859902038e-06,
+      "loss": 0.6192,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1509184408829354,
+      "learning_rate": 9.689854373402492e-06,
+      "loss": 0.7272,
+      "step": 2581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2490415502629686,
+      "learning_rate": 9.689616799190947e-06,
+      "loss": 0.6105,
+      "step": 2582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.184200678104851,
+      "learning_rate": 9.689379137272377e-06,
+      "loss": 0.6781,
+      "step": 2583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0065092354078296,
+      "learning_rate": 9.689141387651758e-06,
+      "loss": 0.4151,
+      "step": 2584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.572828198905685,
+      "learning_rate": 9.688903550334063e-06,
+      "loss": 0.4443,
+      "step": 2585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2156923513313025,
+      "learning_rate": 9.688665625324277e-06,
+      "loss": 0.6862,
+      "step": 2586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0927043594793955,
+      "learning_rate": 9.688427612627376e-06,
+      "loss": 0.5227,
+      "step": 2587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.053705365249417,
+      "learning_rate": 9.688189512248345e-06,
+      "loss": 0.5327,
+      "step": 2588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.289010433323601,
+      "learning_rate": 9.68795132419217e-06,
+      "loss": 0.608,
+      "step": 2589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.068460855793826,
+      "learning_rate": 9.687713048463833e-06,
+      "loss": 0.5692,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9304116361163643,
+      "learning_rate": 9.687474685068324e-06,
+      "loss": 0.583,
+      "step": 2591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9075098106238877,
+      "learning_rate": 9.687236234010634e-06,
+      "loss": 0.5629,
+      "step": 2592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7740512271114732,
+      "learning_rate": 9.686997695295755e-06,
+      "loss": 0.7027,
+      "step": 2593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3850331168319108,
+      "learning_rate": 9.68675906892868e-06,
+      "loss": 0.4882,
+      "step": 2594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.067818547119486,
+      "learning_rate": 9.686520354914404e-06,
+      "loss": 0.5335,
+      "step": 2595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.917582953535379,
+      "learning_rate": 9.686281553257923e-06,
+      "loss": 0.5692,
+      "step": 2596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9067306170654827,
+      "learning_rate": 9.68604266396424e-06,
+      "loss": 0.5651,
+      "step": 2597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1711825000394573,
+      "learning_rate": 9.685803687038352e-06,
+      "loss": 0.6226,
+      "step": 2598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5786901963582838,
+      "learning_rate": 9.685564622485266e-06,
+      "loss": 0.3304,
+      "step": 2599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.840429334111089,
+      "learning_rate": 9.685325470309982e-06,
+      "loss": 0.5561,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5949224233627319,
+      "eval_runtime": 67.9407,
+      "eval_samples_per_second": 50.191,
+      "eval_steps_per_second": 1.575,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3883912777410234,
+      "learning_rate": 9.685086230517511e-06,
+      "loss": 0.7538,
+      "step": 2601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3389103137379976,
+      "learning_rate": 9.684846903112857e-06,
+      "loss": 0.8156,
+      "step": 2602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6918356408536526,
+      "learning_rate": 9.684607488101036e-06,
+      "loss": 0.4842,
+      "step": 2603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9053616505247721,
+      "learning_rate": 9.684367985487053e-06,
+      "loss": 0.5226,
+      "step": 2604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6483777424122381,
+      "learning_rate": 9.684128395275925e-06,
+      "loss": 0.5484,
+      "step": 2605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.847337653928051,
+      "learning_rate": 9.68388871747267e-06,
+      "loss": 0.5224,
+      "step": 2606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8399890940406043,
+      "learning_rate": 9.683648952082303e-06,
+      "loss": 0.504,
+      "step": 2607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8095345665836358,
+      "learning_rate": 9.683409099109844e-06,
+      "loss": 0.5498,
+      "step": 2608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.185664735013071,
+      "learning_rate": 9.683169158560313e-06,
+      "loss": 0.5652,
+      "step": 2609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.060650051067869,
+      "learning_rate": 9.682929130438735e-06,
+      "loss": 0.5372,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.69560935191651,
+      "learning_rate": 9.682689014750133e-06,
+      "loss": 0.7381,
+      "step": 2611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0133214266955766,
+      "learning_rate": 9.682448811499534e-06,
+      "loss": 0.6362,
+      "step": 2612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.816394603612562,
+      "learning_rate": 9.682208520691966e-06,
+      "loss": 0.6366,
+      "step": 2613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3208656931264247,
+      "learning_rate": 9.681968142332464e-06,
+      "loss": 0.6286,
+      "step": 2614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.644984208119318,
+      "learning_rate": 9.681727676426052e-06,
+      "loss": 0.5557,
+      "step": 2615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6386477119502993,
+      "learning_rate": 9.681487122977773e-06,
+      "loss": 0.4944,
+      "step": 2616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.118745538778957,
+      "learning_rate": 9.681246481992656e-06,
+      "loss": 0.6555,
+      "step": 2617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.272165095566166,
+      "learning_rate": 9.681005753475739e-06,
+      "loss": 0.785,
+      "step": 2618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.144936699093302,
+      "learning_rate": 9.680764937432066e-06,
+      "loss": 0.5783,
+      "step": 2619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5681110754108525,
+      "learning_rate": 9.680524033866675e-06,
+      "loss": 0.4205,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.906892002679356,
+      "learning_rate": 9.680283042784611e-06,
+      "loss": 0.4858,
+      "step": 2621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.795578054766045,
+      "learning_rate": 9.680041964190917e-06,
+      "loss": 0.5934,
+      "step": 2622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0138021604424026,
+      "learning_rate": 9.67980079809064e-06,
+      "loss": 0.6248,
+      "step": 2623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7602491961151356,
+      "learning_rate": 9.67955954448883e-06,
+      "loss": 0.4989,
+      "step": 2624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.900474822392104,
+      "learning_rate": 9.679318203390537e-06,
+      "loss": 0.5392,
+      "step": 2625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5631245504007676,
+      "learning_rate": 9.679076774800815e-06,
+      "loss": 0.451,
+      "step": 2626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9373268216989086,
+      "learning_rate": 9.678835258724715e-06,
+      "loss": 0.6236,
+      "step": 2627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1895382356197506,
+      "learning_rate": 9.678593655167294e-06,
+      "loss": 0.6824,
+      "step": 2628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9863473121014472,
+      "learning_rate": 9.678351964133612e-06,
+      "loss": 0.6529,
+      "step": 2629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.13483599465224,
+      "learning_rate": 9.678110185628726e-06,
+      "loss": 0.7944,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9233010021932673,
+      "learning_rate": 9.677868319657697e-06,
+      "loss": 0.4838,
+      "step": 2631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.264985438174819,
+      "learning_rate": 9.677626366225589e-06,
+      "loss": 0.6764,
+      "step": 2632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9609414984404696,
+      "learning_rate": 9.677384325337472e-06,
+      "loss": 0.753,
+      "step": 2633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8011841488375016,
+      "learning_rate": 9.677142196998404e-06,
+      "loss": 0.6042,
+      "step": 2634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.961542935194143,
+      "learning_rate": 9.676899981213462e-06,
+      "loss": 0.6486,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2077829911458884,
+      "learning_rate": 9.676657677987711e-06,
+      "loss": 0.7012,
+      "step": 2636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.880153649438016,
+      "learning_rate": 9.676415287326229e-06,
+      "loss": 0.5234,
+      "step": 2637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1443205690127956,
+      "learning_rate": 9.676172809234084e-06,
+      "loss": 0.5871,
+      "step": 2638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0500906463296094,
+      "learning_rate": 9.675930243716357e-06,
+      "loss": 0.5544,
+      "step": 2639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0394694340139026,
+      "learning_rate": 9.675687590778124e-06,
+      "loss": 0.5482,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.505519197063765,
+      "learning_rate": 9.675444850424463e-06,
+      "loss": 0.7488,
+      "step": 2641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.99485072158079,
+      "learning_rate": 9.67520202266046e-06,
+      "loss": 0.5797,
+      "step": 2642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.198298032243517,
+      "learning_rate": 9.674959107491194e-06,
+      "loss": 0.6957,
+      "step": 2643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8116633702033345,
+      "learning_rate": 9.674716104921752e-06,
+      "loss": 0.6373,
+      "step": 2644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1014475157886467,
+      "learning_rate": 9.674473014957223e-06,
+      "loss": 0.4571,
+      "step": 2645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8791022596289728,
+      "learning_rate": 9.674229837602693e-06,
+      "loss": 0.4436,
+      "step": 2646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2671456890885984,
+      "learning_rate": 9.673986572863255e-06,
+      "loss": 0.6096,
+      "step": 2647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9791031470193223,
+      "learning_rate": 9.673743220744e-06,
+      "loss": 0.6778,
+      "step": 2648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3700571542881366,
+      "learning_rate": 9.673499781250024e-06,
+      "loss": 0.6651,
+      "step": 2649
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.61403124637433,
+      "learning_rate": 9.673256254386422e-06,
+      "loss": 0.6226,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5943422317504883,
+      "eval_runtime": 67.9239,
+      "eval_samples_per_second": 50.203,
+      "eval_steps_per_second": 1.575,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2205609169621456,
+      "learning_rate": 9.67301264015829e-06,
+      "loss": 0.5913,
+      "step": 2651
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9753094172859504,
+      "learning_rate": 9.672768938570735e-06,
+      "loss": 0.5799,
+      "step": 2652
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.779505350976235,
+      "learning_rate": 9.67252514962885e-06,
+      "loss": 0.5511,
+      "step": 2653
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.038565407447997,
+      "learning_rate": 9.672281273337746e-06,
+      "loss": 0.6685,
+      "step": 2654
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7224831860361962,
+      "learning_rate": 9.672037309702523e-06,
+      "loss": 0.4525,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9978254661884436,
+      "learning_rate": 9.671793258728291e-06,
+      "loss": 0.4542,
+      "step": 2656
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.404507398821972,
+      "learning_rate": 9.671549120420156e-06,
+      "loss": 0.7537,
+      "step": 2657
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6046231068824959,
+      "learning_rate": 9.671304894783233e-06,
+      "loss": 0.4341,
+      "step": 2658
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8464816164010494,
+      "learning_rate": 9.671060581822634e-06,
+      "loss": 0.569,
+      "step": 2659
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6069174237045694,
+      "learning_rate": 9.670816181543472e-06,
+      "loss": 0.4018,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1292300482961073,
+      "learning_rate": 9.670571693950862e-06,
+      "loss": 0.6688,
+      "step": 2661
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9048579389411335,
+      "learning_rate": 9.670327119049925e-06,
+      "loss": 0.5544,
+      "step": 2662
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7852445932780676,
+      "learning_rate": 9.67008245684578e-06,
+      "loss": 0.5311,
+      "step": 2663
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.746205344478805,
+      "learning_rate": 9.66983770734355e-06,
+      "loss": 0.4764,
+      "step": 2664
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0272489584755733,
+      "learning_rate": 9.669592870548358e-06,
+      "loss": 0.5235,
+      "step": 2665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.118842688716742,
+      "learning_rate": 9.669347946465326e-06,
+      "loss": 0.6135,
+      "step": 2666
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4397205695246118,
+      "learning_rate": 9.669102935099587e-06,
+      "loss": 0.3698,
+      "step": 2667
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6860081026769833,
+      "learning_rate": 9.668857836456268e-06,
+      "loss": 0.4806,
+      "step": 2668
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3333286532124595,
+      "learning_rate": 9.668612650540498e-06,
+      "loss": 0.5605,
+      "step": 2669
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.107753817127362,
+      "learning_rate": 9.668367377357412e-06,
+      "loss": 0.7227,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6853345458823783,
+      "learning_rate": 9.668122016912143e-06,
+      "loss": 0.4587,
+      "step": 2671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.107609966249131,
+      "learning_rate": 9.667876569209829e-06,
+      "loss": 0.5379,
+      "step": 2672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8521689402828525,
+      "learning_rate": 9.667631034255608e-06,
+      "loss": 0.4007,
+      "step": 2673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.039169760103592,
+      "learning_rate": 9.667385412054621e-06,
+      "loss": 0.5394,
+      "step": 2674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.320009307151562,
+      "learning_rate": 9.667139702612008e-06,
+      "loss": 0.5322,
+      "step": 2675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8339579525581542,
+      "learning_rate": 9.666893905932913e-06,
+      "loss": 0.5808,
+      "step": 2676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9445699442804545,
+      "learning_rate": 9.666648022022483e-06,
+      "loss": 0.623,
+      "step": 2677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9238327293047452,
+      "learning_rate": 9.666402050885865e-06,
+      "loss": 0.4588,
+      "step": 2678
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9191081101984415,
+      "learning_rate": 9.666155992528208e-06,
+      "loss": 0.5752,
+      "step": 2679
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0940439304506127,
+      "learning_rate": 9.66590984695466e-06,
+      "loss": 0.4958,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0826106430788354,
+      "learning_rate": 9.66566361417038e-06,
+      "loss": 0.6031,
+      "step": 2681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5420571205840272,
+      "learning_rate": 9.66541729418052e-06,
+      "loss": 0.533,
+      "step": 2682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.791556877262903,
+      "learning_rate": 9.665170886990236e-06,
+      "loss": 0.4941,
+      "step": 2683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.12468504301467,
+      "learning_rate": 9.664924392604684e-06,
+      "loss": 0.645,
+      "step": 2684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.201053606220207,
+      "learning_rate": 9.664677811029027e-06,
+      "loss": 0.6241,
+      "step": 2685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8391216176997913,
+      "learning_rate": 9.66443114226843e-06,
+      "loss": 0.6871,
+      "step": 2686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.119467307624011,
+      "learning_rate": 9.66418438632805e-06,
+      "loss": 0.6563,
+      "step": 2687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4207079963762035,
+      "learning_rate": 9.66393754321306e-06,
+      "loss": 0.6081,
+      "step": 2688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8341070759331808,
+      "learning_rate": 9.66369061292862e-06,
+      "loss": 0.4732,
+      "step": 2689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.049241446592274,
+      "learning_rate": 9.663443595479906e-06,
+      "loss": 0.5585,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6987219570028975,
+      "learning_rate": 9.663196490872086e-06,
+      "loss": 0.4832,
+      "step": 2691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7770555311626273,
+      "learning_rate": 9.662949299110332e-06,
+      "loss": 0.577,
+      "step": 2692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4229895355126945,
+      "learning_rate": 9.662702020199822e-06,
+      "loss": 0.6848,
+      "step": 2693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9522579305340892,
+      "learning_rate": 9.66245465414573e-06,
+      "loss": 0.5308,
+      "step": 2694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8340594142564686,
+      "learning_rate": 9.662207200953233e-06,
+      "loss": 0.546,
+      "step": 2695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.26665056077254,
+      "learning_rate": 9.661959660627517e-06,
+      "loss": 0.6628,
+      "step": 2696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0538354068516917,
+      "learning_rate": 9.66171203317376e-06,
+      "loss": 0.5991,
+      "step": 2697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.526734862892468,
+      "learning_rate": 9.661464318597144e-06,
+      "loss": 0.5972,
+      "step": 2698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9384039361786007,
+      "learning_rate": 9.66121651690286e-06,
+      "loss": 0.8485,
+      "step": 2699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.179529764460707,
+      "learning_rate": 9.660968628096091e-06,
+      "loss": 0.559,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5927860140800476,
+      "eval_runtime": 67.9381,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4334685123643722,
+      "learning_rate": 9.660720652182029e-06,
+      "loss": 0.5757,
+      "step": 2701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7942378303967574,
+      "learning_rate": 9.660472589165864e-06,
+      "loss": 0.4873,
+      "step": 2702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0646353080024693,
+      "learning_rate": 9.66022443905279e-06,
+      "loss": 0.6254,
+      "step": 2703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.120366150209964,
+      "learning_rate": 9.659976201848002e-06,
+      "loss": 0.676,
+      "step": 2704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2394998850769827,
+      "learning_rate": 9.659727877556695e-06,
+      "loss": 0.7553,
+      "step": 2705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.824835332985818,
+      "learning_rate": 9.659479466184068e-06,
+      "loss": 0.4889,
+      "step": 2706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6966345174555448,
+      "learning_rate": 9.659230967735325e-06,
+      "loss": 0.4718,
+      "step": 2707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.846898944726869,
+      "learning_rate": 9.658982382215661e-06,
+      "loss": 0.5577,
+      "step": 2708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9557410339026888,
+      "learning_rate": 9.658733709630286e-06,
+      "loss": 0.5047,
+      "step": 2709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1771162598134017,
+      "learning_rate": 9.658484949984405e-06,
+      "loss": 0.6743,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8926113194635013,
+      "learning_rate": 9.658236103283224e-06,
+      "loss": 0.5724,
+      "step": 2711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.862454819762349,
+      "learning_rate": 9.657987169531951e-06,
+      "loss": 0.6491,
+      "step": 2712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9532643449158873,
+      "learning_rate": 9.657738148735802e-06,
+      "loss": 0.4901,
+      "step": 2713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.633455402454164,
+      "learning_rate": 9.657489040899986e-06,
+      "loss": 0.4859,
+      "step": 2714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9589206641664816,
+      "learning_rate": 9.657239846029718e-06,
+      "loss": 0.5357,
+      "step": 2715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9447932431733217,
+      "learning_rate": 9.656990564130218e-06,
+      "loss": 0.4664,
+      "step": 2716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0823665537463225,
+      "learning_rate": 9.6567411952067e-06,
+      "loss": 0.5331,
+      "step": 2717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.021161266241039,
+      "learning_rate": 9.656491739264388e-06,
+      "loss": 0.5436,
+      "step": 2718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8003080259831237,
+      "learning_rate": 9.656242196308506e-06,
+      "loss": 0.4769,
+      "step": 2719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.534110940773627,
+      "learning_rate": 9.655992566344272e-06,
+      "loss": 0.5825,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8730091814312226,
+      "learning_rate": 9.655742849376916e-06,
+      "loss": 0.6702,
+      "step": 2721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5987043572277373,
+      "learning_rate": 9.655493045411664e-06,
+      "loss": 0.5388,
+      "step": 2722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8240055343149084,
+      "learning_rate": 9.655243154453745e-06,
+      "loss": 0.4729,
+      "step": 2723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8822121218690635,
+      "learning_rate": 9.654993176508396e-06,
+      "loss": 0.5146,
+      "step": 2724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4907671720682054,
+      "learning_rate": 9.654743111580841e-06,
+      "loss": 0.604,
+      "step": 2725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.181685561685082,
+      "learning_rate": 9.654492959676324e-06,
+      "loss": 0.5182,
+      "step": 2726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.731668435366614,
+      "learning_rate": 9.654242720800074e-06,
+      "loss": 0.631,
+      "step": 2727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8442837579647813,
+      "learning_rate": 9.653992394957332e-06,
+      "loss": 0.4513,
+      "step": 2728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8933030430804152,
+      "learning_rate": 9.653741982153341e-06,
+      "loss": 0.5747,
+      "step": 2729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7672411157800265,
+      "learning_rate": 9.653491482393343e-06,
+      "loss": 0.4482,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.10946880220935,
+      "learning_rate": 9.65324089568258e-06,
+      "loss": 0.6133,
+      "step": 2731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2313142394332584,
+      "learning_rate": 9.652990222026297e-06,
+      "loss": 0.6219,
+      "step": 2732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.226930074094366,
+      "learning_rate": 9.652739461429743e-06,
+      "loss": 0.6285,
+      "step": 2733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2128768331722863,
+      "learning_rate": 9.652488613898169e-06,
+      "loss": 0.6526,
+      "step": 2734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.424318813728591,
+      "learning_rate": 9.652237679436825e-06,
+      "loss": 0.5801,
+      "step": 2735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9427659088056228,
+      "learning_rate": 9.651986658050964e-06,
+      "loss": 0.6554,
+      "step": 2736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6600567064866183,
+      "learning_rate": 9.651735549745842e-06,
+      "loss": 0.432,
+      "step": 2737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3709380983565604,
+      "learning_rate": 9.651484354526713e-06,
+      "loss": 0.6853,
+      "step": 2738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0153137666312717,
+      "learning_rate": 9.651233072398839e-06,
+      "loss": 0.586,
+      "step": 2739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.20432262192459,
+      "learning_rate": 9.650981703367478e-06,
+      "loss": 0.6728,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.708219570179829,
+      "learning_rate": 9.650730247437895e-06,
+      "loss": 0.4082,
+      "step": 2741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0743683034122102,
+      "learning_rate": 9.650478704615352e-06,
+      "loss": 0.5525,
+      "step": 2742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6972446781739416,
+      "learning_rate": 9.650227074905116e-06,
+      "loss": 0.4166,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.737386394661236,
+      "learning_rate": 9.64997535831245e-06,
+      "loss": 0.6589,
+      "step": 2744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8307473645616585,
+      "learning_rate": 9.649723554842633e-06,
+      "loss": 0.5152,
+      "step": 2745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2154657543564973,
+      "learning_rate": 9.649471664500927e-06,
+      "loss": 0.6085,
+      "step": 2746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0260832827595094,
+      "learning_rate": 9.64921968729261e-06,
+      "loss": 0.6254,
+      "step": 2747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5539430904020795,
+      "learning_rate": 9.648967623222958e-06,
+      "loss": 0.7902,
+      "step": 2748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4629464193477684,
+      "learning_rate": 9.648715472297242e-06,
+      "loss": 0.3864,
+      "step": 2749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0367804139438466,
+      "learning_rate": 9.648463234520748e-06,
+      "loss": 0.5387,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.5945780873298645,
+      "eval_runtime": 67.8364,
+      "eval_samples_per_second": 50.268,
+      "eval_steps_per_second": 1.577,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9826097209260256,
+      "learning_rate": 9.648210909898753e-06,
+      "loss": 0.5086,
+      "step": 2751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8770541132319776,
+      "learning_rate": 9.647958498436536e-06,
+      "loss": 0.5038,
+      "step": 2752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.117086065728995,
+      "learning_rate": 9.647706000139387e-06,
+      "loss": 0.5713,
+      "step": 2753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8681061073278595,
+      "learning_rate": 9.647453415012587e-06,
+      "loss": 0.5177,
+      "step": 2754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.668321380361322,
+      "learning_rate": 9.647200743061426e-06,
+      "loss": 0.6249,
+      "step": 2755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.412827036958756,
+      "learning_rate": 9.646947984291195e-06,
+      "loss": 0.6914,
+      "step": 2756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7431122230615501,
+      "learning_rate": 9.646695138707183e-06,
+      "loss": 0.4453,
+      "step": 2757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.810028505645722,
+      "learning_rate": 9.646442206314684e-06,
+      "loss": 0.6054,
+      "step": 2758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9874187343772207,
+      "learning_rate": 9.646189187118993e-06,
+      "loss": 0.5644,
+      "step": 2759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3473541854904973,
+      "learning_rate": 9.645936081125406e-06,
+      "loss": 0.6113,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1639752744781804,
+      "learning_rate": 9.645682888339223e-06,
+      "loss": 0.5945,
+      "step": 2761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7120514375098221,
+      "learning_rate": 9.645429608765742e-06,
+      "loss": 0.4595,
+      "step": 2762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6365581597657073,
+      "learning_rate": 9.645176242410268e-06,
+      "loss": 0.4672,
+      "step": 2763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8953106037506253,
+      "learning_rate": 9.644922789278106e-06,
+      "loss": 0.6038,
+      "step": 2764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7875350862498185,
+      "learning_rate": 9.644669249374558e-06,
+      "loss": 0.7858,
+      "step": 2765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8863539122505584,
+      "learning_rate": 9.644415622704933e-06,
+      "loss": 0.5095,
+      "step": 2766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9051174815977554,
+      "learning_rate": 9.644161909274543e-06,
+      "loss": 0.5087,
+      "step": 2767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6304778105669697,
+      "learning_rate": 9.643908109088695e-06,
+      "loss": 0.4067,
+      "step": 2768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7217958224004466,
+      "learning_rate": 9.643654222152706e-06,
+      "loss": 0.4546,
+      "step": 2769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.271347280992287,
+      "learning_rate": 9.64340024847189e-06,
+      "loss": 0.6102,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.177425392711392,
+      "learning_rate": 9.643146188051563e-06,
+      "loss": 0.6529,
+      "step": 2771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.788582866606593,
+      "learning_rate": 9.642892040897045e-06,
+      "loss": 0.4739,
+      "step": 2772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.571426611552078,
+      "learning_rate": 9.642637807013654e-06,
+      "loss": 0.6898,
+      "step": 2773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9684199046597588,
+      "learning_rate": 9.642383486406713e-06,
+      "loss": 0.5204,
+      "step": 2774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.197203318185824,
+      "learning_rate": 9.642129079081549e-06,
+      "loss": 0.6218,
+      "step": 2775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7943322496304372,
+      "learning_rate": 9.641874585043482e-06,
+      "loss": 0.5069,
+      "step": 2776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.228085627928671,
+      "learning_rate": 9.641620004297846e-06,
+      "loss": 0.7758,
+      "step": 2777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8789301931457696,
+      "learning_rate": 9.641365336849966e-06,
+      "loss": 0.5671,
+      "step": 2778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.959166504262992,
+      "learning_rate": 9.641110582705176e-06,
+      "loss": 0.6182,
+      "step": 2779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8903679197928813,
+      "learning_rate": 9.640855741868809e-06,
+      "loss": 0.5303,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0650483290336825,
+      "learning_rate": 9.640600814346197e-06,
+      "loss": 0.5408,
+      "step": 2781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8546515979895029,
+      "learning_rate": 9.64034580014268e-06,
+      "loss": 0.4887,
+      "step": 2782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3765447568147673,
+      "learning_rate": 9.640090699263596e-06,
+      "loss": 0.6004,
+      "step": 2783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.033635083665837,
+      "learning_rate": 9.639835511714284e-06,
+      "loss": 0.6447,
+      "step": 2784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9608049290843483,
+      "learning_rate": 9.639580237500085e-06,
+      "loss": 0.5553,
+      "step": 2785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0338510422868405,
+      "learning_rate": 9.639324876626347e-06,
+      "loss": 0.6191,
+      "step": 2786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.364753564442246,
+      "learning_rate": 9.639069429098415e-06,
+      "loss": 0.5287,
+      "step": 2787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8904697185393875,
+      "learning_rate": 9.638813894921631e-06,
+      "loss": 0.5331,
+      "step": 2788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8633816884927963,
+      "learning_rate": 9.638558274101352e-06,
+      "loss": 0.5999,
+      "step": 2789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3761540066580187,
+      "learning_rate": 9.638302566642927e-06,
+      "loss": 0.685,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.00439433408992,
+      "learning_rate": 9.638046772551706e-06,
+      "loss": 0.5601,
+      "step": 2791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7860518956898856,
+      "learning_rate": 9.637790891833047e-06,
+      "loss": 0.5334,
+      "step": 2792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.058273149742503,
+      "learning_rate": 9.637534924492306e-06,
+      "loss": 0.6221,
+      "step": 2793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.147529925796314,
+      "learning_rate": 9.63727887053484e-06,
+      "loss": 0.6676,
+      "step": 2794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8588151239000348,
+      "learning_rate": 9.637022729966012e-06,
+      "loss": 0.6383,
+      "step": 2795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8645009983513328,
+      "learning_rate": 9.636766502791183e-06,
+      "loss": 0.5143,
+      "step": 2796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0287107422087223,
+      "learning_rate": 9.636510189015715e-06,
+      "loss": 0.6078,
+      "step": 2797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9370122439061181,
+      "learning_rate": 9.636253788644977e-06,
+      "loss": 0.4927,
+      "step": 2798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.101602274907041,
+      "learning_rate": 9.635997301684334e-06,
+      "loss": 0.6705,
+      "step": 2799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9368763357541043,
+      "learning_rate": 9.635740728139158e-06,
+      "loss": 0.4191,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.5929890871047974,
+      "eval_runtime": 67.8745,
+      "eval_samples_per_second": 50.24,
+      "eval_steps_per_second": 1.576,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.71636802056637,
+      "learning_rate": 9.635484068014817e-06,
+      "loss": 0.4676,
+      "step": 2801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9377802431706581,
+      "learning_rate": 9.635227321316686e-06,
+      "loss": 0.5618,
+      "step": 2802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4998009824080936,
+      "learning_rate": 9.63497048805014e-06,
+      "loss": 0.6905,
+      "step": 2803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.104233449602884,
+      "learning_rate": 9.634713568220555e-06,
+      "loss": 0.551,
+      "step": 2804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7372858622766127,
+      "learning_rate": 9.634456561833309e-06,
+      "loss": 0.4922,
+      "step": 2805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5519690307985257,
+      "learning_rate": 9.634199468893783e-06,
+      "loss": 0.4459,
+      "step": 2806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1344099362327875,
+      "learning_rate": 9.63394228940736e-06,
+      "loss": 0.7017,
+      "step": 2807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7229107500127785,
+      "learning_rate": 9.633685023379422e-06,
+      "loss": 0.4833,
+      "step": 2808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1587859144184125,
+      "learning_rate": 9.633427670815356e-06,
+      "loss": 0.409,
+      "step": 2809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7706846104439637,
+      "learning_rate": 9.63317023172055e-06,
+      "loss": 0.5082,
+      "step": 2810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.411807633319479,
+      "learning_rate": 9.63291270610039e-06,
+      "loss": 0.6009,
+      "step": 2811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.133468131630175,
+      "learning_rate": 9.63265509396027e-06,
+      "loss": 0.7847,
+      "step": 2812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2044734784588202,
+      "learning_rate": 9.632397395305585e-06,
+      "loss": 0.6661,
+      "step": 2813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0866895235983987,
+      "learning_rate": 9.632139610141726e-06,
+      "loss": 0.573,
+      "step": 2814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7300448993690691,
+      "learning_rate": 9.63188173847409e-06,
+      "loss": 0.444,
+      "step": 2815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9079254808536956,
+      "learning_rate": 9.631623780308077e-06,
+      "loss": 0.4411,
+      "step": 2816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9898788756721622,
+      "learning_rate": 9.631365735649085e-06,
+      "loss": 0.565,
+      "step": 2817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.658375907189517,
+      "learning_rate": 9.631107604502518e-06,
+      "loss": 0.5467,
+      "step": 2818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0765665268619653,
+      "learning_rate": 9.630849386873781e-06,
+      "loss": 0.5861,
+      "step": 2819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.055386127341655,
+      "learning_rate": 9.630591082768276e-06,
+      "loss": 0.7017,
+      "step": 2820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7359334371311936,
+      "learning_rate": 9.630332692191414e-06,
+      "loss": 0.4278,
+      "step": 2821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9747678507568878,
+      "learning_rate": 9.6300742151486e-06,
+      "loss": 0.581,
+      "step": 2822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7675414163643777,
+      "learning_rate": 9.629815651645248e-06,
+      "loss": 0.6746,
+      "step": 2823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5238324844413973,
+      "learning_rate": 9.629557001686772e-06,
+      "loss": 0.7396,
+      "step": 2824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.013132303017162,
+      "learning_rate": 9.629298265278583e-06,
+      "loss": 0.4818,
+      "step": 2825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8176591018951773,
+      "learning_rate": 9.629039442426102e-06,
+      "loss": 0.491,
+      "step": 2826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1058296713665023,
+      "learning_rate": 9.628780533134745e-06,
+      "loss": 0.6887,
+      "step": 2827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2044881071953704,
+      "learning_rate": 9.62852153740993e-06,
+      "loss": 0.624,
+      "step": 2828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.441169905321247,
+      "learning_rate": 9.628262455257083e-06,
+      "loss": 0.6519,
+      "step": 2829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.94777646117247,
+      "learning_rate": 9.628003286681625e-06,
+      "loss": 0.6303,
+      "step": 2830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.16675829361963,
+      "learning_rate": 9.62774403168898e-06,
+      "loss": 0.5716,
+      "step": 2831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.077429167764203,
+      "learning_rate": 9.627484690284581e-06,
+      "loss": 0.504,
+      "step": 2832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9353109478077248,
+      "learning_rate": 9.627225262473852e-06,
+      "loss": 0.511,
+      "step": 2833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1338139690459905,
+      "learning_rate": 9.626965748262225e-06,
+      "loss": 0.667,
+      "step": 2834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7196076129299804,
+      "learning_rate": 9.626706147655134e-06,
+      "loss": 0.8858,
+      "step": 2835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1623723521859013,
+      "learning_rate": 9.626446460658014e-06,
+      "loss": 0.5204,
+      "step": 2836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.102941699957138,
+      "learning_rate": 9.626186687276299e-06,
+      "loss": 0.4793,
+      "step": 2837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7014940562445149,
+      "learning_rate": 9.62592682751543e-06,
+      "loss": 0.4444,
+      "step": 2838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.249763328368709,
+      "learning_rate": 9.625666881380843e-06,
+      "loss": 0.6375,
+      "step": 2839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.085056663931363,
+      "learning_rate": 9.625406848877986e-06,
+      "loss": 0.5391,
+      "step": 2840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4598244145836574,
+      "learning_rate": 9.625146730012298e-06,
+      "loss": 0.6241,
+      "step": 2841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2344910195113563,
+      "learning_rate": 9.624886524789222e-06,
+      "loss": 0.5737,
+      "step": 2842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.721250253934503,
+      "learning_rate": 9.624626233214212e-06,
+      "loss": 0.5555,
+      "step": 2843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0127792845766432,
+      "learning_rate": 9.624365855292713e-06,
+      "loss": 0.5841,
+      "step": 2844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.561008447210243,
+      "learning_rate": 9.624105391030177e-06,
+      "loss": 0.5125,
+      "step": 2845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7106186243101797,
+      "learning_rate": 9.623844840432055e-06,
+      "loss": 0.4616,
+      "step": 2846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2232764386105,
+      "learning_rate": 9.623584203503803e-06,
+      "loss": 0.7612,
+      "step": 2847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5626204563237514,
+      "learning_rate": 9.623323480250877e-06,
+      "loss": 0.3661,
+      "step": 2848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0558187439748625,
+      "learning_rate": 9.623062670678735e-06,
+      "loss": 0.6542,
+      "step": 2849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2691662264980264,
+      "learning_rate": 9.622801774792838e-06,
+      "loss": 0.7,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.592528223991394,
+      "eval_runtime": 67.885,
+      "eval_samples_per_second": 50.232,
+      "eval_steps_per_second": 1.576,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.116631021684769,
+      "learning_rate": 9.622540792598645e-06,
+      "loss": 0.5534,
+      "step": 2851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9128659225278108,
+      "learning_rate": 9.622279724101623e-06,
+      "loss": 0.5541,
+      "step": 2852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7040040576385769,
+      "learning_rate": 9.622018569307234e-06,
+      "loss": 0.5219,
+      "step": 2853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6173540157073174,
+      "learning_rate": 9.621757328220946e-06,
+      "loss": 0.527,
+      "step": 2854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.94934915424507,
+      "learning_rate": 9.621496000848229e-06,
+      "loss": 0.4902,
+      "step": 2855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1422412823942505,
+      "learning_rate": 9.621234587194554e-06,
+      "loss": 0.6938,
+      "step": 2856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5374435064348833,
+      "learning_rate": 9.620973087265391e-06,
+      "loss": 0.5452,
+      "step": 2857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7491174542749743,
+      "learning_rate": 9.620711501066217e-06,
+      "loss": 0.4856,
+      "step": 2858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7853026490590973,
+      "learning_rate": 9.620449828602509e-06,
+      "loss": 0.4825,
+      "step": 2859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9278725260153937,
+      "learning_rate": 9.62018806987974e-06,
+      "loss": 0.8749,
+      "step": 2860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8447125063243908,
+      "learning_rate": 9.619926224903396e-06,
+      "loss": 0.5396,
+      "step": 2861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5322548234865658,
+      "learning_rate": 9.619664293678952e-06,
+      "loss": 0.4717,
+      "step": 2862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.640839454739532,
+      "learning_rate": 9.6194022762119e-06,
+      "loss": 0.508,
+      "step": 2863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8064906966191254,
+      "learning_rate": 9.619140172507716e-06,
+      "loss": 0.4723,
+      "step": 2864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1529510175858104,
+      "learning_rate": 9.61887798257189e-06,
+      "loss": 0.6042,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9446495812186337,
+      "learning_rate": 9.618615706409915e-06,
+      "loss": 0.4906,
+      "step": 2866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5782455850976482,
+      "learning_rate": 9.618353344027277e-06,
+      "loss": 0.4469,
+      "step": 2867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.019319150859116,
+      "learning_rate": 9.618090895429469e-06,
+      "loss": 0.5736,
+      "step": 2868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.984488885699852,
+      "learning_rate": 9.617828360621986e-06,
+      "loss": 0.502,
+      "step": 2869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7828362806053966,
+      "learning_rate": 9.617565739610325e-06,
+      "loss": 0.4824,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7158984202340581,
+      "learning_rate": 9.617303032399983e-06,
+      "loss": 0.4926,
+      "step": 2871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3974707968528612,
+      "learning_rate": 9.617040238996459e-06,
+      "loss": 0.7959,
+      "step": 2872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.431559127592057,
+      "learning_rate": 9.616777359405254e-06,
+      "loss": 0.625,
+      "step": 2873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9892099861749095,
+      "learning_rate": 9.616514393631872e-06,
+      "loss": 0.4244,
+      "step": 2874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.09524616773054,
+      "learning_rate": 9.616251341681818e-06,
+      "loss": 0.587,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.416461357641719,
+      "learning_rate": 9.6159882035606e-06,
+      "loss": 0.6821,
+      "step": 2876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3259505688865696,
+      "learning_rate": 9.615724979273726e-06,
+      "loss": 0.5252,
+      "step": 2877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6485206553215828,
+      "learning_rate": 9.615461668826703e-06,
+      "loss": 0.4444,
+      "step": 2878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9899907619501924,
+      "learning_rate": 9.61519827222505e-06,
+      "loss": 0.528,
+      "step": 2879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9935323429770668,
+      "learning_rate": 9.614934789474273e-06,
+      "loss": 0.5177,
+      "step": 2880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.339177174902828,
+      "learning_rate": 9.614671220579896e-06,
+      "loss": 0.6725,
+      "step": 2881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5532728009088874,
+      "learning_rate": 9.614407565547429e-06,
+      "loss": 0.7093,
+      "step": 2882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9898480481151528,
+      "learning_rate": 9.614143824382398e-06,
+      "loss": 0.528,
+      "step": 2883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7593297465245137,
+      "learning_rate": 9.613879997090321e-06,
+      "loss": 0.4841,
+      "step": 2884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5165481541923067,
+      "learning_rate": 9.61361608367672e-06,
+      "loss": 0.6432,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8067797320710024,
+      "learning_rate": 9.613352084147123e-06,
+      "loss": 0.5373,
+      "step": 2886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1073385723009084,
+      "learning_rate": 9.613087998507055e-06,
+      "loss": 0.5201,
+      "step": 2887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3098189584842213,
+      "learning_rate": 9.612823826762043e-06,
+      "loss": 0.6052,
+      "step": 2888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8193702627805663,
+      "learning_rate": 9.61255956891762e-06,
+      "loss": 0.5666,
+      "step": 2889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6876208479769466,
+      "learning_rate": 9.612295224979316e-06,
+      "loss": 0.4801,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0022269833399418,
+      "learning_rate": 9.612030794952667e-06,
+      "loss": 0.6011,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.249838149858072,
+      "learning_rate": 9.611766278843206e-06,
+      "loss": 0.7496,
+      "step": 2892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.668769190969863,
+      "learning_rate": 9.611501676656472e-06,
+      "loss": 0.5458,
+      "step": 2893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9606002087954704,
+      "learning_rate": 9.611236988398003e-06,
+      "loss": 0.6208,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3726026775564906,
+      "learning_rate": 9.610972214073346e-06,
+      "loss": 0.6524,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6364385483537616,
+      "learning_rate": 9.610707353688035e-06,
+      "loss": 0.3802,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1225904342251947,
+      "learning_rate": 9.610442407247619e-06,
+      "loss": 0.5845,
+      "step": 2897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5181235925499172,
+      "learning_rate": 9.610177374757646e-06,
+      "loss": 0.3639,
+      "step": 2898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.947119127591502,
+      "learning_rate": 9.609912256223662e-06,
+      "loss": 0.6533,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.684151351135566,
+      "learning_rate": 9.609647051651216e-06,
+      "loss": 0.5551,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.590482234954834,
+      "eval_runtime": 67.8209,
+      "eval_samples_per_second": 50.279,
+      "eval_steps_per_second": 1.578,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6475069644786937,
+      "learning_rate": 9.609381761045864e-06,
+      "loss": 0.7202,
+      "step": 2901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4804638780484236,
+      "learning_rate": 9.609116384413155e-06,
+      "loss": 0.694,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5920116448067871,
+      "learning_rate": 9.608850921758648e-06,
+      "loss": 0.522,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5063327054694982,
+      "learning_rate": 9.608585373087898e-06,
+      "loss": 0.4292,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1145107980707314,
+      "learning_rate": 9.608319738406467e-06,
+      "loss": 0.6365,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9711336411568678,
+      "learning_rate": 9.608054017719912e-06,
+      "loss": 0.4609,
+      "step": 2906
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.326563707282078,
+      "learning_rate": 9.6077882110338e-06,
+      "loss": 0.5628,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2193467277590506,
+      "learning_rate": 9.607522318353692e-06,
+      "loss": 0.539,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.080177191941391,
+      "learning_rate": 9.607256339685155e-06,
+      "loss": 0.6065,
+      "step": 2909
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5502765578845425,
+      "learning_rate": 9.606990275033757e-06,
+      "loss": 0.4062,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7198611268193258,
+      "learning_rate": 9.606724124405072e-06,
+      "loss": 0.4515,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.039211812403296,
+      "learning_rate": 9.606457887804664e-06,
+      "loss": 0.6021,
+      "step": 2912
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2503734083466562,
+      "learning_rate": 9.606191565238115e-06,
+      "loss": 0.4291,
+      "step": 2913
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.77958880239699,
+      "learning_rate": 9.605925156710993e-06,
+      "loss": 0.8007,
+      "step": 2914
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8763374982438064,
+      "learning_rate": 9.60565866222888e-06,
+      "loss": 0.5011,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9695576652639422,
+      "learning_rate": 9.60539208179735e-06,
+      "loss": 0.6369,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1124158851329295,
+      "learning_rate": 9.60512541542199e-06,
+      "loss": 0.6765,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.673276556205034,
+      "learning_rate": 9.604858663108376e-06,
+      "loss": 0.5691,
+      "step": 2918
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.351587697829975,
+      "learning_rate": 9.604591824862099e-06,
+      "loss": 0.5823,
+      "step": 2919
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.104372115241655,
+      "learning_rate": 9.60432490068874e-06,
+      "loss": 0.6459,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.911261351869348,
+      "learning_rate": 9.604057890593887e-06,
+      "loss": 0.5854,
+      "step": 2921
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9996271250376518,
+      "learning_rate": 9.603790794583134e-06,
+      "loss": 0.5484,
+      "step": 2922
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.111376057286929,
+      "learning_rate": 9.603523612662068e-06,
+      "loss": 0.5518,
+      "step": 2923
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0305612473820807,
+      "learning_rate": 9.603256344836284e-06,
+      "loss": 0.5446,
+      "step": 2924
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2592049184536513,
+      "learning_rate": 9.602988991111378e-06,
+      "loss": 0.6884,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.981768295063063,
+      "learning_rate": 9.602721551492945e-06,
+      "loss": 0.4722,
+      "step": 2926
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4824199948721108,
+      "learning_rate": 9.602454025986585e-06,
+      "loss": 0.4302,
+      "step": 2927
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.98191343241771,
+      "learning_rate": 9.602186414597899e-06,
+      "loss": 0.5962,
+      "step": 2928
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6321234972178884,
+      "learning_rate": 9.601918717332487e-06,
+      "loss": 0.4783,
+      "step": 2929
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.644187818225074,
+      "learning_rate": 9.601650934195954e-06,
+      "loss": 0.66,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.157108187708823,
+      "learning_rate": 9.601383065193908e-06,
+      "loss": 0.5467,
+      "step": 2931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7497572579643204,
+      "learning_rate": 9.601115110331953e-06,
+      "loss": 0.6582,
+      "step": 2932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1600363664955413,
+      "learning_rate": 9.600847069615703e-06,
+      "loss": 0.6209,
+      "step": 2933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8100018009343697,
+      "learning_rate": 9.600578943050765e-06,
+      "loss": 0.4862,
+      "step": 2934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2397342132250886,
+      "learning_rate": 9.600310730642754e-06,
+      "loss": 0.5854,
+      "step": 2935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.805227101869428,
+      "learning_rate": 9.600042432397285e-06,
+      "loss": 0.4637,
+      "step": 2936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.208274808888755,
+      "learning_rate": 9.599774048319973e-06,
+      "loss": 0.4834,
+      "step": 2937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.501518164502458,
+      "learning_rate": 9.599505578416439e-06,
+      "loss": 0.4271,
+      "step": 2938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9279398060164583,
+      "learning_rate": 9.599237022692303e-06,
+      "loss": 0.5781,
+      "step": 2939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.94906723186028,
+      "learning_rate": 9.598968381153185e-06,
+      "loss": 0.5303,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7358877040749168,
+      "learning_rate": 9.598699653804709e-06,
+      "loss": 0.4874,
+      "step": 2941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7897112302562872,
+      "learning_rate": 9.598430840652502e-06,
+      "loss": 0.4501,
+      "step": 2942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.259429939761865,
+      "learning_rate": 9.598161941702192e-06,
+      "loss": 0.6916,
+      "step": 2943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.087660786282608,
+      "learning_rate": 9.597892956959407e-06,
+      "loss": 0.6783,
+      "step": 2944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.922311623833691,
+      "learning_rate": 9.597623886429778e-06,
+      "loss": 0.4996,
+      "step": 2945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.393766504104056,
+      "learning_rate": 9.597354730118938e-06,
+      "loss": 0.5901,
+      "step": 2946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7280045555641603,
+      "learning_rate": 9.59708548803252e-06,
+      "loss": 0.5141,
+      "step": 2947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8194751587847489,
+      "learning_rate": 9.596816160176165e-06,
+      "loss": 0.4446,
+      "step": 2948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3883485466525176,
+      "learning_rate": 9.596546746555507e-06,
+      "loss": 0.6585,
+      "step": 2949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0621943972933043,
+      "learning_rate": 9.596277247176186e-06,
+      "loss": 0.5527,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.5915526151657104,
+      "eval_runtime": 67.8307,
+      "eval_samples_per_second": 50.272,
+      "eval_steps_per_second": 1.577,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.004970254800912,
+      "learning_rate": 9.596007662043847e-06,
+      "loss": 0.469,
+      "step": 2951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8437573370302833,
+      "learning_rate": 9.595737991164131e-06,
+      "loss": 0.5012,
+      "step": 2952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.14093939017376,
+      "learning_rate": 9.595468234542683e-06,
+      "loss": 0.6298,
+      "step": 2953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4568128763661328,
+      "learning_rate": 9.595198392185154e-06,
+      "loss": 0.3617,
+      "step": 2954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.133505197787507,
+      "learning_rate": 9.594928464097189e-06,
+      "loss": 0.5358,
+      "step": 2955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9607569425051754,
+      "learning_rate": 9.59465845028444e-06,
+      "loss": 0.5764,
+      "step": 2956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1422936083013573,
+      "learning_rate": 9.594388350752559e-06,
+      "loss": 0.6673,
+      "step": 2957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6959292677735176,
+      "learning_rate": 9.594118165507203e-06,
+      "loss": 0.5017,
+      "step": 2958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8519393866548488,
+      "learning_rate": 9.593847894554025e-06,
+      "loss": 0.5496,
+      "step": 2959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7295706222370193,
+      "learning_rate": 9.593577537898683e-06,
+      "loss": 0.5406,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.631464480931203,
+      "learning_rate": 9.593307095546839e-06,
+      "loss": 0.5348,
+      "step": 2961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0295939199389803,
+      "learning_rate": 9.593036567504151e-06,
+      "loss": 0.6563,
+      "step": 2962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.419142797181447,
+      "learning_rate": 9.592765953776287e-06,
+      "loss": 0.8345,
+      "step": 2963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.043384885731639,
+      "learning_rate": 9.592495254368912e-06,
+      "loss": 0.6417,
+      "step": 2964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.268084367128611,
+      "learning_rate": 9.592224469287688e-06,
+      "loss": 0.6293,
+      "step": 2965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.130886259919603,
+      "learning_rate": 9.591953598538285e-06,
+      "loss": 0.6951,
+      "step": 2966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.033312697386344,
+      "learning_rate": 9.591682642126378e-06,
+      "loss": 0.5001,
+      "step": 2967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8050942617182104,
+      "learning_rate": 9.591411600057635e-06,
+      "loss": 0.4881,
+      "step": 2968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9106493041448298,
+      "learning_rate": 9.591140472337731e-06,
+      "loss": 0.5566,
+      "step": 2969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.337337671705944,
+      "learning_rate": 9.590869258972345e-06,
+      "loss": 0.6426,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7999639749466996,
+      "learning_rate": 9.590597959967148e-06,
+      "loss": 0.5665,
+      "step": 2971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9471907565085407,
+      "learning_rate": 9.590326575327827e-06,
+      "loss": 0.4638,
+      "step": 2972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8568003398089141,
+      "learning_rate": 9.590055105060058e-06,
+      "loss": 0.5591,
+      "step": 2973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9577470648859587,
+      "learning_rate": 9.589783549169526e-06,
+      "loss": 0.5963,
+      "step": 2974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8837285857973938,
+      "learning_rate": 9.589511907661916e-06,
+      "loss": 0.6078,
+      "step": 2975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8030575599618606,
+      "learning_rate": 9.589240180542913e-06,
+      "loss": 0.5845,
+      "step": 2976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.07806284005832,
+      "learning_rate": 9.588968367818208e-06,
+      "loss": 0.8139,
+      "step": 2977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.764996004684679,
+      "learning_rate": 9.58869646949349e-06,
+      "loss": 0.5477,
+      "step": 2978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5985634489553116,
+      "learning_rate": 9.58842448557445e-06,
+      "loss": 0.609,
+      "step": 2979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9066153639190786,
+      "learning_rate": 9.588152416066784e-06,
+      "loss": 0.7124,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7168226362978392,
+      "learning_rate": 9.587880260976186e-06,
+      "loss": 0.5574,
+      "step": 2981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.054691187823001,
+      "learning_rate": 9.587608020308355e-06,
+      "loss": 0.5883,
+      "step": 2982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1510929069787426,
+      "learning_rate": 9.587335694068986e-06,
+      "loss": 0.849,
+      "step": 2983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9211374960850724,
+      "learning_rate": 9.587063282263786e-06,
+      "loss": 0.4916,
+      "step": 2984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6389998016078844,
+      "learning_rate": 9.586790784898453e-06,
+      "loss": 0.4389,
+      "step": 2985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9095175324394567,
+      "learning_rate": 9.586518201978694e-06,
+      "loss": 0.4557,
+      "step": 2986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0157905654966237,
+      "learning_rate": 9.586245533510215e-06,
+      "loss": 0.5509,
+      "step": 2987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6877354188041684,
+      "learning_rate": 9.585972779498722e-06,
+      "loss": 0.5452,
+      "step": 2988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8294460994011648,
+      "learning_rate": 9.585699939949928e-06,
+      "loss": 0.4797,
+      "step": 2989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.110209316535956,
+      "learning_rate": 9.585427014869543e-06,
+      "loss": 0.6418,
+      "step": 2990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5358772418665259,
+      "learning_rate": 9.585154004263281e-06,
+      "loss": 0.4436,
+      "step": 2991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8925871314920564,
+      "learning_rate": 9.58488090813686e-06,
+      "loss": 0.5272,
+      "step": 2992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7529892548880914,
+      "learning_rate": 9.584607726495992e-06,
+      "loss": 0.5224,
+      "step": 2993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2692662984440153,
+      "learning_rate": 9.584334459346398e-06,
+      "loss": 0.6424,
+      "step": 2994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.08870445309817,
+      "learning_rate": 9.5840611066938e-06,
+      "loss": 0.5318,
+      "step": 2995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4557084244438423,
+      "learning_rate": 9.583787668543919e-06,
+      "loss": 0.7706,
+      "step": 2996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.097764681617332,
+      "learning_rate": 9.58351414490248e-06,
+      "loss": 0.7672,
+      "step": 2997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.080322315271112,
+      "learning_rate": 9.583240535775208e-06,
+      "loss": 0.5351,
+      "step": 2998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7638937966982553,
+      "learning_rate": 9.582966841167832e-06,
+      "loss": 0.5163,
+      "step": 2999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7693723612840977,
+      "learning_rate": 9.58269306108608e-06,
+      "loss": 0.5341,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.5906082987785339,
+      "eval_runtime": 67.8842,
+      "eval_samples_per_second": 50.233,
+      "eval_steps_per_second": 1.576,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5115904534598514,
+      "learning_rate": 9.582419195535686e-06,
+      "loss": 0.6364,
+      "step": 3001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6549787119565413,
+      "learning_rate": 9.582145244522382e-06,
+      "loss": 0.4503,
+      "step": 3002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5346530835277696,
+      "learning_rate": 9.581871208051901e-06,
+      "loss": 0.4886,
+      "step": 3003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.302517186311315,
+      "learning_rate": 9.581597086129985e-06,
+      "loss": 0.5714,
+      "step": 3004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.62927149779114,
+      "learning_rate": 9.581322878762365e-06,
+      "loss": 0.4774,
+      "step": 3005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2655004508204684,
+      "learning_rate": 9.581048585954789e-06,
+      "loss": 0.5755,
+      "step": 3006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6098499763140346,
+      "learning_rate": 9.580774207712994e-06,
+      "loss": 0.7463,
+      "step": 3007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7388126283358316,
+      "learning_rate": 9.580499744042728e-06,
+      "loss": 0.4923,
+      "step": 3008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.245728719746843,
+      "learning_rate": 9.58022519494973e-06,
+      "loss": 0.5035,
+      "step": 3009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9890396672107982,
+      "learning_rate": 9.579950560439756e-06,
+      "loss": 0.4832,
+      "step": 3010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7093005615041954,
+      "learning_rate": 9.579675840518548e-06,
+      "loss": 0.4551,
+      "step": 3011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.716086777115914,
+      "learning_rate": 9.579401035191862e-06,
+      "loss": 0.5073,
+      "step": 3012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0245704687418264,
+      "learning_rate": 9.579126144465449e-06,
+      "loss": 0.6152,
+      "step": 3013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1099459172745867,
+      "learning_rate": 9.578851168345064e-06,
+      "loss": 0.7483,
+      "step": 3014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8052462737070931,
+      "learning_rate": 9.578576106836463e-06,
+      "loss": 0.6165,
+      "step": 3015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9104649290792461,
+      "learning_rate": 9.578300959945403e-06,
+      "loss": 0.4247,
+      "step": 3016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9106076310470483,
+      "learning_rate": 9.578025727677649e-06,
+      "loss": 0.5704,
+      "step": 3017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1104951229339086,
+      "learning_rate": 9.577750410038958e-06,
+      "loss": 0.5487,
+      "step": 3018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8966467988344515,
+      "learning_rate": 9.577475007035092e-06,
+      "loss": 0.5545,
+      "step": 3019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1480848487209645,
+      "learning_rate": 9.577199518671822e-06,
+      "loss": 0.6709,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.80731956343126,
+      "learning_rate": 9.57692394495491e-06,
+      "loss": 0.5001,
+      "step": 3021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.597436812890421,
+      "learning_rate": 9.57664828589013e-06,
+      "loss": 0.3553,
+      "step": 3022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.418369818776088,
+      "learning_rate": 9.57637254148325e-06,
+      "loss": 0.7842,
+      "step": 3023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.508401359461893,
+      "learning_rate": 9.576096711740044e-06,
+      "loss": 0.7592,
+      "step": 3024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.06550263069151,
+      "learning_rate": 9.575820796666283e-06,
+      "loss": 0.6697,
+      "step": 3025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0302579692233396,
+      "learning_rate": 9.575544796267747e-06,
+      "loss": 0.5305,
+      "step": 3026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9031939449015358,
+      "learning_rate": 9.57526871055021e-06,
+      "loss": 0.592,
+      "step": 3027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8791499385909707,
+      "learning_rate": 9.574992539519456e-06,
+      "loss": 0.5951,
+      "step": 3028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7803921347181921,
+      "learning_rate": 9.574716283181262e-06,
+      "loss": 0.596,
+      "step": 3029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5333027225709681,
+      "learning_rate": 9.574439941541415e-06,
+      "loss": 0.5527,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9129524493449332,
+      "learning_rate": 9.574163514605698e-06,
+      "loss": 0.5592,
+      "step": 3031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.336471209367489,
+      "learning_rate": 9.5738870023799e-06,
+      "loss": 0.5555,
+      "step": 3032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.802352473781538,
+      "learning_rate": 9.573610404869806e-06,
+      "loss": 0.5189,
+      "step": 3033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.743689004425902,
+      "learning_rate": 9.573333722081207e-06,
+      "loss": 0.5356,
+      "step": 3034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7775069331131714,
+      "learning_rate": 9.5730569540199e-06,
+      "loss": 0.5428,
+      "step": 3035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8594862473510883,
+      "learning_rate": 9.572780100691673e-06,
+      "loss": 0.5385,
+      "step": 3036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8254123466515308,
+      "learning_rate": 9.572503162102326e-06,
+      "loss": 0.6134,
+      "step": 3037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0626030477249016,
+      "learning_rate": 9.572226138257653e-06,
+      "loss": 0.5955,
+      "step": 3038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9209753066299413,
+      "learning_rate": 9.571949029163458e-06,
+      "loss": 0.5353,
+      "step": 3039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9645162978937496,
+      "learning_rate": 9.571671834825535e-06,
+      "loss": 0.6041,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.404854577020231,
+      "learning_rate": 9.571394555249695e-06,
+      "loss": 0.6288,
+      "step": 3041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5306774635510383,
+      "learning_rate": 9.571117190441737e-06,
+      "loss": 0.7152,
+      "step": 3042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9118339303273284,
+      "learning_rate": 9.570839740407469e-06,
+      "loss": 0.4877,
+      "step": 3043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1849606010277807,
+      "learning_rate": 9.570562205152702e-06,
+      "loss": 0.6418,
+      "step": 3044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2854254735878112,
+      "learning_rate": 9.57028458468324e-06,
+      "loss": 0.5713,
+      "step": 3045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.022658966683422,
+      "learning_rate": 9.570006879004902e-06,
+      "loss": 0.6166,
+      "step": 3046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5858176283161067,
+      "learning_rate": 9.569729088123495e-06,
+      "loss": 0.8528,
+      "step": 3047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1314182201894494,
+      "learning_rate": 9.569451212044839e-06,
+      "loss": 0.6165,
+      "step": 3048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2880050463715325,
+      "learning_rate": 9.569173250774749e-06,
+      "loss": 0.7137,
+      "step": 3049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.013066576488722,
+      "learning_rate": 9.568895204319043e-06,
+      "loss": 0.6929,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.5899938941001892,
+      "eval_runtime": 67.8311,
+      "eval_samples_per_second": 50.272,
+      "eval_steps_per_second": 1.577,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7227860693934882,
+      "learning_rate": 9.568617072683545e-06,
+      "loss": 0.4837,
+      "step": 3051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0585171418930197,
+      "learning_rate": 9.568338855874076e-06,
+      "loss": 0.6888,
+      "step": 3052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7136376868134486,
+      "learning_rate": 9.56806055389646e-06,
+      "loss": 0.5527,
+      "step": 3053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8940494846359124,
+      "learning_rate": 9.56778216675652e-06,
+      "loss": 0.5359,
+      "step": 3054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.755785542472755,
+      "learning_rate": 9.56750369446009e-06,
+      "loss": 0.5154,
+      "step": 3055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9249840019849491,
+      "learning_rate": 9.567225137012996e-06,
+      "loss": 0.5651,
+      "step": 3056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0202935535848305,
+      "learning_rate": 9.56694649442107e-06,
+      "loss": 0.5278,
+      "step": 3057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0113804391499945,
+      "learning_rate": 9.566667766690145e-06,
+      "loss": 0.5609,
+      "step": 3058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0692425145292415,
+      "learning_rate": 9.566388953826058e-06,
+      "loss": 0.606,
+      "step": 3059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0584374822324305,
+      "learning_rate": 9.56611005583464e-06,
+      "loss": 0.6411,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1325306678218516,
+      "learning_rate": 9.565831072721736e-06,
+      "loss": 0.6928,
+      "step": 3061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1328500239859185,
+      "learning_rate": 9.565552004493185e-06,
+      "loss": 0.685,
+      "step": 3062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5384306746142777,
+      "learning_rate": 9.565272851154826e-06,
+      "loss": 0.8685,
+      "step": 3063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2325873993261425,
+      "learning_rate": 9.564993612712506e-06,
+      "loss": 0.7074,
+      "step": 3064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0608526846666044,
+      "learning_rate": 9.564714289172068e-06,
+      "loss": 0.4784,
+      "step": 3065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.024268151111275,
+      "learning_rate": 9.564434880539364e-06,
+      "loss": 0.6579,
+      "step": 3066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2313590199227797,
+      "learning_rate": 9.564155386820237e-06,
+      "loss": 0.8014,
+      "step": 3067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1847625759080893,
+      "learning_rate": 9.563875808020544e-06,
+      "loss": 0.6326,
+      "step": 3068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.51978052002357,
+      "learning_rate": 9.563596144146134e-06,
+      "loss": 0.7387,
+      "step": 3069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8409431498560012,
+      "learning_rate": 9.563316395202862e-06,
+      "loss": 0.451,
+      "step": 3070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.807819570131881,
+      "learning_rate": 9.563036561196585e-06,
+      "loss": 0.5125,
+      "step": 3071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.196748624170899,
+      "learning_rate": 9.562756642133162e-06,
+      "loss": 0.6468,
+      "step": 3072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7687369842007787,
+      "learning_rate": 9.562476638018453e-06,
+      "loss": 0.8087,
+      "step": 3073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0247417937795253,
+      "learning_rate": 9.562196548858317e-06,
+      "loss": 0.6395,
+      "step": 3074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8003995319435122,
+      "learning_rate": 9.561916374658622e-06,
+      "loss": 0.4758,
+      "step": 3075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.227084149504011,
+      "learning_rate": 9.561636115425229e-06,
+      "loss": 0.6447,
+      "step": 3076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5847881714735277,
+      "learning_rate": 9.561355771164006e-06,
+      "loss": 0.4077,
+      "step": 3077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9824420709633288,
+      "learning_rate": 9.561075341880823e-06,
+      "loss": 0.6049,
+      "step": 3078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2616022463467504,
+      "learning_rate": 9.56079482758155e-06,
+      "loss": 0.7018,
+      "step": 3079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8579601594784154,
+      "learning_rate": 9.560514228272062e-06,
+      "loss": 0.6134,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.48811601405222,
+      "learning_rate": 9.560233543958227e-06,
+      "loss": 0.7099,
+      "step": 3081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3729158049766377,
+      "learning_rate": 9.559952774645927e-06,
+      "loss": 0.6068,
+      "step": 3082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7186791129110428,
+      "learning_rate": 9.559671920341037e-06,
+      "loss": 0.4595,
+      "step": 3083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3325494911249773,
+      "learning_rate": 9.559390981049436e-06,
+      "loss": 0.7974,
+      "step": 3084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.000824609850223,
+      "learning_rate": 9.559109956777007e-06,
+      "loss": 0.5845,
+      "step": 3085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1464493865423613,
+      "learning_rate": 9.558828847529631e-06,
+      "loss": 0.5741,
+      "step": 3086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8662334771434332,
+      "learning_rate": 9.558547653313196e-06,
+      "loss": 0.5685,
+      "step": 3087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9893744889874645,
+      "learning_rate": 9.558266374133585e-06,
+      "loss": 0.6033,
+      "step": 3088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7242844942378424,
+      "learning_rate": 9.55798500999669e-06,
+      "loss": 0.5009,
+      "step": 3089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3569171555795796,
+      "learning_rate": 9.557703560908399e-06,
+      "loss": 0.8598,
+      "step": 3090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7710838839095913,
+      "learning_rate": 9.557422026874603e-06,
+      "loss": 0.5501,
+      "step": 3091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7771300973297668,
+      "learning_rate": 9.557140407901197e-06,
+      "loss": 0.5202,
+      "step": 3092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5512573794547735,
+      "learning_rate": 9.556858703994077e-06,
+      "loss": 0.6765,
+      "step": 3093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9798098848035788,
+      "learning_rate": 9.55657691515914e-06,
+      "loss": 0.6006,
+      "step": 3094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5449392374293893,
+      "learning_rate": 9.556295041402284e-06,
+      "loss": 0.4839,
+      "step": 3095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9733756863634724,
+      "learning_rate": 9.556013082729413e-06,
+      "loss": 0.5992,
+      "step": 3096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6883626446193687,
+      "learning_rate": 9.555731039146426e-06,
+      "loss": 0.5259,
+      "step": 3097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.729933838999733,
+      "learning_rate": 9.55544891065923e-06,
+      "loss": 0.4977,
+      "step": 3098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2328025809963195,
+      "learning_rate": 9.555166697273729e-06,
+      "loss": 0.5973,
+      "step": 3099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7794566060275359,
+      "learning_rate": 9.554884398995832e-06,
+      "loss": 0.5453,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.5910373330116272,
+      "eval_runtime": 67.8307,
+      "eval_samples_per_second": 50.272,
+      "eval_steps_per_second": 1.577,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.977867311368933,
+      "learning_rate": 9.554602015831448e-06,
+      "loss": 0.5531,
+      "step": 3101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1124765559117167,
+      "learning_rate": 9.55431954778649e-06,
+      "loss": 0.5533,
+      "step": 3102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8498081410879939,
+      "learning_rate": 9.554036994866872e-06,
+      "loss": 0.5376,
+      "step": 3103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.974889818283197,
+      "learning_rate": 9.553754357078503e-06,
+      "loss": 0.7132,
+      "step": 3104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7228598640149884,
+      "learning_rate": 9.553471634427308e-06,
+      "loss": 0.5156,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0521755400598565,
+      "learning_rate": 9.553188826919202e-06,
+      "loss": 0.6214,
+      "step": 3106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7452172175308134,
+      "learning_rate": 9.552905934560104e-06,
+      "loss": 0.4126,
+      "step": 3107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2241465036012547,
+      "learning_rate": 9.552622957355938e-06,
+      "loss": 0.6679,
+      "step": 3108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9551617576430063,
+      "learning_rate": 9.55233989531263e-06,
+      "loss": 0.4649,
+      "step": 3109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1750479120565362,
+      "learning_rate": 9.552056748436099e-06,
+      "loss": 0.6936,
+      "step": 3110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5916158399734932,
+      "learning_rate": 9.551773516732278e-06,
+      "loss": 0.4452,
+      "step": 3111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.388949682088978,
+      "learning_rate": 9.551490200207097e-06,
+      "loss": 0.6334,
+      "step": 3112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9234916541099591,
+      "learning_rate": 9.551206798866486e-06,
+      "loss": 0.4512,
+      "step": 3113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.144358767769146,
+      "learning_rate": 9.550923312716373e-06,
+      "loss": 0.4833,
+      "step": 3114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8713746579727728,
+      "learning_rate": 9.5506397417627e-06,
+      "loss": 0.6038,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5254531633191286,
+      "learning_rate": 9.550356086011399e-06,
+      "loss": 0.3524,
+      "step": 3116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6416066314501183,
+      "learning_rate": 9.550072345468409e-06,
+      "loss": 0.7601,
+      "step": 3117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.840473130727977,
+      "learning_rate": 9.549788520139668e-06,
+      "loss": 0.5567,
+      "step": 3118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6241794114907944,
+      "learning_rate": 9.549504610031123e-06,
+      "loss": 0.5834,
+      "step": 3119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8865345954987462,
+      "learning_rate": 9.549220615148713e-06,
+      "loss": 0.6497,
+      "step": 3120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9331792494285693,
+      "learning_rate": 9.548936535498384e-06,
+      "loss": 0.4863,
+      "step": 3121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0588457612281825,
+      "learning_rate": 9.548652371086084e-06,
+      "loss": 0.513,
+      "step": 3122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.012287864656277,
+      "learning_rate": 9.54836812191776e-06,
+      "loss": 0.5301,
+      "step": 3123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.155318839340137,
+      "learning_rate": 9.548083787999367e-06,
+      "loss": 0.7569,
+      "step": 3124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8335258800597476,
+      "learning_rate": 9.54779936933685e-06,
+      "loss": 0.4819,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6866309662182626,
+      "learning_rate": 9.547514865936171e-06,
+      "loss": 0.5283,
+      "step": 3126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.044099219741292,
+      "learning_rate": 9.54723027780328e-06,
+      "loss": 0.4046,
+      "step": 3127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2435998361899743,
+      "learning_rate": 9.54694560494414e-06,
+      "loss": 0.4742,
+      "step": 3128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.541767057368725,
+      "learning_rate": 9.546660847364705e-06,
+      "loss": 0.4765,
+      "step": 3129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.926951092362006,
+      "learning_rate": 9.546376005070939e-06,
+      "loss": 0.6437,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.42442665208296,
+      "learning_rate": 9.546091078068804e-06,
+      "loss": 0.7974,
+      "step": 3131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5702976507202844,
+      "learning_rate": 9.545806066364266e-06,
+      "loss": 0.4933,
+      "step": 3132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.541489986199367,
+      "learning_rate": 9.545520969963291e-06,
+      "loss": 0.7114,
+      "step": 3133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8852073518629298,
+      "learning_rate": 9.545235788871847e-06,
+      "loss": 0.5681,
+      "step": 3134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9094826596039047,
+      "learning_rate": 9.544950523095906e-06,
+      "loss": 0.527,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.125985086029629,
+      "learning_rate": 9.544665172641435e-06,
+      "loss": 0.6232,
+      "step": 3136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.215117761463946,
+      "learning_rate": 9.544379737514413e-06,
+      "loss": 0.5003,
+      "step": 3137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9434148595821406,
+      "learning_rate": 9.544094217720813e-06,
+      "loss": 0.6256,
+      "step": 3138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8953003512691426,
+      "learning_rate": 9.543808613266613e-06,
+      "loss": 0.5199,
+      "step": 3139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.661129729217101,
+      "learning_rate": 9.54352292415779e-06,
+      "loss": 0.4477,
+      "step": 3140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0039088215580683,
+      "learning_rate": 9.543237150400327e-06,
+      "loss": 0.5976,
+      "step": 3141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8948688625450247,
+      "learning_rate": 9.542951292000205e-06,
+      "loss": 0.5394,
+      "step": 3142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8527291781917634,
+      "learning_rate": 9.542665348963409e-06,
+      "loss": 0.6266,
+      "step": 3143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9452787482576426,
+      "learning_rate": 9.542379321295926e-06,
+      "loss": 0.5357,
+      "step": 3144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0710806230021337,
+      "learning_rate": 9.54209320900374e-06,
+      "loss": 0.638,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9311987240946606,
+      "learning_rate": 9.541807012092847e-06,
+      "loss": 0.5517,
+      "step": 3146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2272646569093553,
+      "learning_rate": 9.541520730569232e-06,
+      "loss": 0.6448,
+      "step": 3147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1204708399857553,
+      "learning_rate": 9.541234364438891e-06,
+      "loss": 0.6914,
+      "step": 3148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7866158550884148,
+      "learning_rate": 9.54094791370782e-06,
+      "loss": 0.5489,
+      "step": 3149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8863069539225514,
+      "learning_rate": 9.540661378382013e-06,
+      "loss": 0.4566,
+      "step": 3150
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.5891137719154358,
+      "eval_runtime": 67.8947,
+      "eval_samples_per_second": 50.225,
+      "eval_steps_per_second": 1.576,
+      "step": 3150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6745693065400316,
+      "learning_rate": 9.540374758467472e-06,
+      "loss": 0.4827,
+      "step": 3151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.989174387863836,
+      "learning_rate": 9.540088053970194e-06,
+      "loss": 0.6163,
+      "step": 3152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1974943383629864,
+      "learning_rate": 9.539801264896182e-06,
+      "loss": 0.6541,
+      "step": 3153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0776178598508164,
+      "learning_rate": 9.539514391251438e-06,
+      "loss": 0.683,
+      "step": 3154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.801111148613236,
+      "learning_rate": 9.539227433041971e-06,
+      "loss": 0.5225,
+      "step": 3155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.052302668687446,
+      "learning_rate": 9.538940390273785e-06,
+      "loss": 0.732,
+      "step": 3156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.680736151614565,
+      "learning_rate": 9.538653262952892e-06,
+      "loss": 0.5263,
+      "step": 3157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0005398528134894,
+      "learning_rate": 9.538366051085301e-06,
+      "loss": 0.5399,
+      "step": 3158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.447225809358969,
+      "learning_rate": 9.538078754677026e-06,
+      "loss": 0.4633,
+      "step": 3159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0585238645587527,
+      "learning_rate": 9.537791373734078e-06,
+      "loss": 0.5481,
+      "step": 3160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.789444245092961,
+      "learning_rate": 9.537503908262477e-06,
+      "loss": 0.4562,
+      "step": 3161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8845871445591609,
+      "learning_rate": 9.537216358268241e-06,
+      "loss": 0.4692,
+      "step": 3162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8991303081722302,
+      "learning_rate": 9.536928723757387e-06,
+      "loss": 0.4929,
+      "step": 3163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.468974729099657,
+      "learning_rate": 9.536641004735936e-06,
+      "loss": 0.3809,
+      "step": 3164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6579265400132954,
+      "learning_rate": 9.536353201209914e-06,
+      "loss": 0.5557,
+      "step": 3165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9435611435440012,
+      "learning_rate": 9.536065313185346e-06,
+      "loss": 0.5092,
+      "step": 3166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9698276790705405,
+      "learning_rate": 9.535777340668257e-06,
+      "loss": 0.484,
+      "step": 3167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5073661363397854,
+      "learning_rate": 9.535489283664673e-06,
+      "loss": 0.7791,
+      "step": 3168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9541004396652475,
+      "learning_rate": 9.53520114218063e-06,
+      "loss": 0.5124,
+      "step": 3169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.008427482838996,
+      "learning_rate": 9.534912916222158e-06,
+      "loss": 0.6677,
+      "step": 3170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8063655651849397,
+      "learning_rate": 9.534624605795291e-06,
+      "loss": 0.8003,
+      "step": 3171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8818252650579481,
+      "learning_rate": 9.534336210906062e-06,
+      "loss": 0.5076,
+      "step": 3172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9922869358172786,
+      "learning_rate": 9.53404773156051e-06,
+      "loss": 0.5666,
+      "step": 3173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6238801419390096,
+      "learning_rate": 9.533759167764675e-06,
+      "loss": 0.7137,
+      "step": 3174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9334224862875504,
+      "learning_rate": 9.533470519524598e-06,
+      "loss": 0.5753,
+      "step": 3175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8260792119400369,
+      "learning_rate": 9.53318178684632e-06,
+      "loss": 0.5459,
+      "step": 3176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9706197880721792,
+      "learning_rate": 9.532892969735888e-06,
+      "loss": 0.7504,
+      "step": 3177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9399296878106185,
+      "learning_rate": 9.532604068199344e-06,
+      "loss": 0.5406,
+      "step": 3178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.145301288635934,
+      "learning_rate": 9.532315082242741e-06,
+      "loss": 0.5806,
+      "step": 3179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1981151400350036,
+      "learning_rate": 9.532026011872125e-06,
+      "loss": 0.69,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9839736438282203,
+      "learning_rate": 9.53173685709355e-06,
+      "loss": 0.5577,
+      "step": 3181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9544845122485173,
+      "learning_rate": 9.531447617913066e-06,
+      "loss": 0.5995,
+      "step": 3182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7832646010424305,
+      "learning_rate": 9.531158294336732e-06,
+      "loss": 0.4899,
+      "step": 3183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3454789255909647,
+      "learning_rate": 9.530868886370602e-06,
+      "loss": 0.5543,
+      "step": 3184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8859734107921853,
+      "learning_rate": 9.530579394020736e-06,
+      "loss": 0.5326,
+      "step": 3185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.592787282483212,
+      "learning_rate": 9.530289817293193e-06,
+      "loss": 0.5727,
+      "step": 3186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.539936297052289,
+      "learning_rate": 9.530000156194036e-06,
+      "loss": 0.4631,
+      "step": 3187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.206478441011057,
+      "learning_rate": 9.52971041072933e-06,
+      "loss": 0.5438,
+      "step": 3188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8028157058669092,
+      "learning_rate": 9.52942058090514e-06,
+      "loss": 0.4924,
+      "step": 3189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9516238361604288,
+      "learning_rate": 9.529130666727531e-06,
+      "loss": 0.7215,
+      "step": 3190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.472713618011654,
+      "learning_rate": 9.528840668202576e-06,
+      "loss": 0.7237,
+      "step": 3191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.192838233603872,
+      "learning_rate": 9.528550585336344e-06,
+      "loss": 0.6477,
+      "step": 3192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.163358139629063,
+      "learning_rate": 9.528260418134906e-06,
+      "loss": 0.5066,
+      "step": 3193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2455759518536698,
+      "learning_rate": 9.527970166604337e-06,
+      "loss": 0.5681,
+      "step": 3194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1407115355682125,
+      "learning_rate": 9.527679830750719e-06,
+      "loss": 0.5155,
+      "step": 3195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0529782452882674,
+      "learning_rate": 9.527389410580122e-06,
+      "loss": 0.642,
+      "step": 3196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4027424622357456,
+      "learning_rate": 9.52709890609863e-06,
+      "loss": 0.5236,
+      "step": 3197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.59243163767903,
+      "learning_rate": 9.526808317312323e-06,
+      "loss": 0.5111,
+      "step": 3198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6940851402441395,
+      "learning_rate": 9.526517644227285e-06,
+      "loss": 0.5351,
+      "step": 3199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8978096214711677,
+      "learning_rate": 9.526226886849603e-06,
+      "loss": 0.543,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.5886103510856628,
+      "eval_runtime": 67.8419,
+      "eval_samples_per_second": 50.264,
+      "eval_steps_per_second": 1.577,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.196888942924948,
+      "learning_rate": 9.52593604518536e-06,
+      "loss": 0.5557,
+      "step": 3201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8656047234162605,
+      "learning_rate": 9.525645119240649e-06,
+      "loss": 0.4876,
+      "step": 3202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7837639764407827,
+      "learning_rate": 9.525354109021556e-06,
+      "loss": 0.5088,
+      "step": 3203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.914460135268873,
+      "learning_rate": 9.525063014534174e-06,
+      "loss": 0.5747,
+      "step": 3204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0382801875654755,
+      "learning_rate": 9.5247718357846e-06,
+      "loss": 0.5765,
+      "step": 3205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6700087350368202,
+      "learning_rate": 9.524480572778926e-06,
+      "loss": 0.4258,
+      "step": 3206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5514645083066516,
+      "learning_rate": 9.52418922552325e-06,
+      "loss": 0.6933,
+      "step": 3207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2182304209168016,
+      "learning_rate": 9.523897794023673e-06,
+      "loss": 0.6041,
+      "step": 3208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.034885940585191,
+      "learning_rate": 9.523606278286296e-06,
+      "loss": 0.7452,
+      "step": 3209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.044314628656788,
+      "learning_rate": 9.52331467831722e-06,
+      "loss": 0.7016,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.523079134184689,
+      "learning_rate": 9.523022994122552e-06,
+      "loss": 0.5071,
+      "step": 3211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.314990058694135,
+      "learning_rate": 9.522731225708394e-06,
+      "loss": 0.6044,
+      "step": 3212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6547076150669824,
+      "learning_rate": 9.522439373080858e-06,
+      "loss": 0.7109,
+      "step": 3213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1898298879659315,
+      "learning_rate": 9.52214743624605e-06,
+      "loss": 0.6024,
+      "step": 3214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4904352096492368,
+      "learning_rate": 9.521855415210084e-06,
+      "loss": 0.4792,
+      "step": 3215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0295616944562624,
+      "learning_rate": 9.521563309979075e-06,
+      "loss": 0.547,
+      "step": 3216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.841886882560357,
+      "learning_rate": 9.521271120559135e-06,
+      "loss": 0.6312,
+      "step": 3217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0131655140754203,
+      "learning_rate": 9.520978846956381e-06,
+      "loss": 0.5081,
+      "step": 3218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0783117287448665,
+      "learning_rate": 9.520686489176933e-06,
+      "loss": 0.5433,
+      "step": 3219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.701256993213901,
+      "learning_rate": 9.520394047226911e-06,
+      "loss": 0.4631,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6247648944469308,
+      "learning_rate": 9.520101521112436e-06,
+      "loss": 0.484,
+      "step": 3221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9140178524885234,
+      "learning_rate": 9.519808910839633e-06,
+      "loss": 0.5541,
+      "step": 3222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.031919932335614,
+      "learning_rate": 9.519516216414627e-06,
+      "loss": 0.6232,
+      "step": 3223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.898491220227057,
+      "learning_rate": 9.519223437843546e-06,
+      "loss": 0.6055,
+      "step": 3224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8813051706572859,
+      "learning_rate": 9.51893057513252e-06,
+      "loss": 0.4529,
+      "step": 3225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6680617497695343,
+      "learning_rate": 9.518637628287676e-06,
+      "loss": 0.4354,
+      "step": 3226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.809254787281372,
+      "learning_rate": 9.518344597315151e-06,
+      "loss": 0.613,
+      "step": 3227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8237144848670668,
+      "learning_rate": 9.518051482221079e-06,
+      "loss": 0.5104,
+      "step": 3228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1692369803933618,
+      "learning_rate": 9.517758283011593e-06,
+      "loss": 0.5721,
+      "step": 3229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8915072778526651,
+      "learning_rate": 9.517464999692833e-06,
+      "loss": 0.4927,
+      "step": 3230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7747149375723836,
+      "learning_rate": 9.51717163227094e-06,
+      "loss": 0.818,
+      "step": 3231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.927792326242803,
+      "learning_rate": 9.516878180752052e-06,
+      "loss": 0.6395,
+      "step": 3232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.512607656051948,
+      "learning_rate": 9.516584645142315e-06,
+      "loss": 0.4751,
+      "step": 3233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6748961488672574,
+      "learning_rate": 9.516291025447873e-06,
+      "loss": 0.4392,
+      "step": 3234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2521834885617134,
+      "learning_rate": 9.515997321674873e-06,
+      "loss": 0.5725,
+      "step": 3235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5616748793748965,
+      "learning_rate": 9.515703533829465e-06,
+      "loss": 0.8067,
+      "step": 3236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4699267159093448,
+      "learning_rate": 9.515409661917796e-06,
+      "loss": 0.7167,
+      "step": 3237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5214566794456104,
+      "learning_rate": 9.515115705946021e-06,
+      "loss": 0.4831,
+      "step": 3238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.916948275610867,
+      "learning_rate": 9.514821665920291e-06,
+      "loss": 0.6364,
+      "step": 3239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5368602664413769,
+      "learning_rate": 9.514527541846765e-06,
+      "loss": 0.4707,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.031007669715038,
+      "learning_rate": 9.514233333731598e-06,
+      "loss": 0.5115,
+      "step": 3241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.051979447247689,
+      "learning_rate": 9.513939041580948e-06,
+      "loss": 0.5458,
+      "step": 3242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.607157209398919,
+      "learning_rate": 9.513644665400979e-06,
+      "loss": 0.3659,
+      "step": 3243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.240869457302071,
+      "learning_rate": 9.51335020519785e-06,
+      "loss": 0.6376,
+      "step": 3244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2190467197191213,
+      "learning_rate": 9.51305566097773e-06,
+      "loss": 0.6529,
+      "step": 3245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8436323666436392,
+      "learning_rate": 9.51276103274678e-06,
+      "loss": 0.4831,
+      "step": 3246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2238137473509636,
+      "learning_rate": 9.512466320511171e-06,
+      "loss": 0.6429,
+      "step": 3247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1803464543176387,
+      "learning_rate": 9.512171524277073e-06,
+      "loss": 0.5986,
+      "step": 3248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9689273938038339,
+      "learning_rate": 9.511876644050653e-06,
+      "loss": 0.725,
+      "step": 3249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9515313147329225,
+      "learning_rate": 9.511581679838092e-06,
+      "loss": 0.5354,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.5877246260643005,
+      "eval_runtime": 67.8867,
+      "eval_samples_per_second": 50.231,
+      "eval_steps_per_second": 1.576,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8578715042270828,
+      "learning_rate": 9.511286631645557e-06,
+      "loss": 0.6866,
+      "step": 3251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7684096678994168,
+      "learning_rate": 9.510991499479228e-06,
+      "loss": 0.457,
+      "step": 3252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0303495370870444,
+      "learning_rate": 9.510696283345283e-06,
+      "loss": 0.603,
+      "step": 3253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.165735882245229,
+      "learning_rate": 9.510400983249904e-06,
+      "loss": 0.5599,
+      "step": 3254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0469119273563443,
+      "learning_rate": 9.51010559919927e-06,
+      "loss": 0.4899,
+      "step": 3255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.163372246658112,
+      "learning_rate": 9.509810131199566e-06,
+      "loss": 0.5783,
+      "step": 3256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2225456018401712,
+      "learning_rate": 9.509514579256979e-06,
+      "loss": 0.6165,
+      "step": 3257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2181642478383927,
+      "learning_rate": 9.509218943377693e-06,
+      "loss": 0.6645,
+      "step": 3258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.439143033668215,
+      "learning_rate": 9.5089232235679e-06,
+      "loss": 0.5485,
+      "step": 3259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8727571436423232,
+      "learning_rate": 9.508627419833787e-06,
+      "loss": 0.5501,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7881760983434973,
+      "learning_rate": 9.508331532181553e-06,
+      "loss": 0.5944,
+      "step": 3261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8016098392457625,
+      "learning_rate": 9.508035560617384e-06,
+      "loss": 0.5179,
+      "step": 3262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7418500187987398,
+      "learning_rate": 9.507739505147482e-06,
+      "loss": 0.5175,
+      "step": 3263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9337645525904292,
+      "learning_rate": 9.507443365778043e-06,
+      "loss": 0.6761,
+      "step": 3264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7181055525940354,
+      "learning_rate": 9.507147142515265e-06,
+      "loss": 0.5613,
+      "step": 3265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0160564222987887,
+      "learning_rate": 9.506850835365352e-06,
+      "loss": 0.6555,
+      "step": 3266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5841241761178104,
+      "learning_rate": 9.506554444334504e-06,
+      "loss": 0.4388,
+      "step": 3267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.832067390638481,
+      "learning_rate": 9.506257969428928e-06,
+      "loss": 0.4654,
+      "step": 3268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.146507110861188,
+      "learning_rate": 9.50596141065483e-06,
+      "loss": 0.7373,
+      "step": 3269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8179159395477584,
+      "learning_rate": 9.505664768018418e-06,
+      "loss": 0.5358,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.741709902958248,
+      "learning_rate": 9.505368041525902e-06,
+      "loss": 0.5062,
+      "step": 3271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7396010897528518,
+      "learning_rate": 9.505071231183493e-06,
+      "loss": 0.4546,
+      "step": 3272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9674793185663473,
+      "learning_rate": 9.504774336997406e-06,
+      "loss": 0.5614,
+      "step": 3273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9145021487207705,
+      "learning_rate": 9.504477358973856e-06,
+      "loss": 0.4808,
+      "step": 3274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1517404691540536,
+      "learning_rate": 9.50418029711906e-06,
+      "loss": 0.6166,
+      "step": 3275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0238374915042234,
+      "learning_rate": 9.503883151439236e-06,
+      "loss": 0.7032,
+      "step": 3276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9447058852791514,
+      "learning_rate": 9.503585921940606e-06,
+      "loss": 0.4816,
+      "step": 3277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5164948581096214,
+      "learning_rate": 9.503288608629392e-06,
+      "loss": 0.7022,
+      "step": 3278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0726446459059655,
+      "learning_rate": 9.502991211511815e-06,
+      "loss": 0.6902,
+      "step": 3279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.010309917832774,
+      "learning_rate": 9.502693730594103e-06,
+      "loss": 0.5653,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.13772740616453,
+      "learning_rate": 9.502396165882487e-06,
+      "loss": 0.6047,
+      "step": 3281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8112818676027254,
+      "learning_rate": 9.502098517383192e-06,
+      "loss": 0.5787,
+      "step": 3282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9544434828302166,
+      "learning_rate": 9.50180078510245e-06,
+      "loss": 0.5445,
+      "step": 3283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.117245011821629,
+      "learning_rate": 9.501502969046494e-06,
+      "loss": 0.5261,
+      "step": 3284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7775742119107412,
+      "learning_rate": 9.501205069221559e-06,
+      "loss": 0.497,
+      "step": 3285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7150250026070748,
+      "learning_rate": 9.50090708563388e-06,
+      "loss": 0.525,
+      "step": 3286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.978023415645894,
+      "learning_rate": 9.500609018289698e-06,
+      "loss": 0.5484,
+      "step": 3287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1291622852469003,
+      "learning_rate": 9.500310867195252e-06,
+      "loss": 0.519,
+      "step": 3288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6566048263790316,
+      "learning_rate": 9.50001263235678e-06,
+      "loss": 0.4689,
+      "step": 3289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0857793955934882,
+      "learning_rate": 9.49971431378053e-06,
+      "loss": 0.6507,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.004015323827428,
+      "learning_rate": 9.499415911472744e-06,
+      "loss": 0.5941,
+      "step": 3291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8957546446550477,
+      "learning_rate": 9.49911742543967e-06,
+      "loss": 0.6265,
+      "step": 3292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8014090855011369,
+      "learning_rate": 9.498818855687558e-06,
+      "loss": 0.4924,
+      "step": 3293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7653892037137848,
+      "learning_rate": 9.498520202222655e-06,
+      "loss": 0.6249,
+      "step": 3294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9171636618068641,
+      "learning_rate": 9.498221465051218e-06,
+      "loss": 0.5068,
+      "step": 3295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5431981583821592,
+      "learning_rate": 9.497922644179494e-06,
+      "loss": 0.579,
+      "step": 3296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0895639477341357,
+      "learning_rate": 9.497623739613747e-06,
+      "loss": 0.599,
+      "step": 3297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0897307865964616,
+      "learning_rate": 9.497324751360226e-06,
+      "loss": 0.7677,
+      "step": 3298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.600710148120022,
+      "learning_rate": 9.497025679425196e-06,
+      "loss": 0.5007,
+      "step": 3299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1508590091087045,
+      "learning_rate": 9.496726523814917e-06,
+      "loss": 0.5779,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.5860307812690735,
+      "eval_runtime": 67.8463,
+      "eval_samples_per_second": 50.261,
+      "eval_steps_per_second": 1.577,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6417063166107564,
+      "learning_rate": 9.49642728453565e-06,
+      "loss": 0.5155,
+      "step": 3301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6878432457465704,
+      "learning_rate": 9.49612796159366e-06,
+      "loss": 0.5494,
+      "step": 3302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.831434080878817,
+      "learning_rate": 9.495828554995211e-06,
+      "loss": 0.546,
+      "step": 3303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9540130186735762,
+      "learning_rate": 9.495529064746577e-06,
+      "loss": 0.513,
+      "step": 3304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9305539572190966,
+      "learning_rate": 9.49522949085402e-06,
+      "loss": 0.521,
+      "step": 3305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3935181710705635,
+      "learning_rate": 9.494929833323818e-06,
+      "loss": 0.4235,
+      "step": 3306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1526777523129437,
+      "learning_rate": 9.49463009216224e-06,
+      "loss": 0.5032,
+      "step": 3307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6638262084779878,
+      "learning_rate": 9.494330267375564e-06,
+      "loss": 0.3935,
+      "step": 3308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8860727269676687,
+      "learning_rate": 9.494030358970062e-06,
+      "loss": 0.5634,
+      "step": 3309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9120619275230075,
+      "learning_rate": 9.493730366952017e-06,
+      "loss": 0.6517,
+      "step": 3310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7224498448428105,
+      "learning_rate": 9.493430291327707e-06,
+      "loss": 0.7586,
+      "step": 3311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8526546882559358,
+      "learning_rate": 9.493130132103414e-06,
+      "loss": 0.5292,
+      "step": 3312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8330948916196004,
+      "learning_rate": 9.492829889285422e-06,
+      "loss": 0.5207,
+      "step": 3313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.415722296140533,
+      "learning_rate": 9.492529562880016e-06,
+      "loss": 0.6459,
+      "step": 3314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9034347637632898,
+      "learning_rate": 9.492229152893485e-06,
+      "loss": 0.5764,
+      "step": 3315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.072539925197712,
+      "learning_rate": 9.491928659332114e-06,
+      "loss": 0.5592,
+      "step": 3316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0537993934156944,
+      "learning_rate": 9.491628082202197e-06,
+      "loss": 0.4831,
+      "step": 3317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7903548364880644,
+      "learning_rate": 9.491327421510026e-06,
+      "loss": 0.877,
+      "step": 3318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9631854519740344,
+      "learning_rate": 9.491026677261894e-06,
+      "loss": 0.6171,
+      "step": 3319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7270038569124688,
+      "learning_rate": 9.490725849464097e-06,
+      "loss": 0.568,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9484666615110775,
+      "learning_rate": 9.490424938122934e-06,
+      "loss": 0.5588,
+      "step": 3321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7521369751413944,
+      "learning_rate": 9.490123943244703e-06,
+      "loss": 0.3903,
+      "step": 3322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.185509947328603,
+      "learning_rate": 9.489822864835705e-06,
+      "loss": 0.5959,
+      "step": 3323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.192102891801244,
+      "learning_rate": 9.489521702902245e-06,
+      "loss": 0.6859,
+      "step": 3324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8050725942066266,
+      "learning_rate": 9.489220457450625e-06,
+      "loss": 0.5498,
+      "step": 3325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7785918232853666,
+      "learning_rate": 9.488919128487153e-06,
+      "loss": 0.5421,
+      "step": 3326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.121401380876367,
+      "learning_rate": 9.488617716018135e-06,
+      "loss": 0.7728,
+      "step": 3327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8453161248079015,
+      "learning_rate": 9.488316220049885e-06,
+      "loss": 0.6021,
+      "step": 3328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.263795147989963,
+      "learning_rate": 9.48801464058871e-06,
+      "loss": 0.6434,
+      "step": 3329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8640545132269477,
+      "learning_rate": 9.487712977640928e-06,
+      "loss": 0.6237,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.91023186533274,
+      "learning_rate": 9.48741123121285e-06,
+      "loss": 0.6061,
+      "step": 3331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9819217505790512,
+      "learning_rate": 9.487109401310796e-06,
+      "loss": 0.5161,
+      "step": 3332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9167791201529913,
+      "learning_rate": 9.48680748794108e-06,
+      "loss": 0.4181,
+      "step": 3333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0062325028004806,
+      "learning_rate": 9.486505491110028e-06,
+      "loss": 0.5897,
+      "step": 3334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0473010454053178,
+      "learning_rate": 9.48620341082396e-06,
+      "loss": 0.523,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7017131674148958,
+      "learning_rate": 9.4859012470892e-06,
+      "loss": 0.5077,
+      "step": 3336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.757659088024027,
+      "learning_rate": 9.485598999912071e-06,
+      "loss": 0.6952,
+      "step": 3337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0797391490365644,
+      "learning_rate": 9.485296669298904e-06,
+      "loss": 0.5916,
+      "step": 3338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6795223470177845,
+      "learning_rate": 9.484994255256027e-06,
+      "loss": 0.5469,
+      "step": 3339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0639723456008974,
+      "learning_rate": 9.484691757789767e-06,
+      "loss": 0.6338,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.314281235445936,
+      "learning_rate": 9.484389176906463e-06,
+      "loss": 0.7959,
+      "step": 3341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0786354532184736,
+      "learning_rate": 9.484086512612445e-06,
+      "loss": 0.6058,
+      "step": 3342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8167814277611103,
+      "learning_rate": 9.483783764914052e-06,
+      "loss": 0.4374,
+      "step": 3343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9462130140726341,
+      "learning_rate": 9.48348093381762e-06,
+      "loss": 0.5082,
+      "step": 3344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.089713898549303,
+      "learning_rate": 9.483178019329489e-06,
+      "loss": 0.5473,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7398242961376689,
+      "learning_rate": 9.482875021455999e-06,
+      "loss": 0.487,
+      "step": 3346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5018155435476213,
+      "learning_rate": 9.482571940203497e-06,
+      "loss": 0.6367,
+      "step": 3347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.622148414192877,
+      "learning_rate": 9.482268775578323e-06,
+      "loss": 0.5813,
+      "step": 3348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7847103157784092,
+      "learning_rate": 9.481965527586825e-06,
+      "loss": 0.6784,
+      "step": 3349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.018901317582479,
+      "learning_rate": 9.481662196235352e-06,
+      "loss": 0.6243,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.5877712368965149,
+      "eval_runtime": 67.9946,
+      "eval_samples_per_second": 50.151,
+      "eval_steps_per_second": 1.574,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.645294215671235,
+      "learning_rate": 9.481358781530255e-06,
+      "loss": 0.8241,
+      "step": 3351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.094504393245446,
+      "learning_rate": 9.481055283477885e-06,
+      "loss": 0.6393,
+      "step": 3352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.618893278692163,
+      "learning_rate": 9.480751702084593e-06,
+      "loss": 0.6168,
+      "step": 3353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1960442998981518,
+      "learning_rate": 9.480448037356741e-06,
+      "loss": 0.7256,
+      "step": 3354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5902259506127303,
+      "learning_rate": 9.480144289300679e-06,
+      "loss": 0.3606,
+      "step": 3355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.219290860765714,
+      "learning_rate": 9.479840457922767e-06,
+      "loss": 0.5825,
+      "step": 3356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7946190961636486,
+      "learning_rate": 9.47953654322937e-06,
+      "loss": 0.5645,
+      "step": 3357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9616473450239145,
+      "learning_rate": 9.479232545226846e-06,
+      "loss": 0.6396,
+      "step": 3358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2688991051463105,
+      "learning_rate": 9.47892846392156e-06,
+      "loss": 0.7998,
+      "step": 3359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7081338301965867,
+      "learning_rate": 9.478624299319878e-06,
+      "loss": 0.6638,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.946677329305531,
+      "learning_rate": 9.478320051428168e-06,
+      "loss": 0.7414,
+      "step": 3361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8145494462797387,
+      "learning_rate": 9.478015720252799e-06,
+      "loss": 0.4824,
+      "step": 3362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9320085650215773,
+      "learning_rate": 9.47771130580014e-06,
+      "loss": 0.4233,
+      "step": 3363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8944830822855447,
+      "learning_rate": 9.477406808076567e-06,
+      "loss": 0.5521,
+      "step": 3364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.914359036110886,
+      "learning_rate": 9.477102227088452e-06,
+      "loss": 0.5179,
+      "step": 3365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9105866451307365,
+      "learning_rate": 9.476797562842174e-06,
+      "loss": 0.6956,
+      "step": 3366
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9356301212315035,
+      "learning_rate": 9.476492815344108e-06,
+      "loss": 0.5552,
+      "step": 3367
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2349965384674944,
+      "learning_rate": 9.476187984600636e-06,
+      "loss": 0.5638,
+      "step": 3368
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.691679716725464,
+      "learning_rate": 9.475883070618135e-06,
+      "loss": 0.5301,
+      "step": 3369
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9355085340820728,
+      "learning_rate": 9.475578073402994e-06,
+      "loss": 0.6074,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0522024305819646,
+      "learning_rate": 9.475272992961595e-06,
+      "loss": 0.6451,
+      "step": 3371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.395121415187043,
+      "learning_rate": 9.474967829300324e-06,
+      "loss": 0.7798,
+      "step": 3372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6931608033101455,
+      "learning_rate": 9.474662582425572e-06,
+      "loss": 0.4706,
+      "step": 3373
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.901416627392683,
+      "learning_rate": 9.474357252343725e-06,
+      "loss": 0.6402,
+      "step": 3374
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8296754881152557,
+      "learning_rate": 9.47405183906118e-06,
+      "loss": 0.4693,
+      "step": 3375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9086148342164666,
+      "learning_rate": 9.473746342584328e-06,
+      "loss": 0.58,
+      "step": 3376
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1813723525114255,
+      "learning_rate": 9.473440762919563e-06,
+      "loss": 0.7474,
+      "step": 3377
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.194479085562727,
+      "learning_rate": 9.473135100073285e-06,
+      "loss": 0.6538,
+      "step": 3378
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9276073804189646,
+      "learning_rate": 9.47282935405189e-06,
+      "loss": 0.7096,
+      "step": 3379
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0606287762350175,
+      "learning_rate": 9.472523524861781e-06,
+      "loss": 0.5534,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8482050599435444,
+      "learning_rate": 9.472217612509362e-06,
+      "loss": 0.5706,
+      "step": 3381
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9619699762643397,
+      "learning_rate": 9.47191161700103e-06,
+      "loss": 0.5211,
+      "step": 3382
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7520289121875268,
+      "learning_rate": 9.471605538343199e-06,
+      "loss": 0.584,
+      "step": 3383
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.109157137349181,
+      "learning_rate": 9.471299376542272e-06,
+      "loss": 0.6569,
+      "step": 3384
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9945638783204844,
+      "learning_rate": 9.470993131604656e-06,
+      "loss": 0.5103,
+      "step": 3385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.100118355869588,
+      "learning_rate": 9.470686803536768e-06,
+      "loss": 0.5814,
+      "step": 3386
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0797182214089895,
+      "learning_rate": 9.47038039234502e-06,
+      "loss": 0.5889,
+      "step": 3387
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9274836658034653,
+      "learning_rate": 9.47007389803582e-06,
+      "loss": 0.5067,
+      "step": 3388
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8968410667924398,
+      "learning_rate": 9.469767320615591e-06,
+      "loss": 0.5628,
+      "step": 3389
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3714631871340117,
+      "learning_rate": 9.46946066009075e-06,
+      "loss": 0.6148,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.409110275311975,
+      "learning_rate": 9.469153916467715e-06,
+      "loss": 0.4016,
+      "step": 3391
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2211668704076786,
+      "learning_rate": 9.468847089752907e-06,
+      "loss": 0.6311,
+      "step": 3392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7612109221059085,
+      "learning_rate": 9.468540179952751e-06,
+      "loss": 0.5815,
+      "step": 3393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8651825697002609,
+      "learning_rate": 9.468233187073673e-06,
+      "loss": 0.6775,
+      "step": 3394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5987585602921321,
+      "learning_rate": 9.467926111122096e-06,
+      "loss": 0.4846,
+      "step": 3395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0635545122208447,
+      "learning_rate": 9.46761895210445e-06,
+      "loss": 0.5756,
+      "step": 3396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5110939672018975,
+      "learning_rate": 9.467311710027167e-06,
+      "loss": 0.6264,
+      "step": 3397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2271314521201413,
+      "learning_rate": 9.467004384896676e-06,
+      "loss": 0.7201,
+      "step": 3398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9400652859099734,
+      "learning_rate": 9.466696976719413e-06,
+      "loss": 0.477,
+      "step": 3399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.233605696490964,
+      "learning_rate": 9.466389485501814e-06,
+      "loss": 0.7407,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.5908147096633911,
+      "eval_runtime": 67.9107,
+      "eval_samples_per_second": 50.213,
+      "eval_steps_per_second": 1.576,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9270364625209535,
+      "learning_rate": 9.466081911250313e-06,
+      "loss": 0.5141,
+      "step": 3401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3675533326310827,
+      "learning_rate": 9.465774253971351e-06,
+      "loss": 0.6415,
+      "step": 3402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7561876677911623,
+      "learning_rate": 9.465466513671369e-06,
+      "loss": 0.6484,
+      "step": 3403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1516743219107886,
+      "learning_rate": 9.465158690356808e-06,
+      "loss": 0.6299,
+      "step": 3404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4368367086324905,
+      "learning_rate": 9.464850784034113e-06,
+      "loss": 0.5448,
+      "step": 3405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1966605988407744,
+      "learning_rate": 9.464542794709729e-06,
+      "loss": 0.5832,
+      "step": 3406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3727275349357364,
+      "learning_rate": 9.464234722390104e-06,
+      "loss": 0.5245,
+      "step": 3407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.331968845633874,
+      "learning_rate": 9.463926567081686e-06,
+      "loss": 0.7421,
+      "step": 3408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9097147290640464,
+      "learning_rate": 9.46361832879093e-06,
+      "loss": 0.4841,
+      "step": 3409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7761218415455753,
+      "learning_rate": 9.463310007524286e-06,
+      "loss": 0.573,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0799162914138702,
+      "learning_rate": 9.463001603288207e-06,
+      "loss": 0.6422,
+      "step": 3411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.036142251690999,
+      "learning_rate": 9.462693116089151e-06,
+      "loss": 0.5484,
+      "step": 3412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.069374229423299,
+      "learning_rate": 9.462384545933577e-06,
+      "loss": 0.6325,
+      "step": 3413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7219007556153207,
+      "learning_rate": 9.462075892827943e-06,
+      "loss": 0.5774,
+      "step": 3414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6488429384171803,
+      "learning_rate": 9.461767156778712e-06,
+      "loss": 0.5269,
+      "step": 3415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6854278892845154,
+      "learning_rate": 9.461458337792346e-06,
+      "loss": 0.4906,
+      "step": 3416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.06180736759217,
+      "learning_rate": 9.46114943587531e-06,
+      "loss": 0.4965,
+      "step": 3417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.086409796020496,
+      "learning_rate": 9.46084045103407e-06,
+      "loss": 0.527,
+      "step": 3418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6433357124006778,
+      "learning_rate": 9.460531383275097e-06,
+      "loss": 0.6547,
+      "step": 3419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8470304178831007,
+      "learning_rate": 9.46022223260486e-06,
+      "loss": 0.4446,
+      "step": 3420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.215358264392675,
+      "learning_rate": 9.459912999029828e-06,
+      "loss": 0.609,
+      "step": 3421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7904874270128048,
+      "learning_rate": 9.459603682556477e-06,
+      "loss": 0.681,
+      "step": 3422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.120030110423786,
+      "learning_rate": 9.459294283191285e-06,
+      "loss": 0.6646,
+      "step": 3423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0093570940417096,
+      "learning_rate": 9.458984800940724e-06,
+      "loss": 0.501,
+      "step": 3424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.225509489776915,
+      "learning_rate": 9.458675235811277e-06,
+      "loss": 0.6069,
+      "step": 3425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.654956930352674,
+      "learning_rate": 9.45836558780942e-06,
+      "loss": 0.6936,
+      "step": 3426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8909796777392054,
+      "learning_rate": 9.45805585694164e-06,
+      "loss": 0.6223,
+      "step": 3427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1553127647508457,
+      "learning_rate": 9.45774604321442e-06,
+      "loss": 0.6819,
+      "step": 3428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6154138349705647,
+      "learning_rate": 9.457436146634243e-06,
+      "loss": 0.5338,
+      "step": 3429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.323038057196628,
+      "learning_rate": 9.4571261672076e-06,
+      "loss": 0.6206,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1214130518114134,
+      "learning_rate": 9.456816104940976e-06,
+      "loss": 0.6121,
+      "step": 3431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9722312977773189,
+      "learning_rate": 9.456505959840867e-06,
+      "loss": 0.5641,
+      "step": 3432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.85501306499447,
+      "learning_rate": 9.456195731913762e-06,
+      "loss": 0.5727,
+      "step": 3433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9746662153862666,
+      "learning_rate": 9.455885421166158e-06,
+      "loss": 0.8006,
+      "step": 3434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8142213179438427,
+      "learning_rate": 9.455575027604548e-06,
+      "loss": 0.5943,
+      "step": 3435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2191551220931856,
+      "learning_rate": 9.455264551235431e-06,
+      "loss": 0.8311,
+      "step": 3436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9564928616061996,
+      "learning_rate": 9.45495399206531e-06,
+      "loss": 0.481,
+      "step": 3437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5266197261720962,
+      "learning_rate": 9.454643350100683e-06,
+      "loss": 0.5078,
+      "step": 3438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9862330606697407,
+      "learning_rate": 9.454332625348051e-06,
+      "loss": 0.7083,
+      "step": 3439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1104204555806594,
+      "learning_rate": 9.454021817813926e-06,
+      "loss": 0.5978,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9083627428407528,
+      "learning_rate": 9.453710927504807e-06,
+      "loss": 0.6176,
+      "step": 3441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.210878336360078,
+      "learning_rate": 9.453399954427207e-06,
+      "loss": 0.6208,
+      "step": 3442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.788006635557164,
+      "learning_rate": 9.453088898587633e-06,
+      "loss": 0.6713,
+      "step": 3443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0939424253126493,
+      "learning_rate": 9.4527777599926e-06,
+      "loss": 0.485,
+      "step": 3444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6985933576087509,
+      "learning_rate": 9.452466538648619e-06,
+      "loss": 0.367,
+      "step": 3445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.777402392913385,
+      "learning_rate": 9.452155234562206e-06,
+      "loss": 0.5713,
+      "step": 3446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8291794844349192,
+      "learning_rate": 9.451843847739878e-06,
+      "loss": 0.4929,
+      "step": 3447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8850511462019428,
+      "learning_rate": 9.451532378188153e-06,
+      "loss": 0.4319,
+      "step": 3448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7672603774671907,
+      "learning_rate": 9.451220825913551e-06,
+      "loss": 0.4791,
+      "step": 3449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8902598680167917,
+      "learning_rate": 9.450909190922598e-06,
+      "loss": 0.6399,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.5898324251174927,
+      "eval_runtime": 67.9588,
+      "eval_samples_per_second": 50.177,
+      "eval_steps_per_second": 1.574,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1678033378037975,
+      "learning_rate": 9.450597473221814e-06,
+      "loss": 0.6168,
+      "step": 3451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7254963152590226,
+      "learning_rate": 9.450285672817727e-06,
+      "loss": 0.5412,
+      "step": 3452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7195879647517638,
+      "learning_rate": 9.44997378971686e-06,
+      "loss": 0.543,
+      "step": 3453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.798770438104945,
+      "learning_rate": 9.449661823925746e-06,
+      "loss": 0.6483,
+      "step": 3454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.649795761525077,
+      "learning_rate": 9.449349775450916e-06,
+      "loss": 0.515,
+      "step": 3455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6444869845884489,
+      "learning_rate": 9.449037644298901e-06,
+      "loss": 0.3393,
+      "step": 3456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8792237596344454,
+      "learning_rate": 9.448725430476237e-06,
+      "loss": 0.496,
+      "step": 3457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4106216574193513,
+      "learning_rate": 9.448413133989458e-06,
+      "loss": 0.6042,
+      "step": 3458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7944805920386424,
+      "learning_rate": 9.448100754845104e-06,
+      "loss": 0.4215,
+      "step": 3459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.314863654054123,
+      "learning_rate": 9.447788293049708e-06,
+      "loss": 0.5975,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7010601057572037,
+      "learning_rate": 9.44747574860982e-06,
+      "loss": 0.4856,
+      "step": 3461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.824145831742707,
+      "learning_rate": 9.44716312153198e-06,
+      "loss": 0.4841,
+      "step": 3462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.164811979659548,
+      "learning_rate": 9.446850411822729e-06,
+      "loss": 0.6297,
+      "step": 3463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7965243790621794,
+      "learning_rate": 9.446537619488618e-06,
+      "loss": 0.6123,
+      "step": 3464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.204046160380043,
+      "learning_rate": 9.446224744536192e-06,
+      "loss": 0.6435,
+      "step": 3465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.204217219032066,
+      "learning_rate": 9.445911786972001e-06,
+      "loss": 0.65,
+      "step": 3466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8784100611986407,
+      "learning_rate": 9.4455987468026e-06,
+      "loss": 0.5061,
+      "step": 3467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0777500073931017,
+      "learning_rate": 9.44528562403454e-06,
+      "loss": 0.4781,
+      "step": 3468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8748423901747415,
+      "learning_rate": 9.444972418674374e-06,
+      "loss": 0.4486,
+      "step": 3469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.143434052959705,
+      "learning_rate": 9.444659130728664e-06,
+      "loss": 0.6397,
+      "step": 3470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.168516650608013,
+      "learning_rate": 9.44434576020396e-06,
+      "loss": 0.5955,
+      "step": 3471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5324909721022366,
+      "learning_rate": 9.44403230710683e-06,
+      "loss": 0.7992,
+      "step": 3472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9295128524642815,
+      "learning_rate": 9.443718771443835e-06,
+      "loss": 0.4626,
+      "step": 3473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.092727667458908,
+      "learning_rate": 9.443405153221535e-06,
+      "loss": 0.6693,
+      "step": 3474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1604130470285234,
+      "learning_rate": 9.4430914524465e-06,
+      "loss": 0.7655,
+      "step": 3475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8117808529241868,
+      "learning_rate": 9.442777669125292e-06,
+      "loss": 0.6036,
+      "step": 3476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.77328309921414,
+      "learning_rate": 9.442463803264483e-06,
+      "loss": 0.458,
+      "step": 3477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.17731578195489,
+      "learning_rate": 9.442149854870644e-06,
+      "loss": 0.616,
+      "step": 3478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2375236453777476,
+      "learning_rate": 9.441835823950345e-06,
+      "loss": 0.769,
+      "step": 3479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7477580393336574,
+      "learning_rate": 9.441521710510162e-06,
+      "loss": 0.6977,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9401161870836325,
+      "learning_rate": 9.44120751455667e-06,
+      "loss": 0.5161,
+      "step": 3481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6421319197832944,
+      "learning_rate": 9.440893236096448e-06,
+      "loss": 0.4882,
+      "step": 3482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9595133749198967,
+      "learning_rate": 9.440578875136073e-06,
+      "loss": 0.5971,
+      "step": 3483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9740698087553485,
+      "learning_rate": 9.440264431682125e-06,
+      "loss": 0.5297,
+      "step": 3484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8606893088356267,
+      "learning_rate": 9.439949905741192e-06,
+      "loss": 0.4978,
+      "step": 3485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6795243948105791,
+      "learning_rate": 9.439635297319853e-06,
+      "loss": 0.5301,
+      "step": 3486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6594786562685984,
+      "learning_rate": 9.439320606424696e-06,
+      "loss": 0.5339,
+      "step": 3487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.918331948706719,
+      "learning_rate": 9.439005833062309e-06,
+      "loss": 0.4838,
+      "step": 3488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5107112272535261,
+      "learning_rate": 9.438690977239281e-06,
+      "loss": 0.328,
+      "step": 3489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8618038931945406,
+      "learning_rate": 9.438376038962205e-06,
+      "loss": 0.5324,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9991705836088292,
+      "learning_rate": 9.438061018237672e-06,
+      "loss": 0.5261,
+      "step": 3491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.266380292955577,
+      "learning_rate": 9.437745915072277e-06,
+      "loss": 0.545,
+      "step": 3492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0882280244020404,
+      "learning_rate": 9.437430729472619e-06,
+      "loss": 0.7589,
+      "step": 3493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8538636925832057,
+      "learning_rate": 9.437115461445292e-06,
+      "loss": 0.5596,
+      "step": 3494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.312776217904773,
+      "learning_rate": 9.436800110996899e-06,
+      "loss": 0.7049,
+      "step": 3495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9482579772721311,
+      "learning_rate": 9.436484678134042e-06,
+      "loss": 0.5423,
+      "step": 3496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1877733809503064,
+      "learning_rate": 9.436169162863324e-06,
+      "loss": 0.7938,
+      "step": 3497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0324266097276142,
+      "learning_rate": 9.435853565191347e-06,
+      "loss": 0.5653,
+      "step": 3498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.273769020977207,
+      "learning_rate": 9.435537885124721e-06,
+      "loss": 0.8383,
+      "step": 3499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.085765815067492,
+      "learning_rate": 9.435222122670055e-06,
+      "loss": 0.5698,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.5892014503479004,
+      "eval_runtime": 67.9679,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9462718894087254,
+      "learning_rate": 9.434906277833955e-06,
+      "loss": 0.5486,
+      "step": 3501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.16097322397164,
+      "learning_rate": 9.434590350623041e-06,
+      "loss": 0.747,
+      "step": 3502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.856343733696956,
+      "learning_rate": 9.434274341043917e-06,
+      "loss": 0.5058,
+      "step": 3503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1507016977501876,
+      "learning_rate": 9.433958249103206e-06,
+      "loss": 0.7296,
+      "step": 3504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.505820281623048,
+      "learning_rate": 9.433642074807523e-06,
+      "loss": 0.4604,
+      "step": 3505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7875435239104907,
+      "learning_rate": 9.433325818163486e-06,
+      "loss": 0.4515,
+      "step": 3506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.467234819701279,
+      "learning_rate": 9.433009479177715e-06,
+      "loss": 0.4279,
+      "step": 3507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2700322022842876,
+      "learning_rate": 9.432693057856836e-06,
+      "loss": 0.6308,
+      "step": 3508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8217131577726966,
+      "learning_rate": 9.432376554207466e-06,
+      "loss": 0.6756,
+      "step": 3509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.49557474789909,
+      "learning_rate": 9.432059968236239e-06,
+      "loss": 0.4887,
+      "step": 3510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0858953200011854,
+      "learning_rate": 9.431743299949777e-06,
+      "loss": 0.537,
+      "step": 3511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.593613962392707,
+      "learning_rate": 9.431426549354713e-06,
+      "loss": 0.4339,
+      "step": 3512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2526686582270607,
+      "learning_rate": 9.431109716457675e-06,
+      "loss": 0.6942,
+      "step": 3513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9859302763766145,
+      "learning_rate": 9.430792801265297e-06,
+      "loss": 0.7074,
+      "step": 3514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.022935001922362,
+      "learning_rate": 9.430475803784211e-06,
+      "loss": 0.5909,
+      "step": 3515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.090195408373747,
+      "learning_rate": 9.430158724021058e-06,
+      "loss": 0.434,
+      "step": 3516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.783759484990671,
+      "learning_rate": 9.429841561982472e-06,
+      "loss": 0.5313,
+      "step": 3517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.614934848411264,
+      "learning_rate": 9.429524317675096e-06,
+      "loss": 0.7904,
+      "step": 3518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8129513411739895,
+      "learning_rate": 9.429206991105566e-06,
+      "loss": 0.4543,
+      "step": 3519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9556091091494299,
+      "learning_rate": 9.428889582280532e-06,
+      "loss": 0.7417,
+      "step": 3520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7792479758887008,
+      "learning_rate": 9.42857209120663e-06,
+      "loss": 0.5562,
+      "step": 3521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.242292209172046,
+      "learning_rate": 9.428254517890516e-06,
+      "loss": 0.5994,
+      "step": 3522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8590050642690394,
+      "learning_rate": 9.427936862338831e-06,
+      "loss": 0.5707,
+      "step": 3523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8321826124196925,
+      "learning_rate": 9.427619124558229e-06,
+      "loss": 0.5065,
+      "step": 3524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.124822637692711,
+      "learning_rate": 9.427301304555358e-06,
+      "loss": 0.571,
+      "step": 3525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8544206085459376,
+      "learning_rate": 9.426983402336874e-06,
+      "loss": 0.5739,
+      "step": 3526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8293817575642959,
+      "learning_rate": 9.426665417909432e-06,
+      "loss": 0.5873,
+      "step": 3527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7686191997708811,
+      "learning_rate": 9.426347351279691e-06,
+      "loss": 0.5257,
+      "step": 3528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8705974195989772,
+      "learning_rate": 9.426029202454303e-06,
+      "loss": 0.6505,
+      "step": 3529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0083135202274,
+      "learning_rate": 9.425710971439935e-06,
+      "loss": 0.6418,
+      "step": 3530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.161980846602752,
+      "learning_rate": 9.425392658243245e-06,
+      "loss": 0.6665,
+      "step": 3531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.261010060523265,
+      "learning_rate": 9.425074262870898e-06,
+      "loss": 0.629,
+      "step": 3532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0563266218516216,
+      "learning_rate": 9.424755785329558e-06,
+      "loss": 0.481,
+      "step": 3533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5036595829447672,
+      "learning_rate": 9.424437225625896e-06,
+      "loss": 0.5565,
+      "step": 3534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7982969982240282,
+      "learning_rate": 9.424118583766575e-06,
+      "loss": 0.5804,
+      "step": 3535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7825791777899611,
+      "learning_rate": 9.423799859758272e-06,
+      "loss": 0.5268,
+      "step": 3536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.116528886218498,
+      "learning_rate": 9.423481053607657e-06,
+      "loss": 0.5403,
+      "step": 3537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7502842374322698,
+      "learning_rate": 9.4231621653214e-06,
+      "loss": 0.5333,
+      "step": 3538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8507422688459,
+      "learning_rate": 9.422843194906179e-06,
+      "loss": 0.6193,
+      "step": 3539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9026772080285548,
+      "learning_rate": 9.422524142368675e-06,
+      "loss": 0.6018,
+      "step": 3540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7228033156028277,
+      "learning_rate": 9.422205007715562e-06,
+      "loss": 0.5002,
+      "step": 3541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7841890279925308,
+      "learning_rate": 9.421885790953526e-06,
+      "loss": 0.4703,
+      "step": 3542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2012201800576605,
+      "learning_rate": 9.421566492089246e-06,
+      "loss": 0.5909,
+      "step": 3543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7822246393849845,
+      "learning_rate": 9.421247111129407e-06,
+      "loss": 0.5575,
+      "step": 3544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8158259919039073,
+      "learning_rate": 9.420927648080694e-06,
+      "loss": 0.5038,
+      "step": 3545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.120254947981896,
+      "learning_rate": 9.420608102949798e-06,
+      "loss": 0.6135,
+      "step": 3546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8513943471477063,
+      "learning_rate": 9.420288475743406e-06,
+      "loss": 0.5853,
+      "step": 3547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0845557937480677,
+      "learning_rate": 9.41996876646821e-06,
+      "loss": 0.6563,
+      "step": 3548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.040376785850913,
+      "learning_rate": 9.419648975130902e-06,
+      "loss": 0.5578,
+      "step": 3549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.302258951912242,
+      "learning_rate": 9.419329101738177e-06,
+      "loss": 0.6082,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.5875427722930908,
+      "eval_runtime": 67.9467,
+      "eval_samples_per_second": 50.186,
+      "eval_steps_per_second": 1.575,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9650941392913275,
+      "learning_rate": 9.419009146296734e-06,
+      "loss": 0.5921,
+      "step": 3551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4466913033271547,
+      "learning_rate": 9.418689108813264e-06,
+      "loss": 0.4214,
+      "step": 3552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9157268551697753,
+      "learning_rate": 9.418368989294476e-06,
+      "loss": 0.6206,
+      "step": 3553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.138183877873392,
+      "learning_rate": 9.418048787747065e-06,
+      "loss": 0.6357,
+      "step": 3554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5369508692088774,
+      "learning_rate": 9.417728504177737e-06,
+      "loss": 0.7588,
+      "step": 3555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.881735959697069,
+      "learning_rate": 9.417408138593195e-06,
+      "loss": 0.6142,
+      "step": 3556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5953697971974587,
+      "learning_rate": 9.417087691000149e-06,
+      "loss": 0.5001,
+      "step": 3557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0228265777741115,
+      "learning_rate": 9.416767161405304e-06,
+      "loss": 0.6941,
+      "step": 3558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.053176162114944,
+      "learning_rate": 9.41644654981537e-06,
+      "loss": 0.5547,
+      "step": 3559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0409100347093823,
+      "learning_rate": 9.416125856237061e-06,
+      "loss": 0.5292,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4945272768033937,
+      "learning_rate": 9.41580508067709e-06,
+      "loss": 0.4088,
+      "step": 3561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9951826336836795,
+      "learning_rate": 9.415484223142173e-06,
+      "loss": 0.6434,
+      "step": 3562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.897568219559712,
+      "learning_rate": 9.415163283639023e-06,
+      "loss": 0.5131,
+      "step": 3563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1722912673077635,
+      "learning_rate": 9.414842262174363e-06,
+      "loss": 0.7262,
+      "step": 3564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7494947955581186,
+      "learning_rate": 9.414521158754913e-06,
+      "loss": 0.4942,
+      "step": 3565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0164854324655908,
+      "learning_rate": 9.414199973387392e-06,
+      "loss": 0.5239,
+      "step": 3566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3775739270151552,
+      "learning_rate": 9.413878706078526e-06,
+      "loss": 0.6319,
+      "step": 3567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.603398094945688,
+      "learning_rate": 9.413557356835041e-06,
+      "loss": 0.4729,
+      "step": 3568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0371117988845713,
+      "learning_rate": 9.413235925663662e-06,
+      "loss": 0.704,
+      "step": 3569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9625743649423497,
+      "learning_rate": 9.412914412571122e-06,
+      "loss": 0.6037,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5256840304526083,
+      "learning_rate": 9.412592817564148e-06,
+      "loss": 0.4271,
+      "step": 3571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.907294585721979,
+      "learning_rate": 9.412271140649472e-06,
+      "loss": 0.5599,
+      "step": 3572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1526543128412428,
+      "learning_rate": 9.411949381833831e-06,
+      "loss": 0.6985,
+      "step": 3573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.927771103412545,
+      "learning_rate": 9.41162754112396e-06,
+      "loss": 0.5566,
+      "step": 3574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.260438071848397,
+      "learning_rate": 9.411305618526595e-06,
+      "loss": 0.5917,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8505334899034478,
+      "learning_rate": 9.410983614048478e-06,
+      "loss": 0.4985,
+      "step": 3576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0710680526380574,
+      "learning_rate": 9.410661527696347e-06,
+      "loss": 0.505,
+      "step": 3577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7375579843520972,
+      "learning_rate": 9.410339359476946e-06,
+      "loss": 0.5804,
+      "step": 3578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7699751362302547,
+      "learning_rate": 9.410017109397019e-06,
+      "loss": 0.5559,
+      "step": 3579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1923047714997264,
+      "learning_rate": 9.409694777463312e-06,
+      "loss": 0.6885,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.053830190518234,
+      "learning_rate": 9.409372363682575e-06,
+      "loss": 0.5807,
+      "step": 3581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5733939651107962,
+      "learning_rate": 9.409049868061553e-06,
+      "loss": 0.3937,
+      "step": 3582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7330276606562467,
+      "learning_rate": 9.408727290607003e-06,
+      "loss": 0.5003,
+      "step": 3583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1321574654066824,
+      "learning_rate": 9.408404631325674e-06,
+      "loss": 0.7038,
+      "step": 3584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1964150326273675,
+      "learning_rate": 9.408081890224321e-06,
+      "loss": 0.6836,
+      "step": 3585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9883775645927086,
+      "learning_rate": 9.407759067309701e-06,
+      "loss": 0.6124,
+      "step": 3586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0841753187415746,
+      "learning_rate": 9.407436162588573e-06,
+      "loss": 0.7201,
+      "step": 3587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.929021799420239,
+      "learning_rate": 9.407113176067696e-06,
+      "loss": 0.4858,
+      "step": 3588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2660186238926494,
+      "learning_rate": 9.406790107753831e-06,
+      "loss": 0.6074,
+      "step": 3589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0901549247173463,
+      "learning_rate": 9.406466957653742e-06,
+      "loss": 0.6117,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7798065376193763,
+      "learning_rate": 9.406143725774193e-06,
+      "loss": 0.5983,
+      "step": 3591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.242836654933296,
+      "learning_rate": 9.405820412121951e-06,
+      "loss": 0.661,
+      "step": 3592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.691161546337512,
+      "learning_rate": 9.405497016703785e-06,
+      "loss": 0.5831,
+      "step": 3593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.812368419617344,
+      "learning_rate": 9.405173539526466e-06,
+      "loss": 0.7867,
+      "step": 3594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.158055939414594,
+      "learning_rate": 9.404849980596764e-06,
+      "loss": 0.4642,
+      "step": 3595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6233197674032165,
+      "learning_rate": 9.404526339921451e-06,
+      "loss": 0.5733,
+      "step": 3596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.538247162701584,
+      "learning_rate": 9.404202617507306e-06,
+      "loss": 0.3293,
+      "step": 3597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8793765250610528,
+      "learning_rate": 9.403878813361105e-06,
+      "loss": 0.5812,
+      "step": 3598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9429577673503913,
+      "learning_rate": 9.403554927489624e-06,
+      "loss": 0.5044,
+      "step": 3599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5553397166105352,
+      "learning_rate": 9.403230959899646e-06,
+      "loss": 0.5269,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5864858031272888,
+      "eval_runtime": 67.9538,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.060209972941201,
+      "learning_rate": 9.402906910597953e-06,
+      "loss": 0.5296,
+      "step": 3601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3745041560325597,
+      "learning_rate": 9.402582779591328e-06,
+      "loss": 0.6688,
+      "step": 3602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8693956548806079,
+      "learning_rate": 9.402258566886556e-06,
+      "loss": 0.5627,
+      "step": 3603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7524194461089284,
+      "learning_rate": 9.401934272490423e-06,
+      "loss": 0.5491,
+      "step": 3604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1474219688212695,
+      "learning_rate": 9.401609896409721e-06,
+      "loss": 0.6706,
+      "step": 3605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.170195131131733,
+      "learning_rate": 9.40128543865124e-06,
+      "loss": 0.5606,
+      "step": 3606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7746655559620201,
+      "learning_rate": 9.40096089922177e-06,
+      "loss": 0.3854,
+      "step": 3607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.96021616799052,
+      "learning_rate": 9.400636278128106e-06,
+      "loss": 0.6198,
+      "step": 3608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9915462522843783,
+      "learning_rate": 9.400311575377047e-06,
+      "loss": 0.4582,
+      "step": 3609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.152447649808285,
+      "learning_rate": 9.399986790975387e-06,
+      "loss": 0.5201,
+      "step": 3610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4910251877489205,
+      "learning_rate": 9.399661924929923e-06,
+      "loss": 0.3951,
+      "step": 3611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.853154046092541,
+      "learning_rate": 9.399336977247461e-06,
+      "loss": 0.6107,
+      "step": 3612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.68942511082538,
+      "learning_rate": 9.399011947934805e-06,
+      "loss": 0.5551,
+      "step": 3613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.709990486185667,
+      "learning_rate": 9.39868683699875e-06,
+      "loss": 0.5254,
+      "step": 3614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4840029550713982,
+      "learning_rate": 9.398361644446108e-06,
+      "loss": 0.954,
+      "step": 3615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.928598253664864,
+      "learning_rate": 9.39803637028369e-06,
+      "loss": 0.4838,
+      "step": 3616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3754207212739606,
+      "learning_rate": 9.397711014518299e-06,
+      "loss": 0.6588,
+      "step": 3617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.151901687518859,
+      "learning_rate": 9.39738557715675e-06,
+      "loss": 0.487,
+      "step": 3618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.186652335600924,
+      "learning_rate": 9.397060058205854e-06,
+      "loss": 0.7127,
+      "step": 3619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8684667117046094,
+      "learning_rate": 9.396734457672426e-06,
+      "loss": 0.7327,
+      "step": 3620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.509356997368458,
+      "learning_rate": 9.396408775563284e-06,
+      "loss": 0.8529,
+      "step": 3621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1596790504361087,
+      "learning_rate": 9.396083011885243e-06,
+      "loss": 0.6652,
+      "step": 3622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.285072731707485,
+      "learning_rate": 9.395757166645125e-06,
+      "loss": 0.5346,
+      "step": 3623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9702659215632319,
+      "learning_rate": 9.395431239849749e-06,
+      "loss": 0.4987,
+      "step": 3624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1629626694107236,
+      "learning_rate": 9.39510523150594e-06,
+      "loss": 0.5589,
+      "step": 3625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0718980465411425,
+      "learning_rate": 9.394779141620523e-06,
+      "loss": 0.6162,
+      "step": 3626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5279349444758616,
+      "learning_rate": 9.394452970200325e-06,
+      "loss": 0.7939,
+      "step": 3627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1684431296444058,
+      "learning_rate": 9.394126717252172e-06,
+      "loss": 0.5128,
+      "step": 3628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8500971764078749,
+      "learning_rate": 9.393800382782894e-06,
+      "loss": 0.515,
+      "step": 3629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.255047047180345,
+      "learning_rate": 9.393473966799325e-06,
+      "loss": 0.6289,
+      "step": 3630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.051535766770004,
+      "learning_rate": 9.393147469308296e-06,
+      "loss": 0.5831,
+      "step": 3631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1572901517186636,
+      "learning_rate": 9.392820890316642e-06,
+      "loss": 0.5742,
+      "step": 3632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6971503528232792,
+      "learning_rate": 9.3924942298312e-06,
+      "loss": 0.4915,
+      "step": 3633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3967676085717877,
+      "learning_rate": 9.392167487858812e-06,
+      "loss": 0.7674,
+      "step": 3634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9109396930529754,
+      "learning_rate": 9.391840664406313e-06,
+      "loss": 0.5012,
+      "step": 3635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8934577892517863,
+      "learning_rate": 9.391513759480549e-06,
+      "loss": 0.5259,
+      "step": 3636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8289628281508772,
+      "learning_rate": 9.39118677308836e-06,
+      "loss": 0.6101,
+      "step": 3637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8506321280412903,
+      "learning_rate": 9.39085970523659e-06,
+      "loss": 0.5501,
+      "step": 3638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.947312664672043,
+      "learning_rate": 9.390532555932093e-06,
+      "loss": 0.5953,
+      "step": 3639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7379904393934669,
+      "learning_rate": 9.390205325181712e-06,
+      "loss": 0.5455,
+      "step": 3640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7220218347767848,
+      "learning_rate": 9.389878012992299e-06,
+      "loss": 0.7274,
+      "step": 3641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5216348141857712,
+      "learning_rate": 9.389550619370707e-06,
+      "loss": 0.432,
+      "step": 3642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0780601930830906,
+      "learning_rate": 9.389223144323788e-06,
+      "loss": 0.6972,
+      "step": 3643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.863030838801643,
+      "learning_rate": 9.388895587858396e-06,
+      "loss": 0.7255,
+      "step": 3644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9840194303260839,
+      "learning_rate": 9.388567949981393e-06,
+      "loss": 0.7728,
+      "step": 3645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7092388360559647,
+      "learning_rate": 9.388240230699634e-06,
+      "loss": 0.4797,
+      "step": 3646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3300887815408,
+      "learning_rate": 9.38791243001998e-06,
+      "loss": 0.5958,
+      "step": 3647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.854909004333428,
+      "learning_rate": 9.387584547949297e-06,
+      "loss": 0.5461,
+      "step": 3648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6333104346404412,
+      "learning_rate": 9.387256584494444e-06,
+      "loss": 0.6373,
+      "step": 3649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2224278730403526,
+      "learning_rate": 9.386928539662288e-06,
+      "loss": 0.5398,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5843286514282227,
+      "eval_runtime": 67.9812,
+      "eval_samples_per_second": 50.161,
+      "eval_steps_per_second": 1.574,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.163029165972253,
+      "learning_rate": 9.3866004134597e-06,
+      "loss": 0.5895,
+      "step": 3651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8507905040126273,
+      "learning_rate": 9.386272205893545e-06,
+      "loss": 0.5331,
+      "step": 3652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9713735199026208,
+      "learning_rate": 9.385943916970695e-06,
+      "loss": 0.5419,
+      "step": 3653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8567276920948175,
+      "learning_rate": 9.385615546698022e-06,
+      "loss": 0.5063,
+      "step": 3654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9703433630448088,
+      "learning_rate": 9.385287095082402e-06,
+      "loss": 0.5466,
+      "step": 3655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.800471415806526,
+      "learning_rate": 9.38495856213071e-06,
+      "loss": 0.5138,
+      "step": 3656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7635472325335748,
+      "learning_rate": 9.384629947849823e-06,
+      "loss": 0.5521,
+      "step": 3657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7277257147641993,
+      "learning_rate": 9.384301252246622e-06,
+      "loss": 0.5615,
+      "step": 3658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.007488701952963,
+      "learning_rate": 9.383972475327985e-06,
+      "loss": 0.6545,
+      "step": 3659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1022427786863673,
+      "learning_rate": 9.383643617100799e-06,
+      "loss": 0.6325,
+      "step": 3660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.777265301239716,
+      "learning_rate": 9.383314677571945e-06,
+      "loss": 0.5601,
+      "step": 3661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0356026374542195,
+      "learning_rate": 9.382985656748309e-06,
+      "loss": 0.6337,
+      "step": 3662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6886150687234271,
+      "learning_rate": 9.382656554636781e-06,
+      "loss": 0.5092,
+      "step": 3663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9710258241278968,
+      "learning_rate": 9.38232737124425e-06,
+      "loss": 0.4099,
+      "step": 3664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7492603594561595,
+      "learning_rate": 9.381998106577606e-06,
+      "loss": 0.6037,
+      "step": 3665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.001509756056166,
+      "learning_rate": 9.381668760643745e-06,
+      "loss": 0.5398,
+      "step": 3666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7587656717589208,
+      "learning_rate": 9.381339333449557e-06,
+      "loss": 0.4905,
+      "step": 3667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7560820830664152,
+      "learning_rate": 9.381009825001942e-06,
+      "loss": 0.6022,
+      "step": 3668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1522565675682763,
+      "learning_rate": 9.380680235307798e-06,
+      "loss": 0.694,
+      "step": 3669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6201466908452784,
+      "learning_rate": 9.380350564374022e-06,
+      "loss": 0.498,
+      "step": 3670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7156192929369434,
+      "learning_rate": 9.380020812207518e-06,
+      "loss": 0.4972,
+      "step": 3671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7481119026058927,
+      "learning_rate": 9.379690978815187e-06,
+      "loss": 0.5133,
+      "step": 3672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7555117716630375,
+      "learning_rate": 9.379361064203937e-06,
+      "loss": 0.461,
+      "step": 3673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.722178742061427,
+      "learning_rate": 9.379031068380671e-06,
+      "loss": 0.5819,
+      "step": 3674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9368802395816844,
+      "learning_rate": 9.3787009913523e-06,
+      "loss": 0.6436,
+      "step": 3675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.766010751865493,
+      "learning_rate": 9.378370833125731e-06,
+      "loss": 0.5022,
+      "step": 3676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.64415009338513,
+      "learning_rate": 9.37804059370788e-06,
+      "loss": 0.473,
+      "step": 3677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.035393903122745,
+      "learning_rate": 9.377710273105658e-06,
+      "loss": 0.4322,
+      "step": 3678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.242951051508498,
+      "learning_rate": 9.377379871325976e-06,
+      "loss": 0.6046,
+      "step": 3679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.848425605902073,
+      "learning_rate": 9.37704938837576e-06,
+      "loss": 0.6109,
+      "step": 3680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.453560733870149,
+      "learning_rate": 9.376718824261919e-06,
+      "loss": 0.5056,
+      "step": 3681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7673781499817904,
+      "learning_rate": 9.37638817899138e-06,
+      "loss": 0.45,
+      "step": 3682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.67459655007997,
+      "learning_rate": 9.376057452571063e-06,
+      "loss": 0.4768,
+      "step": 3683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9075733478306907,
+      "learning_rate": 9.375726645007888e-06,
+      "loss": 0.5312,
+      "step": 3684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7784687412842006,
+      "learning_rate": 9.375395756308783e-06,
+      "loss": 0.5133,
+      "step": 3685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6314345698929713,
+      "learning_rate": 9.375064786480677e-06,
+      "loss": 0.4953,
+      "step": 3686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.669322805187621,
+      "learning_rate": 9.374733735530494e-06,
+      "loss": 0.6053,
+      "step": 3687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7513714050160378,
+      "learning_rate": 9.374402603465169e-06,
+      "loss": 0.4859,
+      "step": 3688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8146350659170158,
+      "learning_rate": 9.374071390291631e-06,
+      "loss": 0.4773,
+      "step": 3689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7047084914547626,
+      "learning_rate": 9.373740096016816e-06,
+      "loss": 0.4401,
+      "step": 3690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1663013057449834,
+      "learning_rate": 9.373408720647657e-06,
+      "loss": 0.4524,
+      "step": 3691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.133333727906336,
+      "learning_rate": 9.373077264191093e-06,
+      "loss": 0.6437,
+      "step": 3692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1614155913985162,
+      "learning_rate": 9.372745726654062e-06,
+      "loss": 0.4933,
+      "step": 3693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.984245508397986,
+      "learning_rate": 9.372414108043505e-06,
+      "loss": 0.4969,
+      "step": 3694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.179258199876279,
+      "learning_rate": 9.372082408366361e-06,
+      "loss": 0.5995,
+      "step": 3695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.06608670679541,
+      "learning_rate": 9.371750627629581e-06,
+      "loss": 0.5944,
+      "step": 3696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.286118875420831,
+      "learning_rate": 9.371418765840105e-06,
+      "loss": 0.5966,
+      "step": 3697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.767446896888357,
+      "learning_rate": 9.37108682300488e-06,
+      "loss": 0.5158,
+      "step": 3698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9314836357084746,
+      "learning_rate": 9.370754799130858e-06,
+      "loss": 0.5158,
+      "step": 3699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.818476807832134,
+      "learning_rate": 9.37042269422499e-06,
+      "loss": 0.6675,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5861907005310059,
+      "eval_runtime": 67.9156,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4056634456154649,
+      "learning_rate": 9.370090508294226e-06,
+      "loss": 0.3378,
+      "step": 3701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.110291868310698,
+      "learning_rate": 9.36975824134552e-06,
+      "loss": 0.6187,
+      "step": 3702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3906108669571133,
+      "learning_rate": 9.36942589338583e-06,
+      "loss": 0.447,
+      "step": 3703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4335928498510757,
+      "learning_rate": 9.369093464422111e-06,
+      "loss": 0.5674,
+      "step": 3704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9918283504773986,
+      "learning_rate": 9.368760954461324e-06,
+      "loss": 0.4736,
+      "step": 3705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.13047916007435,
+      "learning_rate": 9.36842836351043e-06,
+      "loss": 0.7048,
+      "step": 3706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.944278731304873,
+      "learning_rate": 9.36809569157639e-06,
+      "loss": 0.7236,
+      "step": 3707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.681357976382639,
+      "learning_rate": 9.367762938666169e-06,
+      "loss": 0.4983,
+      "step": 3708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9278603105844374,
+      "learning_rate": 9.367430104786734e-06,
+      "loss": 0.5863,
+      "step": 3709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8363910465049256,
+      "learning_rate": 9.367097189945052e-06,
+      "loss": 0.4261,
+      "step": 3710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9909106238056047,
+      "learning_rate": 9.366764194148093e-06,
+      "loss": 0.4997,
+      "step": 3711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0997921445559227,
+      "learning_rate": 9.366431117402826e-06,
+      "loss": 0.4512,
+      "step": 3712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8864492825235564,
+      "learning_rate": 9.366097959716227e-06,
+      "loss": 0.7342,
+      "step": 3713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5462221997932086,
+      "learning_rate": 9.365764721095268e-06,
+      "loss": 0.4847,
+      "step": 3714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.258434159818986,
+      "learning_rate": 9.365431401546924e-06,
+      "loss": 0.6542,
+      "step": 3715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.296183081874394,
+      "learning_rate": 9.365098001078174e-06,
+      "loss": 0.5152,
+      "step": 3716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3362884703411533,
+      "learning_rate": 9.364764519696e-06,
+      "loss": 0.6801,
+      "step": 3717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.891443341813374,
+      "learning_rate": 9.364430957407381e-06,
+      "loss": 0.6506,
+      "step": 3718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.086555197053113,
+      "learning_rate": 9.364097314219299e-06,
+      "loss": 0.7778,
+      "step": 3719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7033660688627086,
+      "learning_rate": 9.363763590138741e-06,
+      "loss": 0.6002,
+      "step": 3720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.218545846332353,
+      "learning_rate": 9.363429785172693e-06,
+      "loss": 0.7361,
+      "step": 3721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.834162682133079,
+      "learning_rate": 9.363095899328141e-06,
+      "loss": 0.6142,
+      "step": 3722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6491928720867253,
+      "learning_rate": 9.362761932612075e-06,
+      "loss": 0.5106,
+      "step": 3723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7305237730623477,
+      "learning_rate": 9.362427885031488e-06,
+      "loss": 0.6281,
+      "step": 3724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1468701594054993,
+      "learning_rate": 9.362093756593371e-06,
+      "loss": 0.6886,
+      "step": 3725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.919977516096027,
+      "learning_rate": 9.361759547304723e-06,
+      "loss": 0.5381,
+      "step": 3726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7128933776600437,
+      "learning_rate": 9.361425257172535e-06,
+      "loss": 0.5398,
+      "step": 3727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9016225396283957,
+      "learning_rate": 9.361090886203808e-06,
+      "loss": 0.4967,
+      "step": 3728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.980473575415399,
+      "learning_rate": 9.360756434405543e-06,
+      "loss": 0.6008,
+      "step": 3729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7401654514012774,
+      "learning_rate": 9.360421901784739e-06,
+      "loss": 0.5582,
+      "step": 3730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0294663073462793,
+      "learning_rate": 9.360087288348399e-06,
+      "loss": 0.6772,
+      "step": 3731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7643155245191746,
+      "learning_rate": 9.35975259410353e-06,
+      "loss": 0.4127,
+      "step": 3732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6982116549530435,
+      "learning_rate": 9.35941781905714e-06,
+      "loss": 0.4733,
+      "step": 3733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.800065698796575,
+      "learning_rate": 9.359082963216232e-06,
+      "loss": 0.5385,
+      "step": 3734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3255870458479397,
+      "learning_rate": 9.35874802658782e-06,
+      "loss": 0.5943,
+      "step": 3735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8806425641463815,
+      "learning_rate": 9.358413009178916e-06,
+      "loss": 0.4708,
+      "step": 3736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8859595373139248,
+      "learning_rate": 9.358077910996532e-06,
+      "loss": 0.6103,
+      "step": 3737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.197383674096259,
+      "learning_rate": 9.357742732047683e-06,
+      "loss": 0.638,
+      "step": 3738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2180292452110066,
+      "learning_rate": 9.357407472339385e-06,
+      "loss": 0.6863,
+      "step": 3739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9210342806188505,
+      "learning_rate": 9.35707213187866e-06,
+      "loss": 0.536,
+      "step": 3740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7844189161296096,
+      "learning_rate": 9.356736710672523e-06,
+      "loss": 0.5203,
+      "step": 3741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9453983006843636,
+      "learning_rate": 9.356401208728e-06,
+      "loss": 0.5681,
+      "step": 3742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.103337600696327,
+      "learning_rate": 9.356065626052113e-06,
+      "loss": 0.5647,
+      "step": 3743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8269591537758247,
+      "learning_rate": 9.355729962651886e-06,
+      "loss": 0.4715,
+      "step": 3744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.160765517802622,
+      "learning_rate": 9.355394218534348e-06,
+      "loss": 0.7705,
+      "step": 3745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8122538999060407,
+      "learning_rate": 9.355058393706526e-06,
+      "loss": 0.6317,
+      "step": 3746
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6086755444681946,
+      "learning_rate": 9.354722488175453e-06,
+      "loss": 0.5483,
+      "step": 3747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6557331958917128,
+      "learning_rate": 9.354386501948157e-06,
+      "loss": 0.5563,
+      "step": 3748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0220091711101857,
+      "learning_rate": 9.354050435031674e-06,
+      "loss": 0.4896,
+      "step": 3749
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.080640608673691,
+      "learning_rate": 9.353714287433038e-06,
+      "loss": 0.4579,
+      "step": 3750
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5846731662750244,
+      "eval_runtime": 67.9617,
+      "eval_samples_per_second": 50.175,
+      "eval_steps_per_second": 1.574,
+      "step": 3750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.887145869846854,
+      "learning_rate": 9.353378059159289e-06,
+      "loss": 0.516,
+      "step": 3751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4792327012984323,
+      "learning_rate": 9.353041750217461e-06,
+      "loss": 0.706,
+      "step": 3752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9664239863015587,
+      "learning_rate": 9.3527053606146e-06,
+      "loss": 0.6162,
+      "step": 3753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.179666183809481,
+      "learning_rate": 9.352368890357745e-06,
+      "loss": 0.5224,
+      "step": 3754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8051995411398236,
+      "learning_rate": 9.35203233945394e-06,
+      "loss": 0.499,
+      "step": 3755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5777424957043462,
+      "learning_rate": 9.351695707910229e-06,
+      "loss": 0.4372,
+      "step": 3756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9662850888696972,
+      "learning_rate": 9.351358995733662e-06,
+      "loss": 0.4973,
+      "step": 3757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8198245952666094,
+      "learning_rate": 9.351022202931285e-06,
+      "loss": 0.5723,
+      "step": 3758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2292087260858633,
+      "learning_rate": 9.350685329510152e-06,
+      "loss": 0.6247,
+      "step": 3759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9714457264149639,
+      "learning_rate": 9.350348375477312e-06,
+      "loss": 0.5452,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1277033560747762,
+      "learning_rate": 9.350011340839822e-06,
+      "loss": 0.7325,
+      "step": 3761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7507222476566844,
+      "learning_rate": 9.349674225604735e-06,
+      "loss": 0.6192,
+      "step": 3762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8728176376869108,
+      "learning_rate": 9.34933702977911e-06,
+      "loss": 0.5679,
+      "step": 3763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6285089339187167,
+      "learning_rate": 9.348999753370004e-06,
+      "loss": 0.538,
+      "step": 3764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.98939017117873,
+      "learning_rate": 9.348662396384479e-06,
+      "loss": 0.5538,
+      "step": 3765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3419152486864734,
+      "learning_rate": 9.348324958829596e-06,
+      "loss": 0.7072,
+      "step": 3766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8653265172332034,
+      "learning_rate": 9.347987440712422e-06,
+      "loss": 0.6169,
+      "step": 3767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9066263253749973,
+      "learning_rate": 9.347649842040023e-06,
+      "loss": 0.6952,
+      "step": 3768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.153529179126615,
+      "learning_rate": 9.34731216281946e-06,
+      "loss": 0.5803,
+      "step": 3769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4369060454421614,
+      "learning_rate": 9.346974403057808e-06,
+      "loss": 0.4476,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9866554770773996,
+      "learning_rate": 9.346636562762139e-06,
+      "loss": 0.7617,
+      "step": 3771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9260183147367067,
+      "learning_rate": 9.346298641939522e-06,
+      "loss": 0.5037,
+      "step": 3772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.56029868758942,
+      "learning_rate": 9.34596064059703e-06,
+      "loss": 0.3978,
+      "step": 3773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.315690852675625,
+      "learning_rate": 9.34562255874174e-06,
+      "loss": 0.7333,
+      "step": 3774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8873188774010339,
+      "learning_rate": 9.345284396380733e-06,
+      "loss": 0.507,
+      "step": 3775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1073689147934354,
+      "learning_rate": 9.344946153521085e-06,
+      "loss": 0.606,
+      "step": 3776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9483523050241642,
+      "learning_rate": 9.344607830169876e-06,
+      "loss": 0.5434,
+      "step": 3777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0526966349703217,
+      "learning_rate": 9.344269426334191e-06,
+      "loss": 0.4857,
+      "step": 3778
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.831082091212042,
+      "learning_rate": 9.343930942021113e-06,
+      "loss": 0.5944,
+      "step": 3779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1034121965792703,
+      "learning_rate": 9.34359237723773e-06,
+      "loss": 0.5996,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0926090549134853,
+      "learning_rate": 9.343253731991125e-06,
+      "loss": 0.5837,
+      "step": 3781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9177156882322535,
+      "learning_rate": 9.342915006288392e-06,
+      "loss": 0.3979,
+      "step": 3782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.384780111296147,
+      "learning_rate": 9.34257620013662e-06,
+      "loss": 0.4538,
+      "step": 3783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1386881393585955,
+      "learning_rate": 9.342237313542903e-06,
+      "loss": 0.5251,
+      "step": 3784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6531900496689527,
+      "learning_rate": 9.341898346514334e-06,
+      "loss": 0.4489,
+      "step": 3785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8537681159869945,
+      "learning_rate": 9.341559299058008e-06,
+      "loss": 0.5772,
+      "step": 3786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9557445827540143,
+      "learning_rate": 9.341220171181023e-06,
+      "loss": 0.5696,
+      "step": 3787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2520878720392714,
+      "learning_rate": 9.340880962890482e-06,
+      "loss": 0.7056,
+      "step": 3788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.640414785238385,
+      "learning_rate": 9.340541674193484e-06,
+      "loss": 0.3837,
+      "step": 3789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6005756543601344,
+      "learning_rate": 9.34020230509713e-06,
+      "loss": 0.4629,
+      "step": 3790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6060257297259377,
+      "learning_rate": 9.339862855608527e-06,
+      "loss": 0.4581,
+      "step": 3791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0098887615819305,
+      "learning_rate": 9.339523325734779e-06,
+      "loss": 0.6588,
+      "step": 3792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9459870304931361,
+      "learning_rate": 9.339183715482998e-06,
+      "loss": 0.6009,
+      "step": 3793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2272516690331905,
+      "learning_rate": 9.338844024860288e-06,
+      "loss": 0.6357,
+      "step": 3794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.433647082664086,
+      "learning_rate": 9.338504253873763e-06,
+      "loss": 0.5178,
+      "step": 3795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2836440771756097,
+      "learning_rate": 9.338164402530534e-06,
+      "loss": 0.8207,
+      "step": 3796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8064519426699277,
+      "learning_rate": 9.33782447083772e-06,
+      "loss": 0.513,
+      "step": 3797
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9333414385890821,
+      "learning_rate": 9.337484458802433e-06,
+      "loss": 0.6672,
+      "step": 3798
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9788828080300078,
+      "learning_rate": 9.337144366431792e-06,
+      "loss": 0.4849,
+      "step": 3799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8896951977861416,
+      "learning_rate": 9.336804193732918e-06,
+      "loss": 0.599,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5849971175193787,
+      "eval_runtime": 67.9394,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6322506929261245,
+      "learning_rate": 9.336463940712933e-06,
+      "loss": 0.5676,
+      "step": 3801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.851707869553651,
+      "learning_rate": 9.336123607378956e-06,
+      "loss": 0.4835,
+      "step": 3802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3512100484152807,
+      "learning_rate": 9.335783193738114e-06,
+      "loss": 0.4177,
+      "step": 3803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5393608796252574,
+      "learning_rate": 9.335442699797534e-06,
+      "loss": 0.7972,
+      "step": 3804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.011083907916172,
+      "learning_rate": 9.335102125564342e-06,
+      "loss": 0.5313,
+      "step": 3805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7733657599954586,
+      "learning_rate": 9.33476147104567e-06,
+      "loss": 0.533,
+      "step": 3806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9026605092400024,
+      "learning_rate": 9.334420736248649e-06,
+      "loss": 0.3896,
+      "step": 3807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8959644972941694,
+      "learning_rate": 9.334079921180413e-06,
+      "loss": 0.5981,
+      "step": 3808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.092134204263416,
+      "learning_rate": 9.333739025848091e-06,
+      "loss": 0.6104,
+      "step": 3809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9162296925694045,
+      "learning_rate": 9.333398050258826e-06,
+      "loss": 0.5442,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8180047533272143,
+      "learning_rate": 9.333056994419757e-06,
+      "loss": 0.5509,
+      "step": 3811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.257886248402672,
+      "learning_rate": 9.332715858338015e-06,
+      "loss": 0.6338,
+      "step": 3812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6348775240515123,
+      "learning_rate": 9.332374642020751e-06,
+      "loss": 0.654,
+      "step": 3813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9834238430898832,
+      "learning_rate": 9.332033345475102e-06,
+      "loss": 0.5301,
+      "step": 3814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7857843461225598,
+      "learning_rate": 9.331691968708215e-06,
+      "loss": 0.5124,
+      "step": 3815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7420641337555383,
+      "learning_rate": 9.33135051172724e-06,
+      "loss": 0.5175,
+      "step": 3816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.046429478541367,
+      "learning_rate": 9.331008974539318e-06,
+      "loss": 0.5047,
+      "step": 3817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.929159148084434,
+      "learning_rate": 9.330667357151603e-06,
+      "loss": 0.6212,
+      "step": 3818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9284965165486485,
+      "learning_rate": 9.330325659571249e-06,
+      "loss": 0.5037,
+      "step": 3819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.498170716951437,
+      "learning_rate": 9.329983881805404e-06,
+      "loss": 0.4231,
+      "step": 3820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1648890925056326,
+      "learning_rate": 9.329642023861226e-06,
+      "loss": 0.6819,
+      "step": 3821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.954867468462907,
+      "learning_rate": 9.329300085745871e-06,
+      "loss": 0.6331,
+      "step": 3822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6835522067801016,
+      "learning_rate": 9.328958067466496e-06,
+      "loss": 0.4459,
+      "step": 3823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7492296788449644,
+      "learning_rate": 9.328615969030264e-06,
+      "loss": 0.5345,
+      "step": 3824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.903183947106352,
+      "learning_rate": 9.328273790444334e-06,
+      "loss": 0.5608,
+      "step": 3825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0854211459051797,
+      "learning_rate": 9.32793153171587e-06,
+      "loss": 0.5309,
+      "step": 3826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.206492988220673,
+      "learning_rate": 9.327589192852038e-06,
+      "loss": 0.7666,
+      "step": 3827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8314544658930472,
+      "learning_rate": 9.327246773860002e-06,
+      "loss": 0.5908,
+      "step": 3828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8107007493546623,
+      "learning_rate": 9.326904274746932e-06,
+      "loss": 0.4412,
+      "step": 3829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8344192544768367,
+      "learning_rate": 9.32656169552e-06,
+      "loss": 0.403,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.09691556025662,
+      "learning_rate": 9.326219036186373e-06,
+      "loss": 0.6356,
+      "step": 3831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7694587581349774,
+      "learning_rate": 9.325876296753229e-06,
+      "loss": 0.5433,
+      "step": 3832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.638231707394143,
+      "learning_rate": 9.32553347722774e-06,
+      "loss": 0.5031,
+      "step": 3833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8139462890275457,
+      "learning_rate": 9.325190577617086e-06,
+      "loss": 0.5273,
+      "step": 3834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.254710608989368,
+      "learning_rate": 9.324847597928438e-06,
+      "loss": 0.5454,
+      "step": 3835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9645020887917795,
+      "learning_rate": 9.324504538168986e-06,
+      "loss": 0.4589,
+      "step": 3836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0473548786898688,
+      "learning_rate": 9.324161398345905e-06,
+      "loss": 0.6236,
+      "step": 3837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.104056747660407,
+      "learning_rate": 9.323818178466382e-06,
+      "loss": 0.6436,
+      "step": 3838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9729100669211588,
+      "learning_rate": 9.323474878537598e-06,
+      "loss": 0.5593,
+      "step": 3839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8119400176629867,
+      "learning_rate": 9.323131498566742e-06,
+      "loss": 0.5484,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.748157467450038,
+      "learning_rate": 9.322788038561006e-06,
+      "loss": 0.4952,
+      "step": 3841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.74989921356216,
+      "learning_rate": 9.322444498527575e-06,
+      "loss": 0.4659,
+      "step": 3842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7924419639055267,
+      "learning_rate": 9.322100878473645e-06,
+      "loss": 0.4662,
+      "step": 3843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5644676609047439,
+      "learning_rate": 9.321757178406403e-06,
+      "loss": 0.459,
+      "step": 3844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7613170001785008,
+      "learning_rate": 9.32141339833305e-06,
+      "loss": 0.6316,
+      "step": 3845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.984686842902743,
+      "learning_rate": 9.321069538260783e-06,
+      "loss": 0.4652,
+      "step": 3846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7802199113529036,
+      "learning_rate": 9.320725598196794e-06,
+      "loss": 0.5846,
+      "step": 3847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7918782399924458,
+      "learning_rate": 9.320381578148292e-06,
+      "loss": 0.6775,
+      "step": 3848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1236073930487804,
+      "learning_rate": 9.320037478122472e-06,
+      "loss": 0.5549,
+      "step": 3849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9974147186611255,
+      "learning_rate": 9.319693298126542e-06,
+      "loss": 0.4902,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5841235518455505,
+      "eval_runtime": 67.9481,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.078959380050842,
+      "learning_rate": 9.319349038167704e-06,
+      "loss": 0.6653,
+      "step": 3851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5516685315252625,
+      "learning_rate": 9.319004698253168e-06,
+      "loss": 0.4901,
+      "step": 3852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9361102583798206,
+      "learning_rate": 9.318660278390138e-06,
+      "loss": 0.6038,
+      "step": 3853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9576213450307451,
+      "learning_rate": 9.318315778585829e-06,
+      "loss": 0.5618,
+      "step": 3854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3652278573340517,
+      "learning_rate": 9.317971198847451e-06,
+      "loss": 0.4472,
+      "step": 3855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2347174042920672,
+      "learning_rate": 9.317626539182216e-06,
+      "loss": 0.6199,
+      "step": 3856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2246909266688792,
+      "learning_rate": 9.317281799597341e-06,
+      "loss": 0.6002,
+      "step": 3857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.728938391290499,
+      "learning_rate": 9.316936980100043e-06,
+      "loss": 0.4788,
+      "step": 3858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.642679815547786,
+      "learning_rate": 9.31659208069754e-06,
+      "loss": 0.4057,
+      "step": 3859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7794055536797162,
+      "learning_rate": 9.316247101397053e-06,
+      "loss": 0.52,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8496958283516385,
+      "learning_rate": 9.315902042205802e-06,
+      "loss": 0.5793,
+      "step": 3861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8610198616942548,
+      "learning_rate": 9.315556903131012e-06,
+      "loss": 0.6006,
+      "step": 3862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4192960879112757,
+      "learning_rate": 9.31521168417991e-06,
+      "loss": 0.7284,
+      "step": 3863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9170829880353717,
+      "learning_rate": 9.314866385359718e-06,
+      "loss": 0.6681,
+      "step": 3864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4512031125931422,
+      "learning_rate": 9.314521006677668e-06,
+      "loss": 0.5204,
+      "step": 3865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5720182885830085,
+      "learning_rate": 9.314175548140992e-06,
+      "loss": 0.6001,
+      "step": 3866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.469631712048848,
+      "learning_rate": 9.313830009756921e-06,
+      "loss": 0.3703,
+      "step": 3867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0766112759504907,
+      "learning_rate": 9.313484391532685e-06,
+      "loss": 0.5655,
+      "step": 3868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0252694205239257,
+      "learning_rate": 9.313138693475523e-06,
+      "loss": 0.5844,
+      "step": 3869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1422099262682304,
+      "learning_rate": 9.312792915592671e-06,
+      "loss": 0.6904,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3304537792931725,
+      "learning_rate": 9.312447057891369e-06,
+      "loss": 0.7431,
+      "step": 3871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4379592875680225,
+      "learning_rate": 9.312101120378853e-06,
+      "loss": 0.6364,
+      "step": 3872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7704977761153404,
+      "learning_rate": 9.311755103062368e-06,
+      "loss": 0.5143,
+      "step": 3873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1255279752106593,
+      "learning_rate": 9.311409005949158e-06,
+      "loss": 0.4529,
+      "step": 3874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8612504253155613,
+      "learning_rate": 9.31106282904647e-06,
+      "loss": 0.573,
+      "step": 3875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.83870068055082,
+      "learning_rate": 9.310716572361546e-06,
+      "loss": 0.5649,
+      "step": 3876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.297400334465539,
+      "learning_rate": 9.310370235901638e-06,
+      "loss": 0.4924,
+      "step": 3877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.900488773128254,
+      "learning_rate": 9.310023819673997e-06,
+      "loss": 0.4238,
+      "step": 3878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1154713118825463,
+      "learning_rate": 9.309677323685872e-06,
+      "loss": 0.5968,
+      "step": 3879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6498031052530175,
+      "learning_rate": 9.30933074794452e-06,
+      "loss": 0.4657,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0289401191138827,
+      "learning_rate": 9.308984092457195e-06,
+      "loss": 0.5833,
+      "step": 3881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.055425471764517,
+      "learning_rate": 9.308637357231154e-06,
+      "loss": 0.5568,
+      "step": 3882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5913664208746297,
+      "learning_rate": 9.308290542273655e-06,
+      "loss": 0.6911,
+      "step": 3883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0102911134927597,
+      "learning_rate": 9.307943647591959e-06,
+      "loss": 0.7137,
+      "step": 3884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0315067249883927,
+      "learning_rate": 9.307596673193326e-06,
+      "loss": 0.6239,
+      "step": 3885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6393277826157466,
+      "learning_rate": 9.307249619085024e-06,
+      "loss": 0.4939,
+      "step": 3886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8502401071501264,
+      "learning_rate": 9.306902485274318e-06,
+      "loss": 0.4535,
+      "step": 3887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7062360940243864,
+      "learning_rate": 9.30655527176847e-06,
+      "loss": 0.501,
+      "step": 3888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7969561110004855,
+      "learning_rate": 9.30620797857475e-06,
+      "loss": 0.5058,
+      "step": 3889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9788184936182402,
+      "learning_rate": 9.305860605700434e-06,
+      "loss": 0.7069,
+      "step": 3890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.465033557682648,
+      "learning_rate": 9.30551315315279e-06,
+      "loss": 0.8128,
+      "step": 3891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.915862688002775,
+      "learning_rate": 9.305165620939089e-06,
+      "loss": 0.54,
+      "step": 3892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1367680562287306,
+      "learning_rate": 9.304818009066609e-06,
+      "loss": 0.5739,
+      "step": 3893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0672486100000755,
+      "learning_rate": 9.304470317542628e-06,
+      "loss": 0.5765,
+      "step": 3894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1485522570382005,
+      "learning_rate": 9.304122546374425e-06,
+      "loss": 0.6703,
+      "step": 3895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.063720882758252,
+      "learning_rate": 9.30377469556928e-06,
+      "loss": 0.6528,
+      "step": 3896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4890265405815823,
+      "learning_rate": 9.303426765134473e-06,
+      "loss": 0.5309,
+      "step": 3897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3558161558743014,
+      "learning_rate": 9.30307875507729e-06,
+      "loss": 0.6347,
+      "step": 3898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8615080462747053,
+      "learning_rate": 9.302730665405014e-06,
+      "loss": 0.5518,
+      "step": 3899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9442627618546247,
+      "learning_rate": 9.302382496124934e-06,
+      "loss": 0.5747,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5831843614578247,
+      "eval_runtime": 67.935,
+      "eval_samples_per_second": 50.195,
+      "eval_steps_per_second": 1.575,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7558361658316919,
+      "learning_rate": 9.30203424724434e-06,
+      "loss": 0.5526,
+      "step": 3901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4659739506525153,
+      "learning_rate": 9.301685918770517e-06,
+      "loss": 0.4726,
+      "step": 3902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.797034338626684,
+      "learning_rate": 9.301337510710763e-06,
+      "loss": 0.5446,
+      "step": 3903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8207523083218606,
+      "learning_rate": 9.30098902307237e-06,
+      "loss": 0.5325,
+      "step": 3904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7021750011453711,
+      "learning_rate": 9.30064045586263e-06,
+      "loss": 0.5494,
+      "step": 3905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8217873322946616,
+      "learning_rate": 9.300291809088846e-06,
+      "loss": 0.6836,
+      "step": 3906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.39364199297935,
+      "learning_rate": 9.299943082758312e-06,
+      "loss": 0.5541,
+      "step": 3907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6026650193981136,
+      "learning_rate": 9.29959427687833e-06,
+      "loss": 0.4468,
+      "step": 3908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.223198201953757,
+      "learning_rate": 9.2992453914562e-06,
+      "loss": 0.6032,
+      "step": 3909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6847272129514452,
+      "learning_rate": 9.298896426499231e-06,
+      "loss": 0.4427,
+      "step": 3910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8291530386446755,
+      "learning_rate": 9.298547382014722e-06,
+      "loss": 0.579,
+      "step": 3911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4833449519493456,
+      "learning_rate": 9.298198258009983e-06,
+      "loss": 0.3939,
+      "step": 3912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7528462968953857,
+      "learning_rate": 9.297849054492323e-06,
+      "loss": 0.4667,
+      "step": 3913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5283217666556095,
+      "learning_rate": 9.297499771469053e-06,
+      "loss": 0.3832,
+      "step": 3914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5746283222359878,
+      "learning_rate": 9.297150408947484e-06,
+      "loss": 0.5421,
+      "step": 3915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8467747338453988,
+      "learning_rate": 9.296800966934929e-06,
+      "loss": 0.7001,
+      "step": 3916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7884462001498547,
+      "learning_rate": 9.296451445438705e-06,
+      "loss": 0.5019,
+      "step": 3917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4750273779899534,
+      "learning_rate": 9.296101844466126e-06,
+      "loss": 0.6933,
+      "step": 3918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.097352928466962,
+      "learning_rate": 9.295752164024515e-06,
+      "loss": 0.5063,
+      "step": 3919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.067740356412659,
+      "learning_rate": 9.295402404121187e-06,
+      "loss": 0.6129,
+      "step": 3920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0409297391572614,
+      "learning_rate": 9.29505256476347e-06,
+      "loss": 0.5701,
+      "step": 3921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.078083220840903,
+      "learning_rate": 9.294702645958682e-06,
+      "loss": 0.6761,
+      "step": 3922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.919992302047224,
+      "learning_rate": 9.294352647714154e-06,
+      "loss": 0.4979,
+      "step": 3923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.44966465515875,
+      "learning_rate": 9.29400257003721e-06,
+      "loss": 0.6403,
+      "step": 3924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9937888498535052,
+      "learning_rate": 9.293652412935176e-06,
+      "loss": 0.6229,
+      "step": 3925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7781999737507252,
+      "learning_rate": 9.293302176415389e-06,
+      "loss": 0.4574,
+      "step": 3926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8087801187324295,
+      "learning_rate": 9.292951860485174e-06,
+      "loss": 0.5031,
+      "step": 3927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3469570491614684,
+      "learning_rate": 9.292601465151867e-06,
+      "loss": 0.5987,
+      "step": 3928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.199694294053787,
+      "learning_rate": 9.292250990422807e-06,
+      "loss": 0.5994,
+      "step": 3929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1120663206028305,
+      "learning_rate": 9.291900436305326e-06,
+      "loss": 0.6976,
+      "step": 3930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.442657658868948,
+      "learning_rate": 9.291549802806767e-06,
+      "loss": 0.523,
+      "step": 3931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.828701800778058,
+      "learning_rate": 9.291199089934466e-06,
+      "loss": 0.586,
+      "step": 3932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8147191462038121,
+      "learning_rate": 9.290848297695766e-06,
+      "loss": 0.6475,
+      "step": 3933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7126677770743974,
+      "learning_rate": 9.290497426098012e-06,
+      "loss": 0.3959,
+      "step": 3934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.930643995994619,
+      "learning_rate": 9.290146475148548e-06,
+      "loss": 0.5076,
+      "step": 3935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9877501245026472,
+      "learning_rate": 9.289795444854723e-06,
+      "loss": 0.5394,
+      "step": 3936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0023464458265954,
+      "learning_rate": 9.289444335223883e-06,
+      "loss": 0.6214,
+      "step": 3937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.281091414997116,
+      "learning_rate": 9.28909314626338e-06,
+      "loss": 0.5275,
+      "step": 3938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8104908351787674,
+      "learning_rate": 9.288741877980568e-06,
+      "loss": 0.5497,
+      "step": 3939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.570556572913583,
+      "learning_rate": 9.288390530382797e-06,
+      "loss": 0.5817,
+      "step": 3940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6200949939115799,
+      "learning_rate": 9.288039103477422e-06,
+      "loss": 0.5197,
+      "step": 3941
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6756386396051521,
+      "learning_rate": 9.2876875972718e-06,
+      "loss": 0.4564,
+      "step": 3942
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.464839086143698,
+      "learning_rate": 9.287336011773294e-06,
+      "loss": 0.605,
+      "step": 3943
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1823032015401562,
+      "learning_rate": 9.286984346989258e-06,
+      "loss": 0.4764,
+      "step": 3944
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5765885440076928,
+      "learning_rate": 9.286632602927059e-06,
+      "loss": 0.7884,
+      "step": 3945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.817365588135969,
+      "learning_rate": 9.286280779594057e-06,
+      "loss": 0.5698,
+      "step": 3946
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0176078207159107,
+      "learning_rate": 9.285928876997619e-06,
+      "loss": 0.5666,
+      "step": 3947
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8346044860974524,
+      "learning_rate": 9.285576895145112e-06,
+      "loss": 0.5285,
+      "step": 3948
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.946243123558707,
+      "learning_rate": 9.285224834043905e-06,
+      "loss": 0.5595,
+      "step": 3949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1849465013435507,
+      "learning_rate": 9.284872693701366e-06,
+      "loss": 0.6994,
+      "step": 3950
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5822182297706604,
+      "eval_runtime": 67.9457,
+      "eval_samples_per_second": 50.187,
+      "eval_steps_per_second": 1.575,
+      "step": 3950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.956563781731029,
+      "learning_rate": 9.284520474124868e-06,
+      "loss": 0.528,
+      "step": 3951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3726025442368677,
+      "learning_rate": 9.284168175321786e-06,
+      "loss": 0.6049,
+      "step": 3952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8974130723864935,
+      "learning_rate": 9.283815797299491e-06,
+      "loss": 0.5616,
+      "step": 3953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0630135128392246,
+      "learning_rate": 9.283463340065364e-06,
+      "loss": 0.5835,
+      "step": 3954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2140926811312927,
+      "learning_rate": 9.283110803626781e-06,
+      "loss": 0.5922,
+      "step": 3955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.179790345973615,
+      "learning_rate": 9.282758187991123e-06,
+      "loss": 0.5317,
+      "step": 3956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8105200011327023,
+      "learning_rate": 9.282405493165773e-06,
+      "loss": 0.5005,
+      "step": 3957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7625136648371558,
+      "learning_rate": 9.282052719158114e-06,
+      "loss": 0.4799,
+      "step": 3958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7654687935496853,
+      "learning_rate": 9.281699865975528e-06,
+      "loss": 0.513,
+      "step": 3959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.593225175080379,
+      "learning_rate": 9.281346933625406e-06,
+      "loss": 0.4598,
+      "step": 3960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.716829584204485,
+      "learning_rate": 9.280993922115136e-06,
+      "loss": 0.577,
+      "step": 3961
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0228770993084484,
+      "learning_rate": 9.280640831452103e-06,
+      "loss": 0.6217,
+      "step": 3962
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0036470652970353,
+      "learning_rate": 9.280287661643706e-06,
+      "loss": 0.5655,
+      "step": 3963
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.638209824143626,
+      "learning_rate": 9.279934412697334e-06,
+      "loss": 0.5564,
+      "step": 3964
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.118592406466575,
+      "learning_rate": 9.27958108462038e-06,
+      "loss": 0.6101,
+      "step": 3965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9065263462046123,
+      "learning_rate": 9.279227677420246e-06,
+      "loss": 0.4793,
+      "step": 3966
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6313007101136274,
+      "learning_rate": 9.278874191104328e-06,
+      "loss": 0.4733,
+      "step": 3967
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7350977414193784,
+      "learning_rate": 9.278520625680027e-06,
+      "loss": 0.5378,
+      "step": 3968
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8739411716214642,
+      "learning_rate": 9.278166981154742e-06,
+      "loss": 0.579,
+      "step": 3969
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0622288825650936,
+      "learning_rate": 9.277813257535877e-06,
+      "loss": 0.7848,
+      "step": 3970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7486974597736908,
+      "learning_rate": 9.27745945483084e-06,
+      "loss": 0.5892,
+      "step": 3971
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9501407148143286,
+      "learning_rate": 9.277105573047033e-06,
+      "loss": 0.3954,
+      "step": 3972
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5555268397854756,
+      "learning_rate": 9.27675161219187e-06,
+      "loss": 0.5339,
+      "step": 3973
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9302739208724813,
+      "learning_rate": 9.276397572272756e-06,
+      "loss": 0.6522,
+      "step": 3974
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6329256728304418,
+      "learning_rate": 9.276043453297105e-06,
+      "loss": 0.6226,
+      "step": 3975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9583211991723086,
+      "learning_rate": 9.27568925527233e-06,
+      "loss": 0.6457,
+      "step": 3976
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7018126613760691,
+      "learning_rate": 9.275334978205846e-06,
+      "loss": 0.5119,
+      "step": 3977
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7871221379569309,
+      "learning_rate": 9.274980622105066e-06,
+      "loss": 0.4953,
+      "step": 3978
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0225199603415915,
+      "learning_rate": 9.274626186977416e-06,
+      "loss": 0.4317,
+      "step": 3979
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.405962966596497,
+      "learning_rate": 9.274271672830307e-06,
+      "loss": 0.6757,
+      "step": 3980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.405851601722218,
+      "learning_rate": 9.273917079671168e-06,
+      "loss": 0.744,
+      "step": 3981
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9460607151435334,
+      "learning_rate": 9.273562407507417e-06,
+      "loss": 0.5787,
+      "step": 3982
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.403235959838248,
+      "learning_rate": 9.273207656346482e-06,
+      "loss": 0.7359,
+      "step": 3983
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6598131502230384,
+      "learning_rate": 9.272852826195789e-06,
+      "loss": 0.5048,
+      "step": 3984
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.858297573647229,
+      "learning_rate": 9.272497917062763e-06,
+      "loss": 0.5057,
+      "step": 3985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3951855840489746,
+      "learning_rate": 9.272142928954839e-06,
+      "loss": 0.6831,
+      "step": 3986
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.094618540386623,
+      "learning_rate": 9.271787861879444e-06,
+      "loss": 0.5441,
+      "step": 3987
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6894576986776997,
+      "learning_rate": 9.271432715844013e-06,
+      "loss": 0.5987,
+      "step": 3988
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.922924775342951,
+      "learning_rate": 9.271077490855981e-06,
+      "loss": 0.5803,
+      "step": 3989
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1703577267305008,
+      "learning_rate": 9.270722186922783e-06,
+      "loss": 0.791,
+      "step": 3990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.498976662605158,
+      "learning_rate": 9.270366804051862e-06,
+      "loss": 0.638,
+      "step": 3991
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9095947218457325,
+      "learning_rate": 9.270011342250648e-06,
+      "loss": 0.4121,
+      "step": 3992
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.971291870280982,
+      "learning_rate": 9.269655801526594e-06,
+      "loss": 0.5498,
+      "step": 3993
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7989571170791674,
+      "learning_rate": 9.269300181887133e-06,
+      "loss": 0.4638,
+      "step": 3994
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8331647626470722,
+      "learning_rate": 9.268944483339714e-06,
+      "loss": 0.5153,
+      "step": 3995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.01284467052279,
+      "learning_rate": 9.268588705891786e-06,
+      "loss": 0.523,
+      "step": 3996
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4085007392097597,
+      "learning_rate": 9.268232849550793e-06,
+      "loss": 0.7366,
+      "step": 3997
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7660837270611076,
+      "learning_rate": 9.267876914324185e-06,
+      "loss": 0.5386,
+      "step": 3998
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8024648717702951,
+      "learning_rate": 9.267520900219415e-06,
+      "loss": 0.5406,
+      "step": 3999
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.550632509830378,
+      "learning_rate": 9.267164807243935e-06,
+      "loss": 0.5263,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5831869840621948,
+      "eval_runtime": 67.9113,
+      "eval_samples_per_second": 50.213,
+      "eval_steps_per_second": 1.576,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.953997852834892,
+      "learning_rate": 9.266808635405199e-06,
+      "loss": 0.5698,
+      "step": 4001
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3014245569985325,
+      "learning_rate": 9.266452384710664e-06,
+      "loss": 0.5116,
+      "step": 4002
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.112769816702147,
+      "learning_rate": 9.266096055167788e-06,
+      "loss": 0.5178,
+      "step": 4003
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.868635359631934,
+      "learning_rate": 9.26573964678403e-06,
+      "loss": 0.4985,
+      "step": 4004
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.026325322983133,
+      "learning_rate": 9.265383159566851e-06,
+      "loss": 0.608,
+      "step": 4005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7706717685040603,
+      "learning_rate": 9.265026593523718e-06,
+      "loss": 0.5304,
+      "step": 4006
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8441192108635063,
+      "learning_rate": 9.264669948662086e-06,
+      "loss": 0.5984,
+      "step": 4007
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8883930992510405,
+      "learning_rate": 9.264313224989431e-06,
+      "loss": 0.5096,
+      "step": 4008
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9875746666806784,
+      "learning_rate": 9.263956422513215e-06,
+      "loss": 0.7428,
+      "step": 4009
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2523070459221297,
+      "learning_rate": 9.26359954124091e-06,
+      "loss": 0.655,
+      "step": 4010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8570493613307468,
+      "learning_rate": 9.263242581179986e-06,
+      "loss": 0.513,
+      "step": 4011
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.664774378740834,
+      "learning_rate": 9.262885542337916e-06,
+      "loss": 0.3733,
+      "step": 4012
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8025565714589755,
+      "learning_rate": 9.262528424722174e-06,
+      "loss": 0.3846,
+      "step": 4013
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8931890529119224,
+      "learning_rate": 9.262171228340236e-06,
+      "loss": 0.5366,
+      "step": 4014
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1966053584909644,
+      "learning_rate": 9.261813953199581e-06,
+      "loss": 0.4992,
+      "step": 4015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.961238377837069,
+      "learning_rate": 9.261456599307688e-06,
+      "loss": 0.5124,
+      "step": 4016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9571804055652282,
+      "learning_rate": 9.261099166672038e-06,
+      "loss": 0.6053,
+      "step": 4017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.934058954619608,
+      "learning_rate": 9.26074165530011e-06,
+      "loss": 0.5202,
+      "step": 4018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0162997467848536,
+      "learning_rate": 9.260384065199394e-06,
+      "loss": 0.4454,
+      "step": 4019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2019944238331277,
+      "learning_rate": 9.260026396377373e-06,
+      "loss": 0.4828,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8472853066535753,
+      "learning_rate": 9.259668648841533e-06,
+      "loss": 0.4736,
+      "step": 4021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5724707935980866,
+      "learning_rate": 9.259310822599367e-06,
+      "loss": 0.5125,
+      "step": 4022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.591668289510081,
+      "learning_rate": 9.258952917658362e-06,
+      "loss": 0.548,
+      "step": 4023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.139207102335303,
+      "learning_rate": 9.258594934026014e-06,
+      "loss": 0.8064,
+      "step": 4024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.097710209941526,
+      "learning_rate": 9.258236871709816e-06,
+      "loss": 0.6205,
+      "step": 4025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5381796863953818,
+      "learning_rate": 9.257878730717263e-06,
+      "loss": 0.5116,
+      "step": 4026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8114882375722852,
+      "learning_rate": 9.257520511055852e-06,
+      "loss": 0.6014,
+      "step": 4027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.089317520328327,
+      "learning_rate": 9.257162212733083e-06,
+      "loss": 0.4913,
+      "step": 4028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1849511387352565,
+      "learning_rate": 9.256803835756458e-06,
+      "loss": 0.658,
+      "step": 4029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6301017982268906,
+      "learning_rate": 9.256445380133477e-06,
+      "loss": 0.6025,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9011616691410609,
+      "learning_rate": 9.256086845871646e-06,
+      "loss": 0.5059,
+      "step": 4031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4135435697865892,
+      "learning_rate": 9.25572823297847e-06,
+      "loss": 0.6699,
+      "step": 4032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.151373706788335,
+      "learning_rate": 9.255369541461456e-06,
+      "loss": 0.5889,
+      "step": 4033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4930123931968964,
+      "learning_rate": 9.255010771328111e-06,
+      "loss": 0.4417,
+      "step": 4034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9821330141718423,
+      "learning_rate": 9.25465192258595e-06,
+      "loss": 0.6946,
+      "step": 4035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5689003689311358,
+      "learning_rate": 9.254292995242483e-06,
+      "loss": 0.3756,
+      "step": 4036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8462882352174679,
+      "learning_rate": 9.253933989305224e-06,
+      "loss": 0.6812,
+      "step": 4037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6628471265636922,
+      "learning_rate": 9.253574904781688e-06,
+      "loss": 0.4825,
+      "step": 4038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8686769934576164,
+      "learning_rate": 9.253215741679393e-06,
+      "loss": 0.5771,
+      "step": 4039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.633039649770832,
+      "learning_rate": 9.252856500005858e-06,
+      "loss": 0.6079,
+      "step": 4040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4452801375122568,
+      "learning_rate": 9.252497179768604e-06,
+      "loss": 0.3992,
+      "step": 4041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7005107158943298,
+      "learning_rate": 9.252137780975152e-06,
+      "loss": 0.4894,
+      "step": 4042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9892375488165146,
+      "learning_rate": 9.251778303633024e-06,
+      "loss": 0.5884,
+      "step": 4043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3706166953346157,
+      "learning_rate": 9.251418747749749e-06,
+      "loss": 0.7674,
+      "step": 4044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9856818741780344,
+      "learning_rate": 9.251059113332854e-06,
+      "loss": 0.7123,
+      "step": 4045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7426256457851075,
+      "learning_rate": 9.250699400389864e-06,
+      "loss": 0.5436,
+      "step": 4046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8472027461996579,
+      "learning_rate": 9.250339608928313e-06,
+      "loss": 0.564,
+      "step": 4047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7786725695376404,
+      "learning_rate": 9.24997973895573e-06,
+      "loss": 0.5524,
+      "step": 4048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7747600553610983,
+      "learning_rate": 9.249619790479651e-06,
+      "loss": 0.5076,
+      "step": 4049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7269355003218327,
+      "learning_rate": 9.249259763507614e-06,
+      "loss": 0.4524,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.5835996866226196,
+      "eval_runtime": 67.8959,
+      "eval_samples_per_second": 50.224,
+      "eval_steps_per_second": 1.576,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.756255863358974,
+      "learning_rate": 9.248899658047148e-06,
+      "loss": 0.5195,
+      "step": 4051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1889227569087537,
+      "learning_rate": 9.248539474105797e-06,
+      "loss": 0.7202,
+      "step": 4052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9781555523086902,
+      "learning_rate": 9.248179211691103e-06,
+      "loss": 0.5382,
+      "step": 4053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.855807415959085,
+      "learning_rate": 9.247818870810603e-06,
+      "loss": 0.4721,
+      "step": 4054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.115590527626326,
+      "learning_rate": 9.247458451471843e-06,
+      "loss": 0.6514,
+      "step": 4055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.510219084014672,
+      "learning_rate": 9.24709795368237e-06,
+      "loss": 0.6202,
+      "step": 4056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8580780425145151,
+      "learning_rate": 9.246737377449727e-06,
+      "loss": 0.6266,
+      "step": 4057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.054988009598793,
+      "learning_rate": 9.246376722781463e-06,
+      "loss": 0.5637,
+      "step": 4058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.682842217502365,
+      "learning_rate": 9.246015989685132e-06,
+      "loss": 0.4382,
+      "step": 4059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.051317634273733,
+      "learning_rate": 9.245655178168282e-06,
+      "loss": 0.5717,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8720051040483725,
+      "learning_rate": 9.245294288238466e-06,
+      "loss": 0.5996,
+      "step": 4061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8336352332898125,
+      "learning_rate": 9.244933319903242e-06,
+      "loss": 0.429,
+      "step": 4062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.512107465080195,
+      "learning_rate": 9.244572273170165e-06,
+      "loss": 0.7646,
+      "step": 4063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.124957305768645,
+      "learning_rate": 9.244211148046794e-06,
+      "loss": 0.6801,
+      "step": 4064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.956843050487664,
+      "learning_rate": 9.243849944540687e-06,
+      "loss": 0.4966,
+      "step": 4065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0884034252761947,
+      "learning_rate": 9.243488662659405e-06,
+      "loss": 0.5858,
+      "step": 4066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7718648506387176,
+      "learning_rate": 9.243127302410515e-06,
+      "loss": 0.5278,
+      "step": 4067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2892329019882993,
+      "learning_rate": 9.24276586380158e-06,
+      "loss": 0.6758,
+      "step": 4068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0060957719941777,
+      "learning_rate": 9.242404346840165e-06,
+      "loss": 0.5937,
+      "step": 4069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.504491965372331,
+      "learning_rate": 9.242042751533842e-06,
+      "loss": 0.5857,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.390556940496449,
+      "learning_rate": 9.241681077890175e-06,
+      "loss": 0.5462,
+      "step": 4071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0803446216641985,
+      "learning_rate": 9.24131932591674e-06,
+      "loss": 0.5311,
+      "step": 4072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.852264774879428,
+      "learning_rate": 9.240957495621111e-06,
+      "loss": 0.6639,
+      "step": 4073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.780229049431932,
+      "learning_rate": 9.240595587010858e-06,
+      "loss": 0.4886,
+      "step": 4074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6182478774757196,
+      "learning_rate": 9.240233600093562e-06,
+      "loss": 0.583,
+      "step": 4075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1163948302273434,
+      "learning_rate": 9.239871534876797e-06,
+      "loss": 0.782,
+      "step": 4076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.184584642053287,
+      "learning_rate": 9.239509391368143e-06,
+      "loss": 0.5469,
+      "step": 4077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9601563169153737,
+      "learning_rate": 9.239147169575184e-06,
+      "loss": 0.6469,
+      "step": 4078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.17579422810534,
+      "learning_rate": 9.238784869505502e-06,
+      "loss": 0.6285,
+      "step": 4079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8294233998678633,
+      "learning_rate": 9.23842249116668e-06,
+      "loss": 0.5782,
+      "step": 4080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4975802432341092,
+      "learning_rate": 9.238060034566308e-06,
+      "loss": 0.451,
+      "step": 4081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8941898025348836,
+      "learning_rate": 9.237697499711966e-06,
+      "loss": 0.6154,
+      "step": 4082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9081590686562289,
+      "learning_rate": 9.237334886611253e-06,
+      "loss": 0.4828,
+      "step": 4083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2003599182386013,
+      "learning_rate": 9.236972195271753e-06,
+      "loss": 0.5939,
+      "step": 4084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0722508107482005,
+      "learning_rate": 9.23660942570106e-06,
+      "loss": 0.4932,
+      "step": 4085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.061928693405777,
+      "learning_rate": 9.236246577906772e-06,
+      "loss": 0.802,
+      "step": 4086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.16142433363546,
+      "learning_rate": 9.235883651896483e-06,
+      "loss": 0.5461,
+      "step": 4087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9592807465569715,
+      "learning_rate": 9.23552064767779e-06,
+      "loss": 0.5381,
+      "step": 4088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7627879785094085,
+      "learning_rate": 9.235157565258291e-06,
+      "loss": 0.6041,
+      "step": 4089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7435711038245307,
+      "learning_rate": 9.234794404645588e-06,
+      "loss": 0.4167,
+      "step": 4090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2051032297855713,
+      "learning_rate": 9.234431165847285e-06,
+      "loss": 0.6156,
+      "step": 4091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.306174388488265,
+      "learning_rate": 9.234067848870985e-06,
+      "loss": 0.6067,
+      "step": 4092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6965247126690053,
+      "learning_rate": 9.233704453724293e-06,
+      "loss": 0.4708,
+      "step": 4093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0627817322239714,
+      "learning_rate": 9.23334098041482e-06,
+      "loss": 0.5531,
+      "step": 4094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7563193283254326,
+      "learning_rate": 9.232977428950168e-06,
+      "loss": 0.6025,
+      "step": 4095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7180291261307028,
+      "learning_rate": 9.232613799337956e-06,
+      "loss": 0.5057,
+      "step": 4096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2779100365626714,
+      "learning_rate": 9.232250091585791e-06,
+      "loss": 0.6129,
+      "step": 4097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.451063537408051,
+      "learning_rate": 9.23188630570129e-06,
+      "loss": 0.7856,
+      "step": 4098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.841666255885023,
+      "learning_rate": 9.231522441692064e-06,
+      "loss": 0.8431,
+      "step": 4099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0827710687652194,
+      "learning_rate": 9.231158499565737e-06,
+      "loss": 0.6562,
+      "step": 4100
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.5820208787918091,
+      "eval_runtime": 67.9593,
+      "eval_samples_per_second": 50.177,
+      "eval_steps_per_second": 1.574,
+      "step": 4100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2451338512935344,
+      "learning_rate": 9.230794479329922e-06,
+      "loss": 0.6908,
+      "step": 4101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9611002696297808,
+      "learning_rate": 9.230430380992243e-06,
+      "loss": 0.4541,
+      "step": 4102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7437228948249668,
+      "learning_rate": 9.23006620456032e-06,
+      "loss": 0.644,
+      "step": 4103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.243686631677916,
+      "learning_rate": 9.229701950041778e-06,
+      "loss": 0.6458,
+      "step": 4104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9614741613352926,
+      "learning_rate": 9.229337617444243e-06,
+      "loss": 0.4622,
+      "step": 4105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.400155655434013,
+      "learning_rate": 9.22897320677534e-06,
+      "loss": 0.6299,
+      "step": 4106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2073231506147177,
+      "learning_rate": 9.228608718042701e-06,
+      "loss": 0.702,
+      "step": 4107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7521101183550147,
+      "learning_rate": 9.228244151253953e-06,
+      "loss": 0.579,
+      "step": 4108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.332648093074497,
+      "learning_rate": 9.227879506416731e-06,
+      "loss": 0.6573,
+      "step": 4109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.09537694780407,
+      "learning_rate": 9.227514783538666e-06,
+      "loss": 0.5057,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8940913886790813,
+      "learning_rate": 9.227149982627395e-06,
+      "loss": 0.7054,
+      "step": 4111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.049627391388569,
+      "learning_rate": 9.226785103690551e-06,
+      "loss": 0.6421,
+      "step": 4112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9406335213768282,
+      "learning_rate": 9.22642014673578e-06,
+      "loss": 0.5377,
+      "step": 4113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.272869445650475,
+      "learning_rate": 9.226055111770715e-06,
+      "loss": 0.6624,
+      "step": 4114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0376445029073045,
+      "learning_rate": 9.225689998803002e-06,
+      "loss": 0.56,
+      "step": 4115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7529871734015003,
+      "learning_rate": 9.22532480784028e-06,
+      "loss": 0.4534,
+      "step": 4116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.004846696021747,
+      "learning_rate": 9.224959538890201e-06,
+      "loss": 0.7623,
+      "step": 4117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0451452988743517,
+      "learning_rate": 9.224594191960405e-06,
+      "loss": 0.6092,
+      "step": 4118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.833783314599413,
+      "learning_rate": 9.224228767058545e-06,
+      "loss": 0.5483,
+      "step": 4119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6532390363266734,
+      "learning_rate": 9.223863264192268e-06,
+      "loss": 0.6231,
+      "step": 4120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8736333144889088,
+      "learning_rate": 9.223497683369227e-06,
+      "loss": 0.5921,
+      "step": 4121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2650730456737063,
+      "learning_rate": 9.223132024597075e-06,
+      "loss": 0.5376,
+      "step": 4122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5735587473675086,
+      "learning_rate": 9.222766287883464e-06,
+      "loss": 0.5703,
+      "step": 4123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.182283606403867,
+      "learning_rate": 9.222400473236056e-06,
+      "loss": 0.6018,
+      "step": 4124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7865678464586583,
+      "learning_rate": 9.222034580662506e-06,
+      "loss": 0.4157,
+      "step": 4125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9565903256953672,
+      "learning_rate": 9.221668610170473e-06,
+      "loss": 0.4976,
+      "step": 4126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.084308186916414,
+      "learning_rate": 9.221302561767619e-06,
+      "loss": 0.7618,
+      "step": 4127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9637367498148828,
+      "learning_rate": 9.220936435461605e-06,
+      "loss": 0.5964,
+      "step": 4128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6534166107748653,
+      "learning_rate": 9.220570231260102e-06,
+      "loss": 0.593,
+      "step": 4129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5641208614832782,
+      "learning_rate": 9.22020394917077e-06,
+      "loss": 0.4675,
+      "step": 4130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.627534306764585,
+      "learning_rate": 9.219837589201282e-06,
+      "loss": 0.6712,
+      "step": 4131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.857893884766149,
+      "learning_rate": 9.219471151359302e-06,
+      "loss": 0.5041,
+      "step": 4132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9669210522508709,
+      "learning_rate": 9.219104635652503e-06,
+      "loss": 0.5796,
+      "step": 4133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9456100141785426,
+      "learning_rate": 9.21873804208856e-06,
+      "loss": 0.4704,
+      "step": 4134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.70578061132491,
+      "learning_rate": 9.218371370675144e-06,
+      "loss": 0.5346,
+      "step": 4135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7678983462239726,
+      "learning_rate": 9.218004621419935e-06,
+      "loss": 0.5661,
+      "step": 4136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9287428304332421,
+      "learning_rate": 9.217637794330606e-06,
+      "loss": 0.6581,
+      "step": 4137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6265113832105464,
+      "learning_rate": 9.21727088941484e-06,
+      "loss": 0.464,
+      "step": 4138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2469712334186216,
+      "learning_rate": 9.21690390668032e-06,
+      "loss": 0.4799,
+      "step": 4139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.280770404960507,
+      "learning_rate": 9.216536846134722e-06,
+      "loss": 0.54,
+      "step": 4140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0421639715767723,
+      "learning_rate": 9.216169707785733e-06,
+      "loss": 0.5721,
+      "step": 4141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1553916968847955,
+      "learning_rate": 9.215802491641041e-06,
+      "loss": 0.7502,
+      "step": 4142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8573857035744847,
+      "learning_rate": 9.215435197708333e-06,
+      "loss": 0.4677,
+      "step": 4143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7670932591592061,
+      "learning_rate": 9.215067825995294e-06,
+      "loss": 0.5137,
+      "step": 4144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.166264789964711,
+      "learning_rate": 9.21470037650962e-06,
+      "loss": 0.5943,
+      "step": 4145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8827147296949647,
+      "learning_rate": 9.214332849258999e-06,
+      "loss": 0.4741,
+      "step": 4146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0186118518140326,
+      "learning_rate": 9.213965244251128e-06,
+      "loss": 0.5541,
+      "step": 4147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.853595799541421,
+      "learning_rate": 9.213597561493702e-06,
+      "loss": 0.6153,
+      "step": 4148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.956509080698889,
+      "learning_rate": 9.213229800994417e-06,
+      "loss": 0.5828,
+      "step": 4149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7626655988994222,
+      "learning_rate": 9.212861962760971e-06,
+      "loss": 0.5451,
+      "step": 4150
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.5826732516288757,
+      "eval_runtime": 67.9555,
+      "eval_samples_per_second": 50.18,
+      "eval_steps_per_second": 1.575,
+      "step": 4150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.478860388331222,
+      "learning_rate": 9.212494046801068e-06,
+      "loss": 0.6487,
+      "step": 4151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.138110632961059,
+      "learning_rate": 9.212126053122407e-06,
+      "loss": 0.6112,
+      "step": 4152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8785975152372805,
+      "learning_rate": 9.211757981732693e-06,
+      "loss": 0.5762,
+      "step": 4153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.998715246289021,
+      "learning_rate": 9.211389832639631e-06,
+      "loss": 0.7402,
+      "step": 4154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.041491486395452,
+      "learning_rate": 9.211021605850928e-06,
+      "loss": 0.5742,
+      "step": 4155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.027865488813116,
+      "learning_rate": 9.210653301374293e-06,
+      "loss": 0.5881,
+      "step": 4156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7271755187372932,
+      "learning_rate": 9.210284919217436e-06,
+      "loss": 0.4839,
+      "step": 4157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.018983361332892,
+      "learning_rate": 9.209916459388068e-06,
+      "loss": 0.5639,
+      "step": 4158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0024153398674303,
+      "learning_rate": 9.209547921893906e-06,
+      "loss": 0.6019,
+      "step": 4159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1137415055670306,
+      "learning_rate": 9.209179306742659e-06,
+      "loss": 0.714,
+      "step": 4160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6857450254604,
+      "learning_rate": 9.208810613942048e-06,
+      "loss": 0.5291,
+      "step": 4161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.127274313405465,
+      "learning_rate": 9.20844184349979e-06,
+      "loss": 0.6419,
+      "step": 4162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.737370485023416,
+      "learning_rate": 9.208072995423607e-06,
+      "loss": 0.4759,
+      "step": 4163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2130936727565786,
+      "learning_rate": 9.207704069721219e-06,
+      "loss": 0.604,
+      "step": 4164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5543941749861485,
+      "learning_rate": 9.207335066400347e-06,
+      "loss": 0.6332,
+      "step": 4165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8942821035608335,
+      "learning_rate": 9.20696598546872e-06,
+      "loss": 0.6226,
+      "step": 4166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9560435061901567,
+      "learning_rate": 9.206596826934062e-06,
+      "loss": 0.5733,
+      "step": 4167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.431782801663525,
+      "learning_rate": 9.206227590804102e-06,
+      "loss": 0.766,
+      "step": 4168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9010564735685735,
+      "learning_rate": 9.205858277086567e-06,
+      "loss": 0.5615,
+      "step": 4169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7496896695882667,
+      "learning_rate": 9.205488885789195e-06,
+      "loss": 0.5436,
+      "step": 4170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9311366078211898,
+      "learning_rate": 9.205119416919711e-06,
+      "loss": 0.5052,
+      "step": 4171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.65621850229259,
+      "learning_rate": 9.204749870485856e-06,
+      "loss": 0.5289,
+      "step": 4172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2473913990903855,
+      "learning_rate": 9.204380246495363e-06,
+      "loss": 0.6011,
+      "step": 4173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6626987843095566,
+      "learning_rate": 9.20401054495597e-06,
+      "loss": 0.6707,
+      "step": 4174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9676103702504097,
+      "learning_rate": 9.203640765875416e-06,
+      "loss": 0.4175,
+      "step": 4175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8430496951417301,
+      "learning_rate": 9.203270909261444e-06,
+      "loss": 0.6511,
+      "step": 4176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6628128841060585,
+      "learning_rate": 9.202900975121796e-06,
+      "loss": 0.4592,
+      "step": 4177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8164746739863697,
+      "learning_rate": 9.202530963464215e-06,
+      "loss": 0.4693,
+      "step": 4178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8089533837642748,
+      "learning_rate": 9.202160874296448e-06,
+      "loss": 0.5913,
+      "step": 4179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0727797404729973,
+      "learning_rate": 9.201790707626242e-06,
+      "loss": 0.5283,
+      "step": 4180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8555239789503362,
+      "learning_rate": 9.201420463461349e-06,
+      "loss": 0.5904,
+      "step": 4181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7155669655240289,
+      "learning_rate": 9.201050141809516e-06,
+      "loss": 0.4502,
+      "step": 4182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1773454298054613,
+      "learning_rate": 9.200679742678497e-06,
+      "loss": 0.542,
+      "step": 4183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.044089225144912,
+      "learning_rate": 9.200309266076048e-06,
+      "loss": 0.6274,
+      "step": 4184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9700872126872095,
+      "learning_rate": 9.199938712009923e-06,
+      "loss": 0.5334,
+      "step": 4185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9947824617066765,
+      "learning_rate": 9.199568080487878e-06,
+      "loss": 0.729,
+      "step": 4186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8275325751583644,
+      "learning_rate": 9.199197371517673e-06,
+      "loss": 0.4966,
+      "step": 4187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.488494140879796,
+      "learning_rate": 9.198826585107068e-06,
+      "loss": 0.6129,
+      "step": 4188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9125006636684199,
+      "learning_rate": 9.198455721263825e-06,
+      "loss": 0.6198,
+      "step": 4189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1106920010535664,
+      "learning_rate": 9.198084779995711e-06,
+      "loss": 0.6422,
+      "step": 4190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4016037410070186,
+      "learning_rate": 9.197713761310488e-06,
+      "loss": 0.6759,
+      "step": 4191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9967394690326037,
+      "learning_rate": 9.197342665215925e-06,
+      "loss": 0.5768,
+      "step": 4192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.109600514421551,
+      "learning_rate": 9.19697149171979e-06,
+      "loss": 0.7062,
+      "step": 4193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8737923717393743,
+      "learning_rate": 9.196600240829852e-06,
+      "loss": 0.4885,
+      "step": 4194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.029929606066795,
+      "learning_rate": 9.196228912553883e-06,
+      "loss": 0.5655,
+      "step": 4195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6690911472262744,
+      "learning_rate": 9.195857506899661e-06,
+      "loss": 0.496,
+      "step": 4196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9626293697637074,
+      "learning_rate": 9.195486023874954e-06,
+      "loss": 0.6231,
+      "step": 4197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8406820207992247,
+      "learning_rate": 9.195114463487546e-06,
+      "loss": 0.4885,
+      "step": 4198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7770753376733703,
+      "learning_rate": 9.194742825745207e-06,
+      "loss": 0.5468,
+      "step": 4199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7155154504590138,
+      "learning_rate": 9.194371110655727e-06,
+      "loss": 0.5048,
+      "step": 4200
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.5784969925880432,
+      "eval_runtime": 67.9439,
+      "eval_samples_per_second": 50.188,
+      "eval_steps_per_second": 1.575,
+      "step": 4200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8227139264948302,
+      "learning_rate": 9.19399931822688e-06,
+      "loss": 0.4529,
+      "step": 4201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7051896333600982,
+      "learning_rate": 9.193627448466451e-06,
+      "loss": 0.5163,
+      "step": 4202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.925759873735708,
+      "learning_rate": 9.193255501382226e-06,
+      "loss": 0.6505,
+      "step": 4203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7379957701692468,
+      "learning_rate": 9.192883476981991e-06,
+      "loss": 0.5273,
+      "step": 4204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.884215565308942,
+      "learning_rate": 9.192511375273535e-06,
+      "loss": 0.5226,
+      "step": 4205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4576503477710547,
+      "learning_rate": 9.192139196264645e-06,
+      "loss": 0.5293,
+      "step": 4206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7720080937713558,
+      "learning_rate": 9.191766939963118e-06,
+      "loss": 0.5411,
+      "step": 4207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.267082348917357,
+      "learning_rate": 9.19139460637674e-06,
+      "loss": 0.493,
+      "step": 4208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1258945617308713,
+      "learning_rate": 9.19102219551331e-06,
+      "loss": 0.5361,
+      "step": 4209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1616820453600787,
+      "learning_rate": 9.190649707380622e-06,
+      "loss": 0.5227,
+      "step": 4210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1751039128167613,
+      "learning_rate": 9.190277141986474e-06,
+      "loss": 0.5714,
+      "step": 4211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5879835154022683,
+      "learning_rate": 9.18990449933867e-06,
+      "loss": 0.4938,
+      "step": 4212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7634405479530533,
+      "learning_rate": 9.189531779445002e-06,
+      "loss": 0.5458,
+      "step": 4213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.933274062735518,
+      "learning_rate": 9.189158982313282e-06,
+      "loss": 0.5744,
+      "step": 4214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0840280334631696,
+      "learning_rate": 9.188786107951307e-06,
+      "loss": 0.8165,
+      "step": 4215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.918743607781593,
+      "learning_rate": 9.188413156366887e-06,
+      "loss": 0.5477,
+      "step": 4216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3001662885936884,
+      "learning_rate": 9.18804012756783e-06,
+      "loss": 0.7726,
+      "step": 4217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3125402099118197,
+      "learning_rate": 9.187667021561942e-06,
+      "loss": 0.6712,
+      "step": 4218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7161503623177372,
+      "learning_rate": 9.187293838357036e-06,
+      "loss": 0.4629,
+      "step": 4219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.117368296775059,
+      "learning_rate": 9.186920577960926e-06,
+      "loss": 0.5604,
+      "step": 4220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.997283993340792,
+      "learning_rate": 9.18654724038142e-06,
+      "loss": 0.4999,
+      "step": 4221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0282002629707545,
+      "learning_rate": 9.186173825626339e-06,
+      "loss": 0.6503,
+      "step": 4222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1664980019841047,
+      "learning_rate": 9.185800333703498e-06,
+      "loss": 0.5698,
+      "step": 4223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.954180971952933,
+      "learning_rate": 9.185426764620717e-06,
+      "loss": 0.5641,
+      "step": 4224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0340161250230757,
+      "learning_rate": 9.185053118385817e-06,
+      "loss": 0.5748,
+      "step": 4225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7686877766209346,
+      "learning_rate": 9.184679395006616e-06,
+      "loss": 0.5213,
+      "step": 4226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9966534949373624,
+      "learning_rate": 9.184305594490943e-06,
+      "loss": 0.5469,
+      "step": 4227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.856979410681044,
+      "learning_rate": 9.18393171684662e-06,
+      "loss": 0.4349,
+      "step": 4228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.191573244165781,
+      "learning_rate": 9.183557762081474e-06,
+      "loss": 0.6268,
+      "step": 4229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1529727972622044,
+      "learning_rate": 9.183183730203333e-06,
+      "loss": 0.8956,
+      "step": 4230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4224231283804043,
+      "learning_rate": 9.18280962122003e-06,
+      "loss": 0.5639,
+      "step": 4231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6731652599948286,
+      "learning_rate": 9.182435435139395e-06,
+      "loss": 0.4596,
+      "step": 4232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.749794066171635,
+      "learning_rate": 9.182061171969262e-06,
+      "loss": 0.6315,
+      "step": 4233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7081349890420736,
+      "learning_rate": 9.181686831717466e-06,
+      "loss": 0.5181,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.24362871652964,
+      "learning_rate": 9.181312414391839e-06,
+      "loss": 0.5944,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.01858775224178,
+      "learning_rate": 9.180937920000226e-06,
+      "loss": 0.5852,
+      "step": 4236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0367682246810204,
+      "learning_rate": 9.180563348550465e-06,
+      "loss": 0.6114,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7491468796326417,
+      "learning_rate": 9.180188700050394e-06,
+      "loss": 0.594,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.053919492467177,
+      "learning_rate": 9.179813974507859e-06,
+      "loss": 0.7318,
+      "step": 4239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0818706240990865,
+      "learning_rate": 9.179439171930706e-06,
+      "loss": 0.5821,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.702184691232269,
+      "learning_rate": 9.179064292326777e-06,
+      "loss": 0.7613,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1275586296572233,
+      "learning_rate": 9.178689335703925e-06,
+      "loss": 0.7122,
+      "step": 4242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6051187024212403,
+      "learning_rate": 9.178314302069994e-06,
+      "loss": 0.4988,
+      "step": 4243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1212242959415186,
+      "learning_rate": 9.177939191432838e-06,
+      "loss": 0.5483,
+      "step": 4244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7372839200739731,
+      "learning_rate": 9.177564003800311e-06,
+      "loss": 0.5239,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4102187328286515,
+      "learning_rate": 9.177188739180264e-06,
+      "loss": 0.6585,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2372696571530613,
+      "learning_rate": 9.176813397580556e-06,
+      "loss": 0.6491,
+      "step": 4247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.052137914795964,
+      "learning_rate": 9.176437979009043e-06,
+      "loss": 0.603,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.286384979889516,
+      "learning_rate": 9.176062483473585e-06,
+      "loss": 0.5975,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9533575175273712,
+      "learning_rate": 9.175686910982042e-06,
+      "loss": 0.5631,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5813916921615601,
+      "eval_runtime": 67.9548,
+      "eval_samples_per_second": 50.18,
+      "eval_steps_per_second": 1.575,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7120119023480338,
+      "learning_rate": 9.175311261542275e-06,
+      "loss": 0.5868,
+      "step": 4251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.728187547418132,
+      "learning_rate": 9.17493553516215e-06,
+      "loss": 0.5224,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9183412209386643,
+      "learning_rate": 9.174559731849532e-06,
+      "loss": 0.6227,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7087352823299164,
+      "learning_rate": 9.17418385161229e-06,
+      "loss": 0.412,
+      "step": 4254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8351448760479139,
+      "learning_rate": 9.17380789445829e-06,
+      "loss": 0.5207,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.260981707461987,
+      "learning_rate": 9.173431860395406e-06,
+      "loss": 0.5092,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.213957551915077,
+      "learning_rate": 9.173055749431505e-06,
+      "loss": 0.7069,
+      "step": 4257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9791344173305412,
+      "learning_rate": 9.172679561574464e-06,
+      "loss": 0.4063,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9260076195100353,
+      "learning_rate": 9.172303296832156e-06,
+      "loss": 0.4548,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.196938489357895,
+      "learning_rate": 9.171926955212462e-06,
+      "loss": 0.5631,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.944972314989838,
+      "learning_rate": 9.171550536723257e-06,
+      "loss": 0.6139,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0947165801513017,
+      "learning_rate": 9.171174041372421e-06,
+      "loss": 0.5924,
+      "step": 4262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8354342228065872,
+      "learning_rate": 9.170797469167838e-06,
+      "loss": 0.5863,
+      "step": 4263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9193331438446513,
+      "learning_rate": 9.170420820117389e-06,
+      "loss": 0.5597,
+      "step": 4264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.92803348519159,
+      "learning_rate": 9.170044094228962e-06,
+      "loss": 0.613,
+      "step": 4265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3465458842067313,
+      "learning_rate": 9.169667291510439e-06,
+      "loss": 0.6344,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6233066698467395,
+      "learning_rate": 9.16929041196971e-06,
+      "loss": 0.4527,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.987652979801804,
+      "learning_rate": 9.168913455614669e-06,
+      "loss": 0.7122,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0452861505661724,
+      "learning_rate": 9.1685364224532e-06,
+      "loss": 0.5546,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.942405881133707,
+      "learning_rate": 9.168159312493201e-06,
+      "loss": 0.6731,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.672525857944254,
+      "learning_rate": 9.167782125742564e-06,
+      "loss": 0.4052,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6996212124508492,
+      "learning_rate": 9.16740486220919e-06,
+      "loss": 0.5291,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8328153409497807,
+      "learning_rate": 9.16702752190097e-06,
+      "loss": 0.6365,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0066545205781607,
+      "learning_rate": 9.166650104825807e-06,
+      "loss": 0.715,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.811180556954922,
+      "learning_rate": 9.166272610991601e-06,
+      "loss": 0.5883,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0031887500015544,
+      "learning_rate": 9.165895040406256e-06,
+      "loss": 0.5035,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9759184032219572,
+      "learning_rate": 9.165517393077676e-06,
+      "loss": 0.5129,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9860671279695639,
+      "learning_rate": 9.165139669013765e-06,
+      "loss": 0.7067,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.434350138524439,
+      "learning_rate": 9.164761868222431e-06,
+      "loss": 0.4544,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8126658764792865,
+      "learning_rate": 9.164383990711584e-06,
+      "loss": 0.4387,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7651029874309823,
+      "learning_rate": 9.164006036489133e-06,
+      "loss": 0.497,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4757362146111586,
+      "learning_rate": 9.163628005562994e-06,
+      "loss": 1.057,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.197620667259681,
+      "learning_rate": 9.163249897941077e-06,
+      "loss": 0.5757,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0329158579672315,
+      "learning_rate": 9.1628717136313e-06,
+      "loss": 0.6148,
+      "step": 4284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7931234371963671,
+      "learning_rate": 9.162493452641578e-06,
+      "loss": 0.5643,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8419029593359306,
+      "learning_rate": 9.16211511497983e-06,
+      "loss": 0.595,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8948563427094767,
+      "learning_rate": 9.161736700653976e-06,
+      "loss": 0.658,
+      "step": 4287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.022917069742594,
+      "learning_rate": 9.161358209671939e-06,
+      "loss": 0.7145,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6363703258499356,
+      "learning_rate": 9.160979642041643e-06,
+      "loss": 0.5755,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9725959880225785,
+      "learning_rate": 9.16060099777101e-06,
+      "loss": 0.5351,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0820717540592457,
+      "learning_rate": 9.16022227686797e-06,
+      "loss": 0.6623,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.963564690140017,
+      "learning_rate": 9.159843479340452e-06,
+      "loss": 0.594,
+      "step": 4292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.157824700725861,
+      "learning_rate": 9.159464605196382e-06,
+      "loss": 0.7376,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.865475122480794,
+      "learning_rate": 9.159085654443694e-06,
+      "loss": 0.3968,
+      "step": 4294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9894589964360359,
+      "learning_rate": 9.15870662709032e-06,
+      "loss": 0.5562,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0248842718251834,
+      "learning_rate": 9.158327523144196e-06,
+      "loss": 0.5597,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.882455829870533,
+      "learning_rate": 9.157948342613257e-06,
+      "loss": 0.5227,
+      "step": 4297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2515586290985468,
+      "learning_rate": 9.15756908550544e-06,
+      "loss": 0.7265,
+      "step": 4298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8596763400044882,
+      "learning_rate": 9.15718975182869e-06,
+      "loss": 0.5169,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.179864687124647,
+      "learning_rate": 9.15681034159094e-06,
+      "loss": 0.6567,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5789434313774109,
+      "eval_runtime": 67.9156,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.424526964950774,
+      "learning_rate": 9.156430854800139e-06,
+      "loss": 0.7413,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1985553271624005,
+      "learning_rate": 9.156051291464228e-06,
+      "loss": 0.6085,
+      "step": 4302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.473250261727933,
+      "learning_rate": 9.155671651591155e-06,
+      "loss": 0.7174,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.22516340194687,
+      "learning_rate": 9.155291935188867e-06,
+      "loss": 0.6282,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9135453902854698,
+      "learning_rate": 9.154912142265311e-06,
+      "loss": 0.6173,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.021393098091508,
+      "learning_rate": 9.154532272828441e-06,
+      "loss": 0.6708,
+      "step": 4306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8484310650610674,
+      "learning_rate": 9.15415232688621e-06,
+      "loss": 0.5376,
+      "step": 4307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2002862415058213,
+      "learning_rate": 9.153772304446566e-06,
+      "loss": 0.7207,
+      "step": 4308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6144402104753353,
+      "learning_rate": 9.153392205517468e-06,
+      "loss": 0.5457,
+      "step": 4309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.215608243879253,
+      "learning_rate": 9.153012030106874e-06,
+      "loss": 0.6505,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1123357234375764,
+      "learning_rate": 9.152631778222744e-06,
+      "loss": 0.5673,
+      "step": 4311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.377927759674002,
+      "learning_rate": 9.152251449873035e-06,
+      "loss": 0.5817,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.54569675559369,
+      "learning_rate": 9.15187104506571e-06,
+      "loss": 0.5562,
+      "step": 4313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.137535272876764,
+      "learning_rate": 9.151490563808734e-06,
+      "loss": 0.4824,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2408442164973246,
+      "learning_rate": 9.15111000611007e-06,
+      "loss": 0.6181,
+      "step": 4315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7543357406375577,
+      "learning_rate": 9.150729371977686e-06,
+      "loss": 0.5223,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0448402166105715,
+      "learning_rate": 9.150348661419552e-06,
+      "loss": 0.5539,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.771634098932208,
+      "learning_rate": 9.149967874443635e-06,
+      "loss": 0.5696,
+      "step": 4318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5735957842930792,
+      "learning_rate": 9.149587011057907e-06,
+      "loss": 0.447,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.173209123633177,
+      "learning_rate": 9.149206071270342e-06,
+      "loss": 0.5749,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7388576110313119,
+      "learning_rate": 9.148825055088915e-06,
+      "loss": 0.4556,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0710865334149666,
+      "learning_rate": 9.148443962521602e-06,
+      "loss": 0.6201,
+      "step": 4322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9485695948478057,
+      "learning_rate": 9.148062793576379e-06,
+      "loss": 0.6259,
+      "step": 4323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5902482650049494,
+      "learning_rate": 9.14768154826123e-06,
+      "loss": 0.5285,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8063042853397195,
+      "learning_rate": 9.147300226584132e-06,
+      "loss": 0.5465,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8915939647233893,
+      "learning_rate": 9.146918828553068e-06,
+      "loss": 0.4197,
+      "step": 4326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0079801535035875,
+      "learning_rate": 9.146537354176025e-06,
+      "loss": 0.501,
+      "step": 4327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9317472820821158,
+      "learning_rate": 9.146155803460987e-06,
+      "loss": 0.7074,
+      "step": 4328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1323378820901278,
+      "learning_rate": 9.145774176415943e-06,
+      "loss": 0.6773,
+      "step": 4329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.081109345545547,
+      "learning_rate": 9.14539247304888e-06,
+      "loss": 0.5443,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8913942126118481,
+      "learning_rate": 9.14501069336779e-06,
+      "loss": 0.5324,
+      "step": 4331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6096314384350194,
+      "learning_rate": 9.144628837380663e-06,
+      "loss": 0.5459,
+      "step": 4332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3570752396134407,
+      "learning_rate": 9.144246905095498e-06,
+      "loss": 0.382,
+      "step": 4333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6394591843378226,
+      "learning_rate": 9.143864896520287e-06,
+      "loss": 0.5435,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1382917778280115,
+      "learning_rate": 9.143482811663027e-06,
+      "loss": 0.6522,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9179985013771148,
+      "learning_rate": 9.14310065053172e-06,
+      "loss": 0.5448,
+      "step": 4336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.836387779000709,
+      "learning_rate": 9.142718413134361e-06,
+      "loss": 0.5275,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5995745039695137,
+      "learning_rate": 9.142336099478955e-06,
+      "loss": 0.4398,
+      "step": 4338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5725675330774995,
+      "learning_rate": 9.141953709573504e-06,
+      "loss": 0.3739,
+      "step": 4339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.560106505532726,
+      "learning_rate": 9.141571243426015e-06,
+      "loss": 0.5402,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7500400969492969,
+      "learning_rate": 9.141188701044495e-06,
+      "loss": 0.4972,
+      "step": 4341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.171778428643274,
+      "learning_rate": 9.14080608243695e-06,
+      "loss": 0.5883,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.763110642369031,
+      "learning_rate": 9.140423387611392e-06,
+      "loss": 0.5172,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8387190285228745,
+      "learning_rate": 9.14004061657583e-06,
+      "loss": 0.4426,
+      "step": 4344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8702714754461474,
+      "learning_rate": 9.13965776933828e-06,
+      "loss": 0.608,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2300074320559786,
+      "learning_rate": 9.139274845906756e-06,
+      "loss": 0.6906,
+      "step": 4346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7017528879540504,
+      "learning_rate": 9.138891846289273e-06,
+      "loss": 0.4733,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7414603202666152,
+      "learning_rate": 9.138508770493848e-06,
+      "loss": 0.5463,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8389557483023686,
+      "learning_rate": 9.138125618528503e-06,
+      "loss": 0.5473,
+      "step": 4349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.660668017977543,
+      "learning_rate": 9.137742390401257e-06,
+      "loss": 0.5198,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.578808069229126,
+      "eval_runtime": 67.9409,
+      "eval_samples_per_second": 50.191,
+      "eval_steps_per_second": 1.575,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6922103808599223,
+      "learning_rate": 9.137359086120135e-06,
+      "loss": 0.5388,
+      "step": 4351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0622082388224507,
+      "learning_rate": 9.136975705693158e-06,
+      "loss": 0.7283,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4138173212012313,
+      "learning_rate": 9.136592249128354e-06,
+      "loss": 0.5488,
+      "step": 4353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5813305796050565,
+      "learning_rate": 9.136208716433751e-06,
+      "loss": 0.4828,
+      "step": 4354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6941971285180195,
+      "learning_rate": 9.135825107617376e-06,
+      "loss": 0.5072,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8917831430861982,
+      "learning_rate": 9.135441422687262e-06,
+      "loss": 0.5589,
+      "step": 4356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.911120847693861,
+      "learning_rate": 9.135057661651439e-06,
+      "loss": 0.6684,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1587605599172655,
+      "learning_rate": 9.134673824517941e-06,
+      "loss": 0.6153,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1291857121369433,
+      "learning_rate": 9.134289911294806e-06,
+      "loss": 0.6488,
+      "step": 4359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.148209162338253,
+      "learning_rate": 9.133905921990067e-06,
+      "loss": 0.5762,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1368314654022282,
+      "learning_rate": 9.133521856611766e-06,
+      "loss": 0.4309,
+      "step": 4361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7971322470717395,
+      "learning_rate": 9.13313771516794e-06,
+      "loss": 0.5503,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9675364434613947,
+      "learning_rate": 9.132753497666635e-06,
+      "loss": 0.6097,
+      "step": 4363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9961369403429454,
+      "learning_rate": 9.132369204115892e-06,
+      "loss": 0.6011,
+      "step": 4364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.032159513458243,
+      "learning_rate": 9.131984834523754e-06,
+      "loss": 0.5427,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8982554102806468,
+      "learning_rate": 9.13160038889827e-06,
+      "loss": 0.6123,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7685444997879667,
+      "learning_rate": 9.131215867247488e-06,
+      "loss": 0.4094,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5882858968018614,
+      "learning_rate": 9.130831269579456e-06,
+      "loss": 0.7022,
+      "step": 4368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0097438347431993,
+      "learning_rate": 9.130446595902228e-06,
+      "loss": 0.5713,
+      "step": 4369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9433359925998601,
+      "learning_rate": 9.130061846223854e-06,
+      "loss": 0.6489,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6717925848796267,
+      "learning_rate": 9.129677020552393e-06,
+      "loss": 0.5271,
+      "step": 4371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7619573657086522,
+      "learning_rate": 9.129292118895894e-06,
+      "loss": 0.4406,
+      "step": 4372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6316481917153478,
+      "learning_rate": 9.128907141262421e-06,
+      "loss": 0.4946,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5848213860140152,
+      "learning_rate": 9.128522087660032e-06,
+      "loss": 0.3974,
+      "step": 4374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.001971289751937,
+      "learning_rate": 9.128136958096785e-06,
+      "loss": 0.6352,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4820051517636732,
+      "learning_rate": 9.127751752580744e-06,
+      "loss": 0.471,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.972711106119243,
+      "learning_rate": 9.127366471119973e-06,
+      "loss": 0.5774,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4066891719395573,
+      "learning_rate": 9.126981113722539e-06,
+      "loss": 0.5861,
+      "step": 4378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0326314419024816,
+      "learning_rate": 9.12659568039651e-06,
+      "loss": 0.6323,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2807039996820273,
+      "learning_rate": 9.126210171149949e-06,
+      "loss": 0.7095,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.399573821441971,
+      "learning_rate": 9.125824585990934e-06,
+      "loss": 0.7116,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5519024232237866,
+      "learning_rate": 9.125438924927532e-06,
+      "loss": 0.4789,
+      "step": 4382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.841383162068827,
+      "learning_rate": 9.125053187967819e-06,
+      "loss": 0.5011,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.827310746231553,
+      "learning_rate": 9.124667375119868e-06,
+      "loss": 0.5622,
+      "step": 4384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9607181196682741,
+      "learning_rate": 9.124281486391758e-06,
+      "loss": 0.6281,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9075165294005443,
+      "learning_rate": 9.123895521791564e-06,
+      "loss": 0.5958,
+      "step": 4386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.787959182602479,
+      "learning_rate": 9.12350948132737e-06,
+      "loss": 0.5137,
+      "step": 4387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6365779235915363,
+      "learning_rate": 9.123123365007255e-06,
+      "loss": 0.3992,
+      "step": 4388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7609819539657339,
+      "learning_rate": 9.122737172839302e-06,
+      "loss": 0.5878,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.760471555418242,
+      "learning_rate": 9.122350904831597e-06,
+      "loss": 0.6165,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.217630563246285,
+      "learning_rate": 9.121964560992227e-06,
+      "loss": 0.3925,
+      "step": 4391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5792408396887898,
+      "learning_rate": 9.121578141329276e-06,
+      "loss": 0.5258,
+      "step": 4392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6274237855033384,
+      "learning_rate": 9.121191645850837e-06,
+      "loss": 0.4392,
+      "step": 4393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.713615265654716,
+      "learning_rate": 9.120805074565001e-06,
+      "loss": 0.5591,
+      "step": 4394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9215016955052644,
+      "learning_rate": 9.12041842747986e-06,
+      "loss": 0.5803,
+      "step": 4395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.995432802497749,
+      "learning_rate": 9.120031704603507e-06,
+      "loss": 0.6017,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8880681528466745,
+      "learning_rate": 9.11964490594404e-06,
+      "loss": 0.5178,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7220293494776908,
+      "learning_rate": 9.119258031509554e-06,
+      "loss": 0.4041,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.692970253415135,
+      "learning_rate": 9.118871081308149e-06,
+      "loss": 0.4136,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1653221438713333,
+      "learning_rate": 9.118484055347927e-06,
+      "loss": 0.5433,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5782193541526794,
+      "eval_runtime": 67.9778,
+      "eval_samples_per_second": 50.163,
+      "eval_steps_per_second": 1.574,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8334574538310713,
+      "learning_rate": 9.118096953636986e-06,
+      "loss": 0.4854,
+      "step": 4401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.339783705072547,
+      "learning_rate": 9.117709776183434e-06,
+      "loss": 0.6525,
+      "step": 4402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0064879492948475,
+      "learning_rate": 9.117322522995377e-06,
+      "loss": 0.7674,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.05262639628719,
+      "learning_rate": 9.116935194080917e-06,
+      "loss": 0.5244,
+      "step": 4404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3038275006626883,
+      "learning_rate": 9.116547789448166e-06,
+      "loss": 0.6924,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9401155161973433,
+      "learning_rate": 9.116160309105235e-06,
+      "loss": 0.6274,
+      "step": 4406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5397649133388052,
+      "learning_rate": 9.115772753060232e-06,
+      "loss": 0.4268,
+      "step": 4407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9378371112990438,
+      "learning_rate": 9.115385121321273e-06,
+      "loss": 0.6112,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1634930322754733,
+      "learning_rate": 9.114997413896472e-06,
+      "loss": 0.4886,
+      "step": 4409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3125659718773126,
+      "learning_rate": 9.114609630793946e-06,
+      "loss": 0.4147,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.128710097410051,
+      "learning_rate": 9.114221772021812e-06,
+      "loss": 0.5298,
+      "step": 4411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.139256923128494,
+      "learning_rate": 9.11383383758819e-06,
+      "loss": 0.8528,
+      "step": 4412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7529099584410723,
+      "learning_rate": 9.1134458275012e-06,
+      "loss": 0.4946,
+      "step": 4413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.888590609490273,
+      "learning_rate": 9.113057741768968e-06,
+      "loss": 0.5682,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5895029173557873,
+      "learning_rate": 9.112669580399617e-06,
+      "loss": 0.3973,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.053478754515145,
+      "learning_rate": 9.112281343401269e-06,
+      "loss": 0.5962,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0553224220817308,
+      "learning_rate": 9.111893030782057e-06,
+      "loss": 0.6362,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4514139170451126,
+      "learning_rate": 9.111504642550107e-06,
+      "loss": 0.3663,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0208992150314677,
+      "learning_rate": 9.111116178713552e-06,
+      "loss": 0.6706,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7522455490925313,
+      "learning_rate": 9.110727639280522e-06,
+      "loss": 0.6249,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6120420547472034,
+      "learning_rate": 9.110339024259152e-06,
+      "loss": 0.5178,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9617186501031567,
+      "learning_rate": 9.109950333657575e-06,
+      "loss": 0.6949,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.587703613338579,
+      "learning_rate": 9.109561567483931e-06,
+      "loss": 0.4386,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9014551079202555,
+      "learning_rate": 9.10917272574636e-06,
+      "loss": 0.6429,
+      "step": 4424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.827421010970891,
+      "learning_rate": 9.108783808452997e-06,
+      "loss": 0.719,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.86664878276118,
+      "learning_rate": 9.108394815611987e-06,
+      "loss": 0.4938,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4967244323232416,
+      "learning_rate": 9.108005747231472e-06,
+      "loss": 0.6155,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.154278297634596,
+      "learning_rate": 9.107616603319599e-06,
+      "loss": 0.757,
+      "step": 4428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.54028329047591,
+      "learning_rate": 9.107227383884512e-06,
+      "loss": 0.6917,
+      "step": 4429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.334044233821694,
+      "learning_rate": 9.106838088934361e-06,
+      "loss": 0.3047,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.01509039915079,
+      "learning_rate": 9.106448718477296e-06,
+      "loss": 0.6206,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6525215151042805,
+      "learning_rate": 9.106059272521465e-06,
+      "loss": 0.4869,
+      "step": 4432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1756031530220854,
+      "learning_rate": 9.105669751075024e-06,
+      "loss": 0.5552,
+      "step": 4433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7165466153353284,
+      "learning_rate": 9.105280154146126e-06,
+      "loss": 0.3827,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7034842698844703,
+      "learning_rate": 9.104890481742927e-06,
+      "loss": 0.5022,
+      "step": 4435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.083151595310631,
+      "learning_rate": 9.104500733873586e-06,
+      "loss": 0.4714,
+      "step": 4436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.040937763058449,
+      "learning_rate": 9.104110910546261e-06,
+      "loss": 0.5414,
+      "step": 4437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8722280519242431,
+      "learning_rate": 9.103721011769113e-06,
+      "loss": 0.5089,
+      "step": 4438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7551947892433157,
+      "learning_rate": 9.103331037550302e-06,
+      "loss": 0.6142,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7425852059731233,
+      "learning_rate": 9.102940987897994e-06,
+      "loss": 0.5493,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1075467070903313,
+      "learning_rate": 9.102550862820357e-06,
+      "loss": 0.6607,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.288683907446357,
+      "learning_rate": 9.102160662325553e-06,
+      "loss": 0.8205,
+      "step": 4442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2023782383644126,
+      "learning_rate": 9.101770386421753e-06,
+      "loss": 0.5779,
+      "step": 4443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.008730417617452,
+      "learning_rate": 9.101380035117126e-06,
+      "loss": 0.5227,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7534316248250887,
+      "learning_rate": 9.100989608419847e-06,
+      "loss": 0.5286,
+      "step": 4445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9552974628006843,
+      "learning_rate": 9.100599106338088e-06,
+      "loss": 0.4779,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.207957611120514,
+      "learning_rate": 9.100208528880022e-06,
+      "loss": 0.5969,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2113798450274857,
+      "learning_rate": 9.099817876053826e-06,
+      "loss": 0.5919,
+      "step": 4448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1394795706606677,
+      "learning_rate": 9.099427147867679e-06,
+      "loss": 0.6097,
+      "step": 4449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1848322311516015,
+      "learning_rate": 9.09903634432976e-06,
+      "loss": 0.4764,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5778436660766602,
+      "eval_runtime": 67.9358,
+      "eval_samples_per_second": 50.194,
+      "eval_steps_per_second": 1.575,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.094392420591768,
+      "learning_rate": 9.098645465448251e-06,
+      "loss": 0.462,
+      "step": 4451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1364299543242558,
+      "learning_rate": 9.098254511231333e-06,
+      "loss": 0.5511,
+      "step": 4452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7707746706611551,
+      "learning_rate": 9.097863481687195e-06,
+      "loss": 0.4537,
+      "step": 4453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4843808145806858,
+      "learning_rate": 9.097472376824017e-06,
+      "loss": 0.435,
+      "step": 4454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9838704850973063,
+      "learning_rate": 9.097081196649989e-06,
+      "loss": 0.661,
+      "step": 4455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7062926494438109,
+      "learning_rate": 9.096689941173302e-06,
+      "loss": 0.6272,
+      "step": 4456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.063284219510452,
+      "learning_rate": 9.096298610402144e-06,
+      "loss": 0.7072,
+      "step": 4457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.751886074768719,
+      "learning_rate": 9.09590720434471e-06,
+      "loss": 0.5639,
+      "step": 4458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.083747781791249,
+      "learning_rate": 9.095515723009191e-06,
+      "loss": 0.5424,
+      "step": 4459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.926162715939642,
+      "learning_rate": 9.095124166403784e-06,
+      "loss": 0.4619,
+      "step": 4460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.88420603375627,
+      "learning_rate": 9.094732534536684e-06,
+      "loss": 0.5286,
+      "step": 4461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.808233908640816,
+      "learning_rate": 9.094340827416093e-06,
+      "loss": 0.5129,
+      "step": 4462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6262187564533355,
+      "learning_rate": 9.093949045050208e-06,
+      "loss": 0.5311,
+      "step": 4463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.151412517756477,
+      "learning_rate": 9.093557187447232e-06,
+      "loss": 0.5822,
+      "step": 4464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9762527530126093,
+      "learning_rate": 9.09316525461537e-06,
+      "loss": 0.6595,
+      "step": 4465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7027855832964371,
+      "learning_rate": 9.092773246562826e-06,
+      "loss": 0.439,
+      "step": 4466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.450142307922059,
+      "learning_rate": 9.092381163297802e-06,
+      "loss": 0.5564,
+      "step": 4467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9617246106296389,
+      "learning_rate": 9.091989004828512e-06,
+      "loss": 0.6464,
+      "step": 4468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6656989358267045,
+      "learning_rate": 9.091596771163163e-06,
+      "loss": 0.4492,
+      "step": 4469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.96145583846626,
+      "learning_rate": 9.091204462309966e-06,
+      "loss": 0.684,
+      "step": 4470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.308417492656998,
+      "learning_rate": 9.090812078277134e-06,
+      "loss": 0.5451,
+      "step": 4471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9319082460881314,
+      "learning_rate": 9.090419619072882e-06,
+      "loss": 0.5268,
+      "step": 4472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.221149348859587,
+      "learning_rate": 9.090027084705426e-06,
+      "loss": 0.5264,
+      "step": 4473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0490640401627704,
+      "learning_rate": 9.089634475182981e-06,
+      "loss": 0.5868,
+      "step": 4474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8602910714456866,
+      "learning_rate": 9.089241790513769e-06,
+      "loss": 0.5225,
+      "step": 4475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9219855403609691,
+      "learning_rate": 9.088849030706009e-06,
+      "loss": 0.5181,
+      "step": 4476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6014940579122687,
+      "learning_rate": 9.088456195767926e-06,
+      "loss": 0.3527,
+      "step": 4477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3820534648928797,
+      "learning_rate": 9.088063285707738e-06,
+      "loss": 0.6445,
+      "step": 4478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.437006920208686,
+      "learning_rate": 9.087670300533675e-06,
+      "loss": 0.4174,
+      "step": 4479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5257584148840069,
+      "learning_rate": 9.087277240253964e-06,
+      "loss": 0.394,
+      "step": 4480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7490321948835115,
+      "learning_rate": 9.08688410487683e-06,
+      "loss": 0.475,
+      "step": 4481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4686348960828692,
+      "learning_rate": 9.086490894410506e-06,
+      "loss": 0.4218,
+      "step": 4482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5546698968802122,
+      "learning_rate": 9.086097608863223e-06,
+      "loss": 0.4887,
+      "step": 4483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.840958982677198,
+      "learning_rate": 9.085704248243213e-06,
+      "loss": 0.5595,
+      "step": 4484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.734361802908784,
+      "learning_rate": 9.085310812558715e-06,
+      "loss": 0.6044,
+      "step": 4485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6136984726556425,
+      "learning_rate": 9.084917301817958e-06,
+      "loss": 0.3895,
+      "step": 4486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.93211397098811,
+      "learning_rate": 9.084523716029184e-06,
+      "loss": 0.5654,
+      "step": 4487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8222556815257354,
+      "learning_rate": 9.084130055200636e-06,
+      "loss": 0.5711,
+      "step": 4488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.218204473995909,
+      "learning_rate": 9.08373631934055e-06,
+      "loss": 0.6199,
+      "step": 4489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9977372901586525,
+      "learning_rate": 9.08334250845717e-06,
+      "loss": 0.4708,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.029943899116224,
+      "learning_rate": 9.082948622558739e-06,
+      "loss": 0.6379,
+      "step": 4491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.659069941156736,
+      "learning_rate": 9.082554661653504e-06,
+      "loss": 0.4961,
+      "step": 4492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8973833417974393,
+      "learning_rate": 9.082160625749713e-06,
+      "loss": 0.4638,
+      "step": 4493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6503249650519332,
+      "learning_rate": 9.081766514855616e-06,
+      "loss": 0.4561,
+      "step": 4494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5304771767733307,
+      "learning_rate": 9.08137232897946e-06,
+      "loss": 0.3886,
+      "step": 4495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9012211264684682,
+      "learning_rate": 9.080978068129497e-06,
+      "loss": 0.5098,
+      "step": 4496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6998675629448778,
+      "learning_rate": 9.080583732313987e-06,
+      "loss": 0.4252,
+      "step": 4497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7620698161102506,
+      "learning_rate": 9.080189321541177e-06,
+      "loss": 0.4842,
+      "step": 4498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0951272423283354,
+      "learning_rate": 9.079794835819328e-06,
+      "loss": 0.521,
+      "step": 4499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8934030757288252,
+      "learning_rate": 9.079400275156698e-06,
+      "loss": 0.4041,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.576078474521637,
+      "eval_runtime": 67.8489,
+      "eval_samples_per_second": 50.259,
+      "eval_steps_per_second": 1.577,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8690153018436317,
+      "learning_rate": 9.079005639561545e-06,
+      "loss": 0.461,
+      "step": 4501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4178050369180615,
+      "learning_rate": 9.078610929042136e-06,
+      "loss": 0.694,
+      "step": 4502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0009561171968318,
+      "learning_rate": 9.078216143606726e-06,
+      "loss": 0.5274,
+      "step": 4503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2043179365517958,
+      "learning_rate": 9.077821283263585e-06,
+      "loss": 0.7762,
+      "step": 4504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4688026093300803,
+      "learning_rate": 9.077426348020978e-06,
+      "loss": 0.6145,
+      "step": 4505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.867370145296535,
+      "learning_rate": 9.07703133788717e-06,
+      "loss": 0.6436,
+      "step": 4506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1083139264538713,
+      "learning_rate": 9.076636252870437e-06,
+      "loss": 0.5296,
+      "step": 4507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9147595464930955,
+      "learning_rate": 9.07624109297904e-06,
+      "loss": 0.657,
+      "step": 4508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8069079752067558,
+      "learning_rate": 9.075845858221262e-06,
+      "loss": 0.4937,
+      "step": 4509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8363648161254567,
+      "learning_rate": 9.07545054860537e-06,
+      "loss": 0.6021,
+      "step": 4510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.651421088149266,
+      "learning_rate": 9.075055164139643e-06,
+      "loss": 0.4478,
+      "step": 4511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8512998996672267,
+      "learning_rate": 9.074659704832356e-06,
+      "loss": 0.4833,
+      "step": 4512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.793379957139933,
+      "learning_rate": 9.074264170691786e-06,
+      "loss": 0.5402,
+      "step": 4513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8099355550613685,
+      "learning_rate": 9.073868561726218e-06,
+      "loss": 0.4346,
+      "step": 4514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.834538751931152,
+      "learning_rate": 9.073472877943931e-06,
+      "loss": 0.5175,
+      "step": 4515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6183725326534426,
+      "learning_rate": 9.073077119353209e-06,
+      "loss": 0.5079,
+      "step": 4516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5827935774172104,
+      "learning_rate": 9.072681285962336e-06,
+      "loss": 0.5275,
+      "step": 4517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.101719840513897,
+      "learning_rate": 9.0722853777796e-06,
+      "loss": 0.5935,
+      "step": 4518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2789576214161844,
+      "learning_rate": 9.071889394813289e-06,
+      "loss": 0.7107,
+      "step": 4519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2629554381248225,
+      "learning_rate": 9.071493337071691e-06,
+      "loss": 0.7308,
+      "step": 4520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0599081428747095,
+      "learning_rate": 9.071097204563099e-06,
+      "loss": 0.7186,
+      "step": 4521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4703574683753684,
+      "learning_rate": 9.070700997295805e-06,
+      "loss": 0.3957,
+      "step": 4522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9206296279444943,
+      "learning_rate": 9.070304715278103e-06,
+      "loss": 0.7522,
+      "step": 4523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8783856473934928,
+      "learning_rate": 9.06990835851829e-06,
+      "loss": 0.7279,
+      "step": 4524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9183000018004122,
+      "learning_rate": 9.069511927024662e-06,
+      "loss": 0.5574,
+      "step": 4525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8724907487505582,
+      "learning_rate": 9.06911542080552e-06,
+      "loss": 0.4828,
+      "step": 4526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.96957508447362,
+      "learning_rate": 9.068718839869164e-06,
+      "loss": 0.4819,
+      "step": 4527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9447012751132415,
+      "learning_rate": 9.068322184223894e-06,
+      "loss": 0.7494,
+      "step": 4528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.661912202297503,
+      "learning_rate": 9.067925453878017e-06,
+      "loss": 0.6794,
+      "step": 4529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8312932509388478,
+      "learning_rate": 9.067528648839835e-06,
+      "loss": 0.5811,
+      "step": 4530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.083141038949046,
+      "learning_rate": 9.067131769117659e-06,
+      "loss": 0.7393,
+      "step": 4531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3719750398154535,
+      "learning_rate": 9.066734814719793e-06,
+      "loss": 0.7457,
+      "step": 4532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.023179564646198,
+      "learning_rate": 9.066337785654549e-06,
+      "loss": 0.5941,
+      "step": 4533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7464953642214631,
+      "learning_rate": 9.06594068193024e-06,
+      "loss": 0.6584,
+      "step": 4534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7702837118893235,
+      "learning_rate": 9.065543503555177e-06,
+      "loss": 0.5213,
+      "step": 4535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8855089415398585,
+      "learning_rate": 9.065146250537675e-06,
+      "loss": 0.4193,
+      "step": 4536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.545836294008874,
+      "learning_rate": 9.064748922886052e-06,
+      "loss": 0.4284,
+      "step": 4537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6952883233879146,
+      "learning_rate": 9.064351520608623e-06,
+      "loss": 0.4453,
+      "step": 4538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9671651157547885,
+      "learning_rate": 9.06395404371371e-06,
+      "loss": 0.6233,
+      "step": 4539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0436853313324175,
+      "learning_rate": 9.063556492209631e-06,
+      "loss": 0.6498,
+      "step": 4540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9036429043133503,
+      "learning_rate": 9.063158866104712e-06,
+      "loss": 0.4544,
+      "step": 4541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0441947157291223,
+      "learning_rate": 9.062761165407275e-06,
+      "loss": 0.5359,
+      "step": 4542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1531824972986278,
+      "learning_rate": 9.062363390125646e-06,
+      "loss": 0.6919,
+      "step": 4543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9018337718011527,
+      "learning_rate": 9.061965540268151e-06,
+      "loss": 0.5108,
+      "step": 4544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9365325776782052,
+      "learning_rate": 9.06156761584312e-06,
+      "loss": 0.4592,
+      "step": 4545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0079752930246,
+      "learning_rate": 9.061169616858883e-06,
+      "loss": 0.7574,
+      "step": 4546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8756585129040904,
+      "learning_rate": 9.060771543323774e-06,
+      "loss": 0.5166,
+      "step": 4547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1516461576778734,
+      "learning_rate": 9.060373395246122e-06,
+      "loss": 0.6683,
+      "step": 4548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0056094561032323,
+      "learning_rate": 9.059975172634266e-06,
+      "loss": 0.6721,
+      "step": 4549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5524563659731783,
+      "learning_rate": 9.059576875496539e-06,
+      "loss": 0.4301,
+      "step": 4550
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.5782231092453003,
+      "eval_runtime": 67.8554,
+      "eval_samples_per_second": 50.254,
+      "eval_steps_per_second": 1.577,
+      "step": 4550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8138473626460394,
+      "learning_rate": 9.059178503841283e-06,
+      "loss": 0.665,
+      "step": 4551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2436454995749786,
+      "learning_rate": 9.058780057676836e-06,
+      "loss": 0.5608,
+      "step": 4552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.941181580999664,
+      "learning_rate": 9.058381537011537e-06,
+      "loss": 0.6363,
+      "step": 4553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.219377112255645,
+      "learning_rate": 9.057982941853733e-06,
+      "loss": 0.7297,
+      "step": 4554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8568232881568059,
+      "learning_rate": 9.057584272211765e-06,
+      "loss": 0.6327,
+      "step": 4555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9005998997571643,
+      "learning_rate": 9.05718552809398e-06,
+      "loss": 0.5445,
+      "step": 4556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.893380793405311,
+      "learning_rate": 9.056786709508723e-06,
+      "loss": 0.7285,
+      "step": 4557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8012370192472067,
+      "learning_rate": 9.056387816464348e-06,
+      "loss": 0.6528,
+      "step": 4558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.267019280846801,
+      "learning_rate": 9.055988848969203e-06,
+      "loss": 0.6248,
+      "step": 4559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.130689487564704,
+      "learning_rate": 9.05558980703164e-06,
+      "loss": 0.7409,
+      "step": 4560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1549566886997424,
+      "learning_rate": 9.055190690660011e-06,
+      "loss": 0.7779,
+      "step": 4561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8321865412847405,
+      "learning_rate": 9.054791499862676e-06,
+      "loss": 0.4379,
+      "step": 4562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9459545158475886,
+      "learning_rate": 9.054392234647987e-06,
+      "loss": 0.5198,
+      "step": 4563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7498750644040804,
+      "learning_rate": 9.053992895024306e-06,
+      "loss": 0.6392,
+      "step": 4564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0098429397268145,
+      "learning_rate": 9.05359348099999e-06,
+      "loss": 0.6379,
+      "step": 4565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0540349323245937,
+      "learning_rate": 9.053193992583402e-06,
+      "loss": 0.5579,
+      "step": 4566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0866461067633297,
+      "learning_rate": 9.052794429782906e-06,
+      "loss": 0.6222,
+      "step": 4567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8405686481207963,
+      "learning_rate": 9.052394792606862e-06,
+      "loss": 0.6261,
+      "step": 4568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.885734208905038,
+      "learning_rate": 9.051995081063642e-06,
+      "loss": 0.4479,
+      "step": 4569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6074593736797371,
+      "learning_rate": 9.05159529516161e-06,
+      "loss": 0.4993,
+      "step": 4570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2324418759424844,
+      "learning_rate": 9.051195434909137e-06,
+      "loss": 0.6212,
+      "step": 4571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1272134455364924,
+      "learning_rate": 9.050795500314592e-06,
+      "loss": 0.5926,
+      "step": 4572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8922867312428477,
+      "learning_rate": 9.050395491386353e-06,
+      "loss": 0.6221,
+      "step": 4573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9868624951963687,
+      "learning_rate": 9.049995408132786e-06,
+      "loss": 0.5636,
+      "step": 4574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2377672906146993,
+      "learning_rate": 9.049595250562272e-06,
+      "loss": 0.5678,
+      "step": 4575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.026970881785782,
+      "learning_rate": 9.049195018683183e-06,
+      "loss": 0.3538,
+      "step": 4576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5548784042353254,
+      "learning_rate": 9.048794712503904e-06,
+      "loss": 0.4014,
+      "step": 4577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.02813398463952,
+      "learning_rate": 9.04839433203281e-06,
+      "loss": 0.9299,
+      "step": 4578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1927305767843994,
+      "learning_rate": 9.047993877278284e-06,
+      "loss": 0.6759,
+      "step": 4579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8385074535331463,
+      "learning_rate": 9.047593348248711e-06,
+      "loss": 0.5776,
+      "step": 4580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8194271674783813,
+      "learning_rate": 9.047192744952475e-06,
+      "loss": 0.5501,
+      "step": 4581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.457221919846998,
+      "learning_rate": 9.046792067397962e-06,
+      "loss": 0.3896,
+      "step": 4582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4602049427592676,
+      "learning_rate": 9.04639131559356e-06,
+      "loss": 0.6009,
+      "step": 4583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0033259005810633,
+      "learning_rate": 9.045990489547661e-06,
+      "loss": 0.5446,
+      "step": 4584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1578806144378735,
+      "learning_rate": 9.045589589268651e-06,
+      "loss": 0.6281,
+      "step": 4585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.066817037063964,
+      "learning_rate": 9.045188614764924e-06,
+      "loss": 0.643,
+      "step": 4586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.161072072738963,
+      "learning_rate": 9.044787566044878e-06,
+      "loss": 0.6998,
+      "step": 4587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8252549810035155,
+      "learning_rate": 9.044386443116903e-06,
+      "loss": 0.547,
+      "step": 4588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7757439030423579,
+      "learning_rate": 9.043985245989402e-06,
+      "loss": 0.5169,
+      "step": 4589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.58374639247084,
+      "learning_rate": 9.043583974670771e-06,
+      "loss": 0.4713,
+      "step": 4590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5921688261487699,
+      "learning_rate": 9.04318262916941e-06,
+      "loss": 0.5916,
+      "step": 4591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.378916177855552,
+      "learning_rate": 9.04278120949372e-06,
+      "loss": 0.333,
+      "step": 4592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7736546939543871,
+      "learning_rate": 9.042379715652107e-06,
+      "loss": 0.4442,
+      "step": 4593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0774005384334946,
+      "learning_rate": 9.041978147652975e-06,
+      "loss": 0.652,
+      "step": 4594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1058619291945107,
+      "learning_rate": 9.041576505504732e-06,
+      "loss": 0.5371,
+      "step": 4595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.155596976112452,
+      "learning_rate": 9.041174789215782e-06,
+      "loss": 0.7119,
+      "step": 4596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7553569221057241,
+      "learning_rate": 9.04077299879454e-06,
+      "loss": 0.519,
+      "step": 4597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7728454389056232,
+      "learning_rate": 9.040371134249411e-06,
+      "loss": 0.5954,
+      "step": 4598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.232711097442537,
+      "learning_rate": 9.039969195588815e-06,
+      "loss": 0.5653,
+      "step": 4599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9781705654993378,
+      "learning_rate": 9.039567182821159e-06,
+      "loss": 0.4989,
+      "step": 4600
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.5779901742935181,
+      "eval_runtime": 67.9243,
+      "eval_samples_per_second": 50.203,
+      "eval_steps_per_second": 1.575,
+      "step": 4600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7409036832709177,
+      "learning_rate": 9.039165095954864e-06,
+      "loss": 0.5074,
+      "step": 4601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.094281893510301,
+      "learning_rate": 9.038762934998349e-06,
+      "loss": 0.5451,
+      "step": 4602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0582708272896615,
+      "learning_rate": 9.038360699960027e-06,
+      "loss": 0.4371,
+      "step": 4603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.581085103398753,
+      "learning_rate": 9.037958390848321e-06,
+      "loss": 0.4244,
+      "step": 4604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7400871321477565,
+      "learning_rate": 9.037556007671655e-06,
+      "loss": 0.6335,
+      "step": 4605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7921029782096598,
+      "learning_rate": 9.03715355043845e-06,
+      "loss": 0.5468,
+      "step": 4606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.977664084204341,
+      "learning_rate": 9.036751019157135e-06,
+      "loss": 0.6737,
+      "step": 4607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2943126265435447,
+      "learning_rate": 9.03634841383613e-06,
+      "loss": 0.787,
+      "step": 4608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9967256528268122,
+      "learning_rate": 9.03594573448387e-06,
+      "loss": 0.4576,
+      "step": 4609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.052455640942186,
+      "learning_rate": 9.035542981108783e-06,
+      "loss": 0.7398,
+      "step": 4610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8629432585314734,
+      "learning_rate": 9.035140153719299e-06,
+      "loss": 0.471,
+      "step": 4611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8555211474772133,
+      "learning_rate": 9.034737252323852e-06,
+      "loss": 0.4798,
+      "step": 4612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4920852056852687,
+      "learning_rate": 9.034334276930875e-06,
+      "loss": 0.6745,
+      "step": 4613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.210066313182855,
+      "learning_rate": 9.033931227548806e-06,
+      "loss": 0.6873,
+      "step": 4614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7444543227185842,
+      "learning_rate": 9.03352810418608e-06,
+      "loss": 0.6358,
+      "step": 4615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8200778062013183,
+      "learning_rate": 9.03312490685114e-06,
+      "loss": 0.5563,
+      "step": 4616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.024265058659937,
+      "learning_rate": 9.032721635552424e-06,
+      "loss": 0.7402,
+      "step": 4617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8167731243234762,
+      "learning_rate": 9.032318290298375e-06,
+      "loss": 0.4627,
+      "step": 4618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.182666738257856,
+      "learning_rate": 9.031914871097437e-06,
+      "loss": 0.611,
+      "step": 4619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.206225659545503,
+      "learning_rate": 9.031511377958053e-06,
+      "loss": 0.6803,
+      "step": 4620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.062812025958702,
+      "learning_rate": 9.031107810888674e-06,
+      "loss": 0.6809,
+      "step": 4621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.805128519725163,
+      "learning_rate": 9.030704169897746e-06,
+      "loss": 0.6269,
+      "step": 4622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9890637989588713,
+      "learning_rate": 9.030300454993718e-06,
+      "loss": 0.5551,
+      "step": 4623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.372729540127483,
+      "learning_rate": 9.029896666185042e-06,
+      "loss": 0.7547,
+      "step": 4624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8687198206152422,
+      "learning_rate": 9.029492803480174e-06,
+      "loss": 0.569,
+      "step": 4625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9139467898544869,
+      "learning_rate": 9.029088866887566e-06,
+      "loss": 0.6149,
+      "step": 4626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.043504084458244,
+      "learning_rate": 9.028684856415672e-06,
+      "loss": 0.6128,
+      "step": 4627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6711840108966278,
+      "learning_rate": 9.028280772072956e-06,
+      "loss": 0.4269,
+      "step": 4628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7555740192281477,
+      "learning_rate": 9.02787661386787e-06,
+      "loss": 0.4544,
+      "step": 4629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6853261248626634,
+      "learning_rate": 9.027472381808881e-06,
+      "loss": 0.4828,
+      "step": 4630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6232317132093546,
+      "learning_rate": 9.02706807590445e-06,
+      "loss": 0.5367,
+      "step": 4631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9623279255195727,
+      "learning_rate": 9.026663696163037e-06,
+      "loss": 0.6126,
+      "step": 4632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.066891596393154,
+      "learning_rate": 9.026259242593109e-06,
+      "loss": 0.6752,
+      "step": 4633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5082142682723987,
+      "learning_rate": 9.025854715203137e-06,
+      "loss": 0.6717,
+      "step": 4634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6241378964201865,
+      "learning_rate": 9.025450114001583e-06,
+      "loss": 0.4372,
+      "step": 4635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.570457937617633,
+      "learning_rate": 9.025045438996923e-06,
+      "loss": 0.3675,
+      "step": 4636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.424632993690636,
+      "learning_rate": 9.024640690197625e-06,
+      "loss": 0.4051,
+      "step": 4637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6523717830450526,
+      "learning_rate": 9.024235867612165e-06,
+      "loss": 0.6212,
+      "step": 4638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6304762477726615,
+      "learning_rate": 9.023830971249016e-06,
+      "loss": 0.4491,
+      "step": 4639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8789632909525515,
+      "learning_rate": 9.023426001116652e-06,
+      "loss": 0.5678,
+      "step": 4640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6406317671022443,
+      "learning_rate": 9.023020957223556e-06,
+      "loss": 0.4566,
+      "step": 4641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1214199421988456,
+      "learning_rate": 9.022615839578202e-06,
+      "loss": 0.5981,
+      "step": 4642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0012074921419925,
+      "learning_rate": 9.022210648189077e-06,
+      "loss": 0.6165,
+      "step": 4643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8362838796433036,
+      "learning_rate": 9.021805383064656e-06,
+      "loss": 0.4943,
+      "step": 4644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2381743094291893,
+      "learning_rate": 9.021400044213427e-06,
+      "loss": 0.6283,
+      "step": 4645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3235779644701404,
+      "learning_rate": 9.020994631643877e-06,
+      "loss": 0.7901,
+      "step": 4646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.617413792178472,
+      "learning_rate": 9.02058914536449e-06,
+      "loss": 0.6143,
+      "step": 4647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.711230104152889,
+      "learning_rate": 9.020183585383759e-06,
+      "loss": 0.4002,
+      "step": 4648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6695779129022572,
+      "learning_rate": 9.019777951710168e-06,
+      "loss": 0.5337,
+      "step": 4649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1150095456988813,
+      "learning_rate": 9.019372244352214e-06,
+      "loss": 0.733,
+      "step": 4650
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.5769872665405273,
+      "eval_runtime": 67.8689,
+      "eval_samples_per_second": 50.244,
+      "eval_steps_per_second": 1.577,
+      "step": 4650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8605159657772428,
+      "learning_rate": 9.018966463318386e-06,
+      "loss": 0.7509,
+      "step": 4651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9038091569812348,
+      "learning_rate": 9.01856060861718e-06,
+      "loss": 0.4587,
+      "step": 4652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8837392605070913,
+      "learning_rate": 9.018154680257095e-06,
+      "loss": 0.6522,
+      "step": 4653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.325840022892212,
+      "learning_rate": 9.017748678246628e-06,
+      "loss": 0.6308,
+      "step": 4654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.149198333064099,
+      "learning_rate": 9.017342602594276e-06,
+      "loss": 0.6336,
+      "step": 4655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5975538894178931,
+      "learning_rate": 9.016936453308542e-06,
+      "loss": 0.4575,
+      "step": 4656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8744756828423828,
+      "learning_rate": 9.016530230397925e-06,
+      "loss": 0.7068,
+      "step": 4657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.033586032209006,
+      "learning_rate": 9.016123933870936e-06,
+      "loss": 0.5598,
+      "step": 4658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.564446243140042,
+      "learning_rate": 9.015717563736074e-06,
+      "loss": 0.4355,
+      "step": 4659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9636145875075346,
+      "learning_rate": 9.015311120001848e-06,
+      "loss": 0.7143,
+      "step": 4660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7744293300904812,
+      "learning_rate": 9.014904602676768e-06,
+      "loss": 0.6462,
+      "step": 4661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7157223694696702,
+      "learning_rate": 9.014498011769344e-06,
+      "loss": 0.5702,
+      "step": 4662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8809509428331235,
+      "learning_rate": 9.014091347288085e-06,
+      "loss": 0.5832,
+      "step": 4663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2650053281687126,
+      "learning_rate": 9.013684609241508e-06,
+      "loss": 0.5607,
+      "step": 4664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5999430955056797,
+      "learning_rate": 9.013277797638127e-06,
+      "loss": 0.5029,
+      "step": 4665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.136776791332689,
+      "learning_rate": 9.012870912486455e-06,
+      "loss": 0.76,
+      "step": 4666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0620967450387315,
+      "learning_rate": 9.012463953795015e-06,
+      "loss": 0.6214,
+      "step": 4667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8603445512787664,
+      "learning_rate": 9.012056921572324e-06,
+      "loss": 0.6528,
+      "step": 4668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5433021560288966,
+      "learning_rate": 9.0116498158269e-06,
+      "loss": 0.3265,
+      "step": 4669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.134465333305093,
+      "learning_rate": 9.011242636567268e-06,
+      "loss": 0.6214,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0534933649747904,
+      "learning_rate": 9.010835383801954e-06,
+      "loss": 0.5545,
+      "step": 4671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2955286326601656,
+      "learning_rate": 9.010428057539483e-06,
+      "loss": 0.7744,
+      "step": 4672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9722423875231452,
+      "learning_rate": 9.01002065778838e-06,
+      "loss": 0.6497,
+      "step": 4673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7253688056614607,
+      "learning_rate": 9.009613184557173e-06,
+      "loss": 0.55,
+      "step": 4674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.419608493756481,
+      "learning_rate": 9.009205637854395e-06,
+      "loss": 0.7092,
+      "step": 4675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1139674677774583,
+      "learning_rate": 9.008798017688578e-06,
+      "loss": 0.5917,
+      "step": 4676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8258403897470268,
+      "learning_rate": 9.008390324068252e-06,
+      "loss": 0.4772,
+      "step": 4677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2412174982647595,
+      "learning_rate": 9.007982557001956e-06,
+      "loss": 0.5127,
+      "step": 4678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9694214847729035,
+      "learning_rate": 9.007574716498222e-06,
+      "loss": 0.429,
+      "step": 4679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7865660768357983,
+      "learning_rate": 9.00716680256559e-06,
+      "loss": 0.5138,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.124220041049461,
+      "learning_rate": 9.006758815212601e-06,
+      "loss": 0.484,
+      "step": 4681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9542585211809207,
+      "learning_rate": 9.006350754447793e-06,
+      "loss": 0.5316,
+      "step": 4682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6941874986352834,
+      "learning_rate": 9.005942620279711e-06,
+      "loss": 0.4851,
+      "step": 4683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2750240619624025,
+      "learning_rate": 9.005534412716897e-06,
+      "loss": 0.5933,
+      "step": 4684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7853983451017914,
+      "learning_rate": 9.005126131767897e-06,
+      "loss": 0.5312,
+      "step": 4685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.676938440293407,
+      "learning_rate": 9.00471777744126e-06,
+      "loss": 0.4553,
+      "step": 4686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6643985687713359,
+      "learning_rate": 9.00430934974553e-06,
+      "loss": 0.4352,
+      "step": 4687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2200334797907093,
+      "learning_rate": 9.003900848689263e-06,
+      "loss": 0.4979,
+      "step": 4688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1704237868931933,
+      "learning_rate": 9.003492274281007e-06,
+      "loss": 0.6313,
+      "step": 4689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.116400197520601,
+      "learning_rate": 9.003083626529318e-06,
+      "loss": 0.6819,
+      "step": 4690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4815919627822223,
+      "learning_rate": 9.002674905442747e-06,
+      "loss": 0.6387,
+      "step": 4691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.354216397505495,
+      "learning_rate": 9.002266111029853e-06,
+      "loss": 0.6892,
+      "step": 4692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0253415217811845,
+      "learning_rate": 9.001857243299194e-06,
+      "loss": 0.7368,
+      "step": 4693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9365480702420752,
+      "learning_rate": 9.001448302259331e-06,
+      "loss": 0.6435,
+      "step": 4694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.063021498458281,
+      "learning_rate": 9.001039287918819e-06,
+      "loss": 0.5712,
+      "step": 4695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7437565884893829,
+      "learning_rate": 9.000630200286224e-06,
+      "loss": 0.4782,
+      "step": 4696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1893368494470673,
+      "learning_rate": 9.000221039370111e-06,
+      "loss": 0.729,
+      "step": 4697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8592373271915135,
+      "learning_rate": 8.999811805179045e-06,
+      "loss": 0.658,
+      "step": 4698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8107179529319581,
+      "learning_rate": 8.999402497721594e-06,
+      "loss": 0.5976,
+      "step": 4699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.128478392895859,
+      "learning_rate": 8.998993117006323e-06,
+      "loss": 0.6316,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.5755705833435059,
+      "eval_runtime": 67.8493,
+      "eval_samples_per_second": 50.258,
+      "eval_steps_per_second": 1.577,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7581060795922214,
+      "learning_rate": 8.998583663041805e-06,
+      "loss": 0.5656,
+      "step": 4701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.996322972716169,
+      "learning_rate": 8.998174135836612e-06,
+      "loss": 0.538,
+      "step": 4702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7785433623931797,
+      "learning_rate": 8.997764535399316e-06,
+      "loss": 0.5841,
+      "step": 4703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.030230323114437,
+      "learning_rate": 8.997354861738492e-06,
+      "loss": 0.7963,
+      "step": 4704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.055012741515007,
+      "learning_rate": 8.996945114862714e-06,
+      "loss": 0.6248,
+      "step": 4705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.774736821687285,
+      "learning_rate": 8.996535294780564e-06,
+      "loss": 0.5273,
+      "step": 4706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8222439432476485,
+      "learning_rate": 8.996125401500619e-06,
+      "loss": 0.6688,
+      "step": 4707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7713901378034724,
+      "learning_rate": 8.995715435031462e-06,
+      "loss": 0.6271,
+      "step": 4708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8613475986251897,
+      "learning_rate": 8.99530539538167e-06,
+      "loss": 0.5612,
+      "step": 4709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8778393715074293,
+      "learning_rate": 8.994895282559834e-06,
+      "loss": 0.5434,
+      "step": 4710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8287525045581,
+      "learning_rate": 8.994485096574535e-06,
+      "loss": 0.5673,
+      "step": 4711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9791217205993756,
+      "learning_rate": 8.99407483743436e-06,
+      "loss": 0.5994,
+      "step": 4712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9124036930518986,
+      "learning_rate": 8.9936645051479e-06,
+      "loss": 0.6883,
+      "step": 4713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.703274341288442,
+      "learning_rate": 8.993254099723743e-06,
+      "loss": 0.5495,
+      "step": 4714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1063254342159174,
+      "learning_rate": 8.992843621170481e-06,
+      "loss": 0.5381,
+      "step": 4715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6459429097518679,
+      "learning_rate": 8.992433069496708e-06,
+      "loss": 0.4923,
+      "step": 4716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7829967716063833,
+      "learning_rate": 8.992022444711017e-06,
+      "loss": 0.5365,
+      "step": 4717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7638778547024054,
+      "learning_rate": 8.991611746822004e-06,
+      "loss": 0.4678,
+      "step": 4718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4905701579785853,
+      "learning_rate": 8.99120097583827e-06,
+      "loss": 0.4034,
+      "step": 4719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0216488984052208,
+      "learning_rate": 8.990790131768409e-06,
+      "loss": 0.5461,
+      "step": 4720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6186646561259908,
+      "learning_rate": 8.990379214621027e-06,
+      "loss": 0.5429,
+      "step": 4721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8164846050236574,
+      "learning_rate": 8.989968224404726e-06,
+      "loss": 0.5051,
+      "step": 4722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6058847527120022,
+      "learning_rate": 8.989557161128105e-06,
+      "loss": 0.4369,
+      "step": 4723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8059954979972275,
+      "learning_rate": 8.989146024799774e-06,
+      "loss": 0.6795,
+      "step": 4724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.788905493377423,
+      "learning_rate": 8.988734815428336e-06,
+      "loss": 0.4763,
+      "step": 4725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.171524193331409,
+      "learning_rate": 8.988323533022404e-06,
+      "loss": 0.7113,
+      "step": 4726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.516809554696681,
+      "learning_rate": 8.987912177590585e-06,
+      "loss": 0.3894,
+      "step": 4727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.97538403144154,
+      "learning_rate": 8.98750074914149e-06,
+      "loss": 0.4896,
+      "step": 4728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.160966372980551,
+      "learning_rate": 8.987089247683734e-06,
+      "loss": 0.6683,
+      "step": 4729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6053445441385128,
+      "learning_rate": 8.98667767322593e-06,
+      "loss": 0.5794,
+      "step": 4730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.860187340900442,
+      "learning_rate": 8.986266025776694e-06,
+      "loss": 0.6576,
+      "step": 4731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1195643195620426,
+      "learning_rate": 8.985854305344645e-06,
+      "loss": 0.564,
+      "step": 4732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2479070808490227,
+      "learning_rate": 8.9854425119384e-06,
+      "loss": 0.5683,
+      "step": 4733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.548816291936887,
+      "learning_rate": 8.985030645566583e-06,
+      "loss": 0.6052,
+      "step": 4734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.971498939186074,
+      "learning_rate": 8.984618706237812e-06,
+      "loss": 0.5178,
+      "step": 4735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.107128598420581,
+      "learning_rate": 8.984206693960715e-06,
+      "loss": 0.5528,
+      "step": 4736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8347493964108943,
+      "learning_rate": 8.983794608743913e-06,
+      "loss": 0.4311,
+      "step": 4737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5929843914269748,
+      "learning_rate": 8.983382450596036e-06,
+      "loss": 0.4225,
+      "step": 4738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5459824627566359,
+      "learning_rate": 8.98297021952571e-06,
+      "loss": 0.4297,
+      "step": 4739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7072249958728607,
+      "learning_rate": 8.982557915541565e-06,
+      "loss": 0.5151,
+      "step": 4740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.979670100298214,
+      "learning_rate": 8.982145538652235e-06,
+      "loss": 0.6859,
+      "step": 4741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.873970046327718,
+      "learning_rate": 8.981733088866351e-06,
+      "loss": 0.548,
+      "step": 4742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9513645347115633,
+      "learning_rate": 8.981320566192544e-06,
+      "loss": 0.7404,
+      "step": 4743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3990725048480286,
+      "learning_rate": 8.980907970639457e-06,
+      "loss": 0.7387,
+      "step": 4744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9986467635497458,
+      "learning_rate": 8.980495302215722e-06,
+      "loss": 0.5411,
+      "step": 4745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9059522196103516,
+      "learning_rate": 8.980082560929981e-06,
+      "loss": 0.5813,
+      "step": 4746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8806211064522689,
+      "learning_rate": 8.979669746790874e-06,
+      "loss": 0.4509,
+      "step": 4747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0605580259179277,
+      "learning_rate": 8.97925685980704e-06,
+      "loss": 0.5838,
+      "step": 4748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.303417020099469,
+      "learning_rate": 8.978843899987125e-06,
+      "loss": 0.8487,
+      "step": 4749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.982885259999542,
+      "learning_rate": 8.978430867339773e-06,
+      "loss": 0.5724,
+      "step": 4750
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.5758548378944397,
+      "eval_runtime": 67.8496,
+      "eval_samples_per_second": 50.258,
+      "eval_steps_per_second": 1.577,
+      "step": 4750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.541285385504332,
+      "learning_rate": 8.978017761873635e-06,
+      "loss": 0.4648,
+      "step": 4751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1934524625886613,
+      "learning_rate": 8.977604583597352e-06,
+      "loss": 0.5658,
+      "step": 4752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0973169271508225,
+      "learning_rate": 8.977191332519576e-06,
+      "loss": 0.5822,
+      "step": 4753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9157459180098906,
+      "learning_rate": 8.976778008648962e-06,
+      "loss": 0.6013,
+      "step": 4754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9601211413672655,
+      "learning_rate": 8.97636461199416e-06,
+      "loss": 0.5638,
+      "step": 4755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7442304146801024,
+      "learning_rate": 8.975951142563822e-06,
+      "loss": 0.4986,
+      "step": 4756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.782639951544733,
+      "learning_rate": 8.975537600366608e-06,
+      "loss": 0.4836,
+      "step": 4757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6860939009140659,
+      "learning_rate": 8.975123985411171e-06,
+      "loss": 0.5016,
+      "step": 4758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.469152729991845,
+      "learning_rate": 8.974710297706172e-06,
+      "loss": 0.4593,
+      "step": 4759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.609219648377936,
+      "learning_rate": 8.974296537260273e-06,
+      "loss": 0.4422,
+      "step": 4760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7261353606739975,
+      "learning_rate": 8.973882704082132e-06,
+      "loss": 0.6465,
+      "step": 4761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6608623556912812,
+      "learning_rate": 8.973468798180417e-06,
+      "loss": 0.5192,
+      "step": 4762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4613789748840806,
+      "learning_rate": 8.973054819563787e-06,
+      "loss": 0.3102,
+      "step": 4763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8897559580040912,
+      "learning_rate": 8.972640768240913e-06,
+      "loss": 0.6403,
+      "step": 4764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1551290589795626,
+      "learning_rate": 8.972226644220462e-06,
+      "loss": 0.6477,
+      "step": 4765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9276693667039515,
+      "learning_rate": 8.971812447511102e-06,
+      "loss": 0.5831,
+      "step": 4766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6385445354080095,
+      "learning_rate": 8.971398178121504e-06,
+      "loss": 0.4626,
+      "step": 4767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2457772063321535,
+      "learning_rate": 8.970983836060343e-06,
+      "loss": 0.8485,
+      "step": 4768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2472255764407443,
+      "learning_rate": 8.97056942133629e-06,
+      "loss": 0.5022,
+      "step": 4769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0095389606611636,
+      "learning_rate": 8.970154933958025e-06,
+      "loss": 0.5462,
+      "step": 4770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5686668931063548,
+      "learning_rate": 8.969740373934217e-06,
+      "loss": 0.3941,
+      "step": 4771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1901443438077517,
+      "learning_rate": 8.969325741273553e-06,
+      "loss": 0.5288,
+      "step": 4772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.120392861622124,
+      "learning_rate": 8.968911035984709e-06,
+      "loss": 0.6766,
+      "step": 4773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7113310493741842,
+      "learning_rate": 8.968496258076365e-06,
+      "loss": 0.3597,
+      "step": 4774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9845601242229274,
+      "learning_rate": 8.968081407557207e-06,
+      "loss": 0.6723,
+      "step": 4775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.906553265362174,
+      "learning_rate": 8.967666484435921e-06,
+      "loss": 0.4972,
+      "step": 4776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9450585555999194,
+      "learning_rate": 8.967251488721189e-06,
+      "loss": 0.5959,
+      "step": 4777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3806445024381473,
+      "learning_rate": 8.9668364204217e-06,
+      "loss": 0.5382,
+      "step": 4778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2259255201306627,
+      "learning_rate": 8.966421279546145e-06,
+      "loss": 0.5486,
+      "step": 4779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.903388271726539,
+      "learning_rate": 8.966006066103215e-06,
+      "loss": 0.582,
+      "step": 4780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7317298486952688,
+      "learning_rate": 8.965590780101598e-06,
+      "loss": 0.51,
+      "step": 4781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9680972866174022,
+      "learning_rate": 8.965175421549994e-06,
+      "loss": 0.5246,
+      "step": 4782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9714138493203084,
+      "learning_rate": 8.96475999045709e-06,
+      "loss": 0.7269,
+      "step": 4783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.54257307798041,
+      "learning_rate": 8.96434448683159e-06,
+      "loss": 0.7797,
+      "step": 4784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.252903699458257,
+      "learning_rate": 8.963928910682192e-06,
+      "loss": 0.6571,
+      "step": 4785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.040786817437164,
+      "learning_rate": 8.963513262017593e-06,
+      "loss": 0.6034,
+      "step": 4786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8748340903955185,
+      "learning_rate": 8.963097540846494e-06,
+      "loss": 0.5268,
+      "step": 4787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6776982187692961,
+      "learning_rate": 8.962681747177598e-06,
+      "loss": 0.6283,
+      "step": 4788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2201485360922972,
+      "learning_rate": 8.962265881019612e-06,
+      "loss": 0.5462,
+      "step": 4789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.40963182517304,
+      "learning_rate": 8.96184994238124e-06,
+      "loss": 0.3932,
+      "step": 4790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.058039532464517,
+      "learning_rate": 8.96143393127119e-06,
+      "loss": 0.5923,
+      "step": 4791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5485778603279943,
+      "learning_rate": 8.96101784769817e-06,
+      "loss": 0.4234,
+      "step": 4792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1159353378179713,
+      "learning_rate": 8.960601691670892e-06,
+      "loss": 0.5159,
+      "step": 4793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.315829683993382,
+      "learning_rate": 8.960185463198065e-06,
+      "loss": 0.6681,
+      "step": 4794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0486117558336336,
+      "learning_rate": 8.959769162288406e-06,
+      "loss": 0.6867,
+      "step": 4795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6317295093814914,
+      "learning_rate": 8.959352788950629e-06,
+      "loss": 0.6145,
+      "step": 4796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5094150902990455,
+      "learning_rate": 8.958936343193449e-06,
+      "loss": 0.4511,
+      "step": 4797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.710502900644084,
+      "learning_rate": 8.958519825025586e-06,
+      "loss": 0.455,
+      "step": 4798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.545370584858076,
+      "learning_rate": 8.958103234455757e-06,
+      "loss": 0.4242,
+      "step": 4799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.137050731732604,
+      "learning_rate": 8.957686571492686e-06,
+      "loss": 0.5259,
+      "step": 4800
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.5753394961357117,
+      "eval_runtime": 67.9361,
+      "eval_samples_per_second": 50.194,
+      "eval_steps_per_second": 1.575,
+      "step": 4800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0212817113922354,
+      "learning_rate": 8.957269836145094e-06,
+      "loss": 0.5075,
+      "step": 4801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6492960139497483,
+      "learning_rate": 8.956853028421706e-06,
+      "loss": 0.4185,
+      "step": 4802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0470978847534895,
+      "learning_rate": 8.956436148331247e-06,
+      "loss": 0.5826,
+      "step": 4803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5101454831418226,
+      "learning_rate": 8.956019195882445e-06,
+      "loss": 0.4682,
+      "step": 4804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9167169135890756,
+      "learning_rate": 8.955602171084025e-06,
+      "loss": 0.519,
+      "step": 4805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0323412473743945,
+      "learning_rate": 8.95518507394472e-06,
+      "loss": 0.5874,
+      "step": 4806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.636864280899432,
+      "learning_rate": 8.954767904473264e-06,
+      "loss": 0.5313,
+      "step": 4807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1631878795494077,
+      "learning_rate": 8.954350662678386e-06,
+      "loss": 0.7693,
+      "step": 4808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6398270003114126,
+      "learning_rate": 8.953933348568823e-06,
+      "loss": 0.4611,
+      "step": 4809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.069547772010378,
+      "learning_rate": 8.953515962153311e-06,
+      "loss": 0.6022,
+      "step": 4810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.448411594460862,
+      "learning_rate": 8.953098503440587e-06,
+      "loss": 0.3871,
+      "step": 4811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4618358301491514,
+      "learning_rate": 8.952680972439392e-06,
+      "loss": 0.3688,
+      "step": 4812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8674032521754749,
+      "learning_rate": 8.952263369158462e-06,
+      "loss": 0.6103,
+      "step": 4813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7929693610420776,
+      "learning_rate": 8.951845693606545e-06,
+      "loss": 0.4638,
+      "step": 4814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.029728183122345,
+      "learning_rate": 8.951427945792382e-06,
+      "loss": 0.6306,
+      "step": 4815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.03009839116368,
+      "learning_rate": 8.95101012572472e-06,
+      "loss": 0.5532,
+      "step": 4816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0656197605095743,
+      "learning_rate": 8.950592233412302e-06,
+      "loss": 0.4835,
+      "step": 4817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.201069226776293,
+      "learning_rate": 8.950174268863882e-06,
+      "loss": 0.6955,
+      "step": 4818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7066241388288463,
+      "learning_rate": 8.949756232088205e-06,
+      "loss": 0.6468,
+      "step": 4819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2588831854592453,
+      "learning_rate": 8.949338123094023e-06,
+      "loss": 0.7776,
+      "step": 4820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8441665236261042,
+      "learning_rate": 8.94891994189009e-06,
+      "loss": 0.5525,
+      "step": 4821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5028189717961673,
+      "learning_rate": 8.948501688485161e-06,
+      "loss": 0.4454,
+      "step": 4822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2970723001449804,
+      "learning_rate": 8.948083362887992e-06,
+      "loss": 0.6284,
+      "step": 4823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7925617351319965,
+      "learning_rate": 8.947664965107338e-06,
+      "loss": 0.5675,
+      "step": 4824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7858123091933464,
+      "learning_rate": 8.947246495151961e-06,
+      "loss": 0.4931,
+      "step": 4825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.61169070164803,
+      "learning_rate": 8.946827953030617e-06,
+      "loss": 0.5577,
+      "step": 4826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6663048464645107,
+      "learning_rate": 8.946409338752071e-06,
+      "loss": 0.5099,
+      "step": 4827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7695484741567673,
+      "learning_rate": 8.94599065232509e-06,
+      "loss": 0.6038,
+      "step": 4828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8928594038734754,
+      "learning_rate": 8.945571893758432e-06,
+      "loss": 0.4606,
+      "step": 4829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1137631863531188,
+      "learning_rate": 8.945153063060868e-06,
+      "loss": 0.744,
+      "step": 4830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0819730285340934,
+      "learning_rate": 8.944734160241163e-06,
+      "loss": 0.5189,
+      "step": 4831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7449237349492799,
+      "learning_rate": 8.944315185308087e-06,
+      "loss": 0.4901,
+      "step": 4832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7708528762017441,
+      "learning_rate": 8.943896138270415e-06,
+      "loss": 0.7535,
+      "step": 4833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3085009037453093,
+      "learning_rate": 8.943477019136916e-06,
+      "loss": 0.5337,
+      "step": 4834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2733928617458896,
+      "learning_rate": 8.943057827916362e-06,
+      "loss": 0.7849,
+      "step": 4835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9343835571848733,
+      "learning_rate": 8.942638564617533e-06,
+      "loss": 0.5496,
+      "step": 4836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7040754672556586,
+      "learning_rate": 8.942219229249202e-06,
+      "loss": 0.4502,
+      "step": 4837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.185743184557637,
+      "learning_rate": 8.941799821820151e-06,
+      "loss": 0.6286,
+      "step": 4838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.350816838588975,
+      "learning_rate": 8.94138034233916e-06,
+      "loss": 0.6374,
+      "step": 4839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9836692477806848,
+      "learning_rate": 8.940960790815007e-06,
+      "loss": 0.4898,
+      "step": 4840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8796402206875047,
+      "learning_rate": 8.940541167256477e-06,
+      "loss": 0.6237,
+      "step": 4841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7682337172299114,
+      "learning_rate": 8.940121471672357e-06,
+      "loss": 0.6208,
+      "step": 4842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7057839697210626,
+      "learning_rate": 8.939701704071427e-06,
+      "loss": 0.5122,
+      "step": 4843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2672809213089997,
+      "learning_rate": 8.939281864462479e-06,
+      "loss": 0.678,
+      "step": 4844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.263462888274442,
+      "learning_rate": 8.938861952854305e-06,
+      "loss": 0.5661,
+      "step": 4845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.353227838719256,
+      "learning_rate": 8.938441969255688e-06,
+      "loss": 0.6856,
+      "step": 4846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9897875013396027,
+      "learning_rate": 8.938021913675427e-06,
+      "loss": 0.5142,
+      "step": 4847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2462142099977815,
+      "learning_rate": 8.93760178612231e-06,
+      "loss": 0.7062,
+      "step": 4848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.175113620963502,
+      "learning_rate": 8.937181586605135e-06,
+      "loss": 0.6647,
+      "step": 4849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0380618758785243,
+      "learning_rate": 8.9367613151327e-06,
+      "loss": 0.5277,
+      "step": 4850
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.5733934640884399,
+      "eval_runtime": 67.909,
+      "eval_samples_per_second": 50.214,
+      "eval_steps_per_second": 1.576,
+      "step": 4850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5058477816213351,
+      "learning_rate": 8.936340971713802e-06,
+      "loss": 0.384,
+      "step": 4851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.768152271861173,
+      "learning_rate": 8.935920556357239e-06,
+      "loss": 0.6081,
+      "step": 4852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.716248358691281,
+      "learning_rate": 8.935500069071813e-06,
+      "loss": 0.5066,
+      "step": 4853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.700115430892141,
+      "learning_rate": 8.935079509866327e-06,
+      "loss": 0.3762,
+      "step": 4854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.796966820442157,
+      "learning_rate": 8.934658878749586e-06,
+      "loss": 0.4558,
+      "step": 4855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7291277723315508,
+      "learning_rate": 8.934238175730392e-06,
+      "loss": 0.5252,
+      "step": 4856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.76733093173104,
+      "learning_rate": 8.933817400817557e-06,
+      "loss": 0.4668,
+      "step": 4857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1990146711213447,
+      "learning_rate": 8.933396554019889e-06,
+      "loss": 0.6827,
+      "step": 4858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9915314937893605,
+      "learning_rate": 8.932975635346196e-06,
+      "loss": 0.6773,
+      "step": 4859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5237544649870234,
+      "learning_rate": 8.932554644805289e-06,
+      "loss": 0.567,
+      "step": 4860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9170013428084605,
+      "learning_rate": 8.932133582405983e-06,
+      "loss": 0.5657,
+      "step": 4861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9882935486017814,
+      "learning_rate": 8.931712448157091e-06,
+      "loss": 0.4322,
+      "step": 4862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2771332715835126,
+      "learning_rate": 8.931291242067433e-06,
+      "loss": 0.581,
+      "step": 4863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.620878484163941,
+      "learning_rate": 8.930869964145821e-06,
+      "loss": 0.4272,
+      "step": 4864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8390832664881434,
+      "learning_rate": 8.93044861440108e-06,
+      "loss": 0.621,
+      "step": 4865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9183347585757144,
+      "learning_rate": 8.930027192842027e-06,
+      "loss": 0.6504,
+      "step": 4866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.19451479827647,
+      "learning_rate": 8.929605699477487e-06,
+      "loss": 0.7028,
+      "step": 4867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7518811689806988,
+      "learning_rate": 8.929184134316278e-06,
+      "loss": 0.4783,
+      "step": 4868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9539505704336702,
+      "learning_rate": 8.928762497367233e-06,
+      "loss": 0.536,
+      "step": 4869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9911456768677276,
+      "learning_rate": 8.928340788639173e-06,
+      "loss": 0.4518,
+      "step": 4870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0091703490271993,
+      "learning_rate": 8.927919008140927e-06,
+      "loss": 0.4918,
+      "step": 4871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5782859562104246,
+      "learning_rate": 8.927497155881327e-06,
+      "loss": 0.4535,
+      "step": 4872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0215356848314783,
+      "learning_rate": 8.927075231869204e-06,
+      "loss": 0.5782,
+      "step": 4873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8976575588307238,
+      "learning_rate": 8.926653236113387e-06,
+      "loss": 0.5757,
+      "step": 4874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8420662223393787,
+      "learning_rate": 8.926231168622716e-06,
+      "loss": 0.451,
+      "step": 4875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.789258179465613,
+      "learning_rate": 8.92580902940602e-06,
+      "loss": 0.5779,
+      "step": 4876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5340618994653716,
+      "learning_rate": 8.92538681847214e-06,
+      "loss": 0.444,
+      "step": 4877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.677171971628331,
+      "learning_rate": 8.924964535829917e-06,
+      "loss": 0.5112,
+      "step": 4878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6730983459075228,
+      "learning_rate": 8.924542181488189e-06,
+      "loss": 0.4714,
+      "step": 4879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7755837148669884,
+      "learning_rate": 8.924119755455797e-06,
+      "loss": 0.5249,
+      "step": 4880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6693356254531966,
+      "learning_rate": 8.923697257741584e-06,
+      "loss": 0.6777,
+      "step": 4881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.522528307759978,
+      "learning_rate": 8.923274688354396e-06,
+      "loss": 0.4096,
+      "step": 4882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.106408465256389,
+      "learning_rate": 8.92285204730308e-06,
+      "loss": 0.6689,
+      "step": 4883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.970833423694348,
+      "learning_rate": 8.92242933459648e-06,
+      "loss": 0.5849,
+      "step": 4884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.95778759955619,
+      "learning_rate": 8.922006550243448e-06,
+      "loss": 0.4919,
+      "step": 4885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.073121479948156,
+      "learning_rate": 8.921583694252833e-06,
+      "loss": 0.5458,
+      "step": 4886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4287604601096497,
+      "learning_rate": 8.921160766633492e-06,
+      "loss": 0.3994,
+      "step": 4887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.944920833767852,
+      "learning_rate": 8.920737767394273e-06,
+      "loss": 0.7431,
+      "step": 4888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.048029149523829,
+      "learning_rate": 8.920314696544034e-06,
+      "loss": 0.5689,
+      "step": 4889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3108638093461917,
+      "learning_rate": 8.919891554091632e-06,
+      "loss": 0.674,
+      "step": 4890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9247805966441462,
+      "learning_rate": 8.919468340045924e-06,
+      "loss": 0.5184,
+      "step": 4891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8569381444112907,
+      "learning_rate": 8.919045054415768e-06,
+      "loss": 0.6334,
+      "step": 4892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9917933513614559,
+      "learning_rate": 8.918621697210031e-06,
+      "loss": 0.5874,
+      "step": 4893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9078495074029176,
+      "learning_rate": 8.91819826843757e-06,
+      "loss": 0.627,
+      "step": 4894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9178744203159956,
+      "learning_rate": 8.917774768107254e-06,
+      "loss": 0.5545,
+      "step": 4895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1484831391179866,
+      "learning_rate": 8.917351196227943e-06,
+      "loss": 0.6194,
+      "step": 4896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9818185143780125,
+      "learning_rate": 8.91692755280851e-06,
+      "loss": 0.6403,
+      "step": 4897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8759172098241517,
+      "learning_rate": 8.916503837857819e-06,
+      "loss": 0.6216,
+      "step": 4898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1124723401974093,
+      "learning_rate": 8.916080051384742e-06,
+      "loss": 0.6775,
+      "step": 4899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1869986974542255,
+      "learning_rate": 8.915656193398152e-06,
+      "loss": 0.5554,
+      "step": 4900
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5732719898223877,
+      "eval_runtime": 67.8808,
+      "eval_samples_per_second": 50.235,
+      "eval_steps_per_second": 1.576,
+      "step": 4900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9820360942172812,
+      "learning_rate": 8.91523226390692e-06,
+      "loss": 0.5149,
+      "step": 4901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8194721903988624,
+      "learning_rate": 8.914808262919923e-06,
+      "loss": 0.4587,
+      "step": 4902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.774909531125132,
+      "learning_rate": 8.914384190446036e-06,
+      "loss": 0.7247,
+      "step": 4903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8012260016827175,
+      "learning_rate": 8.913960046494137e-06,
+      "loss": 0.5188,
+      "step": 4904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7598915750757613,
+      "learning_rate": 8.913535831073103e-06,
+      "loss": 0.496,
+      "step": 4905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3836848114279627,
+      "learning_rate": 8.913111544191819e-06,
+      "loss": 0.6176,
+      "step": 4906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4147883015039846,
+      "learning_rate": 8.912687185859164e-06,
+      "loss": 0.6579,
+      "step": 4907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6669743765827687,
+      "learning_rate": 8.912262756084022e-06,
+      "loss": 0.4302,
+      "step": 4908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7427151663290124,
+      "learning_rate": 8.911838254875277e-06,
+      "loss": 0.4919,
+      "step": 4909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.011066603160021,
+      "learning_rate": 8.91141368224182e-06,
+      "loss": 0.5669,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4973805474288402,
+      "learning_rate": 8.910989038192534e-06,
+      "loss": 0.3952,
+      "step": 4911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9087756846298223,
+      "learning_rate": 8.910564322736315e-06,
+      "loss": 0.6402,
+      "step": 4912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5894915899747777,
+      "learning_rate": 8.910139535882048e-06,
+      "loss": 0.5384,
+      "step": 4913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7791268882555737,
+      "learning_rate": 8.909714677638627e-06,
+      "loss": 0.38,
+      "step": 4914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.866823760445612,
+      "learning_rate": 8.909289748014948e-06,
+      "loss": 0.5432,
+      "step": 4915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.865480287704516,
+      "learning_rate": 8.908864747019907e-06,
+      "loss": 0.5671,
+      "step": 4916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2598601476836007,
+      "learning_rate": 8.908439674662397e-06,
+      "loss": 0.5481,
+      "step": 4917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0494063669656373,
+      "learning_rate": 8.90801453095132e-06,
+      "loss": 0.6388,
+      "step": 4918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0117722813871843,
+      "learning_rate": 8.907589315895577e-06,
+      "loss": 0.5546,
+      "step": 4919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6502328532838058,
+      "learning_rate": 8.907164029504068e-06,
+      "loss": 0.3869,
+      "step": 4920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1918927136629063,
+      "learning_rate": 8.906738671785696e-06,
+      "loss": 0.6587,
+      "step": 4921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9808270591927513,
+      "learning_rate": 8.906313242749367e-06,
+      "loss": 0.5508,
+      "step": 4922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7334520454051605,
+      "learning_rate": 8.905887742403985e-06,
+      "loss": 0.5781,
+      "step": 4923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.105262280203618,
+      "learning_rate": 8.905462170758458e-06,
+      "loss": 0.7668,
+      "step": 4924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9960149269117322,
+      "learning_rate": 8.905036527821697e-06,
+      "loss": 0.5784,
+      "step": 4925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7118855587297215,
+      "learning_rate": 8.90461081360261e-06,
+      "loss": 0.5536,
+      "step": 4926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9266717183434874,
+      "learning_rate": 8.90418502811011e-06,
+      "loss": 0.5682,
+      "step": 4927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0965026538748353,
+      "learning_rate": 8.903759171353112e-06,
+      "loss": 0.7317,
+      "step": 4928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6871933814247742,
+      "learning_rate": 8.903333243340531e-06,
+      "loss": 0.4513,
+      "step": 4929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.794600913455041,
+      "learning_rate": 8.90290724408128e-06,
+      "loss": 0.4206,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9587056634202358,
+      "learning_rate": 8.902481173584281e-06,
+      "loss": 0.5439,
+      "step": 4931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.47356517947273,
+      "learning_rate": 8.902055031858452e-06,
+      "loss": 0.4597,
+      "step": 4932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5087621076656763,
+      "learning_rate": 8.901628818912712e-06,
+      "loss": 0.433,
+      "step": 4933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3294570080218153,
+      "learning_rate": 8.901202534755987e-06,
+      "loss": 0.6665,
+      "step": 4934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1086582038494286,
+      "learning_rate": 8.9007761793972e-06,
+      "loss": 0.5259,
+      "step": 4935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9979274188859593,
+      "learning_rate": 8.900349752845276e-06,
+      "loss": 0.7272,
+      "step": 4936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7788139565214245,
+      "learning_rate": 8.899923255109141e-06,
+      "loss": 0.5136,
+      "step": 4937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1264066131286805,
+      "learning_rate": 8.899496686197726e-06,
+      "loss": 0.5528,
+      "step": 4938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.298047349319387,
+      "learning_rate": 8.899070046119957e-06,
+      "loss": 0.6387,
+      "step": 4939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.640808483521445,
+      "learning_rate": 8.89864333488477e-06,
+      "loss": 0.4407,
+      "step": 4940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7647891610648678,
+      "learning_rate": 8.898216552501095e-06,
+      "loss": 0.5269,
+      "step": 4941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.146727839779798,
+      "learning_rate": 8.897789698977866e-06,
+      "loss": 0.6452,
+      "step": 4942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5792581098732001,
+      "learning_rate": 8.89736277432402e-06,
+      "loss": 0.5949,
+      "step": 4943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3220909969787282,
+      "learning_rate": 8.896935778548497e-06,
+      "loss": 0.7928,
+      "step": 4944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1363958708937667,
+      "learning_rate": 8.896508711660232e-06,
+      "loss": 0.6901,
+      "step": 4945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9578907722195342,
+      "learning_rate": 8.896081573668167e-06,
+      "loss": 0.5335,
+      "step": 4946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7015969722166528,
+      "learning_rate": 8.895654364581242e-06,
+      "loss": 0.5726,
+      "step": 4947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6148345513979052,
+      "learning_rate": 8.895227084408405e-06,
+      "loss": 0.522,
+      "step": 4948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.073020629352209,
+      "learning_rate": 8.894799733158596e-06,
+      "loss": 0.6246,
+      "step": 4949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.044657166227204,
+      "learning_rate": 8.894372310840762e-06,
+      "loss": 0.651,
+      "step": 4950
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5739993453025818,
+      "eval_runtime": 67.9492,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 4950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.048250142102438,
+      "learning_rate": 8.893944817463854e-06,
+      "loss": 0.7083,
+      "step": 4951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3276753335490423,
+      "learning_rate": 8.893517253036819e-06,
+      "loss": 0.7392,
+      "step": 4952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5898943715970035,
+      "learning_rate": 8.893089617568608e-06,
+      "loss": 0.5485,
+      "step": 4953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6541089813671712,
+      "learning_rate": 8.892661911068173e-06,
+      "loss": 0.5076,
+      "step": 4954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7042591289102234,
+      "learning_rate": 8.892234133544467e-06,
+      "loss": 0.4998,
+      "step": 4955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7846235795024805,
+      "learning_rate": 8.891806285006448e-06,
+      "loss": 0.5536,
+      "step": 4956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1525135213938995,
+      "learning_rate": 8.89137836546307e-06,
+      "loss": 0.8163,
+      "step": 4957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9946626561888163,
+      "learning_rate": 8.890950374923295e-06,
+      "loss": 0.6457,
+      "step": 4958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7781335991724792,
+      "learning_rate": 8.890522313396077e-06,
+      "loss": 0.5563,
+      "step": 4959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.146604833625894,
+      "learning_rate": 8.890094180890384e-06,
+      "loss": 0.6194,
+      "step": 4960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8193411614781985,
+      "learning_rate": 8.889665977415172e-06,
+      "loss": 0.4854,
+      "step": 4961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.292200156283021,
+      "learning_rate": 8.889237702979408e-06,
+      "loss": 0.629,
+      "step": 4962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9028272985606174,
+      "learning_rate": 8.888809357592059e-06,
+      "loss": 0.5466,
+      "step": 4963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3432766976780233,
+      "learning_rate": 8.88838094126209e-06,
+      "loss": 0.6654,
+      "step": 4964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9400136319477106,
+      "learning_rate": 8.887952453998471e-06,
+      "loss": 0.4086,
+      "step": 4965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7095260306364353,
+      "learning_rate": 8.887523895810173e-06,
+      "loss": 0.66,
+      "step": 4966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1021874464353143,
+      "learning_rate": 8.887095266706165e-06,
+      "loss": 0.6932,
+      "step": 4967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7562471014805143,
+      "learning_rate": 8.886666566695421e-06,
+      "loss": 0.4478,
+      "step": 4968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1201200804589186,
+      "learning_rate": 8.886237795786916e-06,
+      "loss": 0.7133,
+      "step": 4969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4995301931571978,
+      "learning_rate": 8.885808953989628e-06,
+      "loss": 0.4078,
+      "step": 4970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8945271423988224,
+      "learning_rate": 8.885380041312531e-06,
+      "loss": 0.5143,
+      "step": 4971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.566019940400466,
+      "learning_rate": 8.884951057764606e-06,
+      "loss": 0.4299,
+      "step": 4972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2524286152712794,
+      "learning_rate": 8.884522003354833e-06,
+      "loss": 0.5661,
+      "step": 4973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.067201237898749,
+      "learning_rate": 8.884092878092194e-06,
+      "loss": 0.7061,
+      "step": 4974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1915369051620863,
+      "learning_rate": 8.883663681985672e-06,
+      "loss": 0.4953,
+      "step": 4975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7478710333377259,
+      "learning_rate": 8.883234415044253e-06,
+      "loss": 0.5836,
+      "step": 4976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7632081773383927,
+      "learning_rate": 8.882805077276923e-06,
+      "loss": 0.6201,
+      "step": 4977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0216414106738174,
+      "learning_rate": 8.882375668692669e-06,
+      "loss": 0.7729,
+      "step": 4978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8671764534143718,
+      "learning_rate": 8.881946189300483e-06,
+      "loss": 0.6255,
+      "step": 4979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8915595608604658,
+      "learning_rate": 8.881516639109352e-06,
+      "loss": 0.5065,
+      "step": 4980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.665678007840954,
+      "learning_rate": 8.881087018128272e-06,
+      "loss": 0.5521,
+      "step": 4981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.697530847870831,
+      "learning_rate": 8.880657326366235e-06,
+      "loss": 0.5209,
+      "step": 4982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9002340879015909,
+      "learning_rate": 8.880227563832236e-06,
+      "loss": 0.6601,
+      "step": 4983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8847195246373287,
+      "learning_rate": 8.879797730535274e-06,
+      "loss": 0.5721,
+      "step": 4984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7758490190333744,
+      "learning_rate": 8.879367826484344e-06,
+      "loss": 0.4938,
+      "step": 4985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7739535732731002,
+      "learning_rate": 8.878937851688449e-06,
+      "loss": 0.688,
+      "step": 4986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4127147081052351,
+      "learning_rate": 8.878507806156586e-06,
+      "loss": 0.2942,
+      "step": 4987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8258360995504257,
+      "learning_rate": 8.878077689897764e-06,
+      "loss": 0.4942,
+      "step": 4988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.68604727316849,
+      "learning_rate": 8.877647502920982e-06,
+      "loss": 0.4418,
+      "step": 4989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.024905328928075,
+      "learning_rate": 8.877217245235249e-06,
+      "loss": 0.526,
+      "step": 4990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8290993061585519,
+      "learning_rate": 8.87678691684957e-06,
+      "loss": 0.4335,
+      "step": 4991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7929346115012177,
+      "learning_rate": 8.876356517772953e-06,
+      "loss": 0.5368,
+      "step": 4992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7132741572895958,
+      "learning_rate": 8.87592604801441e-06,
+      "loss": 0.5295,
+      "step": 4993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4501590120614276,
+      "learning_rate": 8.875495507582953e-06,
+      "loss": 0.6256,
+      "step": 4994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3379219124962294,
+      "learning_rate": 8.875064896487593e-06,
+      "loss": 0.6097,
+      "step": 4995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7564128501699947,
+      "learning_rate": 8.874634214737346e-06,
+      "loss": 0.4731,
+      "step": 4996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5286208766988536,
+      "learning_rate": 8.874203462341229e-06,
+      "loss": 0.3922,
+      "step": 4997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9817489660883798,
+      "learning_rate": 8.873772639308257e-06,
+      "loss": 0.4853,
+      "step": 4998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1230037384074794,
+      "learning_rate": 8.87334174564745e-06,
+      "loss": 0.521,
+      "step": 4999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8124721872326155,
+      "learning_rate": 8.87291078136783e-06,
+      "loss": 0.5185,
+      "step": 5000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5734710097312927,
+      "eval_runtime": 67.9164,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 5000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7921666632656583,
+      "learning_rate": 8.872479746478418e-06,
+      "loss": 0.578,
+      "step": 5001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.737228467808178,
+      "learning_rate": 8.872048640988236e-06,
+      "loss": 0.5168,
+      "step": 5002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7904337404586088,
+      "learning_rate": 8.87161746490631e-06,
+      "loss": 0.5865,
+      "step": 5003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.973338057639732,
+      "learning_rate": 8.871186218241668e-06,
+      "loss": 0.4654,
+      "step": 5004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8733427845255761,
+      "learning_rate": 8.870754901003336e-06,
+      "loss": 0.6666,
+      "step": 5005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3079937471002125,
+      "learning_rate": 8.870323513200342e-06,
+      "loss": 0.6426,
+      "step": 5006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3060005864646396,
+      "learning_rate": 8.869892054841722e-06,
+      "loss": 0.6731,
+      "step": 5007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7042055674834677,
+      "learning_rate": 8.869460525936502e-06,
+      "loss": 0.4635,
+      "step": 5008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.073710532342297,
+      "learning_rate": 8.86902892649372e-06,
+      "loss": 0.6708,
+      "step": 5009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7726129259532446,
+      "learning_rate": 8.86859725652241e-06,
+      "loss": 0.4034,
+      "step": 5010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1379545449764352,
+      "learning_rate": 8.868165516031608e-06,
+      "loss": 0.6481,
+      "step": 5011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5730875075696444,
+      "learning_rate": 8.867733705030354e-06,
+      "loss": 0.5429,
+      "step": 5012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6678379410655404,
+      "learning_rate": 8.867301823527686e-06,
+      "loss": 0.4176,
+      "step": 5013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5953525007516187,
+      "learning_rate": 8.866869871532647e-06,
+      "loss": 0.5003,
+      "step": 5014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1041481914651645,
+      "learning_rate": 8.866437849054276e-06,
+      "loss": 0.5267,
+      "step": 5015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2301310026880312,
+      "learning_rate": 8.86600575610162e-06,
+      "loss": 0.3416,
+      "step": 5016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8135071314507927,
+      "learning_rate": 8.865573592683727e-06,
+      "loss": 0.6163,
+      "step": 5017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9076753074896984,
+      "learning_rate": 8.86514135880964e-06,
+      "loss": 0.5249,
+      "step": 5018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5618491495850424,
+      "learning_rate": 8.864709054488408e-06,
+      "loss": 0.4573,
+      "step": 5019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.137828248734085,
+      "learning_rate": 8.864276679729082e-06,
+      "loss": 0.5676,
+      "step": 5020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4320730523178344,
+      "learning_rate": 8.863844234540714e-06,
+      "loss": 0.4736,
+      "step": 5021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6805352710765,
+      "learning_rate": 8.863411718932355e-06,
+      "loss": 0.5406,
+      "step": 5022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7235327696218654,
+      "learning_rate": 8.862979132913061e-06,
+      "loss": 0.4952,
+      "step": 5023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.778619494432014,
+      "learning_rate": 8.862546476491889e-06,
+      "loss": 0.4916,
+      "step": 5024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2916241479678883,
+      "learning_rate": 8.862113749677894e-06,
+      "loss": 0.5266,
+      "step": 5025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2220459607206413,
+      "learning_rate": 8.861680952480137e-06,
+      "loss": 0.7135,
+      "step": 5026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.714346593530986,
+      "learning_rate": 8.861248084907676e-06,
+      "loss": 0.5934,
+      "step": 5027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.805426334750914,
+      "learning_rate": 8.860815146969575e-06,
+      "loss": 0.7102,
+      "step": 5028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9921234244023192,
+      "learning_rate": 8.860382138674896e-06,
+      "loss": 0.4975,
+      "step": 5029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0599809857758187,
+      "learning_rate": 8.859949060032705e-06,
+      "loss": 0.5971,
+      "step": 5030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6513883255449613,
+      "learning_rate": 8.859515911052065e-06,
+      "loss": 0.5198,
+      "step": 5031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9495886363323938,
+      "learning_rate": 8.859082691742049e-06,
+      "loss": 0.4771,
+      "step": 5032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7173261914900175,
+      "learning_rate": 8.858649402111722e-06,
+      "loss": 0.4564,
+      "step": 5033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7309106054684242,
+      "learning_rate": 8.858216042170155e-06,
+      "loss": 0.485,
+      "step": 5034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0727837393055433,
+      "learning_rate": 8.857782611926423e-06,
+      "loss": 0.4968,
+      "step": 5035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4730114687405347,
+      "learning_rate": 8.857349111389596e-06,
+      "loss": 0.3167,
+      "step": 5036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.769739835065876,
+      "learning_rate": 8.85691554056875e-06,
+      "loss": 0.5935,
+      "step": 5037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0331300955735916,
+      "learning_rate": 8.856481899472964e-06,
+      "loss": 0.4869,
+      "step": 5038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6483232779447579,
+      "learning_rate": 8.856048188111314e-06,
+      "loss": 0.4509,
+      "step": 5039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5728316877114477,
+      "learning_rate": 8.855614406492879e-06,
+      "loss": 0.4094,
+      "step": 5040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7491382768818304,
+      "learning_rate": 8.85518055462674e-06,
+      "loss": 0.6145,
+      "step": 5041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9617877743489232,
+      "learning_rate": 8.854746632521981e-06,
+      "loss": 0.6801,
+      "step": 5042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8557868554257357,
+      "learning_rate": 8.854312640187685e-06,
+      "loss": 0.5144,
+      "step": 5043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.118484175429512,
+      "learning_rate": 8.853878577632937e-06,
+      "loss": 0.4345,
+      "step": 5044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6608106220013412,
+      "learning_rate": 8.853444444866824e-06,
+      "loss": 0.5055,
+      "step": 5045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.259574435573087,
+      "learning_rate": 8.853010241898436e-06,
+      "loss": 0.5321,
+      "step": 5046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1249949314671834,
+      "learning_rate": 8.852575968736861e-06,
+      "loss": 0.6011,
+      "step": 5047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9540389579185489,
+      "learning_rate": 8.852141625391188e-06,
+      "loss": 0.4519,
+      "step": 5048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9720182766398127,
+      "learning_rate": 8.851707211870514e-06,
+      "loss": 0.491,
+      "step": 5049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9285684979965259,
+      "learning_rate": 8.851272728183931e-06,
+      "loss": 0.6249,
+      "step": 5050
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5747730135917664,
+      "eval_runtime": 67.9729,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 5050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4727607941644778,
+      "learning_rate": 8.850838174340535e-06,
+      "loss": 0.5923,
+      "step": 5051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.067518104915077,
+      "learning_rate": 8.850403550349423e-06,
+      "loss": 0.582,
+      "step": 5052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.728517542911032,
+      "learning_rate": 8.849968856219695e-06,
+      "loss": 0.6004,
+      "step": 5053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6595500641712342,
+      "learning_rate": 8.849534091960446e-06,
+      "loss": 0.5711,
+      "step": 5054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8003020550021929,
+      "learning_rate": 8.849099257580785e-06,
+      "loss": 0.4804,
+      "step": 5055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8559521397258336,
+      "learning_rate": 8.848664353089811e-06,
+      "loss": 0.535,
+      "step": 5056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6836756732268452,
+      "learning_rate": 8.848229378496627e-06,
+      "loss": 0.4309,
+      "step": 5057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.013150954218684,
+      "learning_rate": 8.847794333810342e-06,
+      "loss": 0.695,
+      "step": 5058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6954218902529414,
+      "learning_rate": 8.847359219040061e-06,
+      "loss": 0.4824,
+      "step": 5059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7465446623495944,
+      "learning_rate": 8.846924034194893e-06,
+      "loss": 0.5684,
+      "step": 5060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0805844760654715,
+      "learning_rate": 8.846488779283952e-06,
+      "loss": 0.6152,
+      "step": 5061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7036218874063027,
+      "learning_rate": 8.846053454316345e-06,
+      "loss": 0.4689,
+      "step": 5062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.554460910371576,
+      "learning_rate": 8.845618059301188e-06,
+      "loss": 0.4882,
+      "step": 5063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5562509070577106,
+      "learning_rate": 8.845182594247594e-06,
+      "loss": 0.3976,
+      "step": 5064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7708853034074734,
+      "learning_rate": 8.844747059164681e-06,
+      "loss": 0.4408,
+      "step": 5065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2281946327631306,
+      "learning_rate": 8.844311454061565e-06,
+      "loss": 0.8718,
+      "step": 5066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.464965901055931,
+      "learning_rate": 8.843875778947367e-06,
+      "loss": 0.5183,
+      "step": 5067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1857803715894515,
+      "learning_rate": 8.843440033831206e-06,
+      "loss": 0.65,
+      "step": 5068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8455594800167163,
+      "learning_rate": 8.843004218722204e-06,
+      "loss": 0.4743,
+      "step": 5069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0716942211002802,
+      "learning_rate": 8.842568333629486e-06,
+      "loss": 0.5515,
+      "step": 5070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4345424184662843,
+      "learning_rate": 8.842132378562176e-06,
+      "loss": 0.5715,
+      "step": 5071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9940365094115353,
+      "learning_rate": 8.841696353529402e-06,
+      "loss": 0.5152,
+      "step": 5072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0788847666155,
+      "learning_rate": 8.841260258540287e-06,
+      "loss": 0.6711,
+      "step": 5073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.043868820421623,
+      "learning_rate": 8.840824093603967e-06,
+      "loss": 0.6085,
+      "step": 5074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.907146887338972,
+      "learning_rate": 8.840387858729566e-06,
+      "loss": 0.6046,
+      "step": 5075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5892278178607744,
+      "learning_rate": 8.839951553926223e-06,
+      "loss": 0.4118,
+      "step": 5076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.164555844501237,
+      "learning_rate": 8.83951517920307e-06,
+      "loss": 0.4697,
+      "step": 5077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6871496574796387,
+      "learning_rate": 8.839078734569238e-06,
+      "loss": 0.5141,
+      "step": 5078
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.094454388243719,
+      "learning_rate": 8.838642220033868e-06,
+      "loss": 0.6015,
+      "step": 5079
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.227964127034758,
+      "learning_rate": 8.838205635606095e-06,
+      "loss": 0.7074,
+      "step": 5080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8634443432242098,
+      "learning_rate": 8.837768981295062e-06,
+      "loss": 0.7262,
+      "step": 5081
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6076446126511923,
+      "learning_rate": 8.837332257109909e-06,
+      "loss": 0.4616,
+      "step": 5082
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.734711801987449,
+      "learning_rate": 8.836895463059776e-06,
+      "loss": 0.4499,
+      "step": 5083
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0140901503435242,
+      "learning_rate": 8.836458599153811e-06,
+      "loss": 0.5003,
+      "step": 5084
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8796339622816807,
+      "learning_rate": 8.836021665401157e-06,
+      "loss": 0.5085,
+      "step": 5085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8236597412895421,
+      "learning_rate": 8.835584661810964e-06,
+      "loss": 0.6401,
+      "step": 5086
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.60022021562052,
+      "learning_rate": 8.835147588392375e-06,
+      "loss": 0.3815,
+      "step": 5087
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8589327108702038,
+      "learning_rate": 8.834710445154545e-06,
+      "loss": 0.5185,
+      "step": 5088
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0123541776028695,
+      "learning_rate": 8.83427323210662e-06,
+      "loss": 0.5587,
+      "step": 5089
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6600689768728354,
+      "learning_rate": 8.833835949257758e-06,
+      "loss": 0.5215,
+      "step": 5090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8388188927537654,
+      "learning_rate": 8.83339859661711e-06,
+      "loss": 0.5689,
+      "step": 5091
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5152374205565968,
+      "learning_rate": 8.832961174193833e-06,
+      "loss": 0.3981,
+      "step": 5092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.847234593975331,
+      "learning_rate": 8.832523681997084e-06,
+      "loss": 0.5759,
+      "step": 5093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7615207835350948,
+      "learning_rate": 8.832086120036024e-06,
+      "loss": 0.6207,
+      "step": 5094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.130344945521783,
+      "learning_rate": 8.831648488319807e-06,
+      "loss": 0.5106,
+      "step": 5095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8775064250161926,
+      "learning_rate": 8.8312107868576e-06,
+      "loss": 0.5928,
+      "step": 5096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6646928099590454,
+      "learning_rate": 8.830773015658564e-06,
+      "loss": 0.6565,
+      "step": 5097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.656225152360051,
+      "learning_rate": 8.830335174731865e-06,
+      "loss": 0.5492,
+      "step": 5098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.189873899493415,
+      "learning_rate": 8.829897264086667e-06,
+      "loss": 0.652,
+      "step": 5099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.512433238698793,
+      "learning_rate": 8.829459283732138e-06,
+      "loss": 0.6406,
+      "step": 5100
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5736879110336304,
+      "eval_runtime": 68.0123,
+      "eval_samples_per_second": 50.138,
+      "eval_steps_per_second": 1.573,
+      "step": 5100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.772923020078652,
+      "learning_rate": 8.829021233677444e-06,
+      "loss": 0.6372,
+      "step": 5101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.987537733637136,
+      "learning_rate": 8.828583113931763e-06,
+      "loss": 0.5992,
+      "step": 5102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.553624068023624,
+      "learning_rate": 8.828144924504257e-06,
+      "loss": 0.4105,
+      "step": 5103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9073524495551641,
+      "learning_rate": 8.82770666540411e-06,
+      "loss": 0.5119,
+      "step": 5104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5138577348242395,
+      "learning_rate": 8.827268336640487e-06,
+      "loss": 0.4748,
+      "step": 5105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5530604822389844,
+      "learning_rate": 8.826829938222569e-06,
+      "loss": 0.4932,
+      "step": 5106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6324335051616743,
+      "learning_rate": 8.826391470159532e-06,
+      "loss": 0.4696,
+      "step": 5107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8404109099288797,
+      "learning_rate": 8.825952932460555e-06,
+      "loss": 0.5421,
+      "step": 5108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.984930543469972,
+      "learning_rate": 8.825514325134822e-06,
+      "loss": 0.593,
+      "step": 5109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2432591530086032,
+      "learning_rate": 8.82507564819151e-06,
+      "loss": 0.4966,
+      "step": 5110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5233867540125996,
+      "learning_rate": 8.824636901639805e-06,
+      "loss": 0.3983,
+      "step": 5111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.075323668914008,
+      "learning_rate": 8.824198085488891e-06,
+      "loss": 0.5675,
+      "step": 5112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9578317530187612,
+      "learning_rate": 8.823759199747956e-06,
+      "loss": 0.5605,
+      "step": 5113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0831276143820157,
+      "learning_rate": 8.823320244426186e-06,
+      "loss": 0.5847,
+      "step": 5114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5324227714949863,
+      "learning_rate": 8.822881219532769e-06,
+      "loss": 0.368,
+      "step": 5115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.904513053479404,
+      "learning_rate": 8.8224421250769e-06,
+      "loss": 0.6222,
+      "step": 5116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7183261537759498,
+      "learning_rate": 8.82200296106777e-06,
+      "loss": 0.5939,
+      "step": 5117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1280707450962892,
+      "learning_rate": 8.82156372751457e-06,
+      "loss": 0.4667,
+      "step": 5118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.285559927882198,
+      "learning_rate": 8.821124424426495e-06,
+      "loss": 0.604,
+      "step": 5119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8230101923946207,
+      "learning_rate": 8.820685051812744e-06,
+      "loss": 0.5866,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4776754796883524,
+      "learning_rate": 8.820245609682513e-06,
+      "loss": 0.3845,
+      "step": 5121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1910568889805404,
+      "learning_rate": 8.819806098045003e-06,
+      "loss": 0.8364,
+      "step": 5122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2040534384121124,
+      "learning_rate": 8.819366516909416e-06,
+      "loss": 0.5786,
+      "step": 5123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.628666248905149,
+      "learning_rate": 8.818926866284951e-06,
+      "loss": 0.4714,
+      "step": 5124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1596870604932272,
+      "learning_rate": 8.818487146180815e-06,
+      "loss": 0.5694,
+      "step": 5125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.729498996671531,
+      "learning_rate": 8.81804735660621e-06,
+      "loss": 0.6034,
+      "step": 5126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.787063387633055,
+      "learning_rate": 8.817607497570346e-06,
+      "loss": 0.6299,
+      "step": 5127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9797239970604708,
+      "learning_rate": 8.817167569082429e-06,
+      "loss": 0.5161,
+      "step": 5128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9964181301633872,
+      "learning_rate": 8.816727571151669e-06,
+      "loss": 0.4462,
+      "step": 5129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8721440959986506,
+      "learning_rate": 8.816287503787277e-06,
+      "loss": 0.6817,
+      "step": 5130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.086482490457012,
+      "learning_rate": 8.815847366998468e-06,
+      "loss": 0.509,
+      "step": 5131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1338223269817282,
+      "learning_rate": 8.815407160794451e-06,
+      "loss": 0.5347,
+      "step": 5132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8980808282561104,
+      "learning_rate": 8.814966885184446e-06,
+      "loss": 0.5142,
+      "step": 5133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7936868376494113,
+      "learning_rate": 8.814526540177669e-06,
+      "loss": 0.5435,
+      "step": 5134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7185758877250175,
+      "learning_rate": 8.814086125783336e-06,
+      "loss": 0.6205,
+      "step": 5135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2265067630589743,
+      "learning_rate": 8.81364564201067e-06,
+      "loss": 0.6283,
+      "step": 5136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6555636785388415,
+      "learning_rate": 8.81320508886889e-06,
+      "loss": 0.4706,
+      "step": 5137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7770626484047471,
+      "learning_rate": 8.812764466367218e-06,
+      "loss": 0.6007,
+      "step": 5138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7366341270750894,
+      "learning_rate": 8.812323774514882e-06,
+      "loss": 0.4904,
+      "step": 5139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8275338489295414,
+      "learning_rate": 8.811883013321104e-06,
+      "loss": 0.5985,
+      "step": 5140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5713176364110928,
+      "learning_rate": 8.81144218279511e-06,
+      "loss": 0.4148,
+      "step": 5141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.80308869336053,
+      "learning_rate": 8.811001282946135e-06,
+      "loss": 0.5489,
+      "step": 5142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8690155692692212,
+      "learning_rate": 8.810560313783402e-06,
+      "loss": 0.5456,
+      "step": 5143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.879385970529857,
+      "learning_rate": 8.810119275316146e-06,
+      "loss": 0.6082,
+      "step": 5144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7465573537923305,
+      "learning_rate": 8.809678167553598e-06,
+      "loss": 0.6167,
+      "step": 5145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9391413146738234,
+      "learning_rate": 8.809236990504994e-06,
+      "loss": 0.508,
+      "step": 5146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9597713304210316,
+      "learning_rate": 8.808795744179569e-06,
+      "loss": 0.5701,
+      "step": 5147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.545769976739907,
+      "learning_rate": 8.80835442858656e-06,
+      "loss": 0.6505,
+      "step": 5148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6783441270117927,
+      "learning_rate": 8.807913043735206e-06,
+      "loss": 0.5755,
+      "step": 5149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.252136244851095,
+      "learning_rate": 8.807471589634748e-06,
+      "loss": 0.5683,
+      "step": 5150
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.5735135078430176,
+      "eval_runtime": 67.9934,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 5150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.063401447905602,
+      "learning_rate": 8.807030066294425e-06,
+      "loss": 0.5297,
+      "step": 5151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.159475765648661,
+      "learning_rate": 8.806588473723483e-06,
+      "loss": 0.5366,
+      "step": 5152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.764797492444658,
+      "learning_rate": 8.806146811931164e-06,
+      "loss": 0.567,
+      "step": 5153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.878183763872634,
+      "learning_rate": 8.805705080926715e-06,
+      "loss": 0.489,
+      "step": 5154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.224540438926069,
+      "learning_rate": 8.805263280719384e-06,
+      "loss": 0.8537,
+      "step": 5155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8602804733195915,
+      "learning_rate": 8.80482141131842e-06,
+      "loss": 0.6864,
+      "step": 5156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.678332327448654,
+      "learning_rate": 8.804379472733071e-06,
+      "loss": 0.6079,
+      "step": 5157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0835973337404416,
+      "learning_rate": 8.80393746497259e-06,
+      "loss": 0.4863,
+      "step": 5158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.01446180944205,
+      "learning_rate": 8.803495388046231e-06,
+      "loss": 0.6709,
+      "step": 5159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9658752359760252,
+      "learning_rate": 8.803053241963249e-06,
+      "loss": 0.7404,
+      "step": 5160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0191854860596528,
+      "learning_rate": 8.802611026732898e-06,
+      "loss": 0.4697,
+      "step": 5161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.778326691885658,
+      "learning_rate": 8.802168742364435e-06,
+      "loss": 0.4742,
+      "step": 5162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9014405035181785,
+      "learning_rate": 8.801726388867124e-06,
+      "loss": 0.4565,
+      "step": 5163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0152519654595715,
+      "learning_rate": 8.801283966250217e-06,
+      "loss": 0.5937,
+      "step": 5164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8842561864858365,
+      "learning_rate": 8.800841474522984e-06,
+      "loss": 0.5159,
+      "step": 5165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4657357433361216,
+      "learning_rate": 8.800398913694685e-06,
+      "loss": 0.4689,
+      "step": 5166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1070638255968035,
+      "learning_rate": 8.799956283774584e-06,
+      "loss": 0.6609,
+      "step": 5167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.768434375556411,
+      "learning_rate": 8.799513584771948e-06,
+      "loss": 0.5592,
+      "step": 5168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2439974467994857,
+      "learning_rate": 8.799070816696045e-06,
+      "loss": 0.5376,
+      "step": 5169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5413443162693812,
+      "learning_rate": 8.79862797955614e-06,
+      "loss": 0.5004,
+      "step": 5170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0139838993188106,
+      "learning_rate": 8.798185073361512e-06,
+      "loss": 0.7093,
+      "step": 5171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7895416402442756,
+      "learning_rate": 8.797742098121427e-06,
+      "loss": 0.4807,
+      "step": 5172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.472489997152394,
+      "learning_rate": 8.797299053845158e-06,
+      "loss": 0.5733,
+      "step": 5173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5401836018886468,
+      "learning_rate": 8.796855940541984e-06,
+      "loss": 0.4848,
+      "step": 5174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9463618823924622,
+      "learning_rate": 8.796412758221174e-06,
+      "loss": 0.4131,
+      "step": 5175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9338404323172103,
+      "learning_rate": 8.795969506892014e-06,
+      "loss": 0.6358,
+      "step": 5176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.121970347146088,
+      "learning_rate": 8.795526186563779e-06,
+      "loss": 0.5569,
+      "step": 5177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.098758720047085,
+      "learning_rate": 8.795082797245749e-06,
+      "loss": 0.4558,
+      "step": 5178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3343931114530854,
+      "learning_rate": 8.794639338947209e-06,
+      "loss": 0.6759,
+      "step": 5179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9558960303997388,
+      "learning_rate": 8.794195811677439e-06,
+      "loss": 0.4947,
+      "step": 5180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8583849961408598,
+      "learning_rate": 8.793752215445726e-06,
+      "loss": 0.576,
+      "step": 5181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.620053784340576,
+      "learning_rate": 8.793308550261359e-06,
+      "loss": 0.5369,
+      "step": 5182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.226213400342636,
+      "learning_rate": 8.792864816133619e-06,
+      "loss": 0.6393,
+      "step": 5183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7617789102820047,
+      "learning_rate": 8.792421013071802e-06,
+      "loss": 0.4873,
+      "step": 5184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4360594377560356,
+      "learning_rate": 8.791977141085194e-06,
+      "loss": 0.6129,
+      "step": 5185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7563272792529834,
+      "learning_rate": 8.791533200183092e-06,
+      "loss": 0.5802,
+      "step": 5186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3154993504188552,
+      "learning_rate": 8.791089190374785e-06,
+      "loss": 0.6902,
+      "step": 5187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6343859289841693,
+      "learning_rate": 8.790645111669571e-06,
+      "loss": 0.5819,
+      "step": 5188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5543070218757304,
+      "learning_rate": 8.790200964076744e-06,
+      "loss": 0.4777,
+      "step": 5189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0613823393438238,
+      "learning_rate": 8.789756747605604e-06,
+      "loss": 0.545,
+      "step": 5190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8352792864982752,
+      "learning_rate": 8.78931246226545e-06,
+      "loss": 0.5506,
+      "step": 5191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9688900151946747,
+      "learning_rate": 8.788868108065584e-06,
+      "loss": 0.4991,
+      "step": 5192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9113031060174421,
+      "learning_rate": 8.788423685015306e-06,
+      "loss": 0.7486,
+      "step": 5193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0588765390225325,
+      "learning_rate": 8.78797919312392e-06,
+      "loss": 0.6459,
+      "step": 5194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8761823471576067,
+      "learning_rate": 8.787534632400733e-06,
+      "loss": 0.4502,
+      "step": 5195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.445252165426927,
+      "learning_rate": 8.78709000285505e-06,
+      "loss": 0.7058,
+      "step": 5196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5255823937973818,
+      "learning_rate": 8.786645304496179e-06,
+      "loss": 0.398,
+      "step": 5197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8034161305241072,
+      "learning_rate": 8.78620053733343e-06,
+      "loss": 0.5699,
+      "step": 5198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7768445169897618,
+      "learning_rate": 8.785755701376115e-06,
+      "loss": 0.61,
+      "step": 5199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5128674711126853,
+      "learning_rate": 8.785310796633544e-06,
+      "loss": 0.4808,
+      "step": 5200
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.5702080726623535,
+      "eval_runtime": 67.9891,
+      "eval_samples_per_second": 50.155,
+      "eval_steps_per_second": 1.574,
+      "step": 5200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9153762630723206,
+      "learning_rate": 8.784865823115031e-06,
+      "loss": 0.4066,
+      "step": 5201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5556882641978662,
+      "learning_rate": 8.784420780829894e-06,
+      "loss": 0.486,
+      "step": 5202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7128758639055723,
+      "learning_rate": 8.783975669787448e-06,
+      "loss": 0.4204,
+      "step": 5203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.140716804593339,
+      "learning_rate": 8.78353048999701e-06,
+      "loss": 0.71,
+      "step": 5204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8848181519398077,
+      "learning_rate": 8.783085241467902e-06,
+      "loss": 0.6984,
+      "step": 5205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0870285432228775,
+      "learning_rate": 8.782639924209442e-06,
+      "loss": 0.5802,
+      "step": 5206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1328853246504975,
+      "learning_rate": 8.782194538230958e-06,
+      "loss": 0.7159,
+      "step": 5207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.682324795276432,
+      "learning_rate": 8.781749083541767e-06,
+      "loss": 0.6098,
+      "step": 5208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8086313434886931,
+      "learning_rate": 8.781303560151196e-06,
+      "loss": 0.5044,
+      "step": 5209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8642761448429872,
+      "learning_rate": 8.780857968068575e-06,
+      "loss": 0.4702,
+      "step": 5210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0582796120744438,
+      "learning_rate": 8.78041230730323e-06,
+      "loss": 0.5154,
+      "step": 5211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.757296309465441,
+      "learning_rate": 8.77996657786449e-06,
+      "loss": 0.5071,
+      "step": 5212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2539144159773814,
+      "learning_rate": 8.779520779761688e-06,
+      "loss": 0.4637,
+      "step": 5213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9342591024106277,
+      "learning_rate": 8.779074913004155e-06,
+      "loss": 0.5433,
+      "step": 5214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4039090341026736,
+      "learning_rate": 8.778628977601225e-06,
+      "loss": 0.6597,
+      "step": 5215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7952542461834162,
+      "learning_rate": 8.778182973562235e-06,
+      "loss": 0.5738,
+      "step": 5216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6617776195224805,
+      "learning_rate": 8.77773690089652e-06,
+      "loss": 0.606,
+      "step": 5217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3415403869478606,
+      "learning_rate": 8.777290759613419e-06,
+      "loss": 0.614,
+      "step": 5218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6234966569361564,
+      "learning_rate": 8.776844549722273e-06,
+      "loss": 0.5267,
+      "step": 5219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5859495333450275,
+      "learning_rate": 8.776398271232419e-06,
+      "loss": 0.4145,
+      "step": 5220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.168810756719094,
+      "learning_rate": 8.775951924153203e-06,
+      "loss": 0.5751,
+      "step": 5221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7915424323701785,
+      "learning_rate": 8.77550550849397e-06,
+      "loss": 0.526,
+      "step": 5222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8992285030430716,
+      "learning_rate": 8.775059024264061e-06,
+      "loss": 0.6353,
+      "step": 5223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1565360551872175,
+      "learning_rate": 8.774612471472826e-06,
+      "loss": 0.476,
+      "step": 5224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7644095435709013,
+      "learning_rate": 8.774165850129614e-06,
+      "loss": 0.4723,
+      "step": 5225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9820379796996517,
+      "learning_rate": 8.773719160243774e-06,
+      "loss": 0.5436,
+      "step": 5226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2540704122955653,
+      "learning_rate": 8.773272401824656e-06,
+      "loss": 0.5056,
+      "step": 5227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2766429240036246,
+      "learning_rate": 8.772825574881612e-06,
+      "loss": 0.7892,
+      "step": 5228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.690291798642259,
+      "learning_rate": 8.772378679424e-06,
+      "loss": 0.5945,
+      "step": 5229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7412940223440811,
+      "learning_rate": 8.771931715461171e-06,
+      "loss": 0.4879,
+      "step": 5230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2469421248041375,
+      "learning_rate": 8.771484683002483e-06,
+      "loss": 0.6399,
+      "step": 5231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7446061153824637,
+      "learning_rate": 8.771037582057296e-06,
+      "loss": 0.4989,
+      "step": 5232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.049268514664128,
+      "learning_rate": 8.770590412634971e-06,
+      "loss": 0.6281,
+      "step": 5233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.151766466891701,
+      "learning_rate": 8.770143174744866e-06,
+      "loss": 0.611,
+      "step": 5234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1513476728092034,
+      "learning_rate": 8.769695868396344e-06,
+      "loss": 0.6462,
+      "step": 5235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8524393633794074,
+      "learning_rate": 8.76924849359877e-06,
+      "loss": 0.5698,
+      "step": 5236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5693248644488962,
+      "learning_rate": 8.76880105036151e-06,
+      "loss": 0.4663,
+      "step": 5237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1465896172572614,
+      "learning_rate": 8.76835353869393e-06,
+      "loss": 0.5018,
+      "step": 5238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4817004744801796,
+      "learning_rate": 8.7679059586054e-06,
+      "loss": 0.5575,
+      "step": 5239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1901409129705947,
+      "learning_rate": 8.767458310105286e-06,
+      "loss": 0.5532,
+      "step": 5240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.238399484331687,
+      "learning_rate": 8.767010593202963e-06,
+      "loss": 0.6613,
+      "step": 5241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8124577730217561,
+      "learning_rate": 8.766562807907802e-06,
+      "loss": 0.4924,
+      "step": 5242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9579476983742563,
+      "learning_rate": 8.76611495422918e-06,
+      "loss": 0.5187,
+      "step": 5243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0300239188398064,
+      "learning_rate": 8.765667032176467e-06,
+      "loss": 0.654,
+      "step": 5244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.066084420975608,
+      "learning_rate": 8.765219041759045e-06,
+      "loss": 0.6596,
+      "step": 5245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1295858593115566,
+      "learning_rate": 8.76477098298629e-06,
+      "loss": 0.609,
+      "step": 5246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8204954381918705,
+      "learning_rate": 8.764322855867582e-06,
+      "loss": 0.5392,
+      "step": 5247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6027158658432585,
+      "learning_rate": 8.763874660412302e-06,
+      "loss": 0.5607,
+      "step": 5248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.208645188707842,
+      "learning_rate": 8.763426396629834e-06,
+      "loss": 0.4734,
+      "step": 5249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.791869171191904,
+      "learning_rate": 8.76297806452956e-06,
+      "loss": 0.5449,
+      "step": 5250
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.5702080726623535,
+      "eval_runtime": 67.9649,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 5250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0811686557442766,
+      "learning_rate": 8.762529664120869e-06,
+      "loss": 0.6647,
+      "step": 5251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9895335943553538,
+      "learning_rate": 8.762081195413145e-06,
+      "loss": 0.6241,
+      "step": 5252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.608886595778895,
+      "learning_rate": 8.761632658415778e-06,
+      "loss": 0.507,
+      "step": 5253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6357506953184666,
+      "learning_rate": 8.761184053138156e-06,
+      "loss": 0.4384,
+      "step": 5254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6075042928333707,
+      "learning_rate": 8.760735379589672e-06,
+      "loss": 0.4101,
+      "step": 5255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.07493343637937,
+      "learning_rate": 8.760286637779717e-06,
+      "loss": 0.5959,
+      "step": 5256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.213730717491337,
+      "learning_rate": 8.759837827717687e-06,
+      "loss": 0.6028,
+      "step": 5257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.083916687521997,
+      "learning_rate": 8.759388949412978e-06,
+      "loss": 0.5691,
+      "step": 5258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7175723482101966,
+      "learning_rate": 8.758940002874985e-06,
+      "loss": 0.4628,
+      "step": 5259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.032885920158017,
+      "learning_rate": 8.758490988113109e-06,
+      "loss": 0.6423,
+      "step": 5260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7570272615822238,
+      "learning_rate": 8.758041905136745e-06,
+      "loss": 0.5186,
+      "step": 5261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1721387478508154,
+      "learning_rate": 8.757592753955299e-06,
+      "loss": 0.5535,
+      "step": 5262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7593051813586302,
+      "learning_rate": 8.757143534578172e-06,
+      "loss": 0.519,
+      "step": 5263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5058779221529006,
+      "learning_rate": 8.756694247014766e-06,
+      "loss": 0.356,
+      "step": 5264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.973266568587971,
+      "learning_rate": 8.756244891274492e-06,
+      "loss": 0.7271,
+      "step": 5265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7178738154274742,
+      "learning_rate": 8.755795467366752e-06,
+      "loss": 0.4989,
+      "step": 5266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2180795335498655,
+      "learning_rate": 8.755345975300958e-06,
+      "loss": 0.5586,
+      "step": 5267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5018677202831574,
+      "learning_rate": 8.754896415086516e-06,
+      "loss": 0.4129,
+      "step": 5268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9533920703578689,
+      "learning_rate": 8.754446786732839e-06,
+      "loss": 0.6237,
+      "step": 5269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8892164854755353,
+      "learning_rate": 8.753997090249341e-06,
+      "loss": 0.577,
+      "step": 5270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6312535243924218,
+      "learning_rate": 8.753547325645435e-06,
+      "loss": 0.4419,
+      "step": 5271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5445461443284445,
+      "learning_rate": 8.753097492930538e-06,
+      "loss": 0.4898,
+      "step": 5272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9770496104397046,
+      "learning_rate": 8.752647592114062e-06,
+      "loss": 0.4916,
+      "step": 5273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7793895115693128,
+      "learning_rate": 8.752197623205434e-06,
+      "loss": 0.4917,
+      "step": 5274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0816622478674858,
+      "learning_rate": 8.751747586214064e-06,
+      "loss": 0.521,
+      "step": 5275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8635165338385105,
+      "learning_rate": 8.751297481149379e-06,
+      "loss": 0.616,
+      "step": 5276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2928406941808506,
+      "learning_rate": 8.750847308020802e-06,
+      "loss": 0.7671,
+      "step": 5277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8362071947151917,
+      "learning_rate": 8.750397066837754e-06,
+      "loss": 0.5,
+      "step": 5278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7584651099077315,
+      "learning_rate": 8.749946757609665e-06,
+      "loss": 0.4432,
+      "step": 5279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.922657949678596,
+      "learning_rate": 8.749496380345956e-06,
+      "loss": 0.6022,
+      "step": 5280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7287249876105946,
+      "learning_rate": 8.749045935056062e-06,
+      "loss": 0.5078,
+      "step": 5281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9679962842576304,
+      "learning_rate": 8.748595421749408e-06,
+      "loss": 0.6573,
+      "step": 5282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8528917932765734,
+      "learning_rate": 8.748144840435425e-06,
+      "loss": 0.5432,
+      "step": 5283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.163297742000532,
+      "learning_rate": 8.747694191123548e-06,
+      "loss": 0.653,
+      "step": 5284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9668768595695072,
+      "learning_rate": 8.747243473823213e-06,
+      "loss": 0.7247,
+      "step": 5285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0736622713456434,
+      "learning_rate": 8.746792688543849e-06,
+      "loss": 0.65,
+      "step": 5286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9696405719124412,
+      "learning_rate": 8.746341835294899e-06,
+      "loss": 0.5845,
+      "step": 5287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6721099163060824,
+      "learning_rate": 8.745890914085798e-06,
+      "loss": 0.6152,
+      "step": 5288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4331868941456987,
+      "learning_rate": 8.745439924925988e-06,
+      "loss": 0.3939,
+      "step": 5289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6825055302833951,
+      "learning_rate": 8.744988867824908e-06,
+      "loss": 0.5219,
+      "step": 5290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5321433659970563,
+      "learning_rate": 8.744537742792003e-06,
+      "loss": 0.5382,
+      "step": 5291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3183578868617483,
+      "learning_rate": 8.744086549836713e-06,
+      "loss": 0.6228,
+      "step": 5292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7247330048932914,
+      "learning_rate": 8.743635288968488e-06,
+      "loss": 0.4248,
+      "step": 5293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7389676535580851,
+      "learning_rate": 8.743183960196772e-06,
+      "loss": 0.5001,
+      "step": 5294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8648042238098392,
+      "learning_rate": 8.742732563531014e-06,
+      "loss": 0.5759,
+      "step": 5295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8362293918114512,
+      "learning_rate": 8.742281098980666e-06,
+      "loss": 0.5351,
+      "step": 5296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1561823288591304,
+      "learning_rate": 8.741829566555176e-06,
+      "loss": 0.6929,
+      "step": 5297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9211880085056399,
+      "learning_rate": 8.741377966263998e-06,
+      "loss": 0.4815,
+      "step": 5298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.498758327990951,
+      "learning_rate": 8.740926298116584e-06,
+      "loss": 0.4021,
+      "step": 5299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.204086910318439,
+      "learning_rate": 8.740474562122391e-06,
+      "loss": 0.6601,
+      "step": 5300
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.5720779299736023,
+      "eval_runtime": 67.9633,
+      "eval_samples_per_second": 50.174,
+      "eval_steps_per_second": 1.574,
+      "step": 5300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3683624901876659,
+      "learning_rate": 8.740022758290877e-06,
+      "loss": 0.3929,
+      "step": 5301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9522823432787846,
+      "learning_rate": 8.739570886631498e-06,
+      "loss": 0.6657,
+      "step": 5302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0797359400297544,
+      "learning_rate": 8.739118947153715e-06,
+      "loss": 0.5008,
+      "step": 5303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6756214848011275,
+      "learning_rate": 8.73866693986699e-06,
+      "loss": 0.5482,
+      "step": 5304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.746489633149937,
+      "learning_rate": 8.738214864780783e-06,
+      "loss": 0.4764,
+      "step": 5305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8507824963744004,
+      "learning_rate": 8.73776272190456e-06,
+      "loss": 0.65,
+      "step": 5306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4001579687015333,
+      "learning_rate": 8.737310511247783e-06,
+      "loss": 0.5443,
+      "step": 5307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8113947660495582,
+      "learning_rate": 8.736858232819924e-06,
+      "loss": 0.4853,
+      "step": 5308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2928742881193391,
+      "learning_rate": 8.736405886630449e-06,
+      "loss": 0.352,
+      "step": 5309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.43916878878101,
+      "learning_rate": 8.735953472688824e-06,
+      "loss": 0.5698,
+      "step": 5310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7703288980268572,
+      "learning_rate": 8.735500991004524e-06,
+      "loss": 0.5146,
+      "step": 5311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9362725324675036,
+      "learning_rate": 8.735048441587022e-06,
+      "loss": 0.5034,
+      "step": 5312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.751289666618752,
+      "learning_rate": 8.734595824445788e-06,
+      "loss": 0.4463,
+      "step": 5313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7901578354476546,
+      "learning_rate": 8.7341431395903e-06,
+      "loss": 0.4968,
+      "step": 5314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7333563137508623,
+      "learning_rate": 8.733690387030037e-06,
+      "loss": 0.667,
+      "step": 5315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.412394440991007,
+      "learning_rate": 8.73323756677447e-06,
+      "loss": 0.7773,
+      "step": 5316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7756848402440446,
+      "learning_rate": 8.732784678833085e-06,
+      "loss": 0.4912,
+      "step": 5317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6417626840374533,
+      "learning_rate": 8.732331723215361e-06,
+      "loss": 0.4548,
+      "step": 5318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8079402251335437,
+      "learning_rate": 8.73187869993078e-06,
+      "loss": 0.6329,
+      "step": 5319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1125851895585686,
+      "learning_rate": 8.731425608988824e-06,
+      "loss": 0.8302,
+      "step": 5320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.570055013349695,
+      "learning_rate": 8.730972450398982e-06,
+      "loss": 0.4404,
+      "step": 5321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7069529132401102,
+      "learning_rate": 8.730519224170738e-06,
+      "loss": 0.4312,
+      "step": 5322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.015707992074208,
+      "learning_rate": 8.73006593031358e-06,
+      "loss": 0.5416,
+      "step": 5323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9355482467651133,
+      "learning_rate": 8.729612568836998e-06,
+      "loss": 0.7228,
+      "step": 5324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9453868881464138,
+      "learning_rate": 8.729159139750484e-06,
+      "loss": 0.6766,
+      "step": 5325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6381112178493304,
+      "learning_rate": 8.728705643063528e-06,
+      "loss": 0.4279,
+      "step": 5326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9455403896034946,
+      "learning_rate": 8.728252078785626e-06,
+      "loss": 0.732,
+      "step": 5327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.697802797607836,
+      "learning_rate": 8.727798446926271e-06,
+      "loss": 0.5187,
+      "step": 5328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9746463089226065,
+      "learning_rate": 8.72734474749496e-06,
+      "loss": 0.4807,
+      "step": 5329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.630095938285787,
+      "learning_rate": 8.726890980501191e-06,
+      "loss": 0.5442,
+      "step": 5330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4315517334029773,
+      "learning_rate": 8.726437145954464e-06,
+      "loss": 0.4024,
+      "step": 5331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1905270239823835,
+      "learning_rate": 8.725983243864281e-06,
+      "loss": 0.6615,
+      "step": 5332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.05658692382565,
+      "learning_rate": 8.725529274240141e-06,
+      "loss": 0.5152,
+      "step": 5333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.125637347268355,
+      "learning_rate": 8.725075237091551e-06,
+      "loss": 0.9178,
+      "step": 5334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6473393769975357,
+      "learning_rate": 8.724621132428011e-06,
+      "loss": 0.5371,
+      "step": 5335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4142998361764993,
+      "learning_rate": 8.724166960259032e-06,
+      "loss": 0.6305,
+      "step": 5336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.947925111812835,
+      "learning_rate": 8.72371272059412e-06,
+      "loss": 0.5348,
+      "step": 5337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5915760278908728,
+      "learning_rate": 8.723258413442785e-06,
+      "loss": 0.5487,
+      "step": 5338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7625828766445182,
+      "learning_rate": 8.722804038814538e-06,
+      "loss": 0.5149,
+      "step": 5339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.125561539912198,
+      "learning_rate": 8.722349596718887e-06,
+      "loss": 0.5972,
+      "step": 5340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.995088318136307,
+      "learning_rate": 8.721895087165352e-06,
+      "loss": 0.5939,
+      "step": 5341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.84001602420248,
+      "learning_rate": 8.721440510163443e-06,
+      "loss": 0.4912,
+      "step": 5342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9001294352537192,
+      "learning_rate": 8.720985865722679e-06,
+      "loss": 0.6487,
+      "step": 5343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.806986366255124,
+      "learning_rate": 8.720531153852575e-06,
+      "loss": 0.5685,
+      "step": 5344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5335743361117449,
+      "learning_rate": 8.720076374562652e-06,
+      "loss": 0.5102,
+      "step": 5345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.875556361779023,
+      "learning_rate": 8.719621527862431e-06,
+      "loss": 0.6434,
+      "step": 5346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.593440744218156,
+      "learning_rate": 8.719166613761432e-06,
+      "loss": 0.3959,
+      "step": 5347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.149218743626566,
+      "learning_rate": 8.71871163226918e-06,
+      "loss": 0.3982,
+      "step": 5348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.039761877023068,
+      "learning_rate": 8.7182565833952e-06,
+      "loss": 0.6252,
+      "step": 5349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9538964948417716,
+      "learning_rate": 8.717801467149017e-06,
+      "loss": 0.6224,
+      "step": 5350
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5693436861038208,
+      "eval_runtime": 67.9711,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 5350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8988278619248289,
+      "learning_rate": 8.717346283540157e-06,
+      "loss": 0.6119,
+      "step": 5351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0065978450027204,
+      "learning_rate": 8.716891032578153e-06,
+      "loss": 0.5909,
+      "step": 5352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.838566172558831,
+      "learning_rate": 8.716435714272533e-06,
+      "loss": 0.6738,
+      "step": 5353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859062145746818,
+      "learning_rate": 8.715980328632828e-06,
+      "loss": 0.7219,
+      "step": 5354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6390187112120653,
+      "learning_rate": 8.715524875668574e-06,
+      "loss": 0.4794,
+      "step": 5355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6066004376300727,
+      "learning_rate": 8.715069355389303e-06,
+      "loss": 0.5104,
+      "step": 5356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.549097337196213,
+      "learning_rate": 8.714613767804552e-06,
+      "loss": 0.5707,
+      "step": 5357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5833558697293555,
+      "learning_rate": 8.714158112923858e-06,
+      "loss": 0.4523,
+      "step": 5358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.148198577084588,
+      "learning_rate": 8.713702390756763e-06,
+      "loss": 0.5758,
+      "step": 5359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.191579372369697,
+      "learning_rate": 8.7132466013128e-06,
+      "loss": 0.806,
+      "step": 5360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.084393551963866,
+      "learning_rate": 8.712790744601519e-06,
+      "loss": 0.6244,
+      "step": 5361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8964981650873138,
+      "learning_rate": 8.712334820632458e-06,
+      "loss": 0.5997,
+      "step": 5362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5891028309032234,
+      "learning_rate": 8.711878829415162e-06,
+      "loss": 0.4679,
+      "step": 5363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8738703176836837,
+      "learning_rate": 8.711422770959178e-06,
+      "loss": 0.466,
+      "step": 5364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.724191267692445,
+      "learning_rate": 8.710966645274053e-06,
+      "loss": 0.4973,
+      "step": 5365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9167056656369819,
+      "learning_rate": 8.710510452369336e-06,
+      "loss": 0.6762,
+      "step": 5366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9896041478856392,
+      "learning_rate": 8.710054192254579e-06,
+      "loss": 0.6602,
+      "step": 5367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7182814858898772,
+      "learning_rate": 8.709597864939328e-06,
+      "loss": 0.58,
+      "step": 5368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.32173763846219,
+      "learning_rate": 8.70914147043314e-06,
+      "loss": 0.5637,
+      "step": 5369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1635933388314665,
+      "learning_rate": 8.70868500874557e-06,
+      "loss": 0.6647,
+      "step": 5370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9827198055576465,
+      "learning_rate": 8.708228479886171e-06,
+      "loss": 0.5533,
+      "step": 5371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9093640049386542,
+      "learning_rate": 8.707771883864503e-06,
+      "loss": 0.5948,
+      "step": 5372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0818688236833673,
+      "learning_rate": 8.70731522069012e-06,
+      "loss": 0.6192,
+      "step": 5373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8848310161128967,
+      "learning_rate": 8.706858490372587e-06,
+      "loss": 0.5448,
+      "step": 5374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6538945616807896,
+      "learning_rate": 8.706401692921465e-06,
+      "loss": 0.7636,
+      "step": 5375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8734805161182024,
+      "learning_rate": 8.705944828346312e-06,
+      "loss": 0.62,
+      "step": 5376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4205235286299325,
+      "learning_rate": 8.705487896656697e-06,
+      "loss": 0.4071,
+      "step": 5377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9316721996981718,
+      "learning_rate": 8.705030897862184e-06,
+      "loss": 0.5744,
+      "step": 5378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3444372332797345,
+      "learning_rate": 8.704573831972338e-06,
+      "loss": 0.3985,
+      "step": 5379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2456349159117224,
+      "learning_rate": 8.704116698996732e-06,
+      "loss": 0.6934,
+      "step": 5380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7755922670996387,
+      "learning_rate": 8.703659498944933e-06,
+      "loss": 0.5457,
+      "step": 5381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7536226544172253,
+      "learning_rate": 8.70320223182651e-06,
+      "loss": 0.4435,
+      "step": 5382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7460638926823826,
+      "learning_rate": 8.702744897651037e-06,
+      "loss": 0.5263,
+      "step": 5383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.368663181738536,
+      "learning_rate": 8.702287496428093e-06,
+      "loss": 0.6274,
+      "step": 5384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0521291506765658,
+      "learning_rate": 8.701830028167246e-06,
+      "loss": 0.681,
+      "step": 5385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.040248730569858,
+      "learning_rate": 8.701372492878077e-06,
+      "loss": 0.5715,
+      "step": 5386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6627046895269477,
+      "learning_rate": 8.700914890570165e-06,
+      "loss": 0.5703,
+      "step": 5387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8508077445584714,
+      "learning_rate": 8.700457221253085e-06,
+      "loss": 0.5453,
+      "step": 5388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9256023483830944,
+      "learning_rate": 8.69999948493642e-06,
+      "loss": 0.5612,
+      "step": 5389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7529610537783118,
+      "learning_rate": 8.699541681629756e-06,
+      "loss": 0.4919,
+      "step": 5390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6579322691889584,
+      "learning_rate": 8.699083811342675e-06,
+      "loss": 0.4726,
+      "step": 5391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9787151781240448,
+      "learning_rate": 8.69862587408476e-06,
+      "loss": 0.4834,
+      "step": 5392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7240853799482154,
+      "learning_rate": 8.6981678698656e-06,
+      "loss": 0.6029,
+      "step": 5393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8920505046709173,
+      "learning_rate": 8.697709798694782e-06,
+      "loss": 0.4683,
+      "step": 5394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4895047686182064,
+      "learning_rate": 8.697251660581897e-06,
+      "loss": 0.5075,
+      "step": 5395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8608285726164588,
+      "learning_rate": 8.69679345553653e-06,
+      "loss": 0.4567,
+      "step": 5396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.23917929331184,
+      "learning_rate": 8.69633518356828e-06,
+      "loss": 0.5323,
+      "step": 5397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5819252869865355,
+      "learning_rate": 8.695876844686737e-06,
+      "loss": 0.5329,
+      "step": 5398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9262204175307316,
+      "learning_rate": 8.6954184389015e-06,
+      "loss": 0.5253,
+      "step": 5399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9005009553112664,
+      "learning_rate": 8.694959966222162e-06,
+      "loss": 0.7499,
+      "step": 5400
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5698920488357544,
+      "eval_runtime": 67.9869,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 5400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.204060162911664,
+      "learning_rate": 8.69450142665832e-06,
+      "loss": 0.5859,
+      "step": 5401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8428808627155393,
+      "learning_rate": 8.694042820219576e-06,
+      "loss": 0.4999,
+      "step": 5402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6341907729491023,
+      "learning_rate": 8.693584146915527e-06,
+      "loss": 0.4863,
+      "step": 5403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8444512667495772,
+      "learning_rate": 8.693125406755782e-06,
+      "loss": 0.5972,
+      "step": 5404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9009230096253855,
+      "learning_rate": 8.692666599749938e-06,
+      "loss": 0.5191,
+      "step": 5405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8339773560536734,
+      "learning_rate": 8.6922077259076e-06,
+      "loss": 0.5439,
+      "step": 5406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4269557593706748,
+      "learning_rate": 8.691748785238378e-06,
+      "loss": 0.6737,
+      "step": 5407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1317603970143613,
+      "learning_rate": 8.691289777751877e-06,
+      "loss": 0.5502,
+      "step": 5408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.084147119093963,
+      "learning_rate": 8.690830703457705e-06,
+      "loss": 0.6471,
+      "step": 5409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.790969195692917,
+      "learning_rate": 8.690371562365477e-06,
+      "loss": 0.5163,
+      "step": 5410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6139975015646129,
+      "learning_rate": 8.6899123544848e-06,
+      "loss": 0.4137,
+      "step": 5411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0518245494677165,
+      "learning_rate": 8.689453079825291e-06,
+      "loss": 0.7121,
+      "step": 5412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.846239072452221,
+      "learning_rate": 8.688993738396562e-06,
+      "loss": 0.5897,
+      "step": 5413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5949162870263804,
+      "learning_rate": 8.688534330208228e-06,
+      "loss": 0.4203,
+      "step": 5414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4297638608864058,
+      "learning_rate": 8.68807485526991e-06,
+      "loss": 0.4181,
+      "step": 5415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.860789403680215,
+      "learning_rate": 8.687615313591225e-06,
+      "loss": 0.5611,
+      "step": 5416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8611037422105805,
+      "learning_rate": 8.687155705181794e-06,
+      "loss": 0.5406,
+      "step": 5417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7158053457995854,
+      "learning_rate": 8.686696030051237e-06,
+      "loss": 0.4989,
+      "step": 5418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3011052074831166,
+      "learning_rate": 8.686236288209178e-06,
+      "loss": 0.5503,
+      "step": 5419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.154425730373496,
+      "learning_rate": 8.685776479665241e-06,
+      "loss": 0.576,
+      "step": 5420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6384420416042214,
+      "learning_rate": 8.685316604429052e-06,
+      "loss": 0.3851,
+      "step": 5421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9338244423616537,
+      "learning_rate": 8.68485666251024e-06,
+      "loss": 0.6178,
+      "step": 5422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7385416049776297,
+      "learning_rate": 8.684396653918431e-06,
+      "loss": 0.4236,
+      "step": 5423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1035633219831595,
+      "learning_rate": 8.683936578663255e-06,
+      "loss": 0.4979,
+      "step": 5424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1374076826186825,
+      "learning_rate": 8.683476436754345e-06,
+      "loss": 0.5232,
+      "step": 5425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0696715800257306,
+      "learning_rate": 8.683016228201334e-06,
+      "loss": 0.5463,
+      "step": 5426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3330874243795003,
+      "learning_rate": 8.682555953013854e-06,
+      "loss": 0.5825,
+      "step": 5427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6906124030303098,
+      "learning_rate": 8.682095611201544e-06,
+      "loss": 0.4801,
+      "step": 5428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.264147779065386,
+      "learning_rate": 8.681635202774037e-06,
+      "loss": 0.6279,
+      "step": 5429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6707583328354814,
+      "learning_rate": 8.681174727740974e-06,
+      "loss": 0.4051,
+      "step": 5430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.207623056095337,
+      "learning_rate": 8.680714186111993e-06,
+      "loss": 0.6006,
+      "step": 5431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4059355243859297,
+      "learning_rate": 8.680253577896736e-06,
+      "loss": 0.4243,
+      "step": 5432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.066692298180359,
+      "learning_rate": 8.679792903104847e-06,
+      "loss": 0.6006,
+      "step": 5433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.042859856917176,
+      "learning_rate": 8.679332161745969e-06,
+      "loss": 0.5775,
+      "step": 5434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3041543658162924,
+      "learning_rate": 8.678871353829745e-06,
+      "loss": 0.789,
+      "step": 5435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.640579199306264,
+      "learning_rate": 8.678410479365825e-06,
+      "loss": 0.6462,
+      "step": 5436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8278082313401829,
+      "learning_rate": 8.677949538363856e-06,
+      "loss": 0.5431,
+      "step": 5437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.822788326013141,
+      "learning_rate": 8.677488530833485e-06,
+      "loss": 0.5663,
+      "step": 5438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9789783461887733,
+      "learning_rate": 8.677027456784368e-06,
+      "loss": 0.4968,
+      "step": 5439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6655315780386217,
+      "learning_rate": 8.676566316226154e-06,
+      "loss": 0.46,
+      "step": 5440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.496249752662404,
+      "learning_rate": 8.676105109168495e-06,
+      "loss": 0.581,
+      "step": 5441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6074146856844844,
+      "learning_rate": 8.67564383562105e-06,
+      "loss": 0.4562,
+      "step": 5442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.717094572639944,
+      "learning_rate": 8.675182495593475e-06,
+      "loss": 0.4419,
+      "step": 5443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5963956631921061,
+      "learning_rate": 8.674721089095426e-06,
+      "loss": 0.4344,
+      "step": 5444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7024524285730123,
+      "learning_rate": 8.67425961613656e-06,
+      "loss": 0.54,
+      "step": 5445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5696012591567412,
+      "learning_rate": 8.673798076726544e-06,
+      "loss": 0.505,
+      "step": 5446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8756686717052988,
+      "learning_rate": 8.673336470875035e-06,
+      "loss": 0.4942,
+      "step": 5447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7141965704297353,
+      "learning_rate": 8.672874798591699e-06,
+      "loss": 0.5215,
+      "step": 5448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6708483495535715,
+      "learning_rate": 8.672413059886199e-06,
+      "loss": 0.4988,
+      "step": 5449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.383731449361933,
+      "learning_rate": 8.671951254768204e-06,
+      "loss": 0.542,
+      "step": 5450
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5703489184379578,
+      "eval_runtime": 67.9816,
+      "eval_samples_per_second": 50.161,
+      "eval_steps_per_second": 1.574,
+      "step": 5450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6100049795703155,
+      "learning_rate": 8.671489383247377e-06,
+      "loss": 0.5056,
+      "step": 5451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8632586326208762,
+      "learning_rate": 8.671027445333389e-06,
+      "loss": 0.5489,
+      "step": 5452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7768455663378995,
+      "learning_rate": 8.670565441035912e-06,
+      "loss": 0.4587,
+      "step": 5453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9206053740833597,
+      "learning_rate": 8.67010337036462e-06,
+      "loss": 0.521,
+      "step": 5454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.00851738855039,
+      "learning_rate": 8.669641233329179e-06,
+      "loss": 0.4924,
+      "step": 5455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5052581094392354,
+      "learning_rate": 8.66917902993927e-06,
+      "loss": 0.4475,
+      "step": 5456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9725154190594791,
+      "learning_rate": 8.668716760204565e-06,
+      "loss": 0.6478,
+      "step": 5457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.181595786599457,
+      "learning_rate": 8.668254424134742e-06,
+      "loss": 0.6695,
+      "step": 5458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1157286492479694,
+      "learning_rate": 8.667792021739482e-06,
+      "loss": 0.7061,
+      "step": 5459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.011709806197984,
+      "learning_rate": 8.667329553028464e-06,
+      "loss": 0.5033,
+      "step": 5460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6347612835224594,
+      "learning_rate": 8.666867018011368e-06,
+      "loss": 0.5066,
+      "step": 5461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.217189913146608,
+      "learning_rate": 8.666404416697879e-06,
+      "loss": 0.7568,
+      "step": 5462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8007007782791933,
+      "learning_rate": 8.665941749097678e-06,
+      "loss": 0.4696,
+      "step": 5463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1071776183560265,
+      "learning_rate": 8.665479015220455e-06,
+      "loss": 0.6448,
+      "step": 5464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0553380239956107,
+      "learning_rate": 8.665016215075895e-06,
+      "loss": 0.6152,
+      "step": 5465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0174334168025037,
+      "learning_rate": 8.664553348673686e-06,
+      "loss": 0.7159,
+      "step": 5466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9546784355882956,
+      "learning_rate": 8.664090416023517e-06,
+      "loss": 0.6458,
+      "step": 5467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9740845699130367,
+      "learning_rate": 8.663627417135083e-06,
+      "loss": 0.6304,
+      "step": 5468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9701522447260706,
+      "learning_rate": 8.663164352018072e-06,
+      "loss": 0.6693,
+      "step": 5469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.566179226009902,
+      "learning_rate": 8.66270122068218e-06,
+      "loss": 0.4669,
+      "step": 5470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3318970552889626,
+      "learning_rate": 8.662238023137104e-06,
+      "loss": 0.4347,
+      "step": 5471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.925805110567662,
+      "learning_rate": 8.661774759392536e-06,
+      "loss": 0.5274,
+      "step": 5472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8349533310850819,
+      "learning_rate": 8.66131142945818e-06,
+      "loss": 0.4377,
+      "step": 5473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1514474739898097,
+      "learning_rate": 8.660848033343733e-06,
+      "loss": 0.5968,
+      "step": 5474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2260398424288166,
+      "learning_rate": 8.660384571058895e-06,
+      "loss": 0.556,
+      "step": 5475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.245692697084012,
+      "learning_rate": 8.65992104261337e-06,
+      "loss": 0.684,
+      "step": 5476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9530117169734844,
+      "learning_rate": 8.659457448016859e-06,
+      "loss": 0.611,
+      "step": 5477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.406361324946669,
+      "learning_rate": 8.658993787279068e-06,
+      "loss": 0.5817,
+      "step": 5478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.646378697312944,
+      "learning_rate": 8.658530060409706e-06,
+      "loss": 0.5117,
+      "step": 5479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9626931290922764,
+      "learning_rate": 8.658066267418479e-06,
+      "loss": 0.4642,
+      "step": 5480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8382209226637782,
+      "learning_rate": 8.657602408315094e-06,
+      "loss": 0.4016,
+      "step": 5481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0991864498927146,
+      "learning_rate": 8.657138483109268e-06,
+      "loss": 0.6408,
+      "step": 5482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8364335295909675,
+      "learning_rate": 8.656674491810704e-06,
+      "loss": 0.5413,
+      "step": 5483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.602398765712316,
+      "learning_rate": 8.656210434429124e-06,
+      "loss": 0.4469,
+      "step": 5484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8049020478533035,
+      "learning_rate": 8.655746310974237e-06,
+      "loss": 0.5585,
+      "step": 5485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2752315958793674,
+      "learning_rate": 8.655282121455762e-06,
+      "loss": 0.7289,
+      "step": 5486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.710144426312416,
+      "learning_rate": 8.654817865883415e-06,
+      "loss": 0.6321,
+      "step": 5487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3292180448322932,
+      "learning_rate": 8.654353544266915e-06,
+      "loss": 0.393,
+      "step": 5488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6187430394148028,
+      "learning_rate": 8.653889156615984e-06,
+      "loss": 0.4788,
+      "step": 5489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.091284155492677,
+      "learning_rate": 8.653424702940342e-06,
+      "loss": 0.6288,
+      "step": 5490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5363424158184074,
+      "learning_rate": 8.652960183249712e-06,
+      "loss": 0.4013,
+      "step": 5491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5361781739745888,
+      "learning_rate": 8.652495597553821e-06,
+      "loss": 0.4749,
+      "step": 5492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3377781663051342,
+      "learning_rate": 8.652030945862392e-06,
+      "loss": 0.5644,
+      "step": 5493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9721057131082258,
+      "learning_rate": 8.65156622818515e-06,
+      "loss": 0.5911,
+      "step": 5494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0755188034642584,
+      "learning_rate": 8.651101444531829e-06,
+      "loss": 0.4496,
+      "step": 5495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0628784001150366,
+      "learning_rate": 8.650636594912156e-06,
+      "loss": 0.5555,
+      "step": 5496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0777051115306815,
+      "learning_rate": 8.650171679335863e-06,
+      "loss": 0.6364,
+      "step": 5497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4208452468101591,
+      "learning_rate": 8.649706697812684e-06,
+      "loss": 0.3604,
+      "step": 5498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8723540765381728,
+      "learning_rate": 8.64924165035235e-06,
+      "loss": 0.4916,
+      "step": 5499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7726806816325382,
+      "learning_rate": 8.648776536964597e-06,
+      "loss": 0.4489,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5688517689704895,
+      "eval_runtime": 67.9861,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.708808430708936,
+      "learning_rate": 8.648311357659164e-06,
+      "loss": 0.5119,
+      "step": 5501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7323366194761638,
+      "learning_rate": 8.647846112445788e-06,
+      "loss": 0.5348,
+      "step": 5502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5425959643659626,
+      "learning_rate": 8.64738080133421e-06,
+      "loss": 0.4218,
+      "step": 5503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9631079996021275,
+      "learning_rate": 8.646915424334168e-06,
+      "loss": 0.426,
+      "step": 5504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3290118577172465,
+      "learning_rate": 8.646449981455409e-06,
+      "loss": 0.6351,
+      "step": 5505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1046573642630526,
+      "learning_rate": 8.645984472707671e-06,
+      "loss": 0.8162,
+      "step": 5506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6181406989690899,
+      "learning_rate": 8.645518898100704e-06,
+      "loss": 0.471,
+      "step": 5507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6299945788678092,
+      "learning_rate": 8.645053257644252e-06,
+      "loss": 0.5212,
+      "step": 5508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2438684046344477,
+      "learning_rate": 8.644587551348063e-06,
+      "loss": 0.4884,
+      "step": 5509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9319844895173381,
+      "learning_rate": 8.644121779221887e-06,
+      "loss": 0.625,
+      "step": 5510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.218718374488598,
+      "learning_rate": 8.643655941275475e-06,
+      "loss": 0.6102,
+      "step": 5511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4497057891641667,
+      "learning_rate": 8.643190037518579e-06,
+      "loss": 0.5406,
+      "step": 5512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6469264191483048,
+      "learning_rate": 8.64272406796095e-06,
+      "loss": 0.4565,
+      "step": 5513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.06466102854698,
+      "learning_rate": 8.642258032612346e-06,
+      "loss": 0.5604,
+      "step": 5514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.070928529235729,
+      "learning_rate": 8.641791931482521e-06,
+      "loss": 0.587,
+      "step": 5515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0534144879458465,
+      "learning_rate": 8.641325764581235e-06,
+      "loss": 0.5654,
+      "step": 5516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7628819196725556,
+      "learning_rate": 8.640859531918244e-06,
+      "loss": 0.6098,
+      "step": 5517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.587452264213432,
+      "learning_rate": 8.64039323350331e-06,
+      "loss": 0.6936,
+      "step": 5518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.300538832192194,
+      "learning_rate": 8.639926869346193e-06,
+      "loss": 0.5609,
+      "step": 5519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7299460191676763,
+      "learning_rate": 8.63946043945666e-06,
+      "loss": 0.5434,
+      "step": 5520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.90752167550356,
+      "learning_rate": 8.63899394384447e-06,
+      "loss": 0.5116,
+      "step": 5521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.865052041241677,
+      "learning_rate": 8.638527382519392e-06,
+      "loss": 0.6948,
+      "step": 5522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0451168262675368,
+      "learning_rate": 8.638060755491193e-06,
+      "loss": 0.5941,
+      "step": 5523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5491875949389575,
+      "learning_rate": 8.637594062769641e-06,
+      "loss": 0.4727,
+      "step": 5524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8562856152485259,
+      "learning_rate": 8.637127304364506e-06,
+      "loss": 0.4827,
+      "step": 5525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7709538891941992,
+      "learning_rate": 8.63666048028556e-06,
+      "loss": 0.5057,
+      "step": 5526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8792093587413294,
+      "learning_rate": 8.636193590542575e-06,
+      "loss": 0.5459,
+      "step": 5527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7084104128893325,
+      "learning_rate": 8.635726635145325e-06,
+      "loss": 0.478,
+      "step": 5528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0967782431407365,
+      "learning_rate": 8.635259614103586e-06,
+      "loss": 0.5826,
+      "step": 5529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.993256054501943,
+      "learning_rate": 8.634792527427134e-06,
+      "loss": 0.5335,
+      "step": 5530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.277775793596958,
+      "learning_rate": 8.634325375125747e-06,
+      "loss": 0.6028,
+      "step": 5531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.663184848684451,
+      "learning_rate": 8.633858157209205e-06,
+      "loss": 0.5033,
+      "step": 5532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.870898130296369,
+      "learning_rate": 8.63339087368729e-06,
+      "loss": 0.605,
+      "step": 5533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.425599819338875,
+      "learning_rate": 8.632923524569784e-06,
+      "loss": 0.5034,
+      "step": 5534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.597461238959831,
+      "learning_rate": 8.632456109866469e-06,
+      "loss": 0.4942,
+      "step": 5535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0909544471557306,
+      "learning_rate": 8.63198862958713e-06,
+      "loss": 0.6148,
+      "step": 5536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7205998496536337,
+      "learning_rate": 8.631521083741556e-06,
+      "loss": 0.5404,
+      "step": 5537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8774254915284532,
+      "learning_rate": 8.631053472339531e-06,
+      "loss": 0.4052,
+      "step": 5538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.629909800915589,
+      "learning_rate": 8.630585795390848e-06,
+      "loss": 0.5193,
+      "step": 5539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.053643744894397,
+      "learning_rate": 8.630118052905295e-06,
+      "loss": 0.6363,
+      "step": 5540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7243820900301332,
+      "learning_rate": 8.629650244892666e-06,
+      "loss": 0.7283,
+      "step": 5541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.433301432218036,
+      "learning_rate": 8.629182371362754e-06,
+      "loss": 0.6033,
+      "step": 5542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.132633550066256,
+      "learning_rate": 8.62871443232535e-06,
+      "loss": 0.4936,
+      "step": 5543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.092488942986818,
+      "learning_rate": 8.628246427790254e-06,
+      "loss": 0.6272,
+      "step": 5544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1092554253292777,
+      "learning_rate": 8.627778357767261e-06,
+      "loss": 0.5448,
+      "step": 5545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.022554961292017,
+      "learning_rate": 8.627310222266172e-06,
+      "loss": 0.637,
+      "step": 5546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9226393960985793,
+      "learning_rate": 8.626842021296785e-06,
+      "loss": 0.6624,
+      "step": 5547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8065754022852774,
+      "learning_rate": 8.626373754868904e-06,
+      "loss": 0.5189,
+      "step": 5548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6922367154325653,
+      "learning_rate": 8.625905422992328e-06,
+      "loss": 0.4426,
+      "step": 5549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.819222400585412,
+      "learning_rate": 8.625437025676865e-06,
+      "loss": 0.5016,
+      "step": 5550
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.570105254650116,
+      "eval_runtime": 68.0236,
+      "eval_samples_per_second": 50.13,
+      "eval_steps_per_second": 1.573,
+      "step": 5550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8688292845216747,
+      "learning_rate": 8.624968562932319e-06,
+      "loss": 0.505,
+      "step": 5551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0353140379221792,
+      "learning_rate": 8.624500034768495e-06,
+      "loss": 0.532,
+      "step": 5552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6954042798982034,
+      "learning_rate": 8.624031441195207e-06,
+      "loss": 0.4055,
+      "step": 5553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3617697517645924,
+      "learning_rate": 8.62356278222226e-06,
+      "loss": 0.3574,
+      "step": 5554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6585382694355875,
+      "learning_rate": 8.623094057859466e-06,
+      "loss": 0.3923,
+      "step": 5555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.945408028546837,
+      "learning_rate": 8.622625268116637e-06,
+      "loss": 0.4554,
+      "step": 5556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.859962126001307,
+      "learning_rate": 8.62215641300359e-06,
+      "loss": 0.4667,
+      "step": 5557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.755267385024001,
+      "learning_rate": 8.621687492530135e-06,
+      "loss": 0.5203,
+      "step": 5558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9133188162431587,
+      "learning_rate": 8.621218506706092e-06,
+      "loss": 0.5999,
+      "step": 5559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5221730803126334,
+      "learning_rate": 8.620749455541278e-06,
+      "loss": 0.6723,
+      "step": 5560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9469646476115374,
+      "learning_rate": 8.620280339045513e-06,
+      "loss": 0.6663,
+      "step": 5561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0213759757700847,
+      "learning_rate": 8.619811157228617e-06,
+      "loss": 0.4619,
+      "step": 5562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.011736285673713,
+      "learning_rate": 8.619341910100413e-06,
+      "loss": 0.5771,
+      "step": 5563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4210182056674927,
+      "learning_rate": 8.618872597670723e-06,
+      "loss": 0.444,
+      "step": 5564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.285382415321037,
+      "learning_rate": 8.618403219949371e-06,
+      "loss": 0.717,
+      "step": 5565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1417777580453152,
+      "learning_rate": 8.617933776946186e-06,
+      "loss": 0.4396,
+      "step": 5566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.817417329007721,
+      "learning_rate": 8.617464268670995e-06,
+      "loss": 0.5578,
+      "step": 5567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.651647943380257,
+      "learning_rate": 8.616994695133625e-06,
+      "loss": 0.4467,
+      "step": 5568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8255373418658136,
+      "learning_rate": 8.616525056343906e-06,
+      "loss": 0.5551,
+      "step": 5569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.922188791422465,
+      "learning_rate": 8.616055352311671e-06,
+      "loss": 0.5218,
+      "step": 5570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8320159900676067,
+      "learning_rate": 8.615585583046755e-06,
+      "loss": 0.5576,
+      "step": 5571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0965003857369253,
+      "learning_rate": 8.615115748558988e-06,
+      "loss": 0.4816,
+      "step": 5572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8275377153156636,
+      "learning_rate": 8.614645848858207e-06,
+      "loss": 0.5106,
+      "step": 5573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8149999069705207,
+      "learning_rate": 8.61417588395425e-06,
+      "loss": 0.6367,
+      "step": 5574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9759074286164022,
+      "learning_rate": 8.613705853856957e-06,
+      "loss": 0.5929,
+      "step": 5575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.061030716930067,
+      "learning_rate": 8.613235758576163e-06,
+      "loss": 0.8721,
+      "step": 5576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0380535593310816,
+      "learning_rate": 8.612765598121713e-06,
+      "loss": 0.4765,
+      "step": 5577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.201688792059053,
+      "learning_rate": 8.61229537250345e-06,
+      "loss": 0.6664,
+      "step": 5578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4549090770517727,
+      "learning_rate": 8.611825081731217e-06,
+      "loss": 0.6946,
+      "step": 5579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.23875592930089,
+      "learning_rate": 8.611354725814858e-06,
+      "loss": 0.6514,
+      "step": 5580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4933741812113985,
+      "learning_rate": 8.610884304764218e-06,
+      "loss": 0.323,
+      "step": 5581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9892340570029778,
+      "learning_rate": 8.610413818589151e-06,
+      "loss": 0.5098,
+      "step": 5582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1712156667758244,
+      "learning_rate": 8.6099432672995e-06,
+      "loss": 0.5722,
+      "step": 5583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7737052077320155,
+      "learning_rate": 8.609472650905119e-06,
+      "loss": 0.5297,
+      "step": 5584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2596457994146153,
+      "learning_rate": 8.60900196941586e-06,
+      "loss": 0.6195,
+      "step": 5585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.242568336366564,
+      "learning_rate": 8.608531222841575e-06,
+      "loss": 0.5267,
+      "step": 5586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.728865191754132,
+      "learning_rate": 8.608060411192121e-06,
+      "loss": 0.5322,
+      "step": 5587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.01594777703615,
+      "learning_rate": 8.607589534477351e-06,
+      "loss": 0.673,
+      "step": 5588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7691726636561365,
+      "learning_rate": 8.607118592707126e-06,
+      "loss": 0.6014,
+      "step": 5589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2220473243131855,
+      "learning_rate": 8.606647585891302e-06,
+      "loss": 0.5734,
+      "step": 5590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0733423458088556,
+      "learning_rate": 8.606176514039742e-06,
+      "loss": 0.653,
+      "step": 5591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8217564114884885,
+      "learning_rate": 8.605705377162304e-06,
+      "loss": 0.6958,
+      "step": 5592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4406014075256734,
+      "learning_rate": 8.605234175268855e-06,
+      "loss": 0.4208,
+      "step": 5593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9010202421177844,
+      "learning_rate": 8.604762908369255e-06,
+      "loss": 0.5441,
+      "step": 5594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.687032125544531,
+      "learning_rate": 8.604291576473374e-06,
+      "loss": 0.4143,
+      "step": 5595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8608631311098245,
+      "learning_rate": 8.603820179591075e-06,
+      "loss": 0.5878,
+      "step": 5596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7779238666738821,
+      "learning_rate": 8.60334871773223e-06,
+      "loss": 0.4862,
+      "step": 5597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9993154376789106,
+      "learning_rate": 8.602877190906707e-06,
+      "loss": 0.7004,
+      "step": 5598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5024876243251961,
+      "learning_rate": 8.602405599124375e-06,
+      "loss": 0.4704,
+      "step": 5599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.428649067697289,
+      "learning_rate": 8.601933942395111e-06,
+      "loss": 0.4714,
+      "step": 5600
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.5679662823677063,
+      "eval_runtime": 67.9865,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 5600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8986798595824248,
+      "learning_rate": 8.601462220728785e-06,
+      "loss": 0.5214,
+      "step": 5601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1473857420712044,
+      "learning_rate": 8.600990434135276e-06,
+      "loss": 0.6722,
+      "step": 5602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0185348528410723,
+      "learning_rate": 8.600518582624458e-06,
+      "loss": 0.5653,
+      "step": 5603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1836354487106675,
+      "learning_rate": 8.600046666206207e-06,
+      "loss": 0.5084,
+      "step": 5604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7123792248925722,
+      "learning_rate": 8.599574684890406e-06,
+      "loss": 0.6461,
+      "step": 5605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1615216279721414,
+      "learning_rate": 8.599102638686936e-06,
+      "loss": 0.6697,
+      "step": 5606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7015585734267757,
+      "learning_rate": 8.598630527605674e-06,
+      "loss": 0.4842,
+      "step": 5607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0056503073383296,
+      "learning_rate": 8.59815835165651e-06,
+      "loss": 0.6563,
+      "step": 5608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.709631524448072,
+      "learning_rate": 8.597686110849322e-06,
+      "loss": 0.6198,
+      "step": 5609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5402472591057477,
+      "learning_rate": 8.597213805194e-06,
+      "loss": 0.3933,
+      "step": 5610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5546737642750352,
+      "learning_rate": 8.596741434700433e-06,
+      "loss": 0.441,
+      "step": 5611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7395685122183706,
+      "learning_rate": 8.596268999378506e-06,
+      "loss": 0.5598,
+      "step": 5612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9152057573630628,
+      "learning_rate": 8.59579649923811e-06,
+      "loss": 0.6395,
+      "step": 5613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.902360215581647,
+      "learning_rate": 8.595323934289137e-06,
+      "loss": 0.5702,
+      "step": 5614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6717830257797404,
+      "learning_rate": 8.594851304541481e-06,
+      "loss": 0.4691,
+      "step": 5615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7951216835595405,
+      "learning_rate": 8.594378610005036e-06,
+      "loss": 0.5245,
+      "step": 5616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8477976395290832,
+      "learning_rate": 8.593905850689695e-06,
+      "loss": 0.5457,
+      "step": 5617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.895989690636467,
+      "learning_rate": 8.593433026605357e-06,
+      "loss": 0.5709,
+      "step": 5618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9847485226013173,
+      "learning_rate": 8.59296013776192e-06,
+      "loss": 0.571,
+      "step": 5619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5572765748805213,
+      "learning_rate": 8.592487184169285e-06,
+      "loss": 0.4631,
+      "step": 5620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0461183032547363,
+      "learning_rate": 8.59201416583735e-06,
+      "loss": 0.5365,
+      "step": 5621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.574351997934442,
+      "learning_rate": 8.59154108277602e-06,
+      "loss": 0.5249,
+      "step": 5622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4786741498550295,
+      "learning_rate": 8.591067934995196e-06,
+      "loss": 0.4211,
+      "step": 5623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.464924793933012,
+      "learning_rate": 8.590594722504786e-06,
+      "loss": 0.7411,
+      "step": 5624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7523273431307642,
+      "learning_rate": 8.590121445314697e-06,
+      "loss": 0.4191,
+      "step": 5625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4182797398229536,
+      "learning_rate": 8.589648103434832e-06,
+      "loss": 0.3918,
+      "step": 5626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8722109790821337,
+      "learning_rate": 8.589174696875103e-06,
+      "loss": 0.5474,
+      "step": 5627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0030134283100365,
+      "learning_rate": 8.588701225645422e-06,
+      "loss": 0.5669,
+      "step": 5628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5664889927891674,
+      "learning_rate": 8.588227689755698e-06,
+      "loss": 0.3809,
+      "step": 5629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0729849847664448,
+      "learning_rate": 8.587754089215848e-06,
+      "loss": 0.6068,
+      "step": 5630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6444809305134753,
+      "learning_rate": 8.587280424035781e-06,
+      "loss": 0.5777,
+      "step": 5631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4536384647945688,
+      "learning_rate": 8.586806694225419e-06,
+      "loss": 0.5048,
+      "step": 5632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8493469953102153,
+      "learning_rate": 8.586332899794674e-06,
+      "loss": 0.5107,
+      "step": 5633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.763324623774584,
+      "learning_rate": 8.585859040753467e-06,
+      "loss": 0.5179,
+      "step": 5634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.829124712516176,
+      "learning_rate": 8.585385117111717e-06,
+      "loss": 0.5124,
+      "step": 5635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9782854914566979,
+      "learning_rate": 8.584911128879346e-06,
+      "loss": 0.5929,
+      "step": 5636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7873503004941842,
+      "learning_rate": 8.58443707606628e-06,
+      "loss": 0.5562,
+      "step": 5637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8009174575660425,
+      "learning_rate": 8.583962958682436e-06,
+      "loss": 0.4827,
+      "step": 5638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5367109415340594,
+      "learning_rate": 8.583488776737745e-06,
+      "loss": 0.4246,
+      "step": 5639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4589087302251618,
+      "learning_rate": 8.58301453024213e-06,
+      "loss": 0.4728,
+      "step": 5640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1172476909586906,
+      "learning_rate": 8.58254021920552e-06,
+      "loss": 0.7155,
+      "step": 5641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.741902571748127,
+      "learning_rate": 8.582065843637847e-06,
+      "loss": 0.4575,
+      "step": 5642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6836191584828064,
+      "learning_rate": 8.581591403549039e-06,
+      "loss": 0.5888,
+      "step": 5643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.001894562470366,
+      "learning_rate": 8.58111689894903e-06,
+      "loss": 0.6005,
+      "step": 5644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6361865758857936,
+      "learning_rate": 8.580642329847752e-06,
+      "loss": 0.4628,
+      "step": 5645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9399344977629942,
+      "learning_rate": 8.58016769625514e-06,
+      "loss": 0.5371,
+      "step": 5646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0727502019380784,
+      "learning_rate": 8.57969299818113e-06,
+      "loss": 0.5256,
+      "step": 5647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5925625753852166,
+      "learning_rate": 8.579218235635662e-06,
+      "loss": 0.5734,
+      "step": 5648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.715256025326034,
+      "learning_rate": 8.578743408628671e-06,
+      "loss": 0.5759,
+      "step": 5649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5355321724893758,
+      "learning_rate": 8.5782685171701e-06,
+      "loss": 0.3886,
+      "step": 5650
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.5650461316108704,
+      "eval_runtime": 67.9743,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 5650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2345232726480058,
+      "learning_rate": 8.577793561269888e-06,
+      "loss": 0.5566,
+      "step": 5651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7522614386149644,
+      "learning_rate": 8.57731854093798e-06,
+      "loss": 0.5782,
+      "step": 5652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7436885515011988,
+      "learning_rate": 8.57684345618432e-06,
+      "loss": 0.5658,
+      "step": 5653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.227234622664808,
+      "learning_rate": 8.576368307018852e-06,
+      "loss": 0.538,
+      "step": 5654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8721642159574614,
+      "learning_rate": 8.575893093451528e-06,
+      "loss": 0.4665,
+      "step": 5655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.176009311606382,
+      "learning_rate": 8.575417815492289e-06,
+      "loss": 0.5356,
+      "step": 5656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.821703988952457,
+      "learning_rate": 8.574942473151091e-06,
+      "loss": 0.5573,
+      "step": 5657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7295943437390302,
+      "learning_rate": 8.574467066437882e-06,
+      "loss": 0.4766,
+      "step": 5658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7108107757514839,
+      "learning_rate": 8.573991595362613e-06,
+      "loss": 0.4438,
+      "step": 5659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9050843354317617,
+      "learning_rate": 8.573516059935239e-06,
+      "loss": 0.4672,
+      "step": 5660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6559038229088217,
+      "learning_rate": 8.573040460165717e-06,
+      "loss": 0.3566,
+      "step": 5661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.139642182713669,
+      "learning_rate": 8.572564796064e-06,
+      "loss": 0.5023,
+      "step": 5662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5274876516340303,
+      "learning_rate": 8.572089067640048e-06,
+      "loss": 0.4049,
+      "step": 5663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0076803163775003,
+      "learning_rate": 8.571613274903821e-06,
+      "loss": 0.6963,
+      "step": 5664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8915153754295138,
+      "learning_rate": 8.571137417865276e-06,
+      "loss": 0.579,
+      "step": 5665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6541995701524672,
+      "learning_rate": 8.570661496534379e-06,
+      "loss": 0.4254,
+      "step": 5666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.640695947422823,
+      "learning_rate": 8.570185510921087e-06,
+      "loss": 0.5606,
+      "step": 5667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8490569196659397,
+      "learning_rate": 8.569709461035373e-06,
+      "loss": 0.5781,
+      "step": 5668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9353630765390455,
+      "learning_rate": 8.569233346887195e-06,
+      "loss": 0.589,
+      "step": 5669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8531316274413467,
+      "learning_rate": 8.568757168486523e-06,
+      "loss": 0.6289,
+      "step": 5670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.967883456693567,
+      "learning_rate": 8.568280925843326e-06,
+      "loss": 0.6137,
+      "step": 5671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5238508548996306,
+      "learning_rate": 8.567804618967575e-06,
+      "loss": 0.4541,
+      "step": 5672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.710381796141255,
+      "learning_rate": 8.567328247869239e-06,
+      "loss": 0.5171,
+      "step": 5673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9150203831096935,
+      "learning_rate": 8.566851812558291e-06,
+      "loss": 0.699,
+      "step": 5674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.724354433251031,
+      "learning_rate": 8.566375313044705e-06,
+      "loss": 0.5691,
+      "step": 5675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9783608486411648,
+      "learning_rate": 8.565898749338457e-06,
+      "loss": 0.6287,
+      "step": 5676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7050824225899839,
+      "learning_rate": 8.565422121449523e-06,
+      "loss": 0.549,
+      "step": 5677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7045114692994743,
+      "learning_rate": 8.56494542938788e-06,
+      "loss": 0.453,
+      "step": 5678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4358252583222966,
+      "learning_rate": 8.564468673163509e-06,
+      "loss": 0.5457,
+      "step": 5679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0576622883845643,
+      "learning_rate": 8.563991852786391e-06,
+      "loss": 0.5831,
+      "step": 5680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9309022124072786,
+      "learning_rate": 8.563514968266505e-06,
+      "loss": 0.6063,
+      "step": 5681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7183652131770883,
+      "learning_rate": 8.563038019613837e-06,
+      "loss": 0.4366,
+      "step": 5682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5697184503507022,
+      "learning_rate": 8.56256100683837e-06,
+      "loss": 0.528,
+      "step": 5683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.859705298806145,
+      "learning_rate": 8.56208392995009e-06,
+      "loss": 0.5289,
+      "step": 5684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7140936830370266,
+      "learning_rate": 8.561606788958985e-06,
+      "loss": 0.5896,
+      "step": 5685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.003258023075353,
+      "learning_rate": 8.561129583875045e-06,
+      "loss": 0.444,
+      "step": 5686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.697452313060535,
+      "learning_rate": 8.560652314708257e-06,
+      "loss": 0.5314,
+      "step": 5687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.708362066309824,
+      "learning_rate": 8.560174981468615e-06,
+      "loss": 0.5391,
+      "step": 5688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9121866214151801,
+      "learning_rate": 8.559697584166108e-06,
+      "loss": 0.646,
+      "step": 5689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.893534181127041,
+      "learning_rate": 8.559220122810735e-06,
+      "loss": 0.4857,
+      "step": 5690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7151981278525357,
+      "learning_rate": 8.558742597412488e-06,
+      "loss": 0.4908,
+      "step": 5691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7003818329988125,
+      "learning_rate": 8.558265007981366e-06,
+      "loss": 0.576,
+      "step": 5692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7963740220274926,
+      "learning_rate": 8.557787354527364e-06,
+      "loss": 0.604,
+      "step": 5693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0806523506486636,
+      "learning_rate": 8.557309637060483e-06,
+      "loss": 0.5732,
+      "step": 5694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.049829041704574,
+      "learning_rate": 8.556831855590723e-06,
+      "loss": 0.4435,
+      "step": 5695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0381596688082615,
+      "learning_rate": 8.556354010128089e-06,
+      "loss": 0.5189,
+      "step": 5696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.598745596731314,
+      "learning_rate": 8.555876100682581e-06,
+      "loss": 0.4745,
+      "step": 5697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9706485499644322,
+      "learning_rate": 8.555398127264206e-06,
+      "loss": 0.5846,
+      "step": 5698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8069496525676587,
+      "learning_rate": 8.554920089882967e-06,
+      "loss": 0.5481,
+      "step": 5699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1196566368022074,
+      "learning_rate": 8.554441988548872e-06,
+      "loss": 0.7425,
+      "step": 5700
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.5659103393554688,
+      "eval_runtime": 67.9692,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 5700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.268082864110178,
+      "learning_rate": 8.553963823271935e-06,
+      "loss": 0.5608,
+      "step": 5701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.32596585676157,
+      "learning_rate": 8.553485594062159e-06,
+      "loss": 0.5969,
+      "step": 5702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.367473463550752,
+      "learning_rate": 8.55300730092956e-06,
+      "loss": 0.606,
+      "step": 5703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8623152945493922,
+      "learning_rate": 8.552528943884151e-06,
+      "loss": 0.6689,
+      "step": 5704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9115788003661782,
+      "learning_rate": 8.552050522935945e-06,
+      "loss": 0.4622,
+      "step": 5705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6517159612572825,
+      "learning_rate": 8.551572038094952e-06,
+      "loss": 0.5077,
+      "step": 5706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7570203461775022,
+      "learning_rate": 8.5510934893712e-06,
+      "loss": 0.5338,
+      "step": 5707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8275117189820553,
+      "learning_rate": 8.550614876774697e-06,
+      "loss": 0.5647,
+      "step": 5708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9654385432242607,
+      "learning_rate": 8.550136200315467e-06,
+      "loss": 0.4619,
+      "step": 5709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7495156113080994,
+      "learning_rate": 8.54965746000353e-06,
+      "loss": 0.4297,
+      "step": 5710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6442695111480765,
+      "learning_rate": 8.54917865584891e-06,
+      "loss": 0.4131,
+      "step": 5711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9431565668430564,
+      "learning_rate": 8.548699787861627e-06,
+      "loss": 0.57,
+      "step": 5712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9904542802306688,
+      "learning_rate": 8.548220856051709e-06,
+      "loss": 0.5252,
+      "step": 5713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7432452715486595,
+      "learning_rate": 8.54774186042918e-06,
+      "loss": 0.4849,
+      "step": 5714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0181407926185875,
+      "learning_rate": 8.547262801004067e-06,
+      "loss": 0.6103,
+      "step": 5715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8338967211665906,
+      "learning_rate": 8.5467836777864e-06,
+      "loss": 0.5368,
+      "step": 5716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3733257706739983,
+      "learning_rate": 8.546304490786212e-06,
+      "loss": 0.3522,
+      "step": 5717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3247201246943745,
+      "learning_rate": 8.54582524001353e-06,
+      "loss": 0.7927,
+      "step": 5718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.015057781442705,
+      "learning_rate": 8.545345925478387e-06,
+      "loss": 0.4916,
+      "step": 5719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7371650544631816,
+      "learning_rate": 8.54486654719082e-06,
+      "loss": 0.4522,
+      "step": 5720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8528812857130705,
+      "learning_rate": 8.544387105160863e-06,
+      "loss": 0.4701,
+      "step": 5721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2761771859915645,
+      "learning_rate": 8.543907599398552e-06,
+      "loss": 0.7015,
+      "step": 5722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7222566435766518,
+      "learning_rate": 8.543428029913928e-06,
+      "loss": 0.4488,
+      "step": 5723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.050243268789426,
+      "learning_rate": 8.542948396717026e-06,
+      "loss": 0.4958,
+      "step": 5724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.481294264099955,
+      "learning_rate": 8.54246869981789e-06,
+      "loss": 0.4836,
+      "step": 5725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6355210966567866,
+      "learning_rate": 8.541988939226564e-06,
+      "loss": 0.4898,
+      "step": 5726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5872955858636293,
+      "learning_rate": 8.541509114953087e-06,
+      "loss": 0.5539,
+      "step": 5727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7003168513778983,
+      "learning_rate": 8.541029227007504e-06,
+      "loss": 0.6501,
+      "step": 5728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9816429488801908,
+      "learning_rate": 8.540549275399867e-06,
+      "loss": 0.4673,
+      "step": 5729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0186366763666457,
+      "learning_rate": 8.540069260140217e-06,
+      "loss": 0.5549,
+      "step": 5730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4454101781758055,
+      "learning_rate": 8.539589181238605e-06,
+      "loss": 0.6256,
+      "step": 5731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.463751015591277,
+      "learning_rate": 8.539109038705083e-06,
+      "loss": 0.6545,
+      "step": 5732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6200302753363613,
+      "learning_rate": 8.538628832549702e-06,
+      "loss": 0.5413,
+      "step": 5733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2329484560878687,
+      "learning_rate": 8.53814856278251e-06,
+      "loss": 0.5718,
+      "step": 5734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6767579648889854,
+      "learning_rate": 8.537668229413567e-06,
+      "loss": 0.5001,
+      "step": 5735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0133357612791496,
+      "learning_rate": 8.537187832452926e-06,
+      "loss": 0.4622,
+      "step": 5736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2565191602658055,
+      "learning_rate": 8.536707371910644e-06,
+      "loss": 0.7393,
+      "step": 5737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7546674860798175,
+      "learning_rate": 8.53622684779678e-06,
+      "loss": 0.518,
+      "step": 5738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6751063928830914,
+      "learning_rate": 8.535746260121393e-06,
+      "loss": 0.533,
+      "step": 5739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6330985703417487,
+      "learning_rate": 8.535265608894542e-06,
+      "loss": 0.5707,
+      "step": 5740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6509045914629674,
+      "learning_rate": 8.534784894126293e-06,
+      "loss": 0.5723,
+      "step": 5741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6204285503518159,
+      "learning_rate": 8.534304115826705e-06,
+      "loss": 0.5128,
+      "step": 5742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.116368735533964,
+      "learning_rate": 8.533823274005847e-06,
+      "loss": 0.4776,
+      "step": 5743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.228822480674051,
+      "learning_rate": 8.533342368673782e-06,
+      "loss": 0.74,
+      "step": 5744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7440161120816688,
+      "learning_rate": 8.532861399840579e-06,
+      "loss": 0.4684,
+      "step": 5745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6191099748726103,
+      "learning_rate": 8.532380367516307e-06,
+      "loss": 0.5315,
+      "step": 5746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4523243394867484,
+      "learning_rate": 8.531899271711033e-06,
+      "loss": 0.4411,
+      "step": 5747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9692123243242197,
+      "learning_rate": 8.531418112434833e-06,
+      "loss": 0.5088,
+      "step": 5748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.912585138895444,
+      "learning_rate": 8.530936889697778e-06,
+      "loss": 0.622,
+      "step": 5749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7200483360778949,
+      "learning_rate": 8.530455603509943e-06,
+      "loss": 0.5365,
+      "step": 5750
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.5664739012718201,
+      "eval_runtime": 67.9051,
+      "eval_samples_per_second": 50.217,
+      "eval_steps_per_second": 1.576,
+      "step": 5750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9546395614690888,
+      "learning_rate": 8.529974253881397e-06,
+      "loss": 0.5721,
+      "step": 5751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0289013465365504,
+      "learning_rate": 8.529492840822227e-06,
+      "loss": 0.5753,
+      "step": 5752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6510013870893536,
+      "learning_rate": 8.529011364342504e-06,
+      "loss": 0.5735,
+      "step": 5753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9976135630086427,
+      "learning_rate": 8.52852982445231e-06,
+      "loss": 0.6052,
+      "step": 5754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.042041119003604,
+      "learning_rate": 8.528048221161726e-06,
+      "loss": 0.5669,
+      "step": 5755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7666949807517904,
+      "learning_rate": 8.527566554480832e-06,
+      "loss": 0.3853,
+      "step": 5756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.434035509384101,
+      "learning_rate": 8.527084824419713e-06,
+      "loss": 0.429,
+      "step": 5757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2157930455048476,
+      "learning_rate": 8.526603030988454e-06,
+      "loss": 0.6549,
+      "step": 5758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.687552271477341,
+      "learning_rate": 8.526121174197139e-06,
+      "loss": 0.603,
+      "step": 5759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6559679935734573,
+      "learning_rate": 8.525639254055857e-06,
+      "loss": 0.8744,
+      "step": 5760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.304089246964766,
+      "learning_rate": 8.525157270574697e-06,
+      "loss": 0.5377,
+      "step": 5761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9081661116656476,
+      "learning_rate": 8.52467522376375e-06,
+      "loss": 0.6355,
+      "step": 5762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.868482813531231,
+      "learning_rate": 8.524193113633103e-06,
+      "loss": 0.5358,
+      "step": 5763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.860386484231858,
+      "learning_rate": 8.523710940192855e-06,
+      "loss": 0.6211,
+      "step": 5764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.879530515554487,
+      "learning_rate": 8.523228703453095e-06,
+      "loss": 0.7723,
+      "step": 5765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8486987697295916,
+      "learning_rate": 8.52274640342392e-06,
+      "loss": 0.6261,
+      "step": 5766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7970063721438727,
+      "learning_rate": 8.522264040115427e-06,
+      "loss": 0.5276,
+      "step": 5767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.172714479695561,
+      "learning_rate": 8.52178161353771e-06,
+      "loss": 0.6823,
+      "step": 5768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.053689250560901,
+      "learning_rate": 8.521299123700876e-06,
+      "loss": 0.5479,
+      "step": 5769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7902058910247647,
+      "learning_rate": 8.520816570615021e-06,
+      "loss": 0.4396,
+      "step": 5770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.209459717017393,
+      "learning_rate": 8.520333954290245e-06,
+      "loss": 0.5415,
+      "step": 5771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.776111343279877,
+      "learning_rate": 8.519851274736657e-06,
+      "loss": 0.508,
+      "step": 5772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.353253463409019,
+      "learning_rate": 8.519368531964354e-06,
+      "loss": 0.4428,
+      "step": 5773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6138337491580128,
+      "learning_rate": 8.51888572598345e-06,
+      "loss": 0.4165,
+      "step": 5774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.214710403125712,
+      "learning_rate": 8.518402856804045e-06,
+      "loss": 0.5233,
+      "step": 5775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.911304454731652,
+      "learning_rate": 8.517919924436253e-06,
+      "loss": 0.4929,
+      "step": 5776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1222087917566954,
+      "learning_rate": 8.517436928890182e-06,
+      "loss": 0.4625,
+      "step": 5777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.605193788098607,
+      "learning_rate": 8.516953870175942e-06,
+      "loss": 0.3901,
+      "step": 5778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0590085812502332,
+      "learning_rate": 8.51647074830365e-06,
+      "loss": 0.4403,
+      "step": 5779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9232095093398842,
+      "learning_rate": 8.515987563283412e-06,
+      "loss": 0.7234,
+      "step": 5780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.49584623133173,
+      "learning_rate": 8.51550431512535e-06,
+      "loss": 0.6532,
+      "step": 5781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7444180281917787,
+      "learning_rate": 8.515021003839578e-06,
+      "loss": 0.5061,
+      "step": 5782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4855028859176296,
+      "learning_rate": 8.514537629436213e-06,
+      "loss": 0.4601,
+      "step": 5783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5292171462101787,
+      "learning_rate": 8.514054191925377e-06,
+      "loss": 0.455,
+      "step": 5784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8159800154672805,
+      "learning_rate": 8.513570691317187e-06,
+      "loss": 0.5087,
+      "step": 5785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.814799798404294,
+      "learning_rate": 8.513087127621767e-06,
+      "loss": 0.5249,
+      "step": 5786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7709688918584099,
+      "learning_rate": 8.512603500849239e-06,
+      "loss": 0.5693,
+      "step": 5787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5450820977347204,
+      "learning_rate": 8.512119811009728e-06,
+      "loss": 0.4158,
+      "step": 5788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6339425772386094,
+      "learning_rate": 8.511636058113362e-06,
+      "loss": 0.4763,
+      "step": 5789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9088874570323275,
+      "learning_rate": 8.511152242170263e-06,
+      "loss": 0.4753,
+      "step": 5790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2090295488571603,
+      "learning_rate": 8.510668363190562e-06,
+      "loss": 0.4984,
+      "step": 5791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.261702176081684,
+      "learning_rate": 8.51018442118439e-06,
+      "loss": 0.8552,
+      "step": 5792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.180066435911173,
+      "learning_rate": 8.509700416161877e-06,
+      "loss": 0.5095,
+      "step": 5793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.167845863089818,
+      "learning_rate": 8.509216348133154e-06,
+      "loss": 0.5998,
+      "step": 5794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9072691973067473,
+      "learning_rate": 8.508732217108356e-06,
+      "loss": 0.7361,
+      "step": 5795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7351645262350126,
+      "learning_rate": 8.508248023097618e-06,
+      "loss": 0.5156,
+      "step": 5796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9086536114687107,
+      "learning_rate": 8.507763766111078e-06,
+      "loss": 0.5024,
+      "step": 5797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9629953727645313,
+      "learning_rate": 8.507279446158869e-06,
+      "loss": 0.67,
+      "step": 5798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.146571122014766,
+      "learning_rate": 8.506795063251133e-06,
+      "loss": 0.5497,
+      "step": 5799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8257197039256388,
+      "learning_rate": 8.506310617398012e-06,
+      "loss": 0.5726,
+      "step": 5800
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5669873356819153,
+      "eval_runtime": 67.9653,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 5800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6627082494143717,
+      "learning_rate": 8.505826108609645e-06,
+      "loss": 0.5224,
+      "step": 5801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.06637330386014,
+      "learning_rate": 8.505341536896174e-06,
+      "loss": 0.5332,
+      "step": 5802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9626154852617077,
+      "learning_rate": 8.504856902267746e-06,
+      "loss": 0.5542,
+      "step": 5803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6899445965089805,
+      "learning_rate": 8.504372204734502e-06,
+      "loss": 0.5441,
+      "step": 5804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7573322693260913,
+      "learning_rate": 8.503887444306595e-06,
+      "loss": 0.6893,
+      "step": 5805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7948935230478058,
+      "learning_rate": 8.50340262099417e-06,
+      "loss": 0.4527,
+      "step": 5806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2010041614833042,
+      "learning_rate": 8.502917734807376e-06,
+      "loss": 0.6648,
+      "step": 5807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.608368425168335,
+      "learning_rate": 8.502432785756366e-06,
+      "loss": 0.5852,
+      "step": 5808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0986521510926286,
+      "learning_rate": 8.501947773851289e-06,
+      "loss": 0.635,
+      "step": 5809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9557021474555227,
+      "learning_rate": 8.5014626991023e-06,
+      "loss": 0.5489,
+      "step": 5810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2379856042614565,
+      "learning_rate": 8.500977561519553e-06,
+      "loss": 0.6563,
+      "step": 5811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.487867638895503,
+      "learning_rate": 8.500492361113204e-06,
+      "loss": 0.3646,
+      "step": 5812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6928858144729746,
+      "learning_rate": 8.500007097893413e-06,
+      "loss": 0.4198,
+      "step": 5813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9032023689255353,
+      "learning_rate": 8.499521771870337e-06,
+      "loss": 0.6024,
+      "step": 5814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9134591036409685,
+      "learning_rate": 8.499036383054135e-06,
+      "loss": 0.6445,
+      "step": 5815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.148324680164189,
+      "learning_rate": 8.498550931454968e-06,
+      "loss": 0.7173,
+      "step": 5816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6583030545402668,
+      "learning_rate": 8.498065417083001e-06,
+      "loss": 0.4183,
+      "step": 5817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3382915367595498,
+      "learning_rate": 8.497579839948397e-06,
+      "loss": 0.3781,
+      "step": 5818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0785802594365537,
+      "learning_rate": 8.49709420006132e-06,
+      "loss": 0.7541,
+      "step": 5819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.978910374967632,
+      "learning_rate": 8.49660849743194e-06,
+      "loss": 0.4816,
+      "step": 5820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8978785825851565,
+      "learning_rate": 8.496122732070419e-06,
+      "loss": 0.5395,
+      "step": 5821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6881629928900768,
+      "learning_rate": 8.49563690398693e-06,
+      "loss": 0.5102,
+      "step": 5822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.836933933414936,
+      "learning_rate": 8.495151013191644e-06,
+      "loss": 0.5432,
+      "step": 5823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.761579267449031,
+      "learning_rate": 8.494665059694733e-06,
+      "loss": 0.4345,
+      "step": 5824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.293450950915053,
+      "learning_rate": 8.494179043506367e-06,
+      "loss": 0.5532,
+      "step": 5825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.886692018386682,
+      "learning_rate": 8.493692964636722e-06,
+      "loss": 0.6135,
+      "step": 5826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5898557127709976,
+      "learning_rate": 8.493206823095977e-06,
+      "loss": 0.4505,
+      "step": 5827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9915171041358,
+      "learning_rate": 8.492720618894305e-06,
+      "loss": 0.6549,
+      "step": 5828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8453899397718683,
+      "learning_rate": 8.492234352041886e-06,
+      "loss": 0.5519,
+      "step": 5829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7612642470433175,
+      "learning_rate": 8.4917480225489e-06,
+      "loss": 0.435,
+      "step": 5830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.125282107956936,
+      "learning_rate": 8.491261630425526e-06,
+      "loss": 0.5717,
+      "step": 5831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4572855255454336,
+      "learning_rate": 8.490775175681947e-06,
+      "loss": 0.639,
+      "step": 5832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8416550009180372,
+      "learning_rate": 8.490288658328348e-06,
+      "loss": 0.637,
+      "step": 5833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8083121716736832,
+      "learning_rate": 8.489802078374913e-06,
+      "loss": 0.7421,
+      "step": 5834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0370675121680493,
+      "learning_rate": 8.48931543583183e-06,
+      "loss": 0.5711,
+      "step": 5835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.431631520265884,
+      "learning_rate": 8.488828730709285e-06,
+      "loss": 0.4411,
+      "step": 5836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7344573232964167,
+      "learning_rate": 8.488341963017465e-06,
+      "loss": 0.4979,
+      "step": 5837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4026551705391586,
+      "learning_rate": 8.487855132766564e-06,
+      "loss": 0.811,
+      "step": 5838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.85953819474496,
+      "learning_rate": 8.487368239966772e-06,
+      "loss": 0.7109,
+      "step": 5839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9168584233067496,
+      "learning_rate": 8.48688128462828e-06,
+      "loss": 0.635,
+      "step": 5840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9608650710538995,
+      "learning_rate": 8.486394266761283e-06,
+      "loss": 0.5357,
+      "step": 5841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9698813339885881,
+      "learning_rate": 8.485907186375978e-06,
+      "loss": 0.5627,
+      "step": 5842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7682673029815488,
+      "learning_rate": 8.48542004348256e-06,
+      "loss": 0.5257,
+      "step": 5843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6024069667946172,
+      "learning_rate": 8.484932838091229e-06,
+      "loss": 0.521,
+      "step": 5844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.045740188573524,
+      "learning_rate": 8.484445570212181e-06,
+      "loss": 0.807,
+      "step": 5845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9162273973712753,
+      "learning_rate": 8.483958239855621e-06,
+      "loss": 0.4959,
+      "step": 5846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7242710277804987,
+      "learning_rate": 8.483470847031747e-06,
+      "loss": 0.4728,
+      "step": 5847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6590757865196522,
+      "learning_rate": 8.482983391750765e-06,
+      "loss": 0.4609,
+      "step": 5848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.945478787017439,
+      "learning_rate": 8.482495874022876e-06,
+      "loss": 0.5298,
+      "step": 5849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1523790015895132,
+      "learning_rate": 8.482008293858292e-06,
+      "loss": 0.6672,
+      "step": 5850
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5649994611740112,
+      "eval_runtime": 67.9996,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 5850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9155420792374929,
+      "learning_rate": 8.481520651267216e-06,
+      "loss": 0.4559,
+      "step": 5851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7338446611578948,
+      "learning_rate": 8.481032946259857e-06,
+      "loss": 0.4414,
+      "step": 5852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.247731168630442,
+      "learning_rate": 8.480545178846422e-06,
+      "loss": 0.5368,
+      "step": 5853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6466012320419332,
+      "learning_rate": 8.480057349037129e-06,
+      "loss": 0.462,
+      "step": 5854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7985560057660261,
+      "learning_rate": 8.479569456842183e-06,
+      "loss": 0.4561,
+      "step": 5855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9293232388193944,
+      "learning_rate": 8.479081502271803e-06,
+      "loss": 0.7813,
+      "step": 5856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.673125730447514,
+      "learning_rate": 8.4785934853362e-06,
+      "loss": 0.4435,
+      "step": 5857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7754750461544033,
+      "learning_rate": 8.478105406045594e-06,
+      "loss": 0.4346,
+      "step": 5858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3730824436132787,
+      "learning_rate": 8.477617264410201e-06,
+      "loss": 0.6045,
+      "step": 5859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.057544062552799,
+      "learning_rate": 8.47712906044024e-06,
+      "loss": 0.6821,
+      "step": 5860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3711475183903437,
+      "learning_rate": 8.476640794145932e-06,
+      "loss": 0.6548,
+      "step": 5861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.234039825675094,
+      "learning_rate": 8.476152465537498e-06,
+      "loss": 0.7471,
+      "step": 5862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8931449263801818,
+      "learning_rate": 8.47566407462516e-06,
+      "loss": 0.5543,
+      "step": 5863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9958302374714547,
+      "learning_rate": 8.475175621419142e-06,
+      "loss": 0.6142,
+      "step": 5864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6850458838586213,
+      "learning_rate": 8.47468710592967e-06,
+      "loss": 0.3947,
+      "step": 5865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.808487346441859,
+      "learning_rate": 8.474198528166973e-06,
+      "loss": 0.4903,
+      "step": 5866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9149855155586255,
+      "learning_rate": 8.473709888141275e-06,
+      "loss": 0.4609,
+      "step": 5867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.560832256414028,
+      "learning_rate": 8.47322118586281e-06,
+      "loss": 0.6266,
+      "step": 5868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.149649515413081,
+      "learning_rate": 8.472732421341803e-06,
+      "loss": 0.6418,
+      "step": 5869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4064061534172105,
+      "learning_rate": 8.472243594588493e-06,
+      "loss": 0.522,
+      "step": 5870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6916520489712885,
+      "learning_rate": 8.471754705613107e-06,
+      "loss": 0.6154,
+      "step": 5871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9605693641704418,
+      "learning_rate": 8.471265754425884e-06,
+      "loss": 0.5811,
+      "step": 5872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9473752371699307,
+      "learning_rate": 8.470776741037055e-06,
+      "loss": 0.4365,
+      "step": 5873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0938276949683323,
+      "learning_rate": 8.470287665456863e-06,
+      "loss": 0.5717,
+      "step": 5874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.112006277434945,
+      "learning_rate": 8.469798527695542e-06,
+      "loss": 0.6964,
+      "step": 5875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8380264857941861,
+      "learning_rate": 8.469309327763334e-06,
+      "loss": 0.5121,
+      "step": 5876
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.85853809916347,
+      "learning_rate": 8.46882006567048e-06,
+      "loss": 0.4397,
+      "step": 5877
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.913817797490893,
+      "learning_rate": 8.46833074142722e-06,
+      "loss": 0.6302,
+      "step": 5878
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7914009269317035,
+      "learning_rate": 8.467841355043802e-06,
+      "loss": 0.4713,
+      "step": 5879
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9614314873488445,
+      "learning_rate": 8.467351906530468e-06,
+      "loss": 0.4429,
+      "step": 5880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.167152865205736,
+      "learning_rate": 8.466862395897463e-06,
+      "loss": 0.555,
+      "step": 5881
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9351675164543292,
+      "learning_rate": 8.466372823155039e-06,
+      "loss": 0.5755,
+      "step": 5882
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8887053036247594,
+      "learning_rate": 8.465883188313442e-06,
+      "loss": 0.5637,
+      "step": 5883
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.152375458910161,
+      "learning_rate": 8.465393491382921e-06,
+      "loss": 0.708,
+      "step": 5884
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.671416960200921,
+      "learning_rate": 8.46490373237373e-06,
+      "loss": 0.6332,
+      "step": 5885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1921050201179595,
+      "learning_rate": 8.464413911296123e-06,
+      "loss": 0.5871,
+      "step": 5886
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8538919420884894,
+      "learning_rate": 8.463924028160348e-06,
+      "loss": 0.5624,
+      "step": 5887
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0668044749989805,
+      "learning_rate": 8.46343408297667e-06,
+      "loss": 0.5435,
+      "step": 5888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.997770978508609,
+      "learning_rate": 8.462944075755335e-06,
+      "loss": 0.5875,
+      "step": 5889
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6803449202138254,
+      "learning_rate": 8.462454006506609e-06,
+      "loss": 0.543,
+      "step": 5890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8351806712892884,
+      "learning_rate": 8.461963875240747e-06,
+      "loss": 0.6518,
+      "step": 5891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.139860506295016,
+      "learning_rate": 8.461473681968011e-06,
+      "loss": 0.5692,
+      "step": 5892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9039989778560424,
+      "learning_rate": 8.460983426698664e-06,
+      "loss": 0.6074,
+      "step": 5893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2177417741771683,
+      "learning_rate": 8.460493109442966e-06,
+      "loss": 0.7004,
+      "step": 5894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.770088565504791,
+      "learning_rate": 8.460002730211185e-06,
+      "loss": 0.5423,
+      "step": 5895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9760864570586714,
+      "learning_rate": 8.459512289013585e-06,
+      "loss": 0.4743,
+      "step": 5896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6391142980847506,
+      "learning_rate": 8.459021785860433e-06,
+      "loss": 0.4927,
+      "step": 5897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0210443524592736,
+      "learning_rate": 8.458531220761999e-06,
+      "loss": 0.4751,
+      "step": 5898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7844140167002627,
+      "learning_rate": 8.458040593728551e-06,
+      "loss": 0.5427,
+      "step": 5899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8387263338391264,
+      "learning_rate": 8.457549904770361e-06,
+      "loss": 0.5331,
+      "step": 5900
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5642580986022949,
+      "eval_runtime": 67.9007,
+      "eval_samples_per_second": 50.22,
+      "eval_steps_per_second": 1.576,
+      "step": 5900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7051788648399402,
+      "learning_rate": 8.4570591538977e-06,
+      "loss": 0.6078,
+      "step": 5901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1964261891372527,
+      "learning_rate": 8.456568341120843e-06,
+      "loss": 0.7332,
+      "step": 5902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.831584488820827,
+      "learning_rate": 8.456077466450064e-06,
+      "loss": 0.6765,
+      "step": 5903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7493148120656077,
+      "learning_rate": 8.455586529895639e-06,
+      "loss": 0.4682,
+      "step": 5904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8677731401444002,
+      "learning_rate": 8.455095531467846e-06,
+      "loss": 0.6113,
+      "step": 5905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1701401973790992,
+      "learning_rate": 8.454604471176962e-06,
+      "loss": 0.5593,
+      "step": 5906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5905046664558375,
+      "learning_rate": 8.454113349033272e-06,
+      "loss": 0.5188,
+      "step": 5907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7796532329697183,
+      "learning_rate": 8.453622165047051e-06,
+      "loss": 0.5565,
+      "step": 5908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6122428936791098,
+      "learning_rate": 8.453130919228587e-06,
+      "loss": 0.4918,
+      "step": 5909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7390355792296701,
+      "learning_rate": 8.452639611588158e-06,
+      "loss": 0.4634,
+      "step": 5910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9293351540611572,
+      "learning_rate": 8.452148242136055e-06,
+      "loss": 0.4948,
+      "step": 5911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5872708174106447,
+      "learning_rate": 8.451656810882562e-06,
+      "loss": 0.4623,
+      "step": 5912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6935704076732196,
+      "learning_rate": 8.451165317837969e-06,
+      "loss": 0.4672,
+      "step": 5913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.811034123994205,
+      "learning_rate": 8.450673763012559e-06,
+      "loss": 0.559,
+      "step": 5914
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5487300285928498,
+      "learning_rate": 8.450182146416628e-06,
+      "loss": 0.4015,
+      "step": 5915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.927179679605302,
+      "learning_rate": 8.449690468060467e-06,
+      "loss": 0.5263,
+      "step": 5916
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7543620589804854,
+      "learning_rate": 8.449198727954369e-06,
+      "loss": 0.5961,
+      "step": 5917
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8235368932951155,
+      "learning_rate": 8.448706926108625e-06,
+      "loss": 0.4822,
+      "step": 5918
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.712331210080619,
+      "learning_rate": 8.448215062533535e-06,
+      "loss": 0.5215,
+      "step": 5919
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5131011247824377,
+      "learning_rate": 8.447723137239392e-06,
+      "loss": 0.5173,
+      "step": 5920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.87965645621854,
+      "learning_rate": 8.447231150236497e-06,
+      "loss": 0.6871,
+      "step": 5921
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8611449061464032,
+      "learning_rate": 8.446739101535149e-06,
+      "loss": 0.5432,
+      "step": 5922
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8210048642927488,
+      "learning_rate": 8.446246991145649e-06,
+      "loss": 0.5831,
+      "step": 5923
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6565217484436578,
+      "learning_rate": 8.445754819078296e-06,
+      "loss": 0.3857,
+      "step": 5924
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7843904759609546,
+      "learning_rate": 8.445262585343397e-06,
+      "loss": 0.4503,
+      "step": 5925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0390760113401947,
+      "learning_rate": 8.444770289951254e-06,
+      "loss": 0.5814,
+      "step": 5926
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6414656570903512,
+      "learning_rate": 8.444277932912175e-06,
+      "loss": 0.4372,
+      "step": 5927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8109363275915547,
+      "learning_rate": 8.443785514236466e-06,
+      "loss": 0.4891,
+      "step": 5928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5335272973328073,
+      "learning_rate": 8.443293033934435e-06,
+      "loss": 0.4617,
+      "step": 5929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0174363884460162,
+      "learning_rate": 8.442800492016394e-06,
+      "loss": 0.4662,
+      "step": 5930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.277450394742591,
+      "learning_rate": 8.442307888492652e-06,
+      "loss": 0.6085,
+      "step": 5931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.129458090542357,
+      "learning_rate": 8.44181522337352e-06,
+      "loss": 0.5752,
+      "step": 5932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9695422131723752,
+      "learning_rate": 8.441322496669316e-06,
+      "loss": 0.5467,
+      "step": 5933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.472059347929063,
+      "learning_rate": 8.44082970839035e-06,
+      "loss": 0.3706,
+      "step": 5934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.970125557188432,
+      "learning_rate": 8.440336858546944e-06,
+      "loss": 0.61,
+      "step": 5935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7640650384155303,
+      "learning_rate": 8.43984394714941e-06,
+      "loss": 0.5414,
+      "step": 5936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.035444742734095,
+      "learning_rate": 8.439350974208069e-06,
+      "loss": 0.6878,
+      "step": 5937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.091747772347827,
+      "learning_rate": 8.438857939733241e-06,
+      "loss": 0.7332,
+      "step": 5938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.012544359909635,
+      "learning_rate": 8.438364843735248e-06,
+      "loss": 0.6159,
+      "step": 5939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2554910222334814,
+      "learning_rate": 8.437871686224413e-06,
+      "loss": 0.6048,
+      "step": 5940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.752110538213008,
+      "learning_rate": 8.437378467211058e-06,
+      "loss": 0.3997,
+      "step": 5941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.258719921884628,
+      "learning_rate": 8.436885186705509e-06,
+      "loss": 0.6421,
+      "step": 5942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6290424639991545,
+      "learning_rate": 8.436391844718094e-06,
+      "loss": 0.5171,
+      "step": 5943
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8804762161985353,
+      "learning_rate": 8.435898441259137e-06,
+      "loss": 0.4939,
+      "step": 5944
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5892511381794305,
+      "learning_rate": 8.435404976338972e-06,
+      "loss": 0.3894,
+      "step": 5945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5436863235945975,
+      "learning_rate": 8.434911449967926e-06,
+      "loss": 0.4766,
+      "step": 5946
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.932217466483162,
+      "learning_rate": 8.434417862156333e-06,
+      "loss": 0.6749,
+      "step": 5947
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.695812451610656,
+      "learning_rate": 8.433924212914523e-06,
+      "loss": 0.6457,
+      "step": 5948
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8295016487152878,
+      "learning_rate": 8.433430502252833e-06,
+      "loss": 0.5485,
+      "step": 5949
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7912331337512493,
+      "learning_rate": 8.432936730181596e-06,
+      "loss": 0.5546,
+      "step": 5950
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5653014183044434,
+      "eval_runtime": 67.9993,
+      "eval_samples_per_second": 50.148,
+      "eval_steps_per_second": 1.574,
+      "step": 5950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2317690126377494,
+      "learning_rate": 8.43244289671115e-06,
+      "loss": 0.5523,
+      "step": 5951
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7808645819238251,
+      "learning_rate": 8.431949001851838e-06,
+      "loss": 0.4551,
+      "step": 5952
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.303626508005965,
+      "learning_rate": 8.43145504561399e-06,
+      "loss": 0.6178,
+      "step": 5953
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6587401146643463,
+      "learning_rate": 8.430961028007955e-06,
+      "loss": 0.4881,
+      "step": 5954
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6903386570035364,
+      "learning_rate": 8.430466949044068e-06,
+      "loss": 0.5334,
+      "step": 5955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.050891834118776,
+      "learning_rate": 8.429972808732678e-06,
+      "loss": 0.5271,
+      "step": 5956
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9932341097377102,
+      "learning_rate": 8.429478607084125e-06,
+      "loss": 0.5658,
+      "step": 5957
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7721167860083253,
+      "learning_rate": 8.428984344108757e-06,
+      "loss": 0.5178,
+      "step": 5958
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9088440625194583,
+      "learning_rate": 8.428490019816924e-06,
+      "loss": 0.4951,
+      "step": 5959
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6695442134253637,
+      "learning_rate": 8.42799563421897e-06,
+      "loss": 0.4762,
+      "step": 5960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8511714309787497,
+      "learning_rate": 8.427501187325244e-06,
+      "loss": 0.536,
+      "step": 5961
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1067605186845264,
+      "learning_rate": 8.4270066791461e-06,
+      "loss": 0.6556,
+      "step": 5962
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7408325475912916,
+      "learning_rate": 8.42651210969189e-06,
+      "loss": 0.4443,
+      "step": 5963
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8054778505118456,
+      "learning_rate": 8.426017478972966e-06,
+      "loss": 0.5687,
+      "step": 5964
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.537146986255098,
+      "learning_rate": 8.425522786999685e-06,
+      "loss": 0.4462,
+      "step": 5965
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8307148493401055,
+      "learning_rate": 8.4250280337824e-06,
+      "loss": 0.4605,
+      "step": 5966
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.446357151669218,
+      "learning_rate": 8.424533219331471e-06,
+      "loss": 0.5614,
+      "step": 5967
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.495456969087286,
+      "learning_rate": 8.424038343657254e-06,
+      "loss": 0.517,
+      "step": 5968
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.180659635349323,
+      "learning_rate": 8.423543406770111e-06,
+      "loss": 0.5492,
+      "step": 5969
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9346315599088093,
+      "learning_rate": 8.423048408680404e-06,
+      "loss": 0.5557,
+      "step": 5970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7749224036900872,
+      "learning_rate": 8.422553349398491e-06,
+      "loss": 0.6275,
+      "step": 5971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5203278924273285,
+      "learning_rate": 8.422058228934742e-06,
+      "loss": 0.4194,
+      "step": 5972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7673900257241608,
+      "learning_rate": 8.421563047299519e-06,
+      "loss": 0.513,
+      "step": 5973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5550253333576083,
+      "learning_rate": 8.421067804503185e-06,
+      "loss": 0.6961,
+      "step": 5974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6379517826074674,
+      "learning_rate": 8.420572500556114e-06,
+      "loss": 0.5405,
+      "step": 5975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5496945234969621,
+      "learning_rate": 8.420077135468672e-06,
+      "loss": 0.4875,
+      "step": 5976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.225126496950986,
+      "learning_rate": 8.419581709251229e-06,
+      "loss": 0.5865,
+      "step": 5977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7269093024492768,
+      "learning_rate": 8.419086221914154e-06,
+      "loss": 0.52,
+      "step": 5978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9505049300141442,
+      "learning_rate": 8.418590673467821e-06,
+      "loss": 0.5216,
+      "step": 5979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8347432951070468,
+      "learning_rate": 8.418095063922608e-06,
+      "loss": 0.5406,
+      "step": 5980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7166430420354355,
+      "learning_rate": 8.417599393288886e-06,
+      "loss": 0.4777,
+      "step": 5981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8029320593906029,
+      "learning_rate": 8.417103661577034e-06,
+      "loss": 0.6541,
+      "step": 5982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6068238913808253,
+      "learning_rate": 8.416607868797427e-06,
+      "loss": 0.4111,
+      "step": 5983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8848380670168625,
+      "learning_rate": 8.416112014960448e-06,
+      "loss": 0.4406,
+      "step": 5984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.955117581887761,
+      "learning_rate": 8.415616100076475e-06,
+      "loss": 0.5907,
+      "step": 5985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7608905494910383,
+      "learning_rate": 8.415120124155887e-06,
+      "loss": 0.4905,
+      "step": 5986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.203282914591239,
+      "learning_rate": 8.414624087209074e-06,
+      "loss": 0.597,
+      "step": 5987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.569360564018018,
+      "learning_rate": 8.414127989246413e-06,
+      "loss": 0.5286,
+      "step": 5988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9189468791565993,
+      "learning_rate": 8.413631830278295e-06,
+      "loss": 0.5307,
+      "step": 5989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9237955240008728,
+      "learning_rate": 8.413135610315102e-06,
+      "loss": 0.5752,
+      "step": 5990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.812064199951402,
+      "learning_rate": 8.412639329367225e-06,
+      "loss": 0.7155,
+      "step": 5991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9598582498710615,
+      "learning_rate": 8.412142987445054e-06,
+      "loss": 0.5813,
+      "step": 5992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9674850520632556,
+      "learning_rate": 8.411646584558977e-06,
+      "loss": 0.485,
+      "step": 5993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.042340143658169,
+      "learning_rate": 8.411150120719388e-06,
+      "loss": 0.6439,
+      "step": 5994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8915068243584026,
+      "learning_rate": 8.410653595936678e-06,
+      "loss": 0.4672,
+      "step": 5995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9654393331280497,
+      "learning_rate": 8.410157010221245e-06,
+      "loss": 0.8452,
+      "step": 5996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6660360548544577,
+      "learning_rate": 8.409660363583479e-06,
+      "loss": 0.4815,
+      "step": 5997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8608387051547444,
+      "learning_rate": 8.409163656033785e-06,
+      "loss": 0.7974,
+      "step": 5998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5385003342692207,
+      "learning_rate": 8.408666887582554e-06,
+      "loss": 0.4298,
+      "step": 5999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9611869276361937,
+      "learning_rate": 8.408170058240188e-06,
+      "loss": 0.6271,
+      "step": 6000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.5639055371284485,
+      "eval_runtime": 67.9786,
+      "eval_samples_per_second": 50.163,
+      "eval_steps_per_second": 1.574,
+      "step": 6000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0175668346336653,
+      "learning_rate": 8.407673168017091e-06,
+      "loss": 0.5717,
+      "step": 6001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.684410240315562,
+      "learning_rate": 8.40717621692366e-06,
+      "loss": 0.519,
+      "step": 6002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6868758083036401,
+      "learning_rate": 8.406679204970302e-06,
+      "loss": 0.5324,
+      "step": 6003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6216133646525739,
+      "learning_rate": 8.406182132167418e-06,
+      "loss": 0.4523,
+      "step": 6004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9270799835106027,
+      "learning_rate": 8.40568499852542e-06,
+      "loss": 0.5112,
+      "step": 6005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.401341829737669,
+      "learning_rate": 8.405187804054707e-06,
+      "loss": 0.4458,
+      "step": 6006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.282239133185889,
+      "learning_rate": 8.404690548765694e-06,
+      "loss": 0.3542,
+      "step": 6007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9950843801729385,
+      "learning_rate": 8.404193232668788e-06,
+      "loss": 0.6049,
+      "step": 6008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.115239580839871,
+      "learning_rate": 8.403695855774402e-06,
+      "loss": 0.549,
+      "step": 6009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9502146128384064,
+      "learning_rate": 8.403198418092944e-06,
+      "loss": 0.6083,
+      "step": 6010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9746352638666376,
+      "learning_rate": 8.402700919634831e-06,
+      "loss": 0.7005,
+      "step": 6011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.669980648530308,
+      "learning_rate": 8.40220336041048e-06,
+      "loss": 0.5064,
+      "step": 6012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9647114389221123,
+      "learning_rate": 8.401705740430303e-06,
+      "loss": 0.4843,
+      "step": 6013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.862930679928457,
+      "learning_rate": 8.401208059704717e-06,
+      "loss": 0.5844,
+      "step": 6014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7247272470069839,
+      "learning_rate": 8.400710318244144e-06,
+      "loss": 0.5286,
+      "step": 6015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9213366826799336,
+      "learning_rate": 8.400212516059001e-06,
+      "loss": 0.6602,
+      "step": 6016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9756336053444983,
+      "learning_rate": 8.399714653159711e-06,
+      "loss": 0.6865,
+      "step": 6017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.610380274270486,
+      "learning_rate": 8.399216729556697e-06,
+      "loss": 0.5224,
+      "step": 6018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5206491703107616,
+      "learning_rate": 8.39871874526038e-06,
+      "loss": 0.3324,
+      "step": 6019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.556210468747234,
+      "learning_rate": 8.398220700281186e-06,
+      "loss": 0.7046,
+      "step": 6020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8683838652453801,
+      "learning_rate": 8.397722594629543e-06,
+      "loss": 0.5661,
+      "step": 6021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6593183905298114,
+      "learning_rate": 8.397224428315876e-06,
+      "loss": 0.5565,
+      "step": 6022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6549662556201015,
+      "learning_rate": 8.396726201350616e-06,
+      "loss": 0.5627,
+      "step": 6023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7965671417669558,
+      "learning_rate": 8.396227913744193e-06,
+      "loss": 0.492,
+      "step": 6024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7655279170513103,
+      "learning_rate": 8.395729565507037e-06,
+      "loss": 0.5196,
+      "step": 6025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.077179174639505,
+      "learning_rate": 8.39523115664958e-06,
+      "loss": 0.6385,
+      "step": 6026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8654490679319649,
+      "learning_rate": 8.394732687182259e-06,
+      "loss": 0.5732,
+      "step": 6027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7505548901686203,
+      "learning_rate": 8.394234157115506e-06,
+      "loss": 0.6754,
+      "step": 6028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7108410726079972,
+      "learning_rate": 8.393735566459758e-06,
+      "loss": 0.5039,
+      "step": 6029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.35878159873853,
+      "learning_rate": 8.393236915225454e-06,
+      "loss": 0.7244,
+      "step": 6030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7831030218599428,
+      "learning_rate": 8.392738203423032e-06,
+      "loss": 0.5245,
+      "step": 6031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.315016118186273,
+      "learning_rate": 8.392239431062935e-06,
+      "loss": 0.5005,
+      "step": 6032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0931989865721063,
+      "learning_rate": 8.391740598155598e-06,
+      "loss": 0.7574,
+      "step": 6033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7002243268854147,
+      "learning_rate": 8.39124170471147e-06,
+      "loss": 0.6797,
+      "step": 6034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7636563958384053,
+      "learning_rate": 8.39074275074099e-06,
+      "loss": 0.4612,
+      "step": 6035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8160725749817626,
+      "learning_rate": 8.39024373625461e-06,
+      "loss": 0.6344,
+      "step": 6036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7075330887097648,
+      "learning_rate": 8.389744661262771e-06,
+      "loss": 0.4995,
+      "step": 6037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7739604634489599,
+      "learning_rate": 8.389245525775923e-06,
+      "loss": 0.5845,
+      "step": 6038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5874295975758768,
+      "learning_rate": 8.388746329804513e-06,
+      "loss": 0.4691,
+      "step": 6039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6933879211921943,
+      "learning_rate": 8.388247073358993e-06,
+      "loss": 0.4316,
+      "step": 6040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6007842844637756,
+      "learning_rate": 8.387747756449814e-06,
+      "loss": 0.4253,
+      "step": 6041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.640476632998518,
+      "learning_rate": 8.38724837908743e-06,
+      "loss": 0.4915,
+      "step": 6042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.669024141271442,
+      "learning_rate": 8.386748941282295e-06,
+      "loss": 0.447,
+      "step": 6043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5453780853609138,
+      "learning_rate": 8.386249443044864e-06,
+      "loss": 0.3772,
+      "step": 6044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.730612505853898,
+      "learning_rate": 8.385749884385593e-06,
+      "loss": 0.5309,
+      "step": 6045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0542267299492614,
+      "learning_rate": 8.385250265314942e-06,
+      "loss": 0.5213,
+      "step": 6046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.958836849436473,
+      "learning_rate": 8.384750585843367e-06,
+      "loss": 0.6321,
+      "step": 6047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7816073045772542,
+      "learning_rate": 8.38425084598133e-06,
+      "loss": 0.5296,
+      "step": 6048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8461828044360844,
+      "learning_rate": 8.383751045739295e-06,
+      "loss": 0.462,
+      "step": 6049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7489540830195587,
+      "learning_rate": 8.383251185127722e-06,
+      "loss": 0.4548,
+      "step": 6050
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.5633514523506165,
+      "eval_runtime": 67.9797,
+      "eval_samples_per_second": 50.162,
+      "eval_steps_per_second": 1.574,
+      "step": 6050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7581853384803177,
+      "learning_rate": 8.382751264157077e-06,
+      "loss": 0.5541,
+      "step": 6051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2040673595521634,
+      "learning_rate": 8.382251282837825e-06,
+      "loss": 0.6888,
+      "step": 6052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.969045038832428,
+      "learning_rate": 8.38175124118043e-06,
+      "loss": 0.4781,
+      "step": 6053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.406844888490128,
+      "learning_rate": 8.381251139195365e-06,
+      "loss": 0.3945,
+      "step": 6054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8575095414064,
+      "learning_rate": 8.380750976893096e-06,
+      "loss": 0.5393,
+      "step": 6055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.065486451139845,
+      "learning_rate": 8.380250754284096e-06,
+      "loss": 0.6285,
+      "step": 6056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.597918518564558,
+      "learning_rate": 8.379750471378833e-06,
+      "loss": 0.509,
+      "step": 6057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2210493482280964,
+      "learning_rate": 8.379250128187785e-06,
+      "loss": 0.67,
+      "step": 6058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.951395811994754,
+      "learning_rate": 8.378749724721423e-06,
+      "loss": 0.5305,
+      "step": 6059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6327761638658647,
+      "learning_rate": 8.378249260990224e-06,
+      "loss": 0.6794,
+      "step": 6060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9323831439146137,
+      "learning_rate": 8.377748737004664e-06,
+      "loss": 0.614,
+      "step": 6061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7243889958861476,
+      "learning_rate": 8.377248152775223e-06,
+      "loss": 0.6491,
+      "step": 6062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4873417178717454,
+      "learning_rate": 8.376747508312375e-06,
+      "loss": 0.5046,
+      "step": 6063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.94073255597327,
+      "learning_rate": 8.376246803626609e-06,
+      "loss": 0.3828,
+      "step": 6064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6432371243409394,
+      "learning_rate": 8.3757460387284e-06,
+      "loss": 0.4358,
+      "step": 6065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8948312328670673,
+      "learning_rate": 8.375245213628235e-06,
+      "loss": 0.6531,
+      "step": 6066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1570670856863035,
+      "learning_rate": 8.374744328336598e-06,
+      "loss": 0.6744,
+      "step": 6067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7497533041332165,
+      "learning_rate": 8.374243382863974e-06,
+      "loss": 0.6124,
+      "step": 6068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5800683410630134,
+      "learning_rate": 8.373742377220848e-06,
+      "loss": 0.4634,
+      "step": 6069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.992478428523687,
+      "learning_rate": 8.373241311417713e-06,
+      "loss": 0.7278,
+      "step": 6070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4226604530901616,
+      "learning_rate": 8.372740185465053e-06,
+      "loss": 0.4182,
+      "step": 6071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3670566397276844,
+      "learning_rate": 8.372238999373363e-06,
+      "loss": 0.5298,
+      "step": 6072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7255605979308113,
+      "learning_rate": 8.371737753153134e-06,
+      "loss": 0.5455,
+      "step": 6073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9651317123877576,
+      "learning_rate": 8.371236446814859e-06,
+      "loss": 0.7309,
+      "step": 6074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9936020945296207,
+      "learning_rate": 8.370735080369032e-06,
+      "loss": 0.501,
+      "step": 6075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.795636195370908,
+      "learning_rate": 8.370233653826148e-06,
+      "loss": 0.4987,
+      "step": 6076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8201045367121977,
+      "learning_rate": 8.369732167196707e-06,
+      "loss": 0.5302,
+      "step": 6077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4115657428014274,
+      "learning_rate": 8.369230620491205e-06,
+      "loss": 0.5556,
+      "step": 6078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9882492677561732,
+      "learning_rate": 8.36872901372014e-06,
+      "loss": 0.5165,
+      "step": 6079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.76544385048538,
+      "learning_rate": 8.368227346894018e-06,
+      "loss": 0.4849,
+      "step": 6080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7864528674022617,
+      "learning_rate": 8.367725620023336e-06,
+      "loss": 0.4878,
+      "step": 6081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1786096453092174,
+      "learning_rate": 8.367223833118601e-06,
+      "loss": 0.5327,
+      "step": 6082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.336999042946636,
+      "learning_rate": 8.366721986190316e-06,
+      "loss": 0.6681,
+      "step": 6083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4707406025759284,
+      "learning_rate": 8.366220079248986e-06,
+      "loss": 0.6048,
+      "step": 6084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.945677200546401,
+      "learning_rate": 8.365718112305118e-06,
+      "loss": 0.6742,
+      "step": 6085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1317593454156394,
+      "learning_rate": 8.365216085369223e-06,
+      "loss": 0.6407,
+      "step": 6086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6756888821092415,
+      "learning_rate": 8.364713998451808e-06,
+      "loss": 0.5263,
+      "step": 6087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7281065891162364,
+      "learning_rate": 8.364211851563385e-06,
+      "loss": 0.4987,
+      "step": 6088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2346171627083695,
+      "learning_rate": 8.363709644714465e-06,
+      "loss": 0.5196,
+      "step": 6089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7001268178422768,
+      "learning_rate": 8.363207377915563e-06,
+      "loss": 0.4757,
+      "step": 6090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8123540255269612,
+      "learning_rate": 8.362705051177193e-06,
+      "loss": 0.4085,
+      "step": 6091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9602233093297883,
+      "learning_rate": 8.362202664509871e-06,
+      "loss": 0.6596,
+      "step": 6092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.500803782353092,
+      "learning_rate": 8.361700217924113e-06,
+      "loss": 0.4333,
+      "step": 6093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.829493506318687,
+      "learning_rate": 8.36119771143044e-06,
+      "loss": 0.6014,
+      "step": 6094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.915795564986628,
+      "learning_rate": 8.36069514503937e-06,
+      "loss": 0.3774,
+      "step": 6095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.17621060644049,
+      "learning_rate": 8.360192518761424e-06,
+      "loss": 0.6318,
+      "step": 6096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7647920526663925,
+      "learning_rate": 8.359689832607121e-06,
+      "loss": 0.4354,
+      "step": 6097
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3018802752179024,
+      "learning_rate": 8.35918708658699e-06,
+      "loss": 0.6196,
+      "step": 6098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4066503533763643,
+      "learning_rate": 8.358684280711553e-06,
+      "loss": 0.3486,
+      "step": 6099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1678874743134475,
+      "learning_rate": 8.358181414991338e-06,
+      "loss": 0.6925,
+      "step": 6100
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.5636129379272461,
+      "eval_runtime": 67.9675,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 6100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0166544828079274,
+      "learning_rate": 8.357678489436868e-06,
+      "loss": 0.6403,
+      "step": 6101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.619595517316035,
+      "learning_rate": 8.357175504058675e-06,
+      "loss": 0.4882,
+      "step": 6102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6948499050723673,
+      "learning_rate": 8.356672458867286e-06,
+      "loss": 0.5152,
+      "step": 6103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.697460892042658,
+      "learning_rate": 8.356169353873235e-06,
+      "loss": 0.4125,
+      "step": 6104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.742246485325786,
+      "learning_rate": 8.355666189087053e-06,
+      "loss": 0.5149,
+      "step": 6105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8294159289356633,
+      "learning_rate": 8.355162964519272e-06,
+      "loss": 0.5389,
+      "step": 6106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7916831851349222,
+      "learning_rate": 8.354659680180427e-06,
+      "loss": 0.5789,
+      "step": 6107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.944158165478318,
+      "learning_rate": 8.354156336081056e-06,
+      "loss": 0.5605,
+      "step": 6108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0595868598010436,
+      "learning_rate": 8.353652932231693e-06,
+      "loss": 0.4793,
+      "step": 6109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7283326440304838,
+      "learning_rate": 8.353149468642881e-06,
+      "loss": 0.4597,
+      "step": 6110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8784688849217135,
+      "learning_rate": 8.352645945325156e-06,
+      "loss": 0.7014,
+      "step": 6111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9539438709451113,
+      "learning_rate": 8.35214236228906e-06,
+      "loss": 0.6643,
+      "step": 6112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.740717944973984,
+      "learning_rate": 8.351638719545136e-06,
+      "loss": 0.5272,
+      "step": 6113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6713775818394951,
+      "learning_rate": 8.351135017103924e-06,
+      "loss": 0.4757,
+      "step": 6114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5140421742052645,
+      "learning_rate": 8.350631254975974e-06,
+      "loss": 0.4493,
+      "step": 6115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9795543683111148,
+      "learning_rate": 8.350127433171828e-06,
+      "loss": 0.5724,
+      "step": 6116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5457360146622519,
+      "learning_rate": 8.349623551702034e-06,
+      "loss": 0.4271,
+      "step": 6117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6651813409553387,
+      "learning_rate": 8.34911961057714e-06,
+      "loss": 0.398,
+      "step": 6118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.858773792046295,
+      "learning_rate": 8.3486156098077e-06,
+      "loss": 0.5117,
+      "step": 6119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7029414670834855,
+      "learning_rate": 8.34811154940426e-06,
+      "loss": 0.5855,
+      "step": 6120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.555793718483846,
+      "learning_rate": 8.347607429377373e-06,
+      "loss": 0.5405,
+      "step": 6121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.672433297232168,
+      "learning_rate": 8.347103249737593e-06,
+      "loss": 0.4296,
+      "step": 6122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9417341958769023,
+      "learning_rate": 8.346599010495476e-06,
+      "loss": 0.4588,
+      "step": 6123
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.269038084256631,
+      "learning_rate": 8.346094711661575e-06,
+      "loss": 0.7542,
+      "step": 6124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7378620597857046,
+      "learning_rate": 8.345590353246452e-06,
+      "loss": 0.4448,
+      "step": 6125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8496248863825826,
+      "learning_rate": 8.345085935260661e-06,
+      "loss": 0.6068,
+      "step": 6126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0231098744711145,
+      "learning_rate": 8.344581457714762e-06,
+      "loss": 0.6248,
+      "step": 6127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.207711341522056,
+      "learning_rate": 8.344076920619317e-06,
+      "loss": 0.6668,
+      "step": 6128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.072711526515017,
+      "learning_rate": 8.34357232398489e-06,
+      "loss": 0.6168,
+      "step": 6129
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7627380208880523,
+      "learning_rate": 8.34306766782204e-06,
+      "loss": 0.4487,
+      "step": 6130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9309285165001362,
+      "learning_rate": 8.342562952141338e-06,
+      "loss": 0.6651,
+      "step": 6131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0857107418022465,
+      "learning_rate": 8.342058176953342e-06,
+      "loss": 0.5295,
+      "step": 6132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0618870698841563,
+      "learning_rate": 8.341553342268627e-06,
+      "loss": 0.6141,
+      "step": 6133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8037210046364622,
+      "learning_rate": 8.341048448097756e-06,
+      "loss": 0.4709,
+      "step": 6134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5345555781597304,
+      "learning_rate": 8.3405434944513e-06,
+      "loss": 0.4083,
+      "step": 6135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6592818269649736,
+      "learning_rate": 8.340038481339832e-06,
+      "loss": 0.5885,
+      "step": 6136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.763387918114044,
+      "learning_rate": 8.339533408773922e-06,
+      "loss": 0.5865,
+      "step": 6137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8567805582209176,
+      "learning_rate": 8.339028276764142e-06,
+      "loss": 0.6482,
+      "step": 6138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.505090508077971,
+      "learning_rate": 8.33852308532107e-06,
+      "loss": 0.4901,
+      "step": 6139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7678861379061417,
+      "learning_rate": 8.338017834455279e-06,
+      "loss": 0.5549,
+      "step": 6140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2466717955689126,
+      "learning_rate": 8.337512524177348e-06,
+      "loss": 0.5153,
+      "step": 6141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6747546013808516,
+      "learning_rate": 8.337007154497855e-06,
+      "loss": 0.4134,
+      "step": 6142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7847135071473024,
+      "learning_rate": 8.33650172542738e-06,
+      "loss": 0.4603,
+      "step": 6143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8641731257090155,
+      "learning_rate": 8.3359962369765e-06,
+      "loss": 0.5927,
+      "step": 6144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7033651991042034,
+      "learning_rate": 8.335490689155804e-06,
+      "loss": 0.53,
+      "step": 6145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4931836453041964,
+      "learning_rate": 8.334985081975869e-06,
+      "loss": 0.4981,
+      "step": 6146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6164489922563412,
+      "learning_rate": 8.334479415447284e-06,
+      "loss": 0.3489,
+      "step": 6147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7085176246460423,
+      "learning_rate": 8.333973689580632e-06,
+      "loss": 0.5148,
+      "step": 6148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.056238823756897,
+      "learning_rate": 8.333467904386502e-06,
+      "loss": 0.44,
+      "step": 6149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0930442382589414,
+      "learning_rate": 8.33296205987548e-06,
+      "loss": 0.6773,
+      "step": 6150
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.5609571933746338,
+      "eval_runtime": 68.014,
+      "eval_samples_per_second": 50.137,
+      "eval_steps_per_second": 1.573,
+      "step": 6150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0662495174871163,
+      "learning_rate": 8.332456156058158e-06,
+      "loss": 0.6951,
+      "step": 6151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0629557909846623,
+      "learning_rate": 8.331950192945126e-06,
+      "loss": 0.5354,
+      "step": 6152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4604055405390484,
+      "learning_rate": 8.331444170546975e-06,
+      "loss": 0.5019,
+      "step": 6153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.956848317148135,
+      "learning_rate": 8.3309380888743e-06,
+      "loss": 0.6827,
+      "step": 6154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5593843012614523,
+      "learning_rate": 8.330431947937692e-06,
+      "loss": 0.505,
+      "step": 6155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3397143068649104,
+      "learning_rate": 8.329925747747751e-06,
+      "loss": 0.3982,
+      "step": 6156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6544474055759715,
+      "learning_rate": 8.329419488315074e-06,
+      "loss": 0.4338,
+      "step": 6157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8572999693741123,
+      "learning_rate": 8.328913169650255e-06,
+      "loss": 0.4963,
+      "step": 6158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7283599550529047,
+      "learning_rate": 8.328406791763896e-06,
+      "loss": 0.5038,
+      "step": 6159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.849352353438399,
+      "learning_rate": 8.327900354666599e-06,
+      "loss": 0.533,
+      "step": 6160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.680635422552298,
+      "learning_rate": 8.327393858368962e-06,
+      "loss": 0.4941,
+      "step": 6161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8212721855264693,
+      "learning_rate": 8.326887302881592e-06,
+      "loss": 0.5677,
+      "step": 6162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0645971879410667,
+      "learning_rate": 8.326380688215092e-06,
+      "loss": 0.5839,
+      "step": 6163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9257402826045023,
+      "learning_rate": 8.325874014380068e-06,
+      "loss": 0.5452,
+      "step": 6164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8615974400760518,
+      "learning_rate": 8.325367281387126e-06,
+      "loss": 0.5811,
+      "step": 6165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.212354431082115,
+      "learning_rate": 8.324860489246877e-06,
+      "loss": 0.6756,
+      "step": 6166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.049056587443416,
+      "learning_rate": 8.324353637969926e-06,
+      "loss": 0.5152,
+      "step": 6167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6538779693312686,
+      "learning_rate": 8.323846727566887e-06,
+      "loss": 0.4247,
+      "step": 6168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7909933879627882,
+      "learning_rate": 8.32333975804837e-06,
+      "loss": 0.6278,
+      "step": 6169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8287831096017695,
+      "learning_rate": 8.32283272942499e-06,
+      "loss": 0.5847,
+      "step": 6170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6365365386827893,
+      "learning_rate": 8.32232564170736e-06,
+      "loss": 0.5364,
+      "step": 6171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4730320428799593,
+      "learning_rate": 8.321818494906096e-06,
+      "loss": 0.4472,
+      "step": 6172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9554475321418041,
+      "learning_rate": 8.321311289031812e-06,
+      "loss": 0.5003,
+      "step": 6173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5013429378696201,
+      "learning_rate": 8.32080402409513e-06,
+      "loss": 0.4663,
+      "step": 6174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7522857469510267,
+      "learning_rate": 8.32029670010667e-06,
+      "loss": 0.6008,
+      "step": 6175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0146384476061674,
+      "learning_rate": 8.319789317077049e-06,
+      "loss": 0.6965,
+      "step": 6176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9648050787673674,
+      "learning_rate": 8.319281875016888e-06,
+      "loss": 0.5618,
+      "step": 6177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8873316601261685,
+      "learning_rate": 8.318774373936812e-06,
+      "loss": 0.5357,
+      "step": 6178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9277780369205457,
+      "learning_rate": 8.318266813847448e-06,
+      "loss": 0.3777,
+      "step": 6179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5434584398133648,
+      "learning_rate": 8.317759194759419e-06,
+      "loss": 0.6092,
+      "step": 6180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7405192621053793,
+      "learning_rate": 8.317251516683349e-06,
+      "loss": 0.4325,
+      "step": 6181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.39988926682804,
+      "learning_rate": 8.31674377962987e-06,
+      "loss": 0.704,
+      "step": 6182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7516485252043719,
+      "learning_rate": 8.316235983609608e-06,
+      "loss": 0.5434,
+      "step": 6183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7362512591676111,
+      "learning_rate": 8.315728128633195e-06,
+      "loss": 0.5573,
+      "step": 6184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8990899490547297,
+      "learning_rate": 8.315220214711264e-06,
+      "loss": 0.6389,
+      "step": 6185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.153019115750991,
+      "learning_rate": 8.314712241854447e-06,
+      "loss": 0.5777,
+      "step": 6186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6987462109614728,
+      "learning_rate": 8.314204210073375e-06,
+      "loss": 0.4914,
+      "step": 6187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.752306088997274,
+      "learning_rate": 8.313696119378688e-06,
+      "loss": 0.5058,
+      "step": 6188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7912264070039499,
+      "learning_rate": 8.31318796978102e-06,
+      "loss": 0.5081,
+      "step": 6189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1967044724613345,
+      "learning_rate": 8.312679761291009e-06,
+      "loss": 0.7375,
+      "step": 6190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4356287136659118,
+      "learning_rate": 8.312171493919293e-06,
+      "loss": 0.3191,
+      "step": 6191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.744278236901389,
+      "learning_rate": 8.311663167676515e-06,
+      "loss": 0.5049,
+      "step": 6192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5581878771593853,
+      "learning_rate": 8.311154782573313e-06,
+      "loss": 0.3606,
+      "step": 6193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.655819916271842,
+      "learning_rate": 8.310646338620332e-06,
+      "loss": 0.6096,
+      "step": 6194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6807286506841055,
+      "learning_rate": 8.310137835828217e-06,
+      "loss": 0.5351,
+      "step": 6195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7636002263652375,
+      "learning_rate": 8.309629274207613e-06,
+      "loss": 0.6779,
+      "step": 6196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5317570416617292,
+      "learning_rate": 8.309120653769164e-06,
+      "loss": 0.4743,
+      "step": 6197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.056727787124728,
+      "learning_rate": 8.308611974523519e-06,
+      "loss": 0.6178,
+      "step": 6198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1343408181976575,
+      "learning_rate": 8.308103236481325e-06,
+      "loss": 0.5519,
+      "step": 6199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.203204446996683,
+      "learning_rate": 8.307594439653236e-06,
+      "loss": 0.7619,
+      "step": 6200
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5598859786987305,
+      "eval_runtime": 67.9544,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 6200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.735534673924137,
+      "learning_rate": 8.307085584049903e-06,
+      "loss": 0.442,
+      "step": 6201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8794553301275,
+      "learning_rate": 8.306576669681974e-06,
+      "loss": 0.5612,
+      "step": 6202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7460445021582696,
+      "learning_rate": 8.306067696560105e-06,
+      "loss": 0.529,
+      "step": 6203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6638411075605406,
+      "learning_rate": 8.305558664694952e-06,
+      "loss": 0.5261,
+      "step": 6204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5524283815458852,
+      "learning_rate": 8.305049574097173e-06,
+      "loss": 0.4359,
+      "step": 6205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.948591471884387,
+      "learning_rate": 8.304540424777422e-06,
+      "loss": 0.6193,
+      "step": 6206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.956211014445417,
+      "learning_rate": 8.304031216746358e-06,
+      "loss": 0.498,
+      "step": 6207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0543334749098356,
+      "learning_rate": 8.303521950014644e-06,
+      "loss": 0.516,
+      "step": 6208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7595354938502492,
+      "learning_rate": 8.30301262459294e-06,
+      "loss": 0.5472,
+      "step": 6209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8486450285785299,
+      "learning_rate": 8.302503240491906e-06,
+      "loss": 0.4856,
+      "step": 6210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.49805449322497,
+      "learning_rate": 8.301993797722208e-06,
+      "loss": 0.3991,
+      "step": 6211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7442880461214814,
+      "learning_rate": 8.301484296294512e-06,
+      "loss": 0.4659,
+      "step": 6212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8719446348034623,
+      "learning_rate": 8.300974736219479e-06,
+      "loss": 0.6317,
+      "step": 6213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7161067115206885,
+      "learning_rate": 8.300465117507782e-06,
+      "loss": 0.5457,
+      "step": 6214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.053056624191781,
+      "learning_rate": 8.299955440170085e-06,
+      "loss": 0.5442,
+      "step": 6215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.755838081328985,
+      "learning_rate": 8.299445704217062e-06,
+      "loss": 0.5124,
+      "step": 6216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9256906380995507,
+      "learning_rate": 8.298935909659382e-06,
+      "loss": 0.5801,
+      "step": 6217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1597006850095197,
+      "learning_rate": 8.298426056507715e-06,
+      "loss": 0.6834,
+      "step": 6218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.773360886488971,
+      "learning_rate": 8.297916144772738e-06,
+      "loss": 0.5568,
+      "step": 6219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7794012954937861,
+      "learning_rate": 8.297406174465125e-06,
+      "loss": 0.5048,
+      "step": 6220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9379521736290937,
+      "learning_rate": 8.296896145595552e-06,
+      "loss": 0.5176,
+      "step": 6221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7750452400051604,
+      "learning_rate": 8.296386058174693e-06,
+      "loss": 0.6637,
+      "step": 6222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8582823186517412,
+      "learning_rate": 8.29587591221323e-06,
+      "loss": 0.5916,
+      "step": 6223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.090561936884097,
+      "learning_rate": 8.295365707721841e-06,
+      "loss": 0.6105,
+      "step": 6224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7475869481614599,
+      "learning_rate": 8.29485544471121e-06,
+      "loss": 0.436,
+      "step": 6225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1324930937302287,
+      "learning_rate": 8.294345123192012e-06,
+      "loss": 0.594,
+      "step": 6226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6738915918551214,
+      "learning_rate": 8.293834743174937e-06,
+      "loss": 0.5007,
+      "step": 6227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7280047967535317,
+      "learning_rate": 8.293324304670666e-06,
+      "loss": 0.56,
+      "step": 6228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.998476926826881,
+      "learning_rate": 8.292813807689887e-06,
+      "loss": 0.5683,
+      "step": 6229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8605242217072837,
+      "learning_rate": 8.292303252243283e-06,
+      "loss": 0.5594,
+      "step": 6230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.556858010821047,
+      "learning_rate": 8.291792638341546e-06,
+      "loss": 0.4536,
+      "step": 6231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9630239630873767,
+      "learning_rate": 8.291281965995366e-06,
+      "loss": 0.6385,
+      "step": 6232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7931096269591804,
+      "learning_rate": 8.29077123521543e-06,
+      "loss": 0.626,
+      "step": 6233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3261718474290143,
+      "learning_rate": 8.290260446012432e-06,
+      "loss": 0.4435,
+      "step": 6234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8377872451127215,
+      "learning_rate": 8.289749598397065e-06,
+      "loss": 0.6855,
+      "step": 6235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1641830023609607,
+      "learning_rate": 8.28923869238002e-06,
+      "loss": 0.6738,
+      "step": 6236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.686914460033865,
+      "learning_rate": 8.288727727971999e-06,
+      "loss": 0.4406,
+      "step": 6237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8266817719454345,
+      "learning_rate": 8.288216705183696e-06,
+      "loss": 0.5648,
+      "step": 6238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7855841677514293,
+      "learning_rate": 8.287705624025805e-06,
+      "loss": 0.4332,
+      "step": 6239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.898696644750432,
+      "learning_rate": 8.287194484509029e-06,
+      "loss": 0.479,
+      "step": 6240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7934413596521495,
+      "learning_rate": 8.286683286644069e-06,
+      "loss": 0.5063,
+      "step": 6241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7002777050623203,
+      "learning_rate": 8.286172030441623e-06,
+      "loss": 0.4848,
+      "step": 6242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.814684631201477,
+      "learning_rate": 8.285660715912398e-06,
+      "loss": 0.4029,
+      "step": 6243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7750595249378778,
+      "learning_rate": 8.285149343067096e-06,
+      "loss": 0.5443,
+      "step": 6244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.901714947565702,
+      "learning_rate": 8.284637911916423e-06,
+      "loss": 0.5402,
+      "step": 6245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1522932086936906,
+      "learning_rate": 8.284126422471081e-06,
+      "loss": 0.6611,
+      "step": 6246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8397342369882521,
+      "learning_rate": 8.283614874741785e-06,
+      "loss": 0.5976,
+      "step": 6247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7765014949809783,
+      "learning_rate": 8.283103268739241e-06,
+      "loss": 0.5024,
+      "step": 6248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1633138710138287,
+      "learning_rate": 8.282591604474158e-06,
+      "loss": 0.5149,
+      "step": 6249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7196156379502812,
+      "learning_rate": 8.282079881957248e-06,
+      "loss": 0.5388,
+      "step": 6250
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5607283115386963,
+      "eval_runtime": 67.9676,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 6250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8754793744459033,
+      "learning_rate": 8.281568101199224e-06,
+      "loss": 0.5597,
+      "step": 6251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.690063398313421,
+      "learning_rate": 8.281056262210798e-06,
+      "loss": 0.571,
+      "step": 6252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2333868444200504,
+      "learning_rate": 8.280544365002689e-06,
+      "loss": 0.3321,
+      "step": 6253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9835430248716899,
+      "learning_rate": 8.280032409585609e-06,
+      "loss": 0.5467,
+      "step": 6254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6314231324542832,
+      "learning_rate": 8.279520395970277e-06,
+      "loss": 0.425,
+      "step": 6255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9187474086655065,
+      "learning_rate": 8.279008324167412e-06,
+      "loss": 0.6806,
+      "step": 6256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7637218632041052,
+      "learning_rate": 8.278496194187734e-06,
+      "loss": 0.5693,
+      "step": 6257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.059336839341318,
+      "learning_rate": 8.277984006041966e-06,
+      "loss": 0.6617,
+      "step": 6258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5283696390722565,
+      "learning_rate": 8.277471759740826e-06,
+      "loss": 0.5454,
+      "step": 6259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3476263949913176,
+      "learning_rate": 8.27695945529504e-06,
+      "loss": 0.471,
+      "step": 6260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3244838127967036,
+      "learning_rate": 8.276447092715335e-06,
+      "loss": 0.7556,
+      "step": 6261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8344485791141796,
+      "learning_rate": 8.275934672012433e-06,
+      "loss": 0.5467,
+      "step": 6262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1462287810273573,
+      "learning_rate": 8.275422193197062e-06,
+      "loss": 0.4646,
+      "step": 6263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9597798972559666,
+      "learning_rate": 8.27490965627995e-06,
+      "loss": 0.7171,
+      "step": 6264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7183033746636884,
+      "learning_rate": 8.27439706127183e-06,
+      "loss": 0.4565,
+      "step": 6265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.559916847115615,
+      "learning_rate": 8.27388440818343e-06,
+      "loss": 0.7051,
+      "step": 6266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7183247962466328,
+      "learning_rate": 8.273371697025484e-06,
+      "loss": 0.5863,
+      "step": 6267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7481010947071727,
+      "learning_rate": 8.272858927808725e-06,
+      "loss": 0.5233,
+      "step": 6268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9761890202009869,
+      "learning_rate": 8.272346100543886e-06,
+      "loss": 0.5427,
+      "step": 6269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6268495481645728,
+      "learning_rate": 8.271833215241702e-06,
+      "loss": 0.5964,
+      "step": 6270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9662800365061865,
+      "learning_rate": 8.27132027191291e-06,
+      "loss": 0.4462,
+      "step": 6271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6436752340316703,
+      "learning_rate": 8.270807270568252e-06,
+      "loss": 0.4879,
+      "step": 6272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1168391178642003,
+      "learning_rate": 8.270294211218465e-06,
+      "loss": 0.6213,
+      "step": 6273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6501241589474005,
+      "learning_rate": 8.269781093874288e-06,
+      "loss": 0.4169,
+      "step": 6274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8967027929730247,
+      "learning_rate": 8.269267918546464e-06,
+      "loss": 0.6022,
+      "step": 6275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8475536113576978,
+      "learning_rate": 8.268754685245736e-06,
+      "loss": 0.5331,
+      "step": 6276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4181546479084386,
+      "learning_rate": 8.268241393982851e-06,
+      "loss": 0.4363,
+      "step": 6277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9044978390761373,
+      "learning_rate": 8.26772804476855e-06,
+      "loss": 0.4858,
+      "step": 6278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9675848702354894,
+      "learning_rate": 8.267214637613581e-06,
+      "loss": 0.5937,
+      "step": 6279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8730959311385484,
+      "learning_rate": 8.266701172528695e-06,
+      "loss": 0.4526,
+      "step": 6280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7931929449417963,
+      "learning_rate": 8.266187649524635e-06,
+      "loss": 0.5454,
+      "step": 6281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1539159911640047,
+      "learning_rate": 8.265674068612156e-06,
+      "loss": 0.6249,
+      "step": 6282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7884717432261972,
+      "learning_rate": 8.265160429802008e-06,
+      "loss": 0.5251,
+      "step": 6283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5501578609964104,
+      "learning_rate": 8.264646733104944e-06,
+      "loss": 0.4676,
+      "step": 6284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.370697084558001,
+      "learning_rate": 8.264132978531718e-06,
+      "loss": 0.4225,
+      "step": 6285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.688065588696221,
+      "learning_rate": 8.263619166093085e-06,
+      "loss": 0.5023,
+      "step": 6286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1694040778168455,
+      "learning_rate": 8.263105295799802e-06,
+      "loss": 0.6953,
+      "step": 6287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0126457369014843,
+      "learning_rate": 8.262591367662625e-06,
+      "loss": 0.5919,
+      "step": 6288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9757355265454748,
+      "learning_rate": 8.262077381692313e-06,
+      "loss": 0.5515,
+      "step": 6289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.964430449656546,
+      "learning_rate": 8.261563337899628e-06,
+      "loss": 0.5182,
+      "step": 6290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.735835153273621,
+      "learning_rate": 8.261049236295328e-06,
+      "loss": 0.5152,
+      "step": 6291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2570056779289054,
+      "learning_rate": 8.260535076890179e-06,
+      "loss": 0.6214,
+      "step": 6292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1964980844473976,
+      "learning_rate": 8.260020859694941e-06,
+      "loss": 0.7658,
+      "step": 6293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5297081864098914,
+      "learning_rate": 8.259506584720383e-06,
+      "loss": 0.543,
+      "step": 6294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4260834243389904,
+      "learning_rate": 8.258992251977267e-06,
+      "loss": 0.6095,
+      "step": 6295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.065472320892206,
+      "learning_rate": 8.258477861476361e-06,
+      "loss": 0.6403,
+      "step": 6296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8813681402444042,
+      "learning_rate": 8.257963413228435e-06,
+      "loss": 0.5118,
+      "step": 6297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8975455820657727,
+      "learning_rate": 8.25744890724426e-06,
+      "loss": 0.5703,
+      "step": 6298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.660109419779786,
+      "learning_rate": 8.256934343534603e-06,
+      "loss": 0.5791,
+      "step": 6299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.828280370625561,
+      "learning_rate": 8.256419722110238e-06,
+      "loss": 0.58,
+      "step": 6300
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.560138463973999,
+      "eval_runtime": 67.991,
+      "eval_samples_per_second": 50.154,
+      "eval_steps_per_second": 1.574,
+      "step": 6300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9574764284370323,
+      "learning_rate": 8.255905042981938e-06,
+      "loss": 0.7002,
+      "step": 6301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6471207468876767,
+      "learning_rate": 8.255390306160479e-06,
+      "loss": 0.4439,
+      "step": 6302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5255693391581884,
+      "learning_rate": 8.254875511656634e-06,
+      "loss": 0.4931,
+      "step": 6303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8796569004265062,
+      "learning_rate": 8.254360659481181e-06,
+      "loss": 0.5693,
+      "step": 6304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8507186743582589,
+      "learning_rate": 8.253845749644899e-06,
+      "loss": 0.5775,
+      "step": 6305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7747468927290577,
+      "learning_rate": 8.253330782158568e-06,
+      "loss": 0.5199,
+      "step": 6306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.781627778040582,
+      "learning_rate": 8.252815757032967e-06,
+      "loss": 0.5476,
+      "step": 6307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7321938797344751,
+      "learning_rate": 8.252300674278877e-06,
+      "loss": 0.5601,
+      "step": 6308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1193087023457804,
+      "learning_rate": 8.251785533907081e-06,
+      "loss": 0.4996,
+      "step": 6309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4156187017378337,
+      "learning_rate": 8.251270335928366e-06,
+      "loss": 0.435,
+      "step": 6310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8147652537764463,
+      "learning_rate": 8.250755080353514e-06,
+      "loss": 0.5855,
+      "step": 6311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7607335446800898,
+      "learning_rate": 8.250239767193313e-06,
+      "loss": 0.5724,
+      "step": 6312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.626498296432988,
+      "learning_rate": 8.249724396458552e-06,
+      "loss": 0.5314,
+      "step": 6313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7844439561918077,
+      "learning_rate": 8.249208968160018e-06,
+      "loss": 0.6035,
+      "step": 6314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4530585121171105,
+      "learning_rate": 8.248693482308499e-06,
+      "loss": 0.5034,
+      "step": 6315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8770318130966979,
+      "learning_rate": 8.248177938914792e-06,
+      "loss": 0.6756,
+      "step": 6316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.883644556299066,
+      "learning_rate": 8.247662337989687e-06,
+      "loss": 0.5672,
+      "step": 6317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.991170794854702,
+      "learning_rate": 8.247146679543978e-06,
+      "loss": 0.6706,
+      "step": 6318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.642350059180139,
+      "learning_rate": 8.246630963588459e-06,
+      "loss": 0.4416,
+      "step": 6319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7923014962331094,
+      "learning_rate": 8.246115190133926e-06,
+      "loss": 0.5687,
+      "step": 6320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8765759096746821,
+      "learning_rate": 8.245599359191177e-06,
+      "loss": 0.5726,
+      "step": 6321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6681262397587004,
+      "learning_rate": 8.245083470771014e-06,
+      "loss": 0.4313,
+      "step": 6322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8804709180994124,
+      "learning_rate": 8.244567524884229e-06,
+      "loss": 0.5078,
+      "step": 6323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7780810962964657,
+      "learning_rate": 8.24405152154163e-06,
+      "loss": 0.55,
+      "step": 6324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.895000641367755,
+      "learning_rate": 8.243535460754018e-06,
+      "loss": 0.7144,
+      "step": 6325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6179877436289525,
+      "learning_rate": 8.243019342532193e-06,
+      "loss": 0.3914,
+      "step": 6326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7199212552262535,
+      "learning_rate": 8.242503166886962e-06,
+      "loss": 0.5553,
+      "step": 6327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6435660531666914,
+      "learning_rate": 8.241986933829132e-06,
+      "loss": 0.5037,
+      "step": 6328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4024314085232026,
+      "learning_rate": 8.241470643369506e-06,
+      "loss": 0.3989,
+      "step": 6329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.729630735880833,
+      "learning_rate": 8.240954295518898e-06,
+      "loss": 0.4777,
+      "step": 6330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.858070025222534,
+      "learning_rate": 8.240437890288114e-06,
+      "loss": 0.4868,
+      "step": 6331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8132775755526744,
+      "learning_rate": 8.239921427687963e-06,
+      "loss": 0.5437,
+      "step": 6332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6159600323876404,
+      "learning_rate": 8.239404907729258e-06,
+      "loss": 0.4991,
+      "step": 6333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0986001142797632,
+      "learning_rate": 8.238888330422815e-06,
+      "loss": 0.6207,
+      "step": 6334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8702077330277853,
+      "learning_rate": 8.238371695779446e-06,
+      "loss": 0.5385,
+      "step": 6335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.182167588541684,
+      "learning_rate": 8.237855003809966e-06,
+      "loss": 0.5463,
+      "step": 6336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.885079567921667,
+      "learning_rate": 8.237338254525192e-06,
+      "loss": 0.5634,
+      "step": 6337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6849940929477882,
+      "learning_rate": 8.236821447935942e-06,
+      "loss": 0.4096,
+      "step": 6338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8262762673612918,
+      "learning_rate": 8.236304584053034e-06,
+      "loss": 0.4963,
+      "step": 6339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.841240868999599,
+      "learning_rate": 8.235787662887288e-06,
+      "loss": 0.5592,
+      "step": 6340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6405870048876527,
+      "learning_rate": 8.235270684449529e-06,
+      "loss": 0.4571,
+      "step": 6341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0396989081220447,
+      "learning_rate": 8.234753648750577e-06,
+      "loss": 0.5783,
+      "step": 6342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1201880972835876,
+      "learning_rate": 8.234236555801255e-06,
+      "loss": 0.6893,
+      "step": 6343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8324896367797363,
+      "learning_rate": 8.23371940561239e-06,
+      "loss": 0.5086,
+      "step": 6344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7016640365345361,
+      "learning_rate": 8.233202198194806e-06,
+      "loss": 0.5951,
+      "step": 6345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.03189859528505,
+      "learning_rate": 8.232684933559333e-06,
+      "loss": 0.5159,
+      "step": 6346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6397711537553277,
+      "learning_rate": 8.232167611716797e-06,
+      "loss": 0.3091,
+      "step": 6347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.026176796610761,
+      "learning_rate": 8.231650232678029e-06,
+      "loss": 0.549,
+      "step": 6348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.598307516070779,
+      "learning_rate": 8.23113279645386e-06,
+      "loss": 0.3491,
+      "step": 6349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9965809901278642,
+      "learning_rate": 8.230615303055122e-06,
+      "loss": 0.5075,
+      "step": 6350
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5609801411628723,
+      "eval_runtime": 67.8926,
+      "eval_samples_per_second": 50.226,
+      "eval_steps_per_second": 1.576,
+      "step": 6350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4749821799661051,
+      "learning_rate": 8.23009775249265e-06,
+      "loss": 0.3228,
+      "step": 6351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9936451058725384,
+      "learning_rate": 8.229580144777279e-06,
+      "loss": 0.5359,
+      "step": 6352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7938781769267678,
+      "learning_rate": 8.22906247991984e-06,
+      "loss": 0.5166,
+      "step": 6353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7273215446192984,
+      "learning_rate": 8.228544757931173e-06,
+      "loss": 0.5326,
+      "step": 6354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4360480901019725,
+      "learning_rate": 8.228026978822118e-06,
+      "loss": 0.5233,
+      "step": 6355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0858185181866244,
+      "learning_rate": 8.227509142603511e-06,
+      "loss": 0.5773,
+      "step": 6356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.939540245710463,
+      "learning_rate": 8.226991249286197e-06,
+      "loss": 0.5709,
+      "step": 6357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7059710016539573,
+      "learning_rate": 8.226473298881011e-06,
+      "loss": 0.5432,
+      "step": 6358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2815639089397135,
+      "learning_rate": 8.225955291398803e-06,
+      "loss": 0.4888,
+      "step": 6359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7953170332106265,
+      "learning_rate": 8.225437226850413e-06,
+      "loss": 0.5243,
+      "step": 6360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6223757514573929,
+      "learning_rate": 8.224919105246688e-06,
+      "loss": 0.6149,
+      "step": 6361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8923444410650432,
+      "learning_rate": 8.224400926598473e-06,
+      "loss": 0.5068,
+      "step": 6362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8658957383173185,
+      "learning_rate": 8.223882690916618e-06,
+      "loss": 0.6082,
+      "step": 6363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9323970254984753,
+      "learning_rate": 8.223364398211969e-06,
+      "loss": 0.432,
+      "step": 6364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.829781430507663,
+      "learning_rate": 8.22284604849538e-06,
+      "loss": 0.4794,
+      "step": 6365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4125756305098103,
+      "learning_rate": 8.2223276417777e-06,
+      "loss": 0.394,
+      "step": 6366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.04477178524046,
+      "learning_rate": 8.221809178069781e-06,
+      "loss": 0.4739,
+      "step": 6367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6822517073318237,
+      "learning_rate": 8.221290657382477e-06,
+      "loss": 0.4841,
+      "step": 6368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.705419595295815,
+      "learning_rate": 8.220772079726645e-06,
+      "loss": 0.5137,
+      "step": 6369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7204062942365514,
+      "learning_rate": 8.220253445113137e-06,
+      "loss": 0.5731,
+      "step": 6370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5241944533980376,
+      "learning_rate": 8.219734753552814e-06,
+      "loss": 0.4587,
+      "step": 6371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7794346849296478,
+      "learning_rate": 8.219216005056532e-06,
+      "loss": 0.4493,
+      "step": 6372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7827552051668958,
+      "learning_rate": 8.218697199635154e-06,
+      "loss": 0.4436,
+      "step": 6373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0160667968002977,
+      "learning_rate": 8.218178337299538e-06,
+      "loss": 0.5142,
+      "step": 6374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.758145143457614,
+      "learning_rate": 8.217659418060547e-06,
+      "loss": 0.5157,
+      "step": 6375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5831092520552053,
+      "learning_rate": 8.217140441929043e-06,
+      "loss": 0.3849,
+      "step": 6376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1553593531877517,
+      "learning_rate": 8.216621408915892e-06,
+      "loss": 0.4915,
+      "step": 6377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4983052934498804,
+      "learning_rate": 8.21610231903196e-06,
+      "loss": 0.3271,
+      "step": 6378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8283694388173466,
+      "learning_rate": 8.215583172288112e-06,
+      "loss": 0.6294,
+      "step": 6379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7185257012813846,
+      "learning_rate": 8.215063968695218e-06,
+      "loss": 0.5694,
+      "step": 6380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9143883833895354,
+      "learning_rate": 8.214544708264145e-06,
+      "loss": 0.5544,
+      "step": 6381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.018597798304275,
+      "learning_rate": 8.214025391005765e-06,
+      "loss": 0.5096,
+      "step": 6382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.599566130452216,
+      "learning_rate": 8.21350601693095e-06,
+      "loss": 0.5544,
+      "step": 6383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.206083278421766,
+      "learning_rate": 8.212986586050572e-06,
+      "loss": 0.7151,
+      "step": 6384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.730819781438441,
+      "learning_rate": 8.212467098375504e-06,
+      "loss": 0.5105,
+      "step": 6385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7526673053538837,
+      "learning_rate": 8.211947553916622e-06,
+      "loss": 0.5617,
+      "step": 6386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6741989690962447,
+      "learning_rate": 8.211427952684803e-06,
+      "loss": 0.5506,
+      "step": 6387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8015443660150945,
+      "learning_rate": 8.210908294690924e-06,
+      "loss": 0.6339,
+      "step": 6388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.846275024491825,
+      "learning_rate": 8.210388579945863e-06,
+      "loss": 0.4636,
+      "step": 6389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6106818338579274,
+      "learning_rate": 8.2098688084605e-06,
+      "loss": 0.3803,
+      "step": 6390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2292474020078306,
+      "learning_rate": 8.209348980245718e-06,
+      "loss": 0.4829,
+      "step": 6391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9925237367272732,
+      "learning_rate": 8.208829095312399e-06,
+      "loss": 0.4651,
+      "step": 6392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6869865013423804,
+      "learning_rate": 8.208309153671421e-06,
+      "loss": 0.5293,
+      "step": 6393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5575223175236117,
+      "learning_rate": 8.207789155333677e-06,
+      "loss": 0.5079,
+      "step": 6394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6766943917650228,
+      "learning_rate": 8.207269100310048e-06,
+      "loss": 0.6027,
+      "step": 6395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3449176462924353,
+      "learning_rate": 8.206748988611423e-06,
+      "loss": 0.7847,
+      "step": 6396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9220669887939292,
+      "learning_rate": 8.206228820248687e-06,
+      "loss": 0.4984,
+      "step": 6397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6660731042952037,
+      "learning_rate": 8.205708595232732e-06,
+      "loss": 0.3538,
+      "step": 6398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8190577579441751,
+      "learning_rate": 8.205188313574451e-06,
+      "loss": 0.5803,
+      "step": 6399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0197897757890875,
+      "learning_rate": 8.204667975284731e-06,
+      "loss": 0.5464,
+      "step": 6400
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5590887069702148,
+      "eval_runtime": 68.0,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 6400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8253974107817856,
+      "learning_rate": 8.204147580374465e-06,
+      "loss": 0.4433,
+      "step": 6401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.123425088324903,
+      "learning_rate": 8.20362712885455e-06,
+      "loss": 0.5949,
+      "step": 6402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7595283846625267,
+      "learning_rate": 8.20310662073588e-06,
+      "loss": 0.5622,
+      "step": 6403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7357206743150893,
+      "learning_rate": 8.202586056029354e-06,
+      "loss": 0.5509,
+      "step": 6404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9633704307929534,
+      "learning_rate": 8.202065434745865e-06,
+      "loss": 0.6841,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8882696110908135,
+      "learning_rate": 8.201544756896317e-06,
+      "loss": 0.7103,
+      "step": 6406
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.008982263818227,
+      "learning_rate": 8.201024022491605e-06,
+      "loss": 0.5328,
+      "step": 6407
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5018446782270867,
+      "learning_rate": 8.200503231542636e-06,
+      "loss": 0.4552,
+      "step": 6408
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.82171164067568,
+      "learning_rate": 8.199982384060307e-06,
+      "loss": 0.6381,
+      "step": 6409
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.803409125109493,
+      "learning_rate": 8.199461480055523e-06,
+      "loss": 0.525,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.107811512376987,
+      "learning_rate": 8.19894051953919e-06,
+      "loss": 0.6068,
+      "step": 6411
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.025876407654156,
+      "learning_rate": 8.198419502522214e-06,
+      "loss": 0.687,
+      "step": 6412
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0428352254719777,
+      "learning_rate": 8.197898429015502e-06,
+      "loss": 0.4663,
+      "step": 6413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9345049603266455,
+      "learning_rate": 8.197377299029963e-06,
+      "loss": 0.5246,
+      "step": 6414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.619221192338941,
+      "learning_rate": 8.196856112576504e-06,
+      "loss": 0.4683,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9713357220467378,
+      "learning_rate": 8.196334869666036e-06,
+      "loss": 0.4256,
+      "step": 6416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.640162447313028,
+      "learning_rate": 8.195813570309475e-06,
+      "loss": 0.5966,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5951172861483647,
+      "learning_rate": 8.195292214517732e-06,
+      "loss": 0.4682,
+      "step": 6418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4231831002697626,
+      "learning_rate": 8.19477080230172e-06,
+      "loss": 0.5835,
+      "step": 6419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4419423957843036,
+      "learning_rate": 8.194249333672352e-06,
+      "loss": 0.2976,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9082889614384728,
+      "learning_rate": 8.193727808640552e-06,
+      "loss": 0.6692,
+      "step": 6421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8059539944883711,
+      "learning_rate": 8.193206227217231e-06,
+      "loss": 0.5217,
+      "step": 6422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1686028540332085,
+      "learning_rate": 8.19268458941331e-06,
+      "loss": 0.7651,
+      "step": 6423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8817643079596742,
+      "learning_rate": 8.192162895239712e-06,
+      "loss": 0.5672,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.799592730885747,
+      "learning_rate": 8.191641144707353e-06,
+      "loss": 0.5541,
+      "step": 6425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7245486575951312,
+      "learning_rate": 8.19111933782716e-06,
+      "loss": 0.5637,
+      "step": 6426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8221782904681438,
+      "learning_rate": 8.190597474610054e-06,
+      "loss": 0.5799,
+      "step": 6427
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9353037345304,
+      "learning_rate": 8.190075555066963e-06,
+      "loss": 0.5937,
+      "step": 6428
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8991684326353102,
+      "learning_rate": 8.18955357920881e-06,
+      "loss": 0.5614,
+      "step": 6429
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.363477188112862,
+      "learning_rate": 8.189031547046524e-06,
+      "loss": 0.6121,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.684457247775329,
+      "learning_rate": 8.188509458591031e-06,
+      "loss": 0.5857,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9346217723272292,
+      "learning_rate": 8.187987313853266e-06,
+      "loss": 0.5272,
+      "step": 6432
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8806013537577766,
+      "learning_rate": 8.187465112844153e-06,
+      "loss": 0.5979,
+      "step": 6433
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.173859849318532,
+      "learning_rate": 8.186942855574628e-06,
+      "loss": 0.6346,
+      "step": 6434
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5982909517716208,
+      "learning_rate": 8.186420542055622e-06,
+      "loss": 0.5566,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4981386654061248,
+      "learning_rate": 8.185898172298072e-06,
+      "loss": 0.3764,
+      "step": 6436
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7627985204826542,
+      "learning_rate": 8.185375746312913e-06,
+      "loss": 0.4819,
+      "step": 6437
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.74231100074312,
+      "learning_rate": 8.184853264111078e-06,
+      "loss": 0.6511,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3779758878704271,
+      "learning_rate": 8.18433072570351e-06,
+      "loss": 0.3408,
+      "step": 6439
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.806161930557281,
+      "learning_rate": 8.183808131101145e-06,
+      "loss": 0.5306,
+      "step": 6440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8737458295942722,
+      "learning_rate": 8.183285480314923e-06,
+      "loss": 0.5848,
+      "step": 6441
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.250337442426504,
+      "learning_rate": 8.182762773355787e-06,
+      "loss": 0.6191,
+      "step": 6442
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8833312396482627,
+      "learning_rate": 8.182240010234678e-06,
+      "loss": 0.4744,
+      "step": 6443
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9626427133466378,
+      "learning_rate": 8.181717190962541e-06,
+      "loss": 0.585,
+      "step": 6444
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.826612471177656,
+      "learning_rate": 8.18119431555032e-06,
+      "loss": 0.4651,
+      "step": 6445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.82697494957159,
+      "learning_rate": 8.18067138400896e-06,
+      "loss": 0.5416,
+      "step": 6446
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9004953705572034,
+      "learning_rate": 8.180148396349411e-06,
+      "loss": 0.6994,
+      "step": 6447
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0327366629657617,
+      "learning_rate": 8.179625352582622e-06,
+      "loss": 0.6094,
+      "step": 6448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2985912425065727,
+      "learning_rate": 8.179102252719537e-06,
+      "loss": 0.5884,
+      "step": 6449
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7655999404042662,
+      "learning_rate": 8.178579096771111e-06,
+      "loss": 0.5932,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5587916970252991,
+      "eval_runtime": 68.0061,
+      "eval_samples_per_second": 50.143,
+      "eval_steps_per_second": 1.573,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6544075937686997,
+      "learning_rate": 8.178055884748298e-06,
+      "loss": 0.5146,
+      "step": 6451
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6525926372654967,
+      "learning_rate": 8.177532616662046e-06,
+      "loss": 0.4802,
+      "step": 6452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6348325794487213,
+      "learning_rate": 8.177009292523312e-06,
+      "loss": 0.6514,
+      "step": 6453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.266936515586245,
+      "learning_rate": 8.176485912343051e-06,
+      "loss": 0.5741,
+      "step": 6454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7019481469640245,
+      "learning_rate": 8.175962476132221e-06,
+      "loss": 0.6809,
+      "step": 6455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7895194291316785,
+      "learning_rate": 8.17543898390178e-06,
+      "loss": 0.4388,
+      "step": 6456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8901178415466129,
+      "learning_rate": 8.174915435662682e-06,
+      "loss": 0.5366,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0931923953824367,
+      "learning_rate": 8.174391831425892e-06,
+      "loss": 0.4005,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8992448310339192,
+      "learning_rate": 8.173868171202371e-06,
+      "loss": 0.5716,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.048436787096475,
+      "learning_rate": 8.17334445500308e-06,
+      "loss": 0.5638,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8055238827888938,
+      "learning_rate": 8.172820682838984e-06,
+      "loss": 0.4857,
+      "step": 6461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9181371298781034,
+      "learning_rate": 8.172296854721047e-06,
+      "loss": 0.5057,
+      "step": 6462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8554474282694144,
+      "learning_rate": 8.171772970660234e-06,
+      "loss": 0.5022,
+      "step": 6463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9175007864076539,
+      "learning_rate": 8.171249030667514e-06,
+      "loss": 0.6411,
+      "step": 6464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.990218634930634,
+      "learning_rate": 8.170725034753855e-06,
+      "loss": 0.6109,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7173622525800367,
+      "learning_rate": 8.170200982930226e-06,
+      "loss": 0.5117,
+      "step": 6466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5067991384055786,
+      "learning_rate": 8.1696768752076e-06,
+      "loss": 0.5985,
+      "step": 6467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6380572979704402,
+      "learning_rate": 8.169152711596945e-06,
+      "loss": 0.5099,
+      "step": 6468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7141857867738277,
+      "learning_rate": 8.168628492109236e-06,
+      "loss": 0.4513,
+      "step": 6469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5861422482340846,
+      "learning_rate": 8.168104216755446e-06,
+      "loss": 0.7339,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.855280811884219,
+      "learning_rate": 8.167579885546552e-06,
+      "loss": 0.5196,
+      "step": 6471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9617646798988977,
+      "learning_rate": 8.16705549849353e-06,
+      "loss": 0.5629,
+      "step": 6472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8450856377246079,
+      "learning_rate": 8.16653105560736e-06,
+      "loss": 0.5997,
+      "step": 6473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.030071950946119,
+      "learning_rate": 8.166006556899015e-06,
+      "loss": 0.6623,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6926828438618808,
+      "learning_rate": 8.16548200237948e-06,
+      "loss": 0.453,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0153200262562616,
+      "learning_rate": 8.164957392059736e-06,
+      "loss": 0.4936,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7507684608021374,
+      "learning_rate": 8.164432725950763e-06,
+      "loss": 0.4559,
+      "step": 6477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9703615653722732,
+      "learning_rate": 8.163908004063547e-06,
+      "loss": 0.6516,
+      "step": 6478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.008402797074059,
+      "learning_rate": 8.16338322640907e-06,
+      "loss": 0.5381,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.184783810754827,
+      "learning_rate": 8.162858392998322e-06,
+      "loss": 0.6441,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1379455137965344,
+      "learning_rate": 8.162333503842285e-06,
+      "loss": 0.5048,
+      "step": 6481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.850275426188624,
+      "learning_rate": 8.16180855895195e-06,
+      "loss": 0.6346,
+      "step": 6482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6123704902697085,
+      "learning_rate": 8.161283558338307e-06,
+      "loss": 0.508,
+      "step": 6483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6804507623556426,
+      "learning_rate": 8.160758502012347e-06,
+      "loss": 0.4396,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.988141114343262,
+      "learning_rate": 8.160233389985057e-06,
+      "loss": 0.6159,
+      "step": 6485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7527507521013932,
+      "learning_rate": 8.159708222267437e-06,
+      "loss": 0.6861,
+      "step": 6486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.042073387896485,
+      "learning_rate": 8.159182998870473e-06,
+      "loss": 0.6098,
+      "step": 6487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.719409695214533,
+      "learning_rate": 8.158657719805167e-06,
+      "loss": 0.4348,
+      "step": 6488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.263508310788714,
+      "learning_rate": 8.158132385082514e-06,
+      "loss": 0.6264,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.861613894241652,
+      "learning_rate": 8.15760699471351e-06,
+      "loss": 0.7388,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3813191950321615,
+      "learning_rate": 8.157081548709153e-06,
+      "loss": 0.5611,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.865454016696861,
+      "learning_rate": 8.156556047080445e-06,
+      "loss": 0.5108,
+      "step": 6492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.063694868788015,
+      "learning_rate": 8.156030489838388e-06,
+      "loss": 0.5867,
+      "step": 6493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8259778556295394,
+      "learning_rate": 8.155504876993981e-06,
+      "loss": 0.4538,
+      "step": 6494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.542559280188038,
+      "learning_rate": 8.154979208558228e-06,
+      "loss": 0.5061,
+      "step": 6495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8107559995979858,
+      "learning_rate": 8.154453484542132e-06,
+      "loss": 0.5411,
+      "step": 6496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7646349395521899,
+      "learning_rate": 8.153927704956706e-06,
+      "loss": 0.5745,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9805505402684505,
+      "learning_rate": 8.15340186981295e-06,
+      "loss": 0.4947,
+      "step": 6498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9837020733611475,
+      "learning_rate": 8.152875979121873e-06,
+      "loss": 0.5645,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0287541654861725,
+      "learning_rate": 8.152350032894486e-06,
+      "loss": 0.6141,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.558673620223999,
+      "eval_runtime": 67.9546,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.268784881395156,
+      "learning_rate": 8.151824031141798e-06,
+      "loss": 0.4213,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0573532768834553,
+      "learning_rate": 8.151297973874821e-06,
+      "loss": 0.5736,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8512872312892898,
+      "learning_rate": 8.150771861104569e-06,
+      "loss": 0.6098,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6357118514308624,
+      "learning_rate": 8.150245692842053e-06,
+      "loss": 0.3139,
+      "step": 6504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9081136978815403,
+      "learning_rate": 8.14971946909829e-06,
+      "loss": 0.7672,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.071136176696475,
+      "learning_rate": 8.149193189884296e-06,
+      "loss": 0.5531,
+      "step": 6506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8292716627917243,
+      "learning_rate": 8.148666855211087e-06,
+      "loss": 0.5498,
+      "step": 6507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8538020434420293,
+      "learning_rate": 8.148140465089683e-06,
+      "loss": 0.5652,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.151002686790586,
+      "learning_rate": 8.147614019531106e-06,
+      "loss": 0.5982,
+      "step": 6509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9323627109080124,
+      "learning_rate": 8.14708751854637e-06,
+      "loss": 0.5332,
+      "step": 6510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7674251714976663,
+      "learning_rate": 8.146560962146503e-06,
+      "loss": 0.5252,
+      "step": 6511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.030298217429431,
+      "learning_rate": 8.146034350342525e-06,
+      "loss": 0.4154,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5287253771883613,
+      "learning_rate": 8.145507683145461e-06,
+      "loss": 0.4334,
+      "step": 6513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.475376438767964,
+      "learning_rate": 8.14498096056634e-06,
+      "loss": 0.4971,
+      "step": 6514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.837125046745633,
+      "learning_rate": 8.144454182616182e-06,
+      "loss": 0.4624,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5469367764387163,
+      "learning_rate": 8.143927349306019e-06,
+      "loss": 0.5044,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.608557105558468,
+      "learning_rate": 8.14340046064688e-06,
+      "loss": 0.3917,
+      "step": 6517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2446786500002127,
+      "learning_rate": 8.142873516649793e-06,
+      "loss": 0.5889,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1777286833613583,
+      "learning_rate": 8.142346517325793e-06,
+      "loss": 0.5795,
+      "step": 6519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9726924899791816,
+      "learning_rate": 8.141819462685909e-06,
+      "loss": 0.7229,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7791586432256863,
+      "learning_rate": 8.141292352741173e-06,
+      "loss": 0.5493,
+      "step": 6521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4717014975880565,
+      "learning_rate": 8.140765187502624e-06,
+      "loss": 0.5859,
+      "step": 6522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1019096309279486,
+      "learning_rate": 8.140237966981296e-06,
+      "loss": 0.5662,
+      "step": 6523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0919239008322776,
+      "learning_rate": 8.139710691188226e-06,
+      "loss": 0.643,
+      "step": 6524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8499826640415509,
+      "learning_rate": 8.139183360134453e-06,
+      "loss": 0.4769,
+      "step": 6525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.105208621255429,
+      "learning_rate": 8.138655973831014e-06,
+      "loss": 0.5689,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.490834527171396,
+      "learning_rate": 8.138128532288952e-06,
+      "loss": 0.4184,
+      "step": 6527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7866931375446637,
+      "learning_rate": 8.137601035519308e-06,
+      "loss": 0.5474,
+      "step": 6528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7686419012076506,
+      "learning_rate": 8.137073483533125e-06,
+      "loss": 0.5333,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8315211364412411,
+      "learning_rate": 8.136545876341447e-06,
+      "loss": 0.6037,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9224613406852664,
+      "learning_rate": 8.136018213955317e-06,
+      "loss": 0.4773,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7954858494078596,
+      "learning_rate": 8.135490496385784e-06,
+      "loss": 0.4662,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.68389246434848,
+      "learning_rate": 8.134962723643893e-06,
+      "loss": 0.4547,
+      "step": 6533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9140333166750312,
+      "learning_rate": 8.134434895740695e-06,
+      "loss": 0.6428,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0535188020976487,
+      "learning_rate": 8.13390701268724e-06,
+      "loss": 0.4856,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7708399264256793,
+      "learning_rate": 8.133379074494576e-06,
+      "loss": 0.5603,
+      "step": 6536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7295633069284648,
+      "learning_rate": 8.132851081173757e-06,
+      "loss": 0.4979,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8392536334530185,
+      "learning_rate": 8.132323032735836e-06,
+      "loss": 0.7046,
+      "step": 6538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2145677111415925,
+      "learning_rate": 8.131794929191866e-06,
+      "loss": 0.5545,
+      "step": 6539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8688762415619862,
+      "learning_rate": 8.131266770552907e-06,
+      "loss": 0.6455,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6934337664906007,
+      "learning_rate": 8.130738556830009e-06,
+      "loss": 0.6011,
+      "step": 6541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.635493095963414,
+      "learning_rate": 8.130210288034236e-06,
+      "loss": 0.4148,
+      "step": 6542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6625544258054057,
+      "learning_rate": 8.129681964176644e-06,
+      "loss": 0.4071,
+      "step": 6543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6318068025061427,
+      "learning_rate": 8.129153585268292e-06,
+      "loss": 0.5354,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8873572728926424,
+      "learning_rate": 8.128625151320244e-06,
+      "loss": 0.6056,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7415718779077443,
+      "learning_rate": 8.128096662343561e-06,
+      "loss": 0.4736,
+      "step": 6546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8206495176258832,
+      "learning_rate": 8.127568118349308e-06,
+      "loss": 0.6011,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4967165939208176,
+      "learning_rate": 8.127039519348547e-06,
+      "loss": 0.5019,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6569478270345248,
+      "learning_rate": 8.126510865352344e-06,
+      "loss": 0.5839,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.901469402392317,
+      "learning_rate": 8.12598215637177e-06,
+      "loss": 0.6389,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5578126311302185,
+      "eval_runtime": 67.9538,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6610749880578335,
+      "learning_rate": 8.12545339241789e-06,
+      "loss": 0.5542,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0186979709501816,
+      "learning_rate": 8.124924573501773e-06,
+      "loss": 0.5666,
+      "step": 6552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6627577123506787,
+      "learning_rate": 8.124395699634489e-06,
+      "loss": 0.483,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7161830153187783,
+      "learning_rate": 8.123866770827113e-06,
+      "loss": 0.4958,
+      "step": 6554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.014519001837863,
+      "learning_rate": 8.123337787090718e-06,
+      "loss": 0.5794,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.071712335347319,
+      "learning_rate": 8.122808748436372e-06,
+      "loss": 0.5928,
+      "step": 6556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.403475796077781,
+      "learning_rate": 8.122279654875156e-06,
+      "loss": 0.5285,
+      "step": 6557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8020811323384696,
+      "learning_rate": 8.121750506418143e-06,
+      "loss": 0.4883,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9953095625515234,
+      "learning_rate": 8.121221303076412e-06,
+      "loss": 0.4982,
+      "step": 6559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9247342006066537,
+      "learning_rate": 8.120692044861042e-06,
+      "loss": 0.5405,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.72911604993158,
+      "learning_rate": 8.12016273178311e-06,
+      "loss": 0.6069,
+      "step": 6561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6752245435552877,
+      "learning_rate": 8.1196333638537e-06,
+      "loss": 0.6289,
+      "step": 6562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2033534716172833,
+      "learning_rate": 8.119103941083892e-06,
+      "loss": 0.6389,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.288290302026024,
+      "learning_rate": 8.118574463484771e-06,
+      "loss": 0.6812,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9131808619643602,
+      "learning_rate": 8.118044931067421e-06,
+      "loss": 0.59,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.147084421054919,
+      "learning_rate": 8.117515343842924e-06,
+      "loss": 0.5766,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5650810070356238,
+      "learning_rate": 8.11698570182237e-06,
+      "loss": 0.4823,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7327409692319733,
+      "learning_rate": 8.116456005016845e-06,
+      "loss": 0.5255,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7028298232488779,
+      "learning_rate": 8.115926253437439e-06,
+      "loss": 0.5145,
+      "step": 6569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8490513362755745,
+      "learning_rate": 8.115396447095244e-06,
+      "loss": 0.6797,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7819898984900546,
+      "learning_rate": 8.114866586001346e-06,
+      "loss": 0.5495,
+      "step": 6571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5975763961339133,
+      "learning_rate": 8.114336670166842e-06,
+      "loss": 0.4573,
+      "step": 6572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5914583218753624,
+      "learning_rate": 8.113806699602822e-06,
+      "loss": 0.4345,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0221759146258664,
+      "learning_rate": 8.113276674320383e-06,
+      "loss": 0.6006,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4371820785162635,
+      "learning_rate": 8.11274659433062e-06,
+      "loss": 0.4556,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1152265145916247,
+      "learning_rate": 8.11221645964463e-06,
+      "loss": 0.6809,
+      "step": 6576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1258477051013815,
+      "learning_rate": 8.11168627027351e-06,
+      "loss": 0.6469,
+      "step": 6577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7639151117890635,
+      "learning_rate": 8.111156026228363e-06,
+      "loss": 0.5307,
+      "step": 6578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4328595343111545,
+      "learning_rate": 8.110625727520283e-06,
+      "loss": 0.494,
+      "step": 6579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.862024355375899,
+      "learning_rate": 8.110095374160376e-06,
+      "loss": 0.6163,
+      "step": 6580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1940746043890496,
+      "learning_rate": 8.109564966159745e-06,
+      "loss": 0.6541,
+      "step": 6581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.02565141501905,
+      "learning_rate": 8.10903450352949e-06,
+      "loss": 0.6245,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6163779549144222,
+      "learning_rate": 8.10850398628072e-06,
+      "loss": 0.4788,
+      "step": 6583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2481356734423903,
+      "learning_rate": 8.10797341442454e-06,
+      "loss": 0.5728,
+      "step": 6584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8508036839834472,
+      "learning_rate": 8.107442787972053e-06,
+      "loss": 0.5402,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9429008089596254,
+      "learning_rate": 8.106912106934374e-06,
+      "loss": 0.6318,
+      "step": 6586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0716525133545196,
+      "learning_rate": 8.106381371322607e-06,
+      "loss": 0.6529,
+      "step": 6587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7844951538663536,
+      "learning_rate": 8.105850581147866e-06,
+      "loss": 0.4379,
+      "step": 6588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8730092265626797,
+      "learning_rate": 8.105319736421261e-06,
+      "loss": 0.5651,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3865799496541285,
+      "learning_rate": 8.104788837153907e-06,
+      "loss": 0.7873,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7825099406943223,
+      "learning_rate": 8.104257883356917e-06,
+      "loss": 0.4733,
+      "step": 6591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5650617460579888,
+      "learning_rate": 8.103726875041404e-06,
+      "loss": 0.6261,
+      "step": 6592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.479006639752995,
+      "learning_rate": 8.103195812218486e-06,
+      "loss": 0.7427,
+      "step": 6593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4244876175464571,
+      "learning_rate": 8.10266469489928e-06,
+      "loss": 0.4033,
+      "step": 6594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0339252566530663,
+      "learning_rate": 8.102133523094907e-06,
+      "loss": 0.5684,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.457458351544401,
+      "learning_rate": 8.101602296816485e-06,
+      "loss": 0.4386,
+      "step": 6596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7700176882041572,
+      "learning_rate": 8.101071016075135e-06,
+      "loss": 0.592,
+      "step": 6597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4565191482379283,
+      "learning_rate": 8.100539680881976e-06,
+      "loss": 0.4068,
+      "step": 6598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7490683699295464,
+      "learning_rate": 8.100008291248138e-06,
+      "loss": 0.5122,
+      "step": 6599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6131408523752022,
+      "learning_rate": 8.09947684718474e-06,
+      "loss": 0.4556,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5561593174934387,
+      "eval_runtime": 67.9211,
+      "eval_samples_per_second": 50.205,
+      "eval_steps_per_second": 1.575,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7069604415838722,
+      "learning_rate": 8.098945348702906e-06,
+      "loss": 0.5094,
+      "step": 6601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2265924019226038,
+      "learning_rate": 8.098413795813769e-06,
+      "loss": 0.841,
+      "step": 6602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7401466655810889,
+      "learning_rate": 8.097882188528451e-06,
+      "loss": 0.5823,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.591918296941156,
+      "learning_rate": 8.097350526858083e-06,
+      "loss": 0.5486,
+      "step": 6604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7558036572163498,
+      "learning_rate": 8.096818810813796e-06,
+      "loss": 0.7394,
+      "step": 6605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0504859001285367,
+      "learning_rate": 8.096287040406721e-06,
+      "loss": 0.5844,
+      "step": 6606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4517775253891205,
+      "learning_rate": 8.095755215647986e-06,
+      "loss": 0.5269,
+      "step": 6607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9234129412009024,
+      "learning_rate": 8.09522333654873e-06,
+      "loss": 0.6715,
+      "step": 6608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8066339553502253,
+      "learning_rate": 8.094691403120086e-06,
+      "loss": 0.6008,
+      "step": 6609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7138087793301622,
+      "learning_rate": 8.09415941537319e-06,
+      "loss": 0.376,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7866222315209743,
+      "learning_rate": 8.093627373319176e-06,
+      "loss": 0.4826,
+      "step": 6611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.893995300060202,
+      "learning_rate": 8.093095276969184e-06,
+      "loss": 0.4843,
+      "step": 6612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7247558536389795,
+      "learning_rate": 8.092563126334352e-06,
+      "loss": 0.4356,
+      "step": 6613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1598667807718246,
+      "learning_rate": 8.092030921425823e-06,
+      "loss": 0.6415,
+      "step": 6614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8461048457073221,
+      "learning_rate": 8.091498662254738e-06,
+      "loss": 0.4991,
+      "step": 6615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.14919305467332,
+      "learning_rate": 8.090966348832237e-06,
+      "loss": 0.5955,
+      "step": 6616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.220708952990314,
+      "learning_rate": 8.090433981169462e-06,
+      "loss": 0.7633,
+      "step": 6617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5267018582967102,
+      "learning_rate": 8.089901559277564e-06,
+      "loss": 0.3975,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.597292866004712,
+      "learning_rate": 8.089369083167684e-06,
+      "loss": 0.6296,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.098679897248448,
+      "learning_rate": 8.08883655285097e-06,
+      "loss": 0.6617,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.712831586235899,
+      "learning_rate": 8.088303968338572e-06,
+      "loss": 0.5065,
+      "step": 6621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5552195773265463,
+      "learning_rate": 8.087771329641638e-06,
+      "loss": 0.4608,
+      "step": 6622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.582958435101186,
+      "learning_rate": 8.087238636771318e-06,
+      "loss": 0.5223,
+      "step": 6623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0429230516400905,
+      "learning_rate": 8.086705889738765e-06,
+      "loss": 0.6571,
+      "step": 6624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.090414903329671,
+      "learning_rate": 8.08617308855513e-06,
+      "loss": 0.5103,
+      "step": 6625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8300249588047368,
+      "learning_rate": 8.08564023323157e-06,
+      "loss": 0.4647,
+      "step": 6626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6520982478011006,
+      "learning_rate": 8.085107323779232e-06,
+      "loss": 0.5646,
+      "step": 6627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.823228749400427,
+      "learning_rate": 8.084574360209285e-06,
+      "loss": 0.6287,
+      "step": 6628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9381399867939868,
+      "learning_rate": 8.084041342532873e-06,
+      "loss": 0.5426,
+      "step": 6629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.042577974371428,
+      "learning_rate": 8.083508270761164e-06,
+      "loss": 0.5543,
+      "step": 6630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7457951569392756,
+      "learning_rate": 8.082975144905315e-06,
+      "loss": 0.5075,
+      "step": 6631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.564485746433646,
+      "learning_rate": 8.082441964976483e-06,
+      "loss": 0.4407,
+      "step": 6632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6116526244364995,
+      "learning_rate": 8.081908730985833e-06,
+      "loss": 0.5178,
+      "step": 6633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9936019579815392,
+      "learning_rate": 8.081375442944526e-06,
+      "loss": 0.5974,
+      "step": 6634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.862707019301954,
+      "learning_rate": 8.08084210086373e-06,
+      "loss": 0.6766,
+      "step": 6635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0858056335805495,
+      "learning_rate": 8.080308704754607e-06,
+      "loss": 0.578,
+      "step": 6636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8114431526779238,
+      "learning_rate": 8.079775254628324e-06,
+      "loss": 0.575,
+      "step": 6637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2441073430094067,
+      "learning_rate": 8.079241750496048e-06,
+      "loss": 0.6077,
+      "step": 6638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6486162065299206,
+      "learning_rate": 8.07870819236895e-06,
+      "loss": 0.5464,
+      "step": 6639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.009718025776994,
+      "learning_rate": 8.078174580258193e-06,
+      "loss": 0.7276,
+      "step": 6640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4995192720314665,
+      "learning_rate": 8.077640914174957e-06,
+      "loss": 0.4998,
+      "step": 6641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.722592794382918,
+      "learning_rate": 8.077107194130407e-06,
+      "loss": 0.4717,
+      "step": 6642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8758122692765542,
+      "learning_rate": 8.07657342013572e-06,
+      "loss": 0.5068,
+      "step": 6643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7067402502405875,
+      "learning_rate": 8.076039592202067e-06,
+      "loss": 0.4501,
+      "step": 6644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6418921006185274,
+      "learning_rate": 8.075505710340628e-06,
+      "loss": 0.5125,
+      "step": 6645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.257699083948802,
+      "learning_rate": 8.074971774562574e-06,
+      "loss": 0.4294,
+      "step": 6646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9117497776233192,
+      "learning_rate": 8.074437784879087e-06,
+      "loss": 0.5924,
+      "step": 6647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5309089037865078,
+      "learning_rate": 8.073903741301344e-06,
+      "loss": 0.3833,
+      "step": 6648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.833222915837585,
+      "learning_rate": 8.073369643840524e-06,
+      "loss": 0.6679,
+      "step": 6649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6039728383388165,
+      "learning_rate": 8.07283549250781e-06,
+      "loss": 0.5494,
+      "step": 6650
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.5568195581436157,
+      "eval_runtime": 67.9405,
+      "eval_samples_per_second": 50.191,
+      "eval_steps_per_second": 1.575,
+      "step": 6650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6454305506898312,
+      "learning_rate": 8.072301287314382e-06,
+      "loss": 0.4779,
+      "step": 6651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1092900841204716,
+      "learning_rate": 8.071767028271423e-06,
+      "loss": 0.5594,
+      "step": 6652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7864134290609963,
+      "learning_rate": 8.07123271539012e-06,
+      "loss": 0.5365,
+      "step": 6653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8577444539024617,
+      "learning_rate": 8.070698348681657e-06,
+      "loss": 0.6716,
+      "step": 6654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7278256359057678,
+      "learning_rate": 8.070163928157221e-06,
+      "loss": 0.6271,
+      "step": 6655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6878679559332144,
+      "learning_rate": 8.069629453828e-06,
+      "loss": 0.5642,
+      "step": 6656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.646919993932024,
+      "learning_rate": 8.069094925705182e-06,
+      "loss": 0.5891,
+      "step": 6657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5589271479412254,
+      "learning_rate": 8.068560343799958e-06,
+      "loss": 0.4883,
+      "step": 6658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5429571285357786,
+      "learning_rate": 8.068025708123518e-06,
+      "loss": 0.4682,
+      "step": 6659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1063775030104432,
+      "learning_rate": 8.067491018687055e-06,
+      "loss": 0.5706,
+      "step": 6660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8210641368331157,
+      "learning_rate": 8.066956275501761e-06,
+      "loss": 0.4357,
+      "step": 6661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1774298294542365,
+      "learning_rate": 8.066421478578833e-06,
+      "loss": 0.6318,
+      "step": 6662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2853622217281475,
+      "learning_rate": 8.065886627929466e-06,
+      "loss": 0.5864,
+      "step": 6663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.614030121690904,
+      "learning_rate": 8.065351723564856e-06,
+      "loss": 0.5354,
+      "step": 6664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7819221140552437,
+      "learning_rate": 8.0648167654962e-06,
+      "loss": 0.5027,
+      "step": 6665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.655742408025332,
+      "learning_rate": 8.064281753734699e-06,
+      "loss": 0.4964,
+      "step": 6666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0209893835822865,
+      "learning_rate": 8.063746688291552e-06,
+      "loss": 0.6773,
+      "step": 6667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5326045511730335,
+      "learning_rate": 8.063211569177961e-06,
+      "loss": 0.5078,
+      "step": 6668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5799171288722957,
+      "learning_rate": 8.062676396405125e-06,
+      "loss": 0.442,
+      "step": 6669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8542265268150573,
+      "learning_rate": 8.062141169984255e-06,
+      "loss": 0.4793,
+      "step": 6670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.829508276812437,
+      "learning_rate": 8.061605889926548e-06,
+      "loss": 0.4873,
+      "step": 6671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6199827284150485,
+      "learning_rate": 8.061070556243214e-06,
+      "loss": 0.4676,
+      "step": 6672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.16905343175087,
+      "learning_rate": 8.060535168945459e-06,
+      "loss": 0.5616,
+      "step": 6673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.031737233536892,
+      "learning_rate": 8.059999728044488e-06,
+      "loss": 0.6106,
+      "step": 6674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7548660077747285,
+      "learning_rate": 8.059464233551515e-06,
+      "loss": 0.4867,
+      "step": 6675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.107918923102106,
+      "learning_rate": 8.058928685477747e-06,
+      "loss": 0.6318,
+      "step": 6676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5918680811106816,
+      "learning_rate": 8.058393083834396e-06,
+      "loss": 0.7176,
+      "step": 6677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4895211039841147,
+      "learning_rate": 8.057857428632677e-06,
+      "loss": 0.4209,
+      "step": 6678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8234069707227212,
+      "learning_rate": 8.057321719883799e-06,
+      "loss": 0.544,
+      "step": 6679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.73067513868425,
+      "learning_rate": 8.056785957598983e-06,
+      "loss": 0.5582,
+      "step": 6680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7363700143963674,
+      "learning_rate": 8.056250141789437e-06,
+      "loss": 0.6115,
+      "step": 6681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.851998187722882,
+      "learning_rate": 8.055714272466384e-06,
+      "loss": 0.5943,
+      "step": 6682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7958073566759927,
+      "learning_rate": 8.055178349641038e-06,
+      "loss": 0.4515,
+      "step": 6683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9331074444742562,
+      "learning_rate": 8.054642373324622e-06,
+      "loss": 0.7407,
+      "step": 6684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.037666759404954,
+      "learning_rate": 8.054106343528355e-06,
+      "loss": 0.4087,
+      "step": 6685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.130336793121467,
+      "learning_rate": 8.053570260263458e-06,
+      "loss": 0.6483,
+      "step": 6686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7479594901917173,
+      "learning_rate": 8.053034123541152e-06,
+      "loss": 0.5619,
+      "step": 6687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6769550209974848,
+      "learning_rate": 8.052497933372665e-06,
+      "loss": 0.4113,
+      "step": 6688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7389100462838407,
+      "learning_rate": 8.051961689769217e-06,
+      "loss": 0.6105,
+      "step": 6689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6534499170248933,
+      "learning_rate": 8.051425392742038e-06,
+      "loss": 0.4868,
+      "step": 6690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0388062514684715,
+      "learning_rate": 8.050889042302353e-06,
+      "loss": 0.6378,
+      "step": 6691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.776999991987122,
+      "learning_rate": 8.05035263846139e-06,
+      "loss": 0.5292,
+      "step": 6692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6148746978891337,
+      "learning_rate": 8.049816181230377e-06,
+      "loss": 0.5112,
+      "step": 6693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7908885130008458,
+      "learning_rate": 8.049279670620547e-06,
+      "loss": 0.5208,
+      "step": 6694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3128894835590477,
+      "learning_rate": 8.048743106643131e-06,
+      "loss": 0.8104,
+      "step": 6695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8658703854399492,
+      "learning_rate": 8.04820648930936e-06,
+      "loss": 0.4522,
+      "step": 6696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.314638667112788,
+      "learning_rate": 8.047669818630472e-06,
+      "loss": 0.5157,
+      "step": 6697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3930032530193284,
+      "learning_rate": 8.047133094617695e-06,
+      "loss": 0.4482,
+      "step": 6698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5985173117748201,
+      "learning_rate": 8.04659631728227e-06,
+      "loss": 0.5045,
+      "step": 6699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4770584958134878,
+      "learning_rate": 8.046059486635433e-06,
+      "loss": 0.4213,
+      "step": 6700
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.5537242889404297,
+      "eval_runtime": 67.9367,
+      "eval_samples_per_second": 50.194,
+      "eval_steps_per_second": 1.575,
+      "step": 6700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8669603848577858,
+      "learning_rate": 8.045522602688423e-06,
+      "loss": 0.5329,
+      "step": 6701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6233132009518685,
+      "learning_rate": 8.044985665452477e-06,
+      "loss": 0.7011,
+      "step": 6702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.212448684609281,
+      "learning_rate": 8.044448674938836e-06,
+      "loss": 0.7161,
+      "step": 6703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0623468837095547,
+      "learning_rate": 8.043911631158745e-06,
+      "loss": 0.6022,
+      "step": 6704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.699733999407912,
+      "learning_rate": 8.043374534123442e-06,
+      "loss": 0.477,
+      "step": 6705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.932655448712741,
+      "learning_rate": 8.042837383844174e-06,
+      "loss": 0.6021,
+      "step": 6706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6834888730891469,
+      "learning_rate": 8.042300180332184e-06,
+      "loss": 0.5105,
+      "step": 6707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0632263034483795,
+      "learning_rate": 8.041762923598721e-06,
+      "loss": 0.6821,
+      "step": 6708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7561625253962336,
+      "learning_rate": 8.041225613655026e-06,
+      "loss": 0.4758,
+      "step": 6709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9443497968317527,
+      "learning_rate": 8.040688250512354e-06,
+      "loss": 0.4948,
+      "step": 6710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3949747716711365,
+      "learning_rate": 8.04015083418195e-06,
+      "loss": 0.5481,
+      "step": 6711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9237967777701659,
+      "learning_rate": 8.039613364675069e-06,
+      "loss": 0.4778,
+      "step": 6712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5545251509179665,
+      "learning_rate": 8.039075842002956e-06,
+      "loss": 0.4972,
+      "step": 6713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.959062624721011,
+      "learning_rate": 8.03853826617687e-06,
+      "loss": 0.5933,
+      "step": 6714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9410998995809579,
+      "learning_rate": 8.038000637208061e-06,
+      "loss": 0.6437,
+      "step": 6715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5911792710630432,
+      "learning_rate": 8.037462955107785e-06,
+      "loss": 0.4166,
+      "step": 6716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7593960306741716,
+      "learning_rate": 8.036925219887298e-06,
+      "loss": 0.4838,
+      "step": 6717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8963842940200826,
+      "learning_rate": 8.036387431557858e-06,
+      "loss": 0.5565,
+      "step": 6718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7729011798637582,
+      "learning_rate": 8.035849590130722e-06,
+      "loss": 0.5762,
+      "step": 6719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5075283256803091,
+      "learning_rate": 8.035311695617149e-06,
+      "loss": 0.4985,
+      "step": 6720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7319311346736586,
+      "learning_rate": 8.034773748028403e-06,
+      "loss": 0.5957,
+      "step": 6721
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9003909827596652,
+      "learning_rate": 8.03423574737574e-06,
+      "loss": 0.4506,
+      "step": 6722
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6165182890617908,
+      "learning_rate": 8.033697693670428e-06,
+      "loss": 0.4429,
+      "step": 6723
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2486273337248615,
+      "learning_rate": 8.033159586923728e-06,
+      "loss": 0.7416,
+      "step": 6724
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7287223311387705,
+      "learning_rate": 8.032621427146905e-06,
+      "loss": 0.4389,
+      "step": 6725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.924004757057991,
+      "learning_rate": 8.032083214351226e-06,
+      "loss": 0.6461,
+      "step": 6726
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4913165716048027,
+      "learning_rate": 8.031544948547955e-06,
+      "loss": 0.5322,
+      "step": 6727
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6089425137712936,
+      "learning_rate": 8.031006629748367e-06,
+      "loss": 0.4519,
+      "step": 6728
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8222772048292735,
+      "learning_rate": 8.030468257963724e-06,
+      "loss": 0.5473,
+      "step": 6729
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6595556005818166,
+      "learning_rate": 8.029929833205302e-06,
+      "loss": 0.3664,
+      "step": 6730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2728409235516143,
+      "learning_rate": 8.02939135548437e-06,
+      "loss": 0.5905,
+      "step": 6731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4478844949381748,
+      "learning_rate": 8.0288528248122e-06,
+      "loss": 0.3548,
+      "step": 6732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3894364321475217,
+      "learning_rate": 8.028314241200068e-06,
+      "loss": 0.5137,
+      "step": 6733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6929586150921325,
+      "learning_rate": 8.027775604659246e-06,
+      "loss": 0.4479,
+      "step": 6734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9557612722971303,
+      "learning_rate": 8.027236915201012e-06,
+      "loss": 0.615,
+      "step": 6735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8812983683397926,
+      "learning_rate": 8.026698172836644e-06,
+      "loss": 0.5422,
+      "step": 6736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9112435808232509,
+      "learning_rate": 8.026159377577419e-06,
+      "loss": 0.5597,
+      "step": 6737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4718372186838835,
+      "learning_rate": 8.025620529434616e-06,
+      "loss": 0.3391,
+      "step": 6738
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.866149897381905,
+      "learning_rate": 8.025081628419517e-06,
+      "loss": 0.4944,
+      "step": 6739
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.084158023107004,
+      "learning_rate": 8.024542674543403e-06,
+      "loss": 0.7184,
+      "step": 6740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8440844842562463,
+      "learning_rate": 8.024003667817555e-06,
+      "loss": 0.4129,
+      "step": 6741
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7543205082120419,
+      "learning_rate": 8.023464608253257e-06,
+      "loss": 0.5456,
+      "step": 6742
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3133681349736404,
+      "learning_rate": 8.022925495861796e-06,
+      "loss": 0.6533,
+      "step": 6743
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3171815814437893,
+      "learning_rate": 8.022386330654456e-06,
+      "loss": 0.4279,
+      "step": 6744
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9768522015069152,
+      "learning_rate": 8.021847112642526e-06,
+      "loss": 0.5438,
+      "step": 6745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7135616036172432,
+      "learning_rate": 8.021307841837293e-06,
+      "loss": 0.6575,
+      "step": 6746
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9194950541485087,
+      "learning_rate": 8.020768518250045e-06,
+      "loss": 0.5226,
+      "step": 6747
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8891543988856914,
+      "learning_rate": 8.020229141892074e-06,
+      "loss": 0.588,
+      "step": 6748
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8362372445290522,
+      "learning_rate": 8.019689712774671e-06,
+      "loss": 0.5682,
+      "step": 6749
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9168966274278867,
+      "learning_rate": 8.019150230909131e-06,
+      "loss": 0.4569,
+      "step": 6750
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.5570847392082214,
+      "eval_runtime": 67.9254,
+      "eval_samples_per_second": 50.202,
+      "eval_steps_per_second": 1.575,
+      "step": 6750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9166902967305532,
+      "learning_rate": 8.018610696306745e-06,
+      "loss": 0.5296,
+      "step": 6751
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8943998387044367,
+      "learning_rate": 8.018071108978806e-06,
+      "loss": 0.6519,
+      "step": 6752
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.018240222209161,
+      "learning_rate": 8.017531468936615e-06,
+      "loss": 0.6087,
+      "step": 6753
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1469613277577624,
+      "learning_rate": 8.016991776191467e-06,
+      "loss": 0.7103,
+      "step": 6754
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6592288179613126,
+      "learning_rate": 8.016452030754657e-06,
+      "loss": 0.4448,
+      "step": 6755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8286805154294505,
+      "learning_rate": 8.015912232637489e-06,
+      "loss": 0.435,
+      "step": 6756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5671680766866904,
+      "learning_rate": 8.01537238185126e-06,
+      "loss": 0.5059,
+      "step": 6757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1275898834602196,
+      "learning_rate": 8.014832478407272e-06,
+      "loss": 0.6293,
+      "step": 6758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1291237204319504,
+      "learning_rate": 8.01429252231683e-06,
+      "loss": 0.5236,
+      "step": 6759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.819611262470864,
+      "learning_rate": 8.013752513591233e-06,
+      "loss": 0.4072,
+      "step": 6760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.613976606075351,
+      "learning_rate": 8.01321245224179e-06,
+      "loss": 0.4536,
+      "step": 6761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.001820092245409,
+      "learning_rate": 8.012672338279804e-06,
+      "loss": 0.5357,
+      "step": 6762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3510416260297435,
+      "learning_rate": 8.012132171716584e-06,
+      "loss": 0.6089,
+      "step": 6763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8583293856336056,
+      "learning_rate": 8.011591952563439e-06,
+      "loss": 0.5763,
+      "step": 6764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.795749016510245,
+      "learning_rate": 8.011051680831676e-06,
+      "loss": 0.4749,
+      "step": 6765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5554429299299721,
+      "learning_rate": 8.010511356532606e-06,
+      "loss": 0.6013,
+      "step": 6766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9722197916818445,
+      "learning_rate": 8.009970979677537e-06,
+      "loss": 0.5982,
+      "step": 6767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7848655387437309,
+      "learning_rate": 8.009430550277787e-06,
+      "loss": 0.4672,
+      "step": 6768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0192424326758123,
+      "learning_rate": 8.008890068344666e-06,
+      "loss": 0.5545,
+      "step": 6769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.034263072508063,
+      "learning_rate": 8.008349533889491e-06,
+      "loss": 0.6159,
+      "step": 6770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.820232238351911,
+      "learning_rate": 8.007808946923575e-06,
+      "loss": 0.5686,
+      "step": 6771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.669629451657144,
+      "learning_rate": 8.00726830745824e-06,
+      "loss": 0.4665,
+      "step": 6772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.485061148077997,
+      "learning_rate": 8.006727615504797e-06,
+      "loss": 0.4896,
+      "step": 6773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8271419783581917,
+      "learning_rate": 8.006186871074568e-06,
+      "loss": 0.5316,
+      "step": 6774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.623628579749736,
+      "learning_rate": 8.005646074178874e-06,
+      "loss": 0.5729,
+      "step": 6775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2977573139366054,
+      "learning_rate": 8.005105224829037e-06,
+      "loss": 0.5237,
+      "step": 6776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7139858962388914,
+      "learning_rate": 8.004564323036377e-06,
+      "loss": 0.6224,
+      "step": 6777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8751534801305536,
+      "learning_rate": 8.004023368812219e-06,
+      "loss": 0.5933,
+      "step": 6778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9077343074897597,
+      "learning_rate": 8.003482362167888e-06,
+      "loss": 0.5844,
+      "step": 6779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.64270297300297,
+      "learning_rate": 8.002941303114707e-06,
+      "loss": 0.4884,
+      "step": 6780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6924060855570866,
+      "learning_rate": 8.002400191664005e-06,
+      "loss": 0.536,
+      "step": 6781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1140909826418417,
+      "learning_rate": 8.001859027827111e-06,
+      "loss": 0.5889,
+      "step": 6782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8803726636766438,
+      "learning_rate": 8.001317811615348e-06,
+      "loss": 0.5402,
+      "step": 6783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.038814366345992,
+      "learning_rate": 8.000776543040053e-06,
+      "loss": 0.596,
+      "step": 6784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4785459652020594,
+      "learning_rate": 8.000235222112555e-06,
+      "loss": 0.4907,
+      "step": 6785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4616056225246294,
+      "learning_rate": 7.999693848844185e-06,
+      "loss": 0.415,
+      "step": 6786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.801070344301124,
+      "learning_rate": 7.999152423246275e-06,
+      "loss": 0.6445,
+      "step": 6787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2277208814357947,
+      "learning_rate": 7.998610945330164e-06,
+      "loss": 0.782,
+      "step": 6788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7487453738675487,
+      "learning_rate": 7.998069415107182e-06,
+      "loss": 0.5202,
+      "step": 6789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7871738090579152,
+      "learning_rate": 7.997527832588672e-06,
+      "loss": 0.5259,
+      "step": 6790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6979816531811722,
+      "learning_rate": 7.996986197785966e-06,
+      "loss": 0.4704,
+      "step": 6791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8995495160422127,
+      "learning_rate": 7.996444510710407e-06,
+      "loss": 0.5118,
+      "step": 6792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1744704797251924,
+      "learning_rate": 7.99590277137333e-06,
+      "loss": 0.7226,
+      "step": 6793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.901040976335248,
+      "learning_rate": 7.99536097978608e-06,
+      "loss": 0.3886,
+      "step": 6794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6924269450404772,
+      "learning_rate": 7.994819135959996e-06,
+      "loss": 0.5059,
+      "step": 6795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7944052461671038,
+      "learning_rate": 7.994277239906425e-06,
+      "loss": 0.6454,
+      "step": 6796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.107009499536074,
+      "learning_rate": 7.993735291636708e-06,
+      "loss": 0.4945,
+      "step": 6797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6257400718853492,
+      "learning_rate": 7.993193291162193e-06,
+      "loss": 0.4389,
+      "step": 6798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2376210621277544,
+      "learning_rate": 7.992651238494225e-06,
+      "loss": 0.6809,
+      "step": 6799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9583629619017113,
+      "learning_rate": 7.992109133644151e-06,
+      "loss": 0.5874,
+      "step": 6800
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.556075930595398,
+      "eval_runtime": 67.9132,
+      "eval_samples_per_second": 50.211,
+      "eval_steps_per_second": 1.576,
+      "step": 6800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.598640178349114,
+      "learning_rate": 7.99156697662332e-06,
+      "loss": 0.3708,
+      "step": 6801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9521437171626652,
+      "learning_rate": 7.991024767443081e-06,
+      "loss": 0.4499,
+      "step": 6802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.849399997247051,
+      "learning_rate": 7.990482506114788e-06,
+      "loss": 0.575,
+      "step": 6803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7292055301089724,
+      "learning_rate": 7.989940192649789e-06,
+      "loss": 0.7737,
+      "step": 6804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6870164930638045,
+      "learning_rate": 7.989397827059438e-06,
+      "loss": 0.5404,
+      "step": 6805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.61609268967767,
+      "learning_rate": 7.98885540935509e-06,
+      "loss": 0.6433,
+      "step": 6806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7396750605835467,
+      "learning_rate": 7.988312939548102e-06,
+      "loss": 0.5034,
+      "step": 6807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.765010808802218,
+      "learning_rate": 7.987770417649826e-06,
+      "loss": 0.5816,
+      "step": 6808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8261533045643075,
+      "learning_rate": 7.987227843671622e-06,
+      "loss": 0.6481,
+      "step": 6809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1054772690581682,
+      "learning_rate": 7.98668521762485e-06,
+      "loss": 0.5083,
+      "step": 6810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8622852993069081,
+      "learning_rate": 7.986142539520864e-06,
+      "loss": 0.5419,
+      "step": 6811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6308310716359844,
+      "learning_rate": 7.98559980937103e-06,
+      "loss": 0.4535,
+      "step": 6812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8029672364026073,
+      "learning_rate": 7.985057027186708e-06,
+      "loss": 0.6258,
+      "step": 6813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.352569946112301,
+      "learning_rate": 7.98451419297926e-06,
+      "loss": 0.4193,
+      "step": 6814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6953368071847268,
+      "learning_rate": 7.98397130676005e-06,
+      "loss": 0.5392,
+      "step": 6815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.817194546724894,
+      "learning_rate": 7.983428368540446e-06,
+      "loss": 0.5284,
+      "step": 6816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9339097628076813,
+      "learning_rate": 7.98288537833181e-06,
+      "loss": 0.467,
+      "step": 6817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.010333570636926,
+      "learning_rate": 7.982342336145511e-06,
+      "loss": 0.5108,
+      "step": 6818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8018108021951975,
+      "learning_rate": 7.981799241992916e-06,
+      "loss": 0.4099,
+      "step": 6819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.62244723961572,
+      "learning_rate": 7.981256095885397e-06,
+      "loss": 0.4629,
+      "step": 6820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7460504282651148,
+      "learning_rate": 7.980712897834322e-06,
+      "loss": 0.4502,
+      "step": 6821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9228991070746033,
+      "learning_rate": 7.980169647851063e-06,
+      "loss": 0.6052,
+      "step": 6822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1107461394201112,
+      "learning_rate": 7.979626345946994e-06,
+      "loss": 0.6002,
+      "step": 6823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7480295091862346,
+      "learning_rate": 7.979082992133488e-06,
+      "loss": 0.5554,
+      "step": 6824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.014905577850175,
+      "learning_rate": 7.978539586421917e-06,
+      "loss": 0.5162,
+      "step": 6825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7285100262194149,
+      "learning_rate": 7.97799612882366e-06,
+      "loss": 0.5221,
+      "step": 6826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9212419035565562,
+      "learning_rate": 7.977452619350094e-06,
+      "loss": 0.5693,
+      "step": 6827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6444062764747591,
+      "learning_rate": 7.976909058012596e-06,
+      "loss": 0.5099,
+      "step": 6828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9531494651290768,
+      "learning_rate": 7.976365444822545e-06,
+      "loss": 0.5454,
+      "step": 6829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7432620980730322,
+      "learning_rate": 7.975821779791323e-06,
+      "loss": 0.4883,
+      "step": 6830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7103758180132933,
+      "learning_rate": 7.975278062930309e-06,
+      "loss": 0.5827,
+      "step": 6831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7988433619372892,
+      "learning_rate": 7.974734294250885e-06,
+      "loss": 0.5694,
+      "step": 6832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8163086307166678,
+      "learning_rate": 7.974190473764437e-06,
+      "loss": 0.64,
+      "step": 6833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1291472187662914,
+      "learning_rate": 7.973646601482348e-06,
+      "loss": 0.5622,
+      "step": 6834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.972137665035518,
+      "learning_rate": 7.973102677416003e-06,
+      "loss": 0.5232,
+      "step": 6835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9699401860684054,
+      "learning_rate": 7.97255870157679e-06,
+      "loss": 0.4984,
+      "step": 6836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3136816953768697,
+      "learning_rate": 7.972014673976096e-06,
+      "loss": 0.7211,
+      "step": 6837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9007544775200305,
+      "learning_rate": 7.97147059462531e-06,
+      "loss": 0.54,
+      "step": 6838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0981348676085183,
+      "learning_rate": 7.970926463535824e-06,
+      "loss": 0.6254,
+      "step": 6839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.370155006138959,
+      "learning_rate": 7.970382280719026e-06,
+      "loss": 0.744,
+      "step": 6840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.709701092427386,
+      "learning_rate": 7.96983804618631e-06,
+      "loss": 0.3916,
+      "step": 6841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.570764829220246,
+      "learning_rate": 7.969293759949066e-06,
+      "loss": 0.418,
+      "step": 6842
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.046243335576091,
+      "learning_rate": 7.968749422018695e-06,
+      "loss": 0.6138,
+      "step": 6843
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1018624673501156,
+      "learning_rate": 7.968205032406586e-06,
+      "loss": 0.5842,
+      "step": 6844
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.405869368828119,
+      "learning_rate": 7.967660591124138e-06,
+      "loss": 0.6797,
+      "step": 6845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.894449547226219,
+      "learning_rate": 7.967116098182747e-06,
+      "loss": 0.5926,
+      "step": 6846
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7129655539596513,
+      "learning_rate": 7.966571553593814e-06,
+      "loss": 0.405,
+      "step": 6847
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7211037943092187,
+      "learning_rate": 7.966026957368737e-06,
+      "loss": 0.5231,
+      "step": 6848
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8387071085887106,
+      "learning_rate": 7.96548230951892e-06,
+      "loss": 0.5658,
+      "step": 6849
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7124374849783457,
+      "learning_rate": 7.96493761005576e-06,
+      "loss": 0.5275,
+      "step": 6850
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5553905367851257,
+      "eval_runtime": 67.9158,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 6850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4371311634431807,
+      "learning_rate": 7.96439285899066e-06,
+      "loss": 0.4567,
+      "step": 6851
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0357019601569344,
+      "learning_rate": 7.96384805633503e-06,
+      "loss": 0.6429,
+      "step": 6852
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6383604375972542,
+      "learning_rate": 7.96330320210027e-06,
+      "loss": 0.5679,
+      "step": 6853
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5406208000912518,
+      "learning_rate": 7.962758296297786e-06,
+      "loss": 0.4629,
+      "step": 6854
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0231591896292733,
+      "learning_rate": 7.962213338938989e-06,
+      "loss": 0.615,
+      "step": 6855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6066322794622532,
+      "learning_rate": 7.961668330035284e-06,
+      "loss": 0.404,
+      "step": 6856
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8781533229390883,
+      "learning_rate": 7.961123269598082e-06,
+      "loss": 0.5954,
+      "step": 6857
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.029702622851699,
+      "learning_rate": 7.960578157638792e-06,
+      "loss": 0.6016,
+      "step": 6858
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9538046311821091,
+      "learning_rate": 7.960032994168827e-06,
+      "loss": 0.5534,
+      "step": 6859
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9554246530736725,
+      "learning_rate": 7.959487779199598e-06,
+      "loss": 0.4616,
+      "step": 6860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7476503734805915,
+      "learning_rate": 7.958942512742522e-06,
+      "loss": 0.4656,
+      "step": 6861
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9105799687751719,
+      "learning_rate": 7.958397194809011e-06,
+      "loss": 0.484,
+      "step": 6862
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.693837764189783,
+      "learning_rate": 7.957851825410481e-06,
+      "loss": 0.5333,
+      "step": 6863
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.86286387701201,
+      "learning_rate": 7.957306404558352e-06,
+      "loss": 0.5493,
+      "step": 6864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5768058701576055,
+      "learning_rate": 7.956760932264039e-06,
+      "loss": 0.4952,
+      "step": 6865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2104439488215935,
+      "learning_rate": 7.95621540853896e-06,
+      "loss": 0.7172,
+      "step": 6866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8987280737991916,
+      "learning_rate": 7.955669833394539e-06,
+      "loss": 0.6171,
+      "step": 6867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6915666798812363,
+      "learning_rate": 7.955124206842195e-06,
+      "loss": 0.4489,
+      "step": 6868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.096244410774859,
+      "learning_rate": 7.95457852889335e-06,
+      "loss": 0.7463,
+      "step": 6869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.075075589547832,
+      "learning_rate": 7.954032799559428e-06,
+      "loss": 0.4877,
+      "step": 6870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0355959144082174,
+      "learning_rate": 7.953487018851854e-06,
+      "loss": 0.5774,
+      "step": 6871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.719974685309651,
+      "learning_rate": 7.952941186782053e-06,
+      "loss": 0.545,
+      "step": 6872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7050211753385147,
+      "learning_rate": 7.952395303361452e-06,
+      "loss": 0.575,
+      "step": 6873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7922750509621532,
+      "learning_rate": 7.951849368601479e-06,
+      "loss": 0.6057,
+      "step": 6874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9590382647668603,
+      "learning_rate": 7.951303382513561e-06,
+      "loss": 0.5327,
+      "step": 6875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.006941795475204,
+      "learning_rate": 7.95075734510913e-06,
+      "loss": 0.394,
+      "step": 6876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9091515597644566,
+      "learning_rate": 7.950211256399616e-06,
+      "loss": 0.6121,
+      "step": 6877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4491256769020533,
+      "learning_rate": 7.94966511639645e-06,
+      "loss": 0.477,
+      "step": 6878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9377426195957657,
+      "learning_rate": 7.94911892511107e-06,
+      "loss": 0.5925,
+      "step": 6879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.032755967438369,
+      "learning_rate": 7.948572682554904e-06,
+      "loss": 0.6542,
+      "step": 6880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9986834770219755,
+      "learning_rate": 7.948026388739388e-06,
+      "loss": 0.591,
+      "step": 6881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4188293847894646,
+      "learning_rate": 7.947480043675963e-06,
+      "loss": 0.3832,
+      "step": 6882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9615783962183195,
+      "learning_rate": 7.946933647376062e-06,
+      "loss": 0.5572,
+      "step": 6883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9356752214550619,
+      "learning_rate": 7.946387199851125e-06,
+      "loss": 0.5562,
+      "step": 6884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7881963498988787,
+      "learning_rate": 7.945840701112593e-06,
+      "loss": 0.5029,
+      "step": 6885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6181143767413912,
+      "learning_rate": 7.945294151171903e-06,
+      "loss": 0.4476,
+      "step": 6886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.895769741668004,
+      "learning_rate": 7.944747550040501e-06,
+      "loss": 0.5749,
+      "step": 6887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1636873247117063,
+      "learning_rate": 7.944200897729827e-06,
+      "loss": 0.6018,
+      "step": 6888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8607286275801251,
+      "learning_rate": 7.943654194251325e-06,
+      "loss": 0.5559,
+      "step": 6889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3962959966878863,
+      "learning_rate": 7.94310743961644e-06,
+      "loss": 0.6995,
+      "step": 6890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.037533905047631,
+      "learning_rate": 7.942560633836618e-06,
+      "loss": 0.6128,
+      "step": 6891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0765109371386616,
+      "learning_rate": 7.942013776923305e-06,
+      "loss": 0.476,
+      "step": 6892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6598208102198264,
+      "learning_rate": 7.941466868887953e-06,
+      "loss": 0.5378,
+      "step": 6893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8860407081615018,
+      "learning_rate": 7.940919909742007e-06,
+      "loss": 0.721,
+      "step": 6894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.005166815563271,
+      "learning_rate": 7.94037289949692e-06,
+      "loss": 0.5862,
+      "step": 6895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.047671011763422,
+      "learning_rate": 7.939825838164141e-06,
+      "loss": 0.6299,
+      "step": 6896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5998229830128412,
+      "learning_rate": 7.939278725755124e-06,
+      "loss": 0.573,
+      "step": 6897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6766099071125389,
+      "learning_rate": 7.938731562281321e-06,
+      "loss": 0.5224,
+      "step": 6898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9300030382364861,
+      "learning_rate": 7.938184347754189e-06,
+      "loss": 0.6209,
+      "step": 6899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0704315066157437,
+      "learning_rate": 7.93763708218518e-06,
+      "loss": 0.5523,
+      "step": 6900
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5558739900588989,
+      "eval_runtime": 67.8968,
+      "eval_samples_per_second": 50.223,
+      "eval_steps_per_second": 1.576,
+      "step": 6900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8156529053153585,
+      "learning_rate": 7.937089765585754e-06,
+      "loss": 0.6655,
+      "step": 6901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8739672303704165,
+      "learning_rate": 7.936542397967369e-06,
+      "loss": 0.5742,
+      "step": 6902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2559040984187533,
+      "learning_rate": 7.935994979341481e-06,
+      "loss": 0.688,
+      "step": 6903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8628684122698618,
+      "learning_rate": 7.93544750971955e-06,
+      "loss": 0.4889,
+      "step": 6904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.062779406113055,
+      "learning_rate": 7.93489998911304e-06,
+      "loss": 0.6257,
+      "step": 6905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9437062561531606,
+      "learning_rate": 7.934352417533409e-06,
+      "loss": 0.5192,
+      "step": 6906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1638037608689986,
+      "learning_rate": 7.933804794992125e-06,
+      "loss": 0.7729,
+      "step": 6907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5289368504917147,
+      "learning_rate": 7.933257121500647e-06,
+      "loss": 0.5968,
+      "step": 6908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.629585104835856,
+      "learning_rate": 7.932709397070444e-06,
+      "loss": 0.4324,
+      "step": 6909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1822542110263274,
+      "learning_rate": 7.93216162171298e-06,
+      "loss": 0.5189,
+      "step": 6910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0541227985763797,
+      "learning_rate": 7.931613795439723e-06,
+      "loss": 0.501,
+      "step": 6911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1169552044565254,
+      "learning_rate": 7.931065918262142e-06,
+      "loss": 0.6095,
+      "step": 6912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8821721710978543,
+      "learning_rate": 7.930517990191705e-06,
+      "loss": 0.5106,
+      "step": 6913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6013125254610654,
+      "learning_rate": 7.929970011239885e-06,
+      "loss": 0.5984,
+      "step": 6914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.239566364319603,
+      "learning_rate": 7.929421981418151e-06,
+      "loss": 0.321,
+      "step": 6915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.369817009277784,
+      "learning_rate": 7.928873900737976e-06,
+      "loss": 0.5252,
+      "step": 6916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4623461241531066,
+      "learning_rate": 7.928325769210838e-06,
+      "loss": 0.602,
+      "step": 6917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.116504848350993,
+      "learning_rate": 7.927777586848205e-06,
+      "loss": 0.4579,
+      "step": 6918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.088062019219388,
+      "learning_rate": 7.927229353661557e-06,
+      "loss": 0.6785,
+      "step": 6919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9568981925090962,
+      "learning_rate": 7.926681069662371e-06,
+      "loss": 0.4464,
+      "step": 6920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7338932442462234,
+      "learning_rate": 7.926132734862122e-06,
+      "loss": 0.7216,
+      "step": 6921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9571170821543589,
+      "learning_rate": 7.925584349272293e-06,
+      "loss": 0.7082,
+      "step": 6922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5935401061393133,
+      "learning_rate": 7.925035912904362e-06,
+      "loss": 0.5628,
+      "step": 6923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0725695490408764,
+      "learning_rate": 7.92448742576981e-06,
+      "loss": 0.7151,
+      "step": 6924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.014951849338032,
+      "learning_rate": 7.92393888788012e-06,
+      "loss": 0.617,
+      "step": 6925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7945677704803378,
+      "learning_rate": 7.923390299246774e-06,
+      "loss": 0.6602,
+      "step": 6926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.162702154747392,
+      "learning_rate": 7.922841659881258e-06,
+      "loss": 0.6658,
+      "step": 6927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.607834282887682,
+      "learning_rate": 7.922292969795057e-06,
+      "loss": 0.4543,
+      "step": 6928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5800489565056577,
+      "learning_rate": 7.921744228999658e-06,
+      "loss": 0.4056,
+      "step": 6929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6932896274878393,
+      "learning_rate": 7.921195437506545e-06,
+      "loss": 0.5118,
+      "step": 6930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4300330491822757,
+      "learning_rate": 7.920646595327211e-06,
+      "loss": 0.3917,
+      "step": 6931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8898726104754409,
+      "learning_rate": 7.920097702473143e-06,
+      "loss": 0.4666,
+      "step": 6932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9043452940290624,
+      "learning_rate": 7.919548758955835e-06,
+      "loss": 0.4502,
+      "step": 6933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7633738775740542,
+      "learning_rate": 7.918999764786775e-06,
+      "loss": 0.6866,
+      "step": 6934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6251897876042651,
+      "learning_rate": 7.918450719977458e-06,
+      "loss": 0.4858,
+      "step": 6935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1187402657131145,
+      "learning_rate": 7.917901624539378e-06,
+      "loss": 0.6243,
+      "step": 6936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9743883613589963,
+      "learning_rate": 7.917352478484028e-06,
+      "loss": 0.5672,
+      "step": 6937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8522816824248296,
+      "learning_rate": 7.916803281822905e-06,
+      "loss": 0.5198,
+      "step": 6938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7989723794290167,
+      "learning_rate": 7.91625403456751e-06,
+      "loss": 0.7523,
+      "step": 6939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6938004806141032,
+      "learning_rate": 7.915704736729332e-06,
+      "loss": 0.4407,
+      "step": 6940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.927907413425461,
+      "learning_rate": 7.91515538831988e-06,
+      "loss": 0.4038,
+      "step": 6941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7571030599643103,
+      "learning_rate": 7.914605989350648e-06,
+      "loss": 0.4548,
+      "step": 6942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.878529854373011,
+      "learning_rate": 7.914056539833141e-06,
+      "loss": 0.6031,
+      "step": 6943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0416165835855424,
+      "learning_rate": 7.913507039778858e-06,
+      "loss": 0.5514,
+      "step": 6944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.265722931092159,
+      "learning_rate": 7.912957489199306e-06,
+      "loss": 0.5505,
+      "step": 6945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.531941933792997,
+      "learning_rate": 7.912407888105986e-06,
+      "loss": 0.7055,
+      "step": 6946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.012411786401637,
+      "learning_rate": 7.911858236510406e-06,
+      "loss": 0.5449,
+      "step": 6947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1542697847153964,
+      "learning_rate": 7.911308534424071e-06,
+      "loss": 0.6105,
+      "step": 6948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5868611057239232,
+      "learning_rate": 7.910758781858492e-06,
+      "loss": 0.4375,
+      "step": 6949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0374094550900326,
+      "learning_rate": 7.910208978825173e-06,
+      "loss": 0.5431,
+      "step": 6950
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5546046495437622,
+      "eval_runtime": 67.8732,
+      "eval_samples_per_second": 50.241,
+      "eval_steps_per_second": 1.576,
+      "step": 6950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9278225908973443,
+      "learning_rate": 7.909659125335627e-06,
+      "loss": 0.5495,
+      "step": 6951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9148470707420964,
+      "learning_rate": 7.909109221401365e-06,
+      "loss": 0.7164,
+      "step": 6952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.156418025792204,
+      "learning_rate": 7.908559267033895e-06,
+      "loss": 0.6632,
+      "step": 6953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6570516256097343,
+      "learning_rate": 7.908009262244735e-06,
+      "loss": 0.534,
+      "step": 6954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6056255275379807,
+      "learning_rate": 7.907459207045394e-06,
+      "loss": 0.4721,
+      "step": 6955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.705628849183296,
+      "learning_rate": 7.906909101447393e-06,
+      "loss": 0.6567,
+      "step": 6956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7234343925694666,
+      "learning_rate": 7.906358945462244e-06,
+      "loss": 0.6259,
+      "step": 6957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7740549393984544,
+      "learning_rate": 7.905808739101465e-06,
+      "loss": 0.5495,
+      "step": 6958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7505566236376475,
+      "learning_rate": 7.905258482376573e-06,
+      "loss": 0.5157,
+      "step": 6959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.055694134581,
+      "learning_rate": 7.90470817529909e-06,
+      "loss": 0.6103,
+      "step": 6960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9271507054639057,
+      "learning_rate": 7.904157817880537e-06,
+      "loss": 0.62,
+      "step": 6961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.766946937556005,
+      "learning_rate": 7.90360741013243e-06,
+      "loss": 0.4746,
+      "step": 6962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.180199577945443,
+      "learning_rate": 7.903056952066297e-06,
+      "loss": 0.6022,
+      "step": 6963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0136030415242074,
+      "learning_rate": 7.902506443693656e-06,
+      "loss": 0.5205,
+      "step": 6964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0487292189151325,
+      "learning_rate": 7.90195588502604e-06,
+      "loss": 0.5894,
+      "step": 6965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9068109627600296,
+      "learning_rate": 7.901405276074965e-06,
+      "loss": 0.5563,
+      "step": 6966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9985599243235719,
+      "learning_rate": 7.900854616851964e-06,
+      "loss": 0.579,
+      "step": 6967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6255935502522996,
+      "learning_rate": 7.900303907368562e-06,
+      "loss": 0.5065,
+      "step": 6968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.66523654945454,
+      "learning_rate": 7.89975314763629e-06,
+      "loss": 0.4487,
+      "step": 6969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.165722331696521,
+      "learning_rate": 7.899202337666675e-06,
+      "loss": 0.5331,
+      "step": 6970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.208532868189339,
+      "learning_rate": 7.898651477471251e-06,
+      "loss": 0.6511,
+      "step": 6971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6582275538956182,
+      "learning_rate": 7.898100567061546e-06,
+      "loss": 0.6666,
+      "step": 6972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.75505834649226,
+      "learning_rate": 7.897549606449096e-06,
+      "loss": 0.5483,
+      "step": 6973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7040006272303418,
+      "learning_rate": 7.896998595645434e-06,
+      "loss": 0.6361,
+      "step": 6974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8611541108540406,
+      "learning_rate": 7.896447534662097e-06,
+      "loss": 0.4952,
+      "step": 6975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6478643056693405,
+      "learning_rate": 7.895896423510618e-06,
+      "loss": 0.4412,
+      "step": 6976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3661818712168214,
+      "learning_rate": 7.895345262202535e-06,
+      "loss": 0.635,
+      "step": 6977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0944502720492695,
+      "learning_rate": 7.894794050749387e-06,
+      "loss": 0.7114,
+      "step": 6978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8426092419828404,
+      "learning_rate": 7.894242789162712e-06,
+      "loss": 0.5648,
+      "step": 6979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.837954304560806,
+      "learning_rate": 7.893691477454054e-06,
+      "loss": 0.5291,
+      "step": 6980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7640225784278214,
+      "learning_rate": 7.89314011563495e-06,
+      "loss": 0.6212,
+      "step": 6981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.151409293437192,
+      "learning_rate": 7.892588703716943e-06,
+      "loss": 0.4944,
+      "step": 6982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9315631592277274,
+      "learning_rate": 7.892037241711578e-06,
+      "loss": 0.6114,
+      "step": 6983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0998633176129347,
+      "learning_rate": 7.891485729630403e-06,
+      "loss": 0.6396,
+      "step": 6984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7610489808677108,
+      "learning_rate": 7.890934167484955e-06,
+      "loss": 0.6784,
+      "step": 6985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5792479575564438,
+      "learning_rate": 7.890382555286788e-06,
+      "loss": 0.5008,
+      "step": 6986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5931901695832658,
+      "learning_rate": 7.889830893047445e-06,
+      "loss": 0.5051,
+      "step": 6987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7861456596045502,
+      "learning_rate": 7.889279180778478e-06,
+      "loss": 0.6285,
+      "step": 6988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.908967902763258,
+      "learning_rate": 7.888727418491434e-06,
+      "loss": 0.6005,
+      "step": 6989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6459824312545441,
+      "learning_rate": 7.888175606197867e-06,
+      "loss": 0.4272,
+      "step": 6990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.021267432892135,
+      "learning_rate": 7.887623743909325e-06,
+      "loss": 0.7101,
+      "step": 6991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8835286014886308,
+      "learning_rate": 7.887071831637365e-06,
+      "loss": 0.5838,
+      "step": 6992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7971491844981198,
+      "learning_rate": 7.886519869393538e-06,
+      "loss": 0.6363,
+      "step": 6993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.920336633603438,
+      "learning_rate": 7.8859678571894e-06,
+      "loss": 0.5989,
+      "step": 6994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8108057615890072,
+      "learning_rate": 7.885415795036505e-06,
+      "loss": 0.5763,
+      "step": 6995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6285296107745386,
+      "learning_rate": 7.884863682946414e-06,
+      "loss": 0.4308,
+      "step": 6996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.892984800478073,
+      "learning_rate": 7.884311520930684e-06,
+      "loss": 0.5936,
+      "step": 6997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.694627034662872,
+      "learning_rate": 7.88375930900087e-06,
+      "loss": 0.5364,
+      "step": 6998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6981022661576048,
+      "learning_rate": 7.88320704716854e-06,
+      "loss": 0.4008,
+      "step": 6999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8642455485931315,
+      "learning_rate": 7.882654735445246e-06,
+      "loss": 0.4865,
+      "step": 7000
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5539634823799133,
+      "eval_runtime": 67.8846,
+      "eval_samples_per_second": 50.232,
+      "eval_steps_per_second": 1.576,
+      "step": 7000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0000415019372086,
+      "learning_rate": 7.882102373842558e-06,
+      "loss": 0.5879,
+      "step": 7001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8159311745802424,
+      "learning_rate": 7.881549962372034e-06,
+      "loss": 0.7072,
+      "step": 7002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8752247055840394,
+      "learning_rate": 7.880997501045241e-06,
+      "loss": 0.6296,
+      "step": 7003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7844787248856582,
+      "learning_rate": 7.880444989873746e-06,
+      "loss": 0.4699,
+      "step": 7004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7878673132737426,
+      "learning_rate": 7.879892428869114e-06,
+      "loss": 0.4686,
+      "step": 7005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2683892133481214,
+      "learning_rate": 7.879339818042912e-06,
+      "loss": 0.621,
+      "step": 7006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2327259976158347,
+      "learning_rate": 7.878787157406709e-06,
+      "loss": 0.6598,
+      "step": 7007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.585093611596535,
+      "learning_rate": 7.878234446972073e-06,
+      "loss": 0.4127,
+      "step": 7008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9403786572814292,
+      "learning_rate": 7.877681686750575e-06,
+      "loss": 0.5722,
+      "step": 7009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6011060055632937,
+      "learning_rate": 7.87712887675379e-06,
+      "loss": 0.4185,
+      "step": 7010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1812433085081837,
+      "learning_rate": 7.87657601699329e-06,
+      "loss": 0.5594,
+      "step": 7011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4648251334743865,
+      "learning_rate": 7.876023107480644e-06,
+      "loss": 0.4129,
+      "step": 7012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0437515455280084,
+      "learning_rate": 7.87547014822743e-06,
+      "loss": 0.6126,
+      "step": 7013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.52578035570811,
+      "learning_rate": 7.874917139245225e-06,
+      "loss": 0.6927,
+      "step": 7014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.760716398617164,
+      "learning_rate": 7.874364080545605e-06,
+      "loss": 0.5284,
+      "step": 7015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.495588378525674,
+      "learning_rate": 7.873810972140149e-06,
+      "loss": 0.4519,
+      "step": 7016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7909719623441978,
+      "learning_rate": 7.873257814040432e-06,
+      "loss": 0.4795,
+      "step": 7017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5658886862079207,
+      "learning_rate": 7.872704606258038e-06,
+      "loss": 0.4171,
+      "step": 7018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.91171541764378,
+      "learning_rate": 7.872151348804546e-06,
+      "loss": 0.6206,
+      "step": 7019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5546108052641652,
+      "learning_rate": 7.87159804169154e-06,
+      "loss": 0.4632,
+      "step": 7020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1627283376871587,
+      "learning_rate": 7.871044684930602e-06,
+      "loss": 0.5938,
+      "step": 7021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5493425960624478,
+      "learning_rate": 7.870491278533316e-06,
+      "loss": 0.5614,
+      "step": 7022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4996916459500778,
+      "learning_rate": 7.869937822511267e-06,
+      "loss": 0.5631,
+      "step": 7023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9075566162399482,
+      "learning_rate": 7.869384316876043e-06,
+      "loss": 0.4996,
+      "step": 7024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7217721881158206,
+      "learning_rate": 7.86883076163923e-06,
+      "loss": 0.465,
+      "step": 7025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8053441080866357,
+      "learning_rate": 7.868277156812415e-06,
+      "loss": 0.425,
+      "step": 7026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7433240250463569,
+      "learning_rate": 7.867723502407188e-06,
+      "loss": 0.5053,
+      "step": 7027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.981517050208963,
+      "learning_rate": 7.867169798435142e-06,
+      "loss": 0.7675,
+      "step": 7028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2024448148966083,
+      "learning_rate": 7.866616044907865e-06,
+      "loss": 0.6608,
+      "step": 7029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6664434302830915,
+      "learning_rate": 7.866062241836951e-06,
+      "loss": 0.4578,
+      "step": 7030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6074260076815823,
+      "learning_rate": 7.865508389233997e-06,
+      "loss": 0.4601,
+      "step": 7031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7782759470812892,
+      "learning_rate": 7.86495448711059e-06,
+      "loss": 0.4516,
+      "step": 7032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8234739135649913,
+      "learning_rate": 7.864400535478333e-06,
+      "loss": 0.6335,
+      "step": 7033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8014032779378795,
+      "learning_rate": 7.863846534348817e-06,
+      "loss": 0.6799,
+      "step": 7034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0371643691405836,
+      "learning_rate": 7.863292483733643e-06,
+      "loss": 0.5963,
+      "step": 7035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7166162300616092,
+      "learning_rate": 7.862738383644408e-06,
+      "loss": 0.4214,
+      "step": 7036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.602974396560931,
+      "learning_rate": 7.862184234092714e-06,
+      "loss": 0.5631,
+      "step": 7037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9162619007057125,
+      "learning_rate": 7.861630035090159e-06,
+      "loss": 0.5908,
+      "step": 7038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8592098748999026,
+      "learning_rate": 7.861075786648348e-06,
+      "loss": 0.6015,
+      "step": 7039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1294505975123807,
+      "learning_rate": 7.86052148877888e-06,
+      "loss": 0.5488,
+      "step": 7040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6291028367193556,
+      "learning_rate": 7.85996714149336e-06,
+      "loss": 0.4682,
+      "step": 7041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7659874839499714,
+      "learning_rate": 7.859412744803395e-06,
+      "loss": 0.6259,
+      "step": 7042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.884255094765855,
+      "learning_rate": 7.858858298720589e-06,
+      "loss": 0.5436,
+      "step": 7043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7767337464545072,
+      "learning_rate": 7.85830380325655e-06,
+      "loss": 0.6048,
+      "step": 7044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5011337051537141,
+      "learning_rate": 7.857749258422888e-06,
+      "loss": 0.4601,
+      "step": 7045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9385883474172785,
+      "learning_rate": 7.857194664231206e-06,
+      "loss": 0.7214,
+      "step": 7046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.542217324253405,
+      "learning_rate": 7.85664002069312e-06,
+      "loss": 0.3828,
+      "step": 7047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2214337824170114,
+      "learning_rate": 7.856085327820238e-06,
+      "loss": 0.6094,
+      "step": 7048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4756063870566387,
+      "learning_rate": 7.855530585624173e-06,
+      "loss": 0.5678,
+      "step": 7049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9325631374486343,
+      "learning_rate": 7.85497579411654e-06,
+      "loss": 0.6415,
+      "step": 7050
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5560624599456787,
+      "eval_runtime": 67.9166,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 7050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.915572632932697,
+      "learning_rate": 7.854420953308948e-06,
+      "loss": 0.8081,
+      "step": 7051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7916274664394243,
+      "learning_rate": 7.85386606321302e-06,
+      "loss": 0.4433,
+      "step": 7052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7100106547292202,
+      "learning_rate": 7.853311123840366e-06,
+      "loss": 0.4271,
+      "step": 7053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.410055048603295,
+      "learning_rate": 7.852756135202605e-06,
+      "loss": 0.3955,
+      "step": 7054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8166199512330126,
+      "learning_rate": 7.852201097311355e-06,
+      "loss": 0.5017,
+      "step": 7055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0163137917767364,
+      "learning_rate": 7.851646010178236e-06,
+      "loss": 0.5981,
+      "step": 7056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8371011610677495,
+      "learning_rate": 7.85109087381487e-06,
+      "loss": 0.478,
+      "step": 7057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7114701509621875,
+      "learning_rate": 7.850535688232877e-06,
+      "loss": 0.6594,
+      "step": 7058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.349025771687362,
+      "learning_rate": 7.849980453443879e-06,
+      "loss": 0.3702,
+      "step": 7059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7797403019243294,
+      "learning_rate": 7.849425169459498e-06,
+      "loss": 0.4456,
+      "step": 7060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.754871963262844,
+      "learning_rate": 7.848869836291362e-06,
+      "loss": 0.4314,
+      "step": 7061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1644819411308838,
+      "learning_rate": 7.848314453951095e-06,
+      "loss": 0.6624,
+      "step": 7062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3104989142349477,
+      "learning_rate": 7.847759022450323e-06,
+      "loss": 0.7257,
+      "step": 7063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8010781455840756,
+      "learning_rate": 7.847203541800674e-06,
+      "loss": 0.5292,
+      "step": 7064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.044372504365349,
+      "learning_rate": 7.846648012013778e-06,
+      "loss": 0.553,
+      "step": 7065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0379538474296286,
+      "learning_rate": 7.846092433101264e-06,
+      "loss": 0.5519,
+      "step": 7066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4770558225766717,
+      "learning_rate": 7.845536805074758e-06,
+      "loss": 0.4561,
+      "step": 7067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0409117793617253,
+      "learning_rate": 7.8449811279459e-06,
+      "loss": 0.7154,
+      "step": 7068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7203341859065058,
+      "learning_rate": 7.844425401726316e-06,
+      "loss": 0.485,
+      "step": 7069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7447900025608634,
+      "learning_rate": 7.843869626427642e-06,
+      "loss": 0.4557,
+      "step": 7070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6762804193810108,
+      "learning_rate": 7.843313802061516e-06,
+      "loss": 0.5879,
+      "step": 7071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.153050332028425,
+      "learning_rate": 7.842757928639569e-06,
+      "loss": 0.7571,
+      "step": 7072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4043020608908066,
+      "learning_rate": 7.842202006173443e-06,
+      "loss": 0.6766,
+      "step": 7073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8286656962492063,
+      "learning_rate": 7.84164603467477e-06,
+      "loss": 0.4978,
+      "step": 7074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7084860558479809,
+      "learning_rate": 7.84109001415519e-06,
+      "loss": 0.5406,
+      "step": 7075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7868303891863109,
+      "learning_rate": 7.840533944626348e-06,
+      "loss": 0.513,
+      "step": 7076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6246061188982748,
+      "learning_rate": 7.83997782609988e-06,
+      "loss": 0.6227,
+      "step": 7077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7442542815705604,
+      "learning_rate": 7.839421658587431e-06,
+      "loss": 0.6228,
+      "step": 7078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6807796359391898,
+      "learning_rate": 7.838865442100642e-06,
+      "loss": 0.4384,
+      "step": 7079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8928427213903285,
+      "learning_rate": 7.838309176651156e-06,
+      "loss": 0.5392,
+      "step": 7080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7358004625478132,
+      "learning_rate": 7.837752862250621e-06,
+      "loss": 0.4198,
+      "step": 7081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5996791511890371,
+      "learning_rate": 7.837196498910682e-06,
+      "loss": 0.4585,
+      "step": 7082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.862443871848503,
+      "learning_rate": 7.836640086642987e-06,
+      "loss": 0.5822,
+      "step": 7083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.613045974413933,
+      "learning_rate": 7.836083625459182e-06,
+      "loss": 0.5761,
+      "step": 7084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.79497669240363,
+      "learning_rate": 7.835527115370917e-06,
+      "loss": 0.5502,
+      "step": 7085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.455198925931877,
+      "learning_rate": 7.834970556389841e-06,
+      "loss": 0.4216,
+      "step": 7086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.711954675660942,
+      "learning_rate": 7.834413948527611e-06,
+      "loss": 0.5545,
+      "step": 7087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.864970515818524,
+      "learning_rate": 7.833857291795871e-06,
+      "loss": 0.6562,
+      "step": 7088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.128537438354063,
+      "learning_rate": 7.83330058620628e-06,
+      "loss": 0.5459,
+      "step": 7089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8220879765805795,
+      "learning_rate": 7.83274383177049e-06,
+      "loss": 0.6334,
+      "step": 7090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9104308328503499,
+      "learning_rate": 7.832187028500157e-06,
+      "loss": 0.5757,
+      "step": 7091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5377601646606123,
+      "learning_rate": 7.831630176406939e-06,
+      "loss": 0.4027,
+      "step": 7092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.733947052037895,
+      "learning_rate": 7.83107327550249e-06,
+      "loss": 0.447,
+      "step": 7093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.626646642526696,
+      "learning_rate": 7.830516325798468e-06,
+      "loss": 0.4709,
+      "step": 7094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.93778354308052,
+      "learning_rate": 7.829959327306535e-06,
+      "loss": 0.4882,
+      "step": 7095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8402195912442387,
+      "learning_rate": 7.829402280038353e-06,
+      "loss": 0.5318,
+      "step": 7096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.944009324279978,
+      "learning_rate": 7.828845184005578e-06,
+      "loss": 0.6579,
+      "step": 7097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7497167180277344,
+      "learning_rate": 7.828288039219878e-06,
+      "loss": 0.6129,
+      "step": 7098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6944858008810777,
+      "learning_rate": 7.827730845692914e-06,
+      "loss": 0.4125,
+      "step": 7099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8099872810639024,
+      "learning_rate": 7.82717360343635e-06,
+      "loss": 0.4773,
+      "step": 7100
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5550462007522583,
+      "eval_runtime": 67.8913,
+      "eval_samples_per_second": 50.227,
+      "eval_steps_per_second": 1.576,
+      "step": 7100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4539122436119387,
+      "learning_rate": 7.82661631246185e-06,
+      "loss": 0.3999,
+      "step": 7101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7816233496323666,
+      "learning_rate": 7.826058972781085e-06,
+      "loss": 0.516,
+      "step": 7102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9554617812473438,
+      "learning_rate": 7.82550158440572e-06,
+      "loss": 0.5849,
+      "step": 7103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0059535356231337,
+      "learning_rate": 7.824944147347421e-06,
+      "loss": 0.6064,
+      "step": 7104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4651795884225922,
+      "learning_rate": 7.824386661617865e-06,
+      "loss": 0.4514,
+      "step": 7105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4334134943032744,
+      "learning_rate": 7.823829127228717e-06,
+      "loss": 0.3517,
+      "step": 7106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.589110736442573,
+      "learning_rate": 7.823271544191646e-06,
+      "loss": 0.4007,
+      "step": 7107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.186905858433966,
+      "learning_rate": 7.822713912518331e-06,
+      "loss": 0.5809,
+      "step": 7108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9660266207723391,
+      "learning_rate": 7.822156232220444e-06,
+      "loss": 0.5389,
+      "step": 7109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8564371178346835,
+      "learning_rate": 7.821598503309655e-06,
+      "loss": 0.5145,
+      "step": 7110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.885098928608879,
+      "learning_rate": 7.821040725797646e-06,
+      "loss": 0.7231,
+      "step": 7111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7519118947856944,
+      "learning_rate": 7.82048289969609e-06,
+      "loss": 0.5547,
+      "step": 7112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0485917068430988,
+      "learning_rate": 7.819925025016667e-06,
+      "loss": 0.4774,
+      "step": 7113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1678356323276913,
+      "learning_rate": 7.819367101771054e-06,
+      "loss": 0.5153,
+      "step": 7114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7986485385108268,
+      "learning_rate": 7.818809129970931e-06,
+      "loss": 0.3786,
+      "step": 7115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9871310723610922,
+      "learning_rate": 7.81825110962798e-06,
+      "loss": 0.58,
+      "step": 7116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.691805164412684,
+      "learning_rate": 7.81769304075388e-06,
+      "loss": 0.4122,
+      "step": 7117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.639410269393526,
+      "learning_rate": 7.817134923360317e-06,
+      "loss": 0.4817,
+      "step": 7118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.584947088424342,
+      "learning_rate": 7.816576757458974e-06,
+      "loss": 0.4817,
+      "step": 7119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9157578905549966,
+      "learning_rate": 7.816018543061535e-06,
+      "loss": 0.4454,
+      "step": 7120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7168872512792774,
+      "learning_rate": 7.815460280179687e-06,
+      "loss": 0.5354,
+      "step": 7121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7761800492901187,
+      "learning_rate": 7.814901968825116e-06,
+      "loss": 0.4811,
+      "step": 7122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6338187692537134,
+      "learning_rate": 7.81434360900951e-06,
+      "loss": 0.4059,
+      "step": 7123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8481010774671136,
+      "learning_rate": 7.813785200744557e-06,
+      "loss": 0.4788,
+      "step": 7124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0016038084436625,
+      "learning_rate": 7.813226744041948e-06,
+      "loss": 0.4802,
+      "step": 7125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0865098965792437,
+      "learning_rate": 7.812668238913374e-06,
+      "loss": 0.5811,
+      "step": 7126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1411432357161444,
+      "learning_rate": 7.81210968537053e-06,
+      "loss": 0.4851,
+      "step": 7127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8399134057185533,
+      "learning_rate": 7.811551083425103e-06,
+      "loss": 0.7823,
+      "step": 7128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8016293463564947,
+      "learning_rate": 7.810992433088788e-06,
+      "loss": 0.4563,
+      "step": 7129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.670973641498962,
+      "learning_rate": 7.810433734373285e-06,
+      "loss": 0.4892,
+      "step": 7130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8315962123080518,
+      "learning_rate": 7.809874987290285e-06,
+      "loss": 0.6553,
+      "step": 7131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8074128163068446,
+      "learning_rate": 7.809316191851489e-06,
+      "loss": 0.6083,
+      "step": 7132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9446259072284695,
+      "learning_rate": 7.808757348068592e-06,
+      "loss": 0.617,
+      "step": 7133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8143018861736195,
+      "learning_rate": 7.808198455953292e-06,
+      "loss": 0.4315,
+      "step": 7134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9689431716395793,
+      "learning_rate": 7.807639515517294e-06,
+      "loss": 0.5012,
+      "step": 7135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9810060189838405,
+      "learning_rate": 7.807080526772294e-06,
+      "loss": 0.4502,
+      "step": 7136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9546638285401325,
+      "learning_rate": 7.806521489729998e-06,
+      "loss": 0.6571,
+      "step": 7137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.088934509146132,
+      "learning_rate": 7.805962404402105e-06,
+      "loss": 0.5158,
+      "step": 7138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.152663052782369,
+      "learning_rate": 7.805403270800323e-06,
+      "loss": 0.639,
+      "step": 7139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9737174633467256,
+      "learning_rate": 7.804844088936354e-06,
+      "loss": 0.6114,
+      "step": 7140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2474171777560543,
+      "learning_rate": 7.804284858821906e-06,
+      "loss": 0.5152,
+      "step": 7141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2873510592469597,
+      "learning_rate": 7.803725580468687e-06,
+      "loss": 0.4085,
+      "step": 7142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6618688217743844,
+      "learning_rate": 7.803166253888404e-06,
+      "loss": 0.4995,
+      "step": 7143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7422568537721448,
+      "learning_rate": 7.802606879092763e-06,
+      "loss": 0.5727,
+      "step": 7144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.874896845572534,
+      "learning_rate": 7.80204745609348e-06,
+      "loss": 0.5254,
+      "step": 7145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7422337173866553,
+      "learning_rate": 7.801487984902265e-06,
+      "loss": 0.6525,
+      "step": 7146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8762348575035,
+      "learning_rate": 7.800928465530824e-06,
+      "loss": 0.5558,
+      "step": 7147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.626473925730573,
+      "learning_rate": 7.800368897990876e-06,
+      "loss": 0.4577,
+      "step": 7148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6068408569190187,
+      "learning_rate": 7.799809282294135e-06,
+      "loss": 0.7281,
+      "step": 7149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7019474904704683,
+      "learning_rate": 7.799249618452316e-06,
+      "loss": 0.4586,
+      "step": 7150
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.554461658000946,
+      "eval_runtime": 67.8431,
+      "eval_samples_per_second": 50.263,
+      "eval_steps_per_second": 1.577,
+      "step": 7150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.535498333879504,
+      "learning_rate": 7.798689906477134e-06,
+      "loss": 0.3884,
+      "step": 7151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5949157987839524,
+      "learning_rate": 7.798130146380303e-06,
+      "loss": 0.4634,
+      "step": 7152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.920941829913956,
+      "learning_rate": 7.79757033817355e-06,
+      "loss": 0.49,
+      "step": 7153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7132634130644417,
+      "learning_rate": 7.797010481868586e-06,
+      "loss": 0.4672,
+      "step": 7154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.63470292935377,
+      "learning_rate": 7.796450577477135e-06,
+      "loss": 0.4474,
+      "step": 7155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4228183599218376,
+      "learning_rate": 7.795890625010918e-06,
+      "loss": 0.4521,
+      "step": 7156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8055890961641132,
+      "learning_rate": 7.795330624481656e-06,
+      "loss": 0.5204,
+      "step": 7157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8484856582243998,
+      "learning_rate": 7.794770575901072e-06,
+      "loss": 0.451,
+      "step": 7158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7248715342333252,
+      "learning_rate": 7.794210479280894e-06,
+      "loss": 0.5849,
+      "step": 7159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.595212260485457,
+      "learning_rate": 7.793650334632842e-06,
+      "loss": 0.4483,
+      "step": 7160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8537191137728133,
+      "learning_rate": 7.793090141968647e-06,
+      "loss": 0.6967,
+      "step": 7161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.902141648830483,
+      "learning_rate": 7.792529901300034e-06,
+      "loss": 0.4499,
+      "step": 7162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9579037941658923,
+      "learning_rate": 7.791969612638732e-06,
+      "loss": 0.5905,
+      "step": 7163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7567924524396696,
+      "learning_rate": 7.791409275996466e-06,
+      "loss": 0.5265,
+      "step": 7164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5935087858209533,
+      "learning_rate": 7.790848891384973e-06,
+      "loss": 0.3829,
+      "step": 7165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5722981654415098,
+      "learning_rate": 7.790288458815982e-06,
+      "loss": 0.4292,
+      "step": 7166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7694506247195658,
+      "learning_rate": 7.789727978301224e-06,
+      "loss": 0.4822,
+      "step": 7167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8709094808436204,
+      "learning_rate": 7.789167449852434e-06,
+      "loss": 0.5319,
+      "step": 7168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.807265784306027,
+      "learning_rate": 7.788606873481343e-06,
+      "loss": 0.6135,
+      "step": 7169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8395101213953913,
+      "learning_rate": 7.788046249199688e-06,
+      "loss": 0.493,
+      "step": 7170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6249738868878643,
+      "learning_rate": 7.78748557701921e-06,
+      "loss": 0.5051,
+      "step": 7171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2153007071936024,
+      "learning_rate": 7.78692485695164e-06,
+      "loss": 0.4184,
+      "step": 7172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7281118119451226,
+      "learning_rate": 7.786364089008716e-06,
+      "loss": 0.5512,
+      "step": 7173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7789734155239807,
+      "learning_rate": 7.785803273202182e-06,
+      "loss": 0.5523,
+      "step": 7174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.792053641486565,
+      "learning_rate": 7.785242409543779e-06,
+      "loss": 0.4266,
+      "step": 7175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9548167282805753,
+      "learning_rate": 7.784681498045242e-06,
+      "loss": 0.6029,
+      "step": 7176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8411455162382733,
+      "learning_rate": 7.784120538718316e-06,
+      "loss": 0.5904,
+      "step": 7177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.908332278958882,
+      "learning_rate": 7.783559531574748e-06,
+      "loss": 0.5748,
+      "step": 7178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2116909754735383,
+      "learning_rate": 7.782998476626279e-06,
+      "loss": 0.5179,
+      "step": 7179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7957635651160992,
+      "learning_rate": 7.782437373884654e-06,
+      "loss": 0.6715,
+      "step": 7180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.397384057389322,
+      "learning_rate": 7.781876223361621e-06,
+      "loss": 0.5153,
+      "step": 7181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9545943281135911,
+      "learning_rate": 7.781315025068927e-06,
+      "loss": 0.5124,
+      "step": 7182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8852296773682629,
+      "learning_rate": 7.78075377901832e-06,
+      "loss": 0.4698,
+      "step": 7183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7520014828723103,
+      "learning_rate": 7.780192485221548e-06,
+      "loss": 0.4899,
+      "step": 7184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8852363168634092,
+      "learning_rate": 7.779631143690363e-06,
+      "loss": 0.5858,
+      "step": 7185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7725605764259482,
+      "learning_rate": 7.779069754436518e-06,
+      "loss": 0.4889,
+      "step": 7186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.014461496886391,
+      "learning_rate": 7.77850831747176e-06,
+      "loss": 0.8376,
+      "step": 7187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7762070448066352,
+      "learning_rate": 7.777946832807849e-06,
+      "loss": 0.5842,
+      "step": 7188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7769220265256023,
+      "learning_rate": 7.777385300456532e-06,
+      "loss": 0.4827,
+      "step": 7189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6937811228112105,
+      "learning_rate": 7.77682372042957e-06,
+      "loss": 0.4546,
+      "step": 7190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.601297897477119,
+      "learning_rate": 7.776262092738718e-06,
+      "loss": 0.5299,
+      "step": 7191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5303508358429032,
+      "learning_rate": 7.775700417395735e-06,
+      "loss": 0.4826,
+      "step": 7192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9173858784612545,
+      "learning_rate": 7.775138694412374e-06,
+      "loss": 0.4977,
+      "step": 7193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7137406816627336,
+      "learning_rate": 7.774576923800398e-06,
+      "loss": 0.4643,
+      "step": 7194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.184290031755473,
+      "learning_rate": 7.774015105571569e-06,
+      "loss": 0.6876,
+      "step": 7195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5432546045889086,
+      "learning_rate": 7.773453239737645e-06,
+      "loss": 0.357,
+      "step": 7196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.07186165125399,
+      "learning_rate": 7.772891326310388e-06,
+      "loss": 0.5637,
+      "step": 7197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0401514682162816,
+      "learning_rate": 7.772329365301566e-06,
+      "loss": 0.7104,
+      "step": 7198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.226189625032721,
+      "learning_rate": 7.771767356722937e-06,
+      "loss": 0.5426,
+      "step": 7199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4794865732462212,
+      "learning_rate": 7.771205300586272e-06,
+      "loss": 0.4403,
+      "step": 7200
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5538336634635925,
+      "eval_runtime": 67.8623,
+      "eval_samples_per_second": 50.249,
+      "eval_steps_per_second": 1.577,
+      "step": 7200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.179043671650493,
+      "learning_rate": 7.770643196903333e-06,
+      "loss": 0.6057,
+      "step": 7201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8159623849747895,
+      "learning_rate": 7.77008104568589e-06,
+      "loss": 0.4844,
+      "step": 7202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7933620156824988,
+      "learning_rate": 7.769518846945708e-06,
+      "loss": 0.4831,
+      "step": 7203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7387643269484017,
+      "learning_rate": 7.768956600694561e-06,
+      "loss": 0.5821,
+      "step": 7204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.584422629175488,
+      "learning_rate": 7.768394306944215e-06,
+      "loss": 0.4868,
+      "step": 7205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5719064883343337,
+      "learning_rate": 7.767831965706444e-06,
+      "loss": 0.4967,
+      "step": 7206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6696644642355383,
+      "learning_rate": 7.767269576993022e-06,
+      "loss": 0.4165,
+      "step": 7207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.164728887356236,
+      "learning_rate": 7.766707140815716e-06,
+      "loss": 0.7204,
+      "step": 7208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8491757775953168,
+      "learning_rate": 7.766144657186306e-06,
+      "loss": 0.6263,
+      "step": 7209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7793465831212965,
+      "learning_rate": 7.765582126116563e-06,
+      "loss": 0.4932,
+      "step": 7210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4218850893821384,
+      "learning_rate": 7.765019547618269e-06,
+      "loss": 0.4326,
+      "step": 7211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6913297193747892,
+      "learning_rate": 7.764456921703196e-06,
+      "loss": 0.5633,
+      "step": 7212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8103414835138416,
+      "learning_rate": 7.763894248383122e-06,
+      "loss": 0.6967,
+      "step": 7213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7260830421914688,
+      "learning_rate": 7.76333152766983e-06,
+      "loss": 0.571,
+      "step": 7214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8974045338251326,
+      "learning_rate": 7.762768759575098e-06,
+      "loss": 0.6518,
+      "step": 7215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0342047427078866,
+      "learning_rate": 7.762205944110706e-06,
+      "loss": 0.632,
+      "step": 7216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0836875445653766,
+      "learning_rate": 7.761643081288439e-06,
+      "loss": 0.6498,
+      "step": 7217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2664074331990394,
+      "learning_rate": 7.761080171120079e-06,
+      "loss": 0.5327,
+      "step": 7218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.030450326007225,
+      "learning_rate": 7.760517213617411e-06,
+      "loss": 0.6426,
+      "step": 7219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.103171789331107,
+      "learning_rate": 7.759954208792218e-06,
+      "loss": 0.6124,
+      "step": 7220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.592886013336181,
+      "learning_rate": 7.759391156656288e-06,
+      "loss": 0.5838,
+      "step": 7221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9582278417639496,
+      "learning_rate": 7.758828057221407e-06,
+      "loss": 0.5364,
+      "step": 7222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6296989121738084,
+      "learning_rate": 7.758264910499362e-06,
+      "loss": 0.5903,
+      "step": 7223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.299001060830947,
+      "learning_rate": 7.757701716501944e-06,
+      "loss": 0.6541,
+      "step": 7224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5815985272029482,
+      "learning_rate": 7.757138475240945e-06,
+      "loss": 0.4386,
+      "step": 7225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7415979821422372,
+      "learning_rate": 7.756575186728152e-06,
+      "loss": 0.4956,
+      "step": 7226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5154334187870175,
+      "learning_rate": 7.75601185097536e-06,
+      "loss": 0.4708,
+      "step": 7227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.419176090233069,
+      "learning_rate": 7.755448467994362e-06,
+      "loss": 0.5304,
+      "step": 7228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.865356643433594,
+      "learning_rate": 7.75488503779695e-06,
+      "loss": 0.5238,
+      "step": 7229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.936218973559792,
+      "learning_rate": 7.754321560394921e-06,
+      "loss": 0.5756,
+      "step": 7230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8671999720242671,
+      "learning_rate": 7.753758035800071e-06,
+      "loss": 0.4898,
+      "step": 7231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9736033045491626,
+      "learning_rate": 7.753194464024195e-06,
+      "loss": 0.5533,
+      "step": 7232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7957211491396523,
+      "learning_rate": 7.752630845079093e-06,
+      "loss": 0.4707,
+      "step": 7233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4767366788935405,
+      "learning_rate": 7.752067178976564e-06,
+      "loss": 0.3688,
+      "step": 7234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.47126915070419,
+      "learning_rate": 7.751503465728406e-06,
+      "loss": 0.4386,
+      "step": 7235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8660917267776915,
+      "learning_rate": 7.750939705346421e-06,
+      "loss": 0.586,
+      "step": 7236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0965774447039878,
+      "learning_rate": 7.750375897842415e-06,
+      "loss": 0.5522,
+      "step": 7237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4148807592546604,
+      "learning_rate": 7.749812043228184e-06,
+      "loss": 0.3463,
+      "step": 7238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.925709597846325,
+      "learning_rate": 7.749248141515536e-06,
+      "loss": 0.4489,
+      "step": 7239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.273849418018363,
+      "learning_rate": 7.748684192716274e-06,
+      "loss": 0.541,
+      "step": 7240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1158658168538307,
+      "learning_rate": 7.748120196842208e-06,
+      "loss": 0.559,
+      "step": 7241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7412804062749416,
+      "learning_rate": 7.747556153905141e-06,
+      "loss": 0.4409,
+      "step": 7242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1391045994117173,
+      "learning_rate": 7.746992063916881e-06,
+      "loss": 0.5903,
+      "step": 7243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8558567067093037,
+      "learning_rate": 7.746427926889238e-06,
+      "loss": 0.6626,
+      "step": 7244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9322404230233032,
+      "learning_rate": 7.745863742834022e-06,
+      "loss": 0.4843,
+      "step": 7245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8735955129640496,
+      "learning_rate": 7.745299511763044e-06,
+      "loss": 0.6332,
+      "step": 7246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5886947314742605,
+      "learning_rate": 7.744735233688114e-06,
+      "loss": 0.518,
+      "step": 7247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7060030020014842,
+      "learning_rate": 7.744170908621047e-06,
+      "loss": 0.4895,
+      "step": 7248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8190647711432237,
+      "learning_rate": 7.743606536573657e-06,
+      "loss": 0.536,
+      "step": 7249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.045439722076662,
+      "learning_rate": 7.743042117557755e-06,
+      "loss": 0.6985,
+      "step": 7250
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5527409315109253,
+      "eval_runtime": 67.8282,
+      "eval_samples_per_second": 50.274,
+      "eval_steps_per_second": 1.578,
+      "step": 7250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8430224689765216,
+      "learning_rate": 7.742477651585161e-06,
+      "loss": 0.3769,
+      "step": 7251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7646626694020726,
+      "learning_rate": 7.741913138667691e-06,
+      "loss": 0.5481,
+      "step": 7252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.825253745869107,
+      "learning_rate": 7.74134857881716e-06,
+      "loss": 0.5742,
+      "step": 7253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.824973780989188,
+      "learning_rate": 7.74078397204539e-06,
+      "loss": 0.4692,
+      "step": 7254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.236937561224019,
+      "learning_rate": 7.7402193183642e-06,
+      "loss": 0.6304,
+      "step": 7255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7871694856705507,
+      "learning_rate": 7.739654617785409e-06,
+      "loss": 0.6569,
+      "step": 7256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4552774527917378,
+      "learning_rate": 7.73908987032084e-06,
+      "loss": 0.4878,
+      "step": 7257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7225435772001951,
+      "learning_rate": 7.738525075982315e-06,
+      "loss": 0.4586,
+      "step": 7258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6723024316269743,
+      "learning_rate": 7.737960234781658e-06,
+      "loss": 0.5723,
+      "step": 7259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5260592540095639,
+      "learning_rate": 7.737395346730693e-06,
+      "loss": 0.4334,
+      "step": 7260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2701613460146235,
+      "learning_rate": 7.736830411841248e-06,
+      "loss": 0.5938,
+      "step": 7261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7835580784707536,
+      "learning_rate": 7.736265430125147e-06,
+      "loss": 0.5108,
+      "step": 7262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6947405275171588,
+      "learning_rate": 7.735700401594217e-06,
+      "loss": 0.4204,
+      "step": 7263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.532816309560225,
+      "learning_rate": 7.735135326260288e-06,
+      "loss": 0.4649,
+      "step": 7264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.669590889279364,
+      "learning_rate": 7.73457020413519e-06,
+      "loss": 0.5247,
+      "step": 7265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.033113887075141,
+      "learning_rate": 7.734005035230751e-06,
+      "loss": 0.5689,
+      "step": 7266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.175084011435151,
+      "learning_rate": 7.733439819558806e-06,
+      "loss": 0.5255,
+      "step": 7267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8166931189289657,
+      "learning_rate": 7.732874557131187e-06,
+      "loss": 0.6633,
+      "step": 7268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8307645055655977,
+      "learning_rate": 7.732309247959724e-06,
+      "loss": 0.4949,
+      "step": 7269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2198551009940197,
+      "learning_rate": 7.731743892056254e-06,
+      "loss": 0.6142,
+      "step": 7270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.519710958241298,
+      "learning_rate": 7.731178489432611e-06,
+      "loss": 0.6927,
+      "step": 7271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.684107866951755,
+      "learning_rate": 7.730613040100633e-06,
+      "loss": 0.6016,
+      "step": 7272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.408057484812123,
+      "learning_rate": 7.730047544072156e-06,
+      "loss": 0.6756,
+      "step": 7273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5621723512393075,
+      "learning_rate": 7.729482001359019e-06,
+      "loss": 0.4701,
+      "step": 7274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.766746967979797,
+      "learning_rate": 7.72891641197306e-06,
+      "loss": 0.4803,
+      "step": 7275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6821284823290374,
+      "learning_rate": 7.728350775926122e-06,
+      "loss": 0.5328,
+      "step": 7276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8698158915855916,
+      "learning_rate": 7.727785093230043e-06,
+      "loss": 0.5444,
+      "step": 7277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8457042545172344,
+      "learning_rate": 7.727219363896669e-06,
+      "loss": 0.4596,
+      "step": 7278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.558754916806558,
+      "learning_rate": 7.726653587937839e-06,
+      "loss": 0.5987,
+      "step": 7279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.13577579653242,
+      "learning_rate": 7.7260877653654e-06,
+      "loss": 0.6674,
+      "step": 7280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2148686668148803,
+      "learning_rate": 7.725521896191195e-06,
+      "loss": 0.7721,
+      "step": 7281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7176975175487035,
+      "learning_rate": 7.724955980427072e-06,
+      "loss": 0.4812,
+      "step": 7282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.767852422832296,
+      "learning_rate": 7.72439001808488e-06,
+      "loss": 0.49,
+      "step": 7283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.087254249503606,
+      "learning_rate": 7.723824009176463e-06,
+      "loss": 0.598,
+      "step": 7284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9501358179439219,
+      "learning_rate": 7.72325795371367e-06,
+      "loss": 0.5811,
+      "step": 7285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8901061857821262,
+      "learning_rate": 7.722691851708354e-06,
+      "loss": 0.4965,
+      "step": 7286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.384195480185117,
+      "learning_rate": 7.722125703172365e-06,
+      "loss": 0.5916,
+      "step": 7287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.595446021107702,
+      "learning_rate": 7.721559508117555e-06,
+      "loss": 0.7022,
+      "step": 7288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8247839065674825,
+      "learning_rate": 7.720993266555777e-06,
+      "loss": 0.3946,
+      "step": 7289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8185682167176018,
+      "learning_rate": 7.720426978498882e-06,
+      "loss": 0.5981,
+      "step": 7290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.556114404672422,
+      "learning_rate": 7.71986064395873e-06,
+      "loss": 0.4902,
+      "step": 7291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7259424266879355,
+      "learning_rate": 7.719294262947171e-06,
+      "loss": 0.5605,
+      "step": 7292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7978624581533291,
+      "learning_rate": 7.718727835476067e-06,
+      "loss": 0.6279,
+      "step": 7293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7647600812243252,
+      "learning_rate": 7.718161361557274e-06,
+      "loss": 0.5781,
+      "step": 7294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6504046247335098,
+      "learning_rate": 7.71759484120265e-06,
+      "loss": 0.5483,
+      "step": 7295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7287817057547865,
+      "learning_rate": 7.717028274424056e-06,
+      "loss": 0.5528,
+      "step": 7296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8997304516575029,
+      "learning_rate": 7.716461661233353e-06,
+      "loss": 0.4727,
+      "step": 7297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9691526921486648,
+      "learning_rate": 7.7158950016424e-06,
+      "loss": 0.5842,
+      "step": 7298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.055184807340354,
+      "learning_rate": 7.715328295663064e-06,
+      "loss": 0.5316,
+      "step": 7299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9575935450966147,
+      "learning_rate": 7.714761543307203e-06,
+      "loss": 0.528,
+      "step": 7300
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.5518983602523804,
+      "eval_runtime": 67.8938,
+      "eval_samples_per_second": 50.226,
+      "eval_steps_per_second": 1.576,
+      "step": 7300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9562417044811662,
+      "learning_rate": 7.714194744586686e-06,
+      "loss": 0.5838,
+      "step": 7301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.938975487344414,
+      "learning_rate": 7.713627899513378e-06,
+      "loss": 0.5882,
+      "step": 7302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7557581515429894,
+      "learning_rate": 7.713061008099143e-06,
+      "loss": 0.4955,
+      "step": 7303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2404286438212893,
+      "learning_rate": 7.712494070355852e-06,
+      "loss": 0.558,
+      "step": 7304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.976278832631872,
+      "learning_rate": 7.711927086295371e-06,
+      "loss": 0.5572,
+      "step": 7305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8148105681038704,
+      "learning_rate": 7.711360055929571e-06,
+      "loss": 0.5509,
+      "step": 7306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.330960882710846,
+      "learning_rate": 7.710792979270322e-06,
+      "loss": 0.6506,
+      "step": 7307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.761168688789674,
+      "learning_rate": 7.710225856329496e-06,
+      "loss": 0.6248,
+      "step": 7308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.892467575548496,
+      "learning_rate": 7.709658687118963e-06,
+      "loss": 0.6073,
+      "step": 7309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.638257936781361,
+      "learning_rate": 7.709091471650599e-06,
+      "loss": 0.5832,
+      "step": 7310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8138587970024003,
+      "learning_rate": 7.708524209936276e-06,
+      "loss": 0.6519,
+      "step": 7311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6484443685465877,
+      "learning_rate": 7.70795690198787e-06,
+      "loss": 0.5911,
+      "step": 7312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6186965885813742,
+      "learning_rate": 7.70738954781726e-06,
+      "loss": 0.4298,
+      "step": 7313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.946847927042801,
+      "learning_rate": 7.706822147436317e-06,
+      "loss": 0.5083,
+      "step": 7314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.514924260608713,
+      "learning_rate": 7.706254700856926e-06,
+      "loss": 0.3962,
+      "step": 7315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8168835400876278,
+      "learning_rate": 7.705687208090961e-06,
+      "loss": 0.54,
+      "step": 7316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7360484226107866,
+      "learning_rate": 7.705119669150308e-06,
+      "loss": 0.6201,
+      "step": 7317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8058337842189962,
+      "learning_rate": 7.704552084046842e-06,
+      "loss": 0.5907,
+      "step": 7318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1052436749769634,
+      "learning_rate": 7.703984452792449e-06,
+      "loss": 0.7255,
+      "step": 7319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8694273394742567,
+      "learning_rate": 7.703416775399007e-06,
+      "loss": 0.4504,
+      "step": 7320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.382602418888228,
+      "learning_rate": 7.702849051878406e-06,
+      "loss": 0.4145,
+      "step": 7321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9230867122598811,
+      "learning_rate": 7.702281282242527e-06,
+      "loss": 0.6378,
+      "step": 7322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8653228301152551,
+      "learning_rate": 7.701713466503259e-06,
+      "loss": 0.5513,
+      "step": 7323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7581330127037988,
+      "learning_rate": 7.701145604672487e-06,
+      "loss": 0.5125,
+      "step": 7324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9003323124996871,
+      "learning_rate": 7.700577696762097e-06,
+      "loss": 0.6172,
+      "step": 7325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4265849195674294,
+      "learning_rate": 7.700009742783981e-06,
+      "loss": 0.4061,
+      "step": 7326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.036043310131912,
+      "learning_rate": 7.699441742750029e-06,
+      "loss": 0.5205,
+      "step": 7327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5150276246141705,
+      "learning_rate": 7.698873696672127e-06,
+      "loss": 0.515,
+      "step": 7328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9555839822239847,
+      "learning_rate": 7.698305604562172e-06,
+      "loss": 0.5146,
+      "step": 7329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.213411736199904,
+      "learning_rate": 7.697737466432054e-06,
+      "loss": 0.5228,
+      "step": 7330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.417418122764745,
+      "learning_rate": 7.697169282293669e-06,
+      "loss": 0.6805,
+      "step": 7331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3817118272030793,
+      "learning_rate": 7.696601052158906e-06,
+      "loss": 0.3987,
+      "step": 7332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0177795636129203,
+      "learning_rate": 7.696032776039667e-06,
+      "loss": 0.6373,
+      "step": 7333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7657413134778979,
+      "learning_rate": 7.695464453947846e-06,
+      "loss": 0.4107,
+      "step": 7334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6928258417281976,
+      "learning_rate": 7.69489608589534e-06,
+      "loss": 0.4665,
+      "step": 7335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7463194895410896,
+      "learning_rate": 7.694327671894046e-06,
+      "loss": 0.7175,
+      "step": 7336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.056160594912056,
+      "learning_rate": 7.693759211955867e-06,
+      "loss": 0.4406,
+      "step": 7337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1248812497753655,
+      "learning_rate": 7.6931907060927e-06,
+      "loss": 0.542,
+      "step": 7338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0224672160917025,
+      "learning_rate": 7.692622154316448e-06,
+      "loss": 0.4655,
+      "step": 7339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0674049855202266,
+      "learning_rate": 7.692053556639016e-06,
+      "loss": 0.6262,
+      "step": 7340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7884670981356363,
+      "learning_rate": 7.691484913072302e-06,
+      "loss": 0.6033,
+      "step": 7341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9555578799567128,
+      "learning_rate": 7.690916223628213e-06,
+      "loss": 0.6006,
+      "step": 7342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9187824799745627,
+      "learning_rate": 7.690347488318653e-06,
+      "loss": 0.5678,
+      "step": 7343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7543755525416975,
+      "learning_rate": 7.68977870715553e-06,
+      "loss": 0.6057,
+      "step": 7344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6796319643732247,
+      "learning_rate": 7.689209880150747e-06,
+      "loss": 0.4393,
+      "step": 7345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.29618804459458,
+      "learning_rate": 7.688641007316218e-06,
+      "loss": 0.6333,
+      "step": 7346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1259134107399666,
+      "learning_rate": 7.688072088663848e-06,
+      "loss": 0.6443,
+      "step": 7347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5455728420215464,
+      "learning_rate": 7.687503124205547e-06,
+      "loss": 0.4667,
+      "step": 7348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.946440842433187,
+      "learning_rate": 7.686934113953228e-06,
+      "loss": 0.6957,
+      "step": 7349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.29702982414239,
+      "learning_rate": 7.686365057918802e-06,
+      "loss": 0.5987,
+      "step": 7350
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.5511100888252258,
+      "eval_runtime": 67.8882,
+      "eval_samples_per_second": 50.23,
+      "eval_steps_per_second": 1.576,
+      "step": 7350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.879861457053756,
+      "learning_rate": 7.68579595611418e-06,
+      "loss": 0.4643,
+      "step": 7351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.988943809750324,
+      "learning_rate": 7.685226808551278e-06,
+      "loss": 0.4866,
+      "step": 7352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.801226254333176,
+      "learning_rate": 7.684657615242012e-06,
+      "loss": 0.5011,
+      "step": 7353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9882275296510814,
+      "learning_rate": 7.684088376198293e-06,
+      "loss": 0.5919,
+      "step": 7354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2268696718906082,
+      "learning_rate": 7.683519091432042e-06,
+      "loss": 0.5587,
+      "step": 7355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9697568687161284,
+      "learning_rate": 7.682949760955173e-06,
+      "loss": 0.6068,
+      "step": 7356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7503301847815333,
+      "learning_rate": 7.68238038477961e-06,
+      "loss": 0.5202,
+      "step": 7357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9597476358589987,
+      "learning_rate": 7.681810962917267e-06,
+      "loss": 0.676,
+      "step": 7358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4956859967312555,
+      "learning_rate": 7.681241495380066e-06,
+      "loss": 0.4857,
+      "step": 7359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9695262263351414,
+      "learning_rate": 7.68067198217993e-06,
+      "loss": 0.5519,
+      "step": 7360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9853390936525315,
+      "learning_rate": 7.68010242332878e-06,
+      "loss": 0.449,
+      "step": 7361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5447535858145285,
+      "learning_rate": 7.679532818838542e-06,
+      "loss": 0.4133,
+      "step": 7362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7091551325816456,
+      "learning_rate": 7.678963168721136e-06,
+      "loss": 0.533,
+      "step": 7363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.180486842136374,
+      "learning_rate": 7.67839347298849e-06,
+      "loss": 0.7234,
+      "step": 7364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8131187718003647,
+      "learning_rate": 7.67782373165253e-06,
+      "loss": 0.5413,
+      "step": 7365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9649708304905613,
+      "learning_rate": 7.677253944725182e-06,
+      "loss": 0.6016,
+      "step": 7366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6876193566392053,
+      "learning_rate": 7.676684112218376e-06,
+      "loss": 0.5727,
+      "step": 7367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.707985016206962,
+      "learning_rate": 7.676114234144041e-06,
+      "loss": 0.5083,
+      "step": 7368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2488527729645797,
+      "learning_rate": 7.675544310514103e-06,
+      "loss": 0.7564,
+      "step": 7369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.671879072305744,
+      "learning_rate": 7.674974341340499e-06,
+      "loss": 0.5951,
+      "step": 7370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1402974167366517,
+      "learning_rate": 7.674404326635157e-06,
+      "loss": 0.672,
+      "step": 7371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0347920017565277,
+      "learning_rate": 7.673834266410011e-06,
+      "loss": 0.4979,
+      "step": 7372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8605329537362583,
+      "learning_rate": 7.673264160676996e-06,
+      "loss": 0.6074,
+      "step": 7373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0157290223075313,
+      "learning_rate": 7.672694009448043e-06,
+      "loss": 0.4526,
+      "step": 7374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7429514998859963,
+      "learning_rate": 7.672123812735093e-06,
+      "loss": 0.5706,
+      "step": 7375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5000831519350029,
+      "learning_rate": 7.67155357055008e-06,
+      "loss": 0.4391,
+      "step": 7376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.664534426599903,
+      "learning_rate": 7.67098328290494e-06,
+      "loss": 0.4492,
+      "step": 7377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7349039910650956,
+      "learning_rate": 7.670412949811615e-06,
+      "loss": 0.5914,
+      "step": 7378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7428056153809393,
+      "learning_rate": 7.669842571282043e-06,
+      "loss": 0.5212,
+      "step": 7379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4726675339405197,
+      "learning_rate": 7.669272147328162e-06,
+      "loss": 0.6563,
+      "step": 7380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.016949489101734,
+      "learning_rate": 7.66870167796192e-06,
+      "loss": 0.7649,
+      "step": 7381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9491452533509879,
+      "learning_rate": 7.668131163195253e-06,
+      "loss": 0.4312,
+      "step": 7382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4740482857996733,
+      "learning_rate": 7.667560603040106e-06,
+      "loss": 0.4728,
+      "step": 7383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5924913075141596,
+      "learning_rate": 7.666989997508424e-06,
+      "loss": 0.5084,
+      "step": 7384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9187170010363588,
+      "learning_rate": 7.666419346612155e-06,
+      "loss": 0.4904,
+      "step": 7385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7897045714802524,
+      "learning_rate": 7.665848650363242e-06,
+      "loss": 0.5018,
+      "step": 7386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8174014978240638,
+      "learning_rate": 7.665277908773632e-06,
+      "loss": 0.5525,
+      "step": 7387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.615939053726187,
+      "learning_rate": 7.664707121855273e-06,
+      "loss": 0.481,
+      "step": 7388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9985649309361044,
+      "learning_rate": 7.664136289620116e-06,
+      "loss": 0.5617,
+      "step": 7389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0970940447895714,
+      "learning_rate": 7.663565412080109e-06,
+      "loss": 0.4503,
+      "step": 7390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6619393100854856,
+      "learning_rate": 7.662994489247206e-06,
+      "loss": 0.3709,
+      "step": 7391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3710801137724948,
+      "learning_rate": 7.662423521133356e-06,
+      "loss": 0.3852,
+      "step": 7392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4550834180986054,
+      "learning_rate": 7.661852507750511e-06,
+      "loss": 0.4191,
+      "step": 7393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9112856023435172,
+      "learning_rate": 7.661281449110628e-06,
+      "loss": 0.6568,
+      "step": 7394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7867202984163013,
+      "learning_rate": 7.660710345225662e-06,
+      "loss": 0.5376,
+      "step": 7395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.362467392282493,
+      "learning_rate": 7.660139196107563e-06,
+      "loss": 0.603,
+      "step": 7396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7312222916295166,
+      "learning_rate": 7.659568001768294e-06,
+      "loss": 0.4056,
+      "step": 7397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.904939383236244,
+      "learning_rate": 7.65899676221981e-06,
+      "loss": 0.4929,
+      "step": 7398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0005854456403998,
+      "learning_rate": 7.65842547747407e-06,
+      "loss": 0.5948,
+      "step": 7399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.021013145037993,
+      "learning_rate": 7.657854147543032e-06,
+      "loss": 0.5525,
+      "step": 7400
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.5500914454460144,
+      "eval_runtime": 67.894,
+      "eval_samples_per_second": 50.225,
+      "eval_steps_per_second": 1.576,
+      "step": 7400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.08856821356341,
+      "learning_rate": 7.657282772438658e-06,
+      "loss": 0.5858,
+      "step": 7401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.079411724677794,
+      "learning_rate": 7.65671135217291e-06,
+      "loss": 0.5295,
+      "step": 7402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7463540892009033,
+      "learning_rate": 7.65613988675775e-06,
+      "loss": 0.606,
+      "step": 7403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.868100073512865,
+      "learning_rate": 7.655568376205141e-06,
+      "loss": 0.5281,
+      "step": 7404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0246057538262328,
+      "learning_rate": 7.654996820527045e-06,
+      "loss": 0.5239,
+      "step": 7405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.608191934180762,
+      "learning_rate": 7.65442521973543e-06,
+      "loss": 0.4145,
+      "step": 7406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9223481893783276,
+      "learning_rate": 7.65385357384226e-06,
+      "loss": 0.6379,
+      "step": 7407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.781058359505532,
+      "learning_rate": 7.653281882859505e-06,
+      "loss": 0.463,
+      "step": 7408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8608180373881407,
+      "learning_rate": 7.65271014679913e-06,
+      "loss": 0.5481,
+      "step": 7409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.671644696001214,
+      "learning_rate": 7.652138365673107e-06,
+      "loss": 0.5291,
+      "step": 7410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8910607710941842,
+      "learning_rate": 7.651566539493402e-06,
+      "loss": 0.543,
+      "step": 7411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6377281299425723,
+      "learning_rate": 7.650994668271989e-06,
+      "loss": 0.5273,
+      "step": 7412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8023200319613732,
+      "learning_rate": 7.650422752020839e-06,
+      "loss": 0.6226,
+      "step": 7413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9272658380094627,
+      "learning_rate": 7.649850790751922e-06,
+      "loss": 0.4138,
+      "step": 7414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.670305908973559,
+      "learning_rate": 7.649278784477216e-06,
+      "loss": 0.4382,
+      "step": 7415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7946327496730936,
+      "learning_rate": 7.648706733208695e-06,
+      "loss": 0.445,
+      "step": 7416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5989043319058913,
+      "learning_rate": 7.648134636958331e-06,
+      "loss": 0.5328,
+      "step": 7417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.570343299064443,
+      "learning_rate": 7.647562495738103e-06,
+      "loss": 0.531,
+      "step": 7418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.183883912883561,
+      "learning_rate": 7.646990309559989e-06,
+      "loss": 0.593,
+      "step": 7419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6596323085847324,
+      "learning_rate": 7.646418078435964e-06,
+      "loss": 0.4821,
+      "step": 7420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5546496790803261,
+      "learning_rate": 7.645845802378013e-06,
+      "loss": 0.3903,
+      "step": 7421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0873934288723244,
+      "learning_rate": 7.64527348139811e-06,
+      "loss": 0.6029,
+      "step": 7422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.729039562500698,
+      "learning_rate": 7.644701115508239e-06,
+      "loss": 0.7502,
+      "step": 7423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.022251293797468,
+      "learning_rate": 7.644128704720386e-06,
+      "loss": 0.5465,
+      "step": 7424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.575545610706034,
+      "learning_rate": 7.643556249046527e-06,
+      "loss": 0.411,
+      "step": 7425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8315184852354263,
+      "learning_rate": 7.64298374849865e-06,
+      "loss": 0.644,
+      "step": 7426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8771569781372075,
+      "learning_rate": 7.642411203088741e-06,
+      "loss": 0.7093,
+      "step": 7427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4675677651652863,
+      "learning_rate": 7.641838612828782e-06,
+      "loss": 0.3565,
+      "step": 7428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0653543036326587,
+      "learning_rate": 7.641265977730763e-06,
+      "loss": 0.5377,
+      "step": 7429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.913440009007694,
+      "learning_rate": 7.640693297806671e-06,
+      "loss": 0.4851,
+      "step": 7430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7808527699641694,
+      "learning_rate": 7.640120573068492e-06,
+      "loss": 0.5731,
+      "step": 7431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4999897669222038,
+      "learning_rate": 7.639547803528223e-06,
+      "loss": 0.5,
+      "step": 7432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8159945019170705,
+      "learning_rate": 7.638974989197846e-06,
+      "loss": 0.5483,
+      "step": 7433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.192674171667968,
+      "learning_rate": 7.638402130089356e-06,
+      "loss": 0.5514,
+      "step": 7434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.775811422409136,
+      "learning_rate": 7.637829226214747e-06,
+      "loss": 0.6613,
+      "step": 7435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5652348517035581,
+      "learning_rate": 7.63725627758601e-06,
+      "loss": 0.4265,
+      "step": 7436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.817060050400747,
+      "learning_rate": 7.636683284215142e-06,
+      "loss": 0.5497,
+      "step": 7437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.00737423923293,
+      "learning_rate": 7.636110246114134e-06,
+      "loss": 0.5269,
+      "step": 7438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5680713883508859,
+      "learning_rate": 7.635537163294987e-06,
+      "loss": 0.4923,
+      "step": 7439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6201936292767791,
+      "learning_rate": 7.634964035769695e-06,
+      "loss": 0.6289,
+      "step": 7440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8475090988010328,
+      "learning_rate": 7.634390863550258e-06,
+      "loss": 0.5589,
+      "step": 7441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6166400846394564,
+      "learning_rate": 7.633817646648672e-06,
+      "loss": 0.375,
+      "step": 7442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9403955434152749,
+      "learning_rate": 7.633244385076942e-06,
+      "loss": 0.714,
+      "step": 7443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.201633079919469,
+      "learning_rate": 7.632671078847061e-06,
+      "loss": 0.7046,
+      "step": 7444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0448834009681,
+      "learning_rate": 7.63209772797104e-06,
+      "loss": 0.5811,
+      "step": 7445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6851929439970625,
+      "learning_rate": 7.631524332460876e-06,
+      "loss": 0.4603,
+      "step": 7446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7751104777305515,
+      "learning_rate": 7.630950892328573e-06,
+      "loss": 0.5733,
+      "step": 7447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7746980138503925,
+      "learning_rate": 7.630377407586138e-06,
+      "loss": 0.5873,
+      "step": 7448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.61141892206903,
+      "learning_rate": 7.629803878245574e-06,
+      "loss": 0.4247,
+      "step": 7449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.110058095715799,
+      "learning_rate": 7.629230304318889e-06,
+      "loss": 0.4871,
+      "step": 7450
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.5485899448394775,
+      "eval_runtime": 67.9389,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 7450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4326256313838117,
+      "learning_rate": 7.628656685818091e-06,
+      "loss": 0.6201,
+      "step": 7451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.075644338685021,
+      "learning_rate": 7.628083022755186e-06,
+      "loss": 0.668,
+      "step": 7452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.97842374743842,
+      "learning_rate": 7.6275093151421874e-06,
+      "loss": 0.6993,
+      "step": 7453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4634767535839073,
+      "learning_rate": 7.626935562991099e-06,
+      "loss": 0.496,
+      "step": 7454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7145544421552954,
+      "learning_rate": 7.626361766313939e-06,
+      "loss": 0.4788,
+      "step": 7455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6664768959087917,
+      "learning_rate": 7.625787925122715e-06,
+      "loss": 0.4532,
+      "step": 7456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9385231791043551,
+      "learning_rate": 7.625214039429441e-06,
+      "loss": 0.6579,
+      "step": 7457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9128081729604685,
+      "learning_rate": 7.624640109246131e-06,
+      "loss": 0.5753,
+      "step": 7458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5198703521053438,
+      "learning_rate": 7.6240661345848e-06,
+      "loss": 0.5444,
+      "step": 7459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.727292543375806,
+      "learning_rate": 7.623492115457464e-06,
+      "loss": 0.5022,
+      "step": 7460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9355058046503575,
+      "learning_rate": 7.622918051876139e-06,
+      "loss": 0.5679,
+      "step": 7461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.959083079763516,
+      "learning_rate": 7.6223439438528424e-06,
+      "loss": 0.7955,
+      "step": 7462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7060029589939092,
+      "learning_rate": 7.621769791399596e-06,
+      "loss": 0.5895,
+      "step": 7463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5714413129714273,
+      "learning_rate": 7.6211955945284155e-06,
+      "loss": 0.5003,
+      "step": 7464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9186611098972333,
+      "learning_rate": 7.620621353251322e-06,
+      "loss": 0.5387,
+      "step": 7465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.369193116091247,
+      "learning_rate": 7.620047067580339e-06,
+      "loss": 0.6196,
+      "step": 7466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6867032943944114,
+      "learning_rate": 7.6194727375274865e-06,
+      "loss": 0.4327,
+      "step": 7467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.069398721486849,
+      "learning_rate": 7.61889836310479e-06,
+      "loss": 0.5015,
+      "step": 7468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1235197549945033,
+      "learning_rate": 7.618323944324272e-06,
+      "loss": 0.7032,
+      "step": 7469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6674080568010818,
+      "learning_rate": 7.617749481197956e-06,
+      "loss": 0.4678,
+      "step": 7470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.933289687351902,
+      "learning_rate": 7.617174973737872e-06,
+      "loss": 0.4226,
+      "step": 7471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2028783857094782,
+      "learning_rate": 7.616600421956044e-06,
+      "loss": 0.5475,
+      "step": 7472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8940765550051277,
+      "learning_rate": 7.616025825864502e-06,
+      "loss": 0.5416,
+      "step": 7473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9488318659935244,
+      "learning_rate": 7.615451185475272e-06,
+      "loss": 0.7093,
+      "step": 7474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5112057420679241,
+      "learning_rate": 7.6148765008003864e-06,
+      "loss": 0.4117,
+      "step": 7475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6017481386400916,
+      "learning_rate": 7.6143017718518744e-06,
+      "loss": 0.5881,
+      "step": 7476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6572056681935052,
+      "learning_rate": 7.613726998641769e-06,
+      "loss": 0.5929,
+      "step": 7477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.887956717402692,
+      "learning_rate": 7.6131521811821005e-06,
+      "loss": 0.601,
+      "step": 7478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.888139472095419,
+      "learning_rate": 7.612577319484905e-06,
+      "loss": 0.5335,
+      "step": 7479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0382963377085392,
+      "learning_rate": 7.612002413562215e-06,
+      "loss": 0.6923,
+      "step": 7480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6309536498436712,
+      "learning_rate": 7.611427463426065e-06,
+      "loss": 0.4714,
+      "step": 7481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.63277034663724,
+      "learning_rate": 7.610852469088494e-06,
+      "loss": 0.5351,
+      "step": 7482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8606171747241054,
+      "learning_rate": 7.610277430561538e-06,
+      "loss": 0.5024,
+      "step": 7483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8441713969544455,
+      "learning_rate": 7.609702347857233e-06,
+      "loss": 0.5526,
+      "step": 7484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.900902111547394,
+      "learning_rate": 7.609127220987621e-06,
+      "loss": 0.5238,
+      "step": 7485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9674712383018338,
+      "learning_rate": 7.60855204996474e-06,
+      "loss": 0.3494,
+      "step": 7486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9095083840103495,
+      "learning_rate": 7.607976834800632e-06,
+      "loss": 0.5137,
+      "step": 7487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6660199308118477,
+      "learning_rate": 7.6074015755073385e-06,
+      "loss": 0.4043,
+      "step": 7488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5526173890729669,
+      "learning_rate": 7.6068262720969024e-06,
+      "loss": 0.3843,
+      "step": 7489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.142486665696338,
+      "learning_rate": 7.606250924581365e-06,
+      "loss": 0.6494,
+      "step": 7490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4959524497993486,
+      "learning_rate": 7.605675532972773e-06,
+      "loss": 0.322,
+      "step": 7491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8617643019701613,
+      "learning_rate": 7.605100097283173e-06,
+      "loss": 0.384,
+      "step": 7492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6539202121892111,
+      "learning_rate": 7.604524617524609e-06,
+      "loss": 0.5547,
+      "step": 7493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.936019933234175,
+      "learning_rate": 7.6039490937091295e-06,
+      "loss": 0.4948,
+      "step": 7494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0106392918851417,
+      "learning_rate": 7.603373525848784e-06,
+      "loss": 0.6244,
+      "step": 7495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6033470015996614,
+      "learning_rate": 7.602797913955618e-06,
+      "loss": 0.4941,
+      "step": 7496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3936716936664357,
+      "learning_rate": 7.602222258041684e-06,
+      "loss": 0.6647,
+      "step": 7497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.028954030310124,
+      "learning_rate": 7.6016465581190334e-06,
+      "loss": 0.6436,
+      "step": 7498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3012396007086355,
+      "learning_rate": 7.601070814199718e-06,
+      "loss": 0.6394,
+      "step": 7499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0755682734819305,
+      "learning_rate": 7.60049502629579e-06,
+      "loss": 0.8035,
+      "step": 7500
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5483761429786682,
+      "eval_runtime": 67.9748,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 7500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0169592163114,
+      "learning_rate": 7.599919194419303e-06,
+      "loss": 0.6412,
+      "step": 7501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6913408291601566,
+      "learning_rate": 7.5993433185823115e-06,
+      "loss": 0.5519,
+      "step": 7502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6605422265048644,
+      "learning_rate": 7.598767398796872e-06,
+      "loss": 0.5436,
+      "step": 7503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.127368042475574,
+      "learning_rate": 7.598191435075041e-06,
+      "loss": 0.5638,
+      "step": 7504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.782213327405287,
+      "learning_rate": 7.597615427428875e-06,
+      "loss": 0.44,
+      "step": 7505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8069420860387035,
+      "learning_rate": 7.597039375870433e-06,
+      "loss": 0.5436,
+      "step": 7506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9863510306109846,
+      "learning_rate": 7.596463280411775e-06,
+      "loss": 0.5571,
+      "step": 7507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6379574640853944,
+      "learning_rate": 7.595887141064958e-06,
+      "loss": 0.4775,
+      "step": 7508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1372941892573323,
+      "learning_rate": 7.595310957842048e-06,
+      "loss": 0.5108,
+      "step": 7509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.081951653589493,
+      "learning_rate": 7.5947347307551035e-06,
+      "loss": 0.5502,
+      "step": 7510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6161478626987633,
+      "learning_rate": 7.5941584598161885e-06,
+      "loss": 0.4767,
+      "step": 7511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7132654404059857,
+      "learning_rate": 7.593582145037368e-06,
+      "loss": 0.4691,
+      "step": 7512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5969665340117716,
+      "learning_rate": 7.593005786430704e-06,
+      "loss": 0.4853,
+      "step": 7513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.054679863042207,
+      "learning_rate": 7.5924293840082656e-06,
+      "loss": 0.6706,
+      "step": 7514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8780490571620474,
+      "learning_rate": 7.591852937782115e-06,
+      "loss": 0.6807,
+      "step": 7515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7803006922823834,
+      "learning_rate": 7.591276447764324e-06,
+      "loss": 0.6049,
+      "step": 7516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.566314046326435,
+      "learning_rate": 7.590699913966961e-06,
+      "loss": 0.408,
+      "step": 7517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7768690953171882,
+      "learning_rate": 7.590123336402092e-06,
+      "loss": 0.7271,
+      "step": 7518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3081040477337478,
+      "learning_rate": 7.5895467150817904e-06,
+      "loss": 0.6951,
+      "step": 7519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8356217478021692,
+      "learning_rate": 7.588970050018124e-06,
+      "loss": 0.5265,
+      "step": 7520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8886736459330262,
+      "learning_rate": 7.588393341223168e-06,
+      "loss": 0.5425,
+      "step": 7521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5767830937525047,
+      "learning_rate": 7.587816588708996e-06,
+      "loss": 0.4816,
+      "step": 7522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5129787751510593,
+      "learning_rate": 7.58723979248768e-06,
+      "loss": 0.4366,
+      "step": 7523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.638378751492619,
+      "learning_rate": 7.586662952571294e-06,
+      "loss": 0.4053,
+      "step": 7524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.496482831403426,
+      "learning_rate": 7.586086068971918e-06,
+      "loss": 0.4008,
+      "step": 7525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8623723974949635,
+      "learning_rate": 7.5855091417016224e-06,
+      "loss": 0.5265,
+      "step": 7526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0481585594551803,
+      "learning_rate": 7.58493217077249e-06,
+      "loss": 0.5865,
+      "step": 7527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6879511165775443,
+      "learning_rate": 7.584355156196598e-06,
+      "loss": 0.5653,
+      "step": 7528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8248359644975296,
+      "learning_rate": 7.583778097986024e-06,
+      "loss": 0.4654,
+      "step": 7529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7073178504201008,
+      "learning_rate": 7.583200996152852e-06,
+      "loss": 0.4436,
+      "step": 7530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8744300942293912,
+      "learning_rate": 7.582623850709159e-06,
+      "loss": 0.685,
+      "step": 7531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1361967799980763,
+      "learning_rate": 7.582046661667031e-06,
+      "loss": 0.4981,
+      "step": 7532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8578523650475542,
+      "learning_rate": 7.581469429038549e-06,
+      "loss": 0.6467,
+      "step": 7533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6916842113125872,
+      "learning_rate": 7.580892152835796e-06,
+      "loss": 0.5856,
+      "step": 7534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8782355153021988,
+      "learning_rate": 7.580314833070859e-06,
+      "loss": 0.4233,
+      "step": 7535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9591772442179445,
+      "learning_rate": 7.579737469755823e-06,
+      "loss": 0.5728,
+      "step": 7536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7033447693587713,
+      "learning_rate": 7.579160062902775e-06,
+      "loss": 0.5221,
+      "step": 7537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.189008349861037,
+      "learning_rate": 7.578582612523802e-06,
+      "loss": 0.5636,
+      "step": 7538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7368941480673104,
+      "learning_rate": 7.578005118630994e-06,
+      "loss": 0.4721,
+      "step": 7539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.219380554571678,
+      "learning_rate": 7.5774275812364396e-06,
+      "loss": 0.5992,
+      "step": 7540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.819882493160625,
+      "learning_rate": 7.576850000352229e-06,
+      "loss": 0.602,
+      "step": 7541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7537975752194834,
+      "learning_rate": 7.576272375990452e-06,
+      "loss": 0.525,
+      "step": 7542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.256292232592902,
+      "learning_rate": 7.575694708163205e-06,
+      "loss": 0.6684,
+      "step": 7543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.577192736264312,
+      "learning_rate": 7.575116996882577e-06,
+      "loss": 0.6022,
+      "step": 7544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7178066829521712,
+      "learning_rate": 7.574539242160663e-06,
+      "loss": 0.4416,
+      "step": 7545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.806127057899388,
+      "learning_rate": 7.573961444009559e-06,
+      "loss": 0.5145,
+      "step": 7546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2788305162129665,
+      "learning_rate": 7.57338360244136e-06,
+      "loss": 0.2868,
+      "step": 7547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.603660793788932,
+      "learning_rate": 7.5728057174681645e-06,
+      "loss": 0.537,
+      "step": 7548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8473434220711786,
+      "learning_rate": 7.572227789102068e-06,
+      "loss": 0.6136,
+      "step": 7549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4866036516285988,
+      "learning_rate": 7.57164981735517e-06,
+      "loss": 0.4217,
+      "step": 7550
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5470616817474365,
+      "eval_runtime": 67.8834,
+      "eval_samples_per_second": 50.233,
+      "eval_steps_per_second": 1.576,
+      "step": 7550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9994443289704145,
+      "learning_rate": 7.57107180223957e-06,
+      "loss": 0.4735,
+      "step": 7551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0686169985313416,
+      "learning_rate": 7.5704937437673674e-06,
+      "loss": 0.495,
+      "step": 7552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.987979953722856,
+      "learning_rate": 7.5699156419506645e-06,
+      "loss": 0.4801,
+      "step": 7553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6248678232214182,
+      "learning_rate": 7.569337496801564e-06,
+      "loss": 0.4553,
+      "step": 7554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.346005020102088,
+      "learning_rate": 7.568759308332168e-06,
+      "loss": 0.3618,
+      "step": 7555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.610956313383435,
+      "learning_rate": 7.568181076554581e-06,
+      "loss": 0.6017,
+      "step": 7556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7126767959903086,
+      "learning_rate": 7.567602801480908e-06,
+      "loss": 0.6171,
+      "step": 7557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1837772892476455,
+      "learning_rate": 7.567024483123253e-06,
+      "loss": 0.7315,
+      "step": 7558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8025366935259057,
+      "learning_rate": 7.566446121493725e-06,
+      "loss": 0.4393,
+      "step": 7559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8112054892312135,
+      "learning_rate": 7.565867716604432e-06,
+      "loss": 0.4084,
+      "step": 7560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8387858241450876,
+      "learning_rate": 7.565289268467481e-06,
+      "loss": 0.5855,
+      "step": 7561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7043940525776966,
+      "learning_rate": 7.564710777094982e-06,
+      "loss": 0.5383,
+      "step": 7562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.448496123369532,
+      "learning_rate": 7.564132242499045e-06,
+      "loss": 0.6215,
+      "step": 7563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9833683959639998,
+      "learning_rate": 7.5635536646917815e-06,
+      "loss": 0.3612,
+      "step": 7564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8453922391463682,
+      "learning_rate": 7.562975043685305e-06,
+      "loss": 0.4956,
+      "step": 7565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5272048705605965,
+      "learning_rate": 7.562396379491727e-06,
+      "loss": 0.4115,
+      "step": 7566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5534233288789532,
+      "learning_rate": 7.561817672123161e-06,
+      "loss": 0.4571,
+      "step": 7567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.740538625743195,
+      "learning_rate": 7.5612389215917226e-06,
+      "loss": 0.5451,
+      "step": 7568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0255759414860344,
+      "learning_rate": 7.560660127909528e-06,
+      "loss": 0.5009,
+      "step": 7569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.316520229119269,
+      "learning_rate": 7.5600812910886935e-06,
+      "loss": 0.3795,
+      "step": 7570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5676500655483911,
+      "learning_rate": 7.559502411141335e-06,
+      "loss": 0.437,
+      "step": 7571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.636069608682799,
+      "learning_rate": 7.558923488079575e-06,
+      "loss": 0.5947,
+      "step": 7572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.814481365098872,
+      "learning_rate": 7.55834452191553e-06,
+      "loss": 0.6244,
+      "step": 7573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6085344964046155,
+      "learning_rate": 7.55776551266132e-06,
+      "loss": 0.5244,
+      "step": 7574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8368566332365766,
+      "learning_rate": 7.557186460329069e-06,
+      "loss": 0.6061,
+      "step": 7575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.910474233954225,
+      "learning_rate": 7.556607364930896e-06,
+      "loss": 0.5669,
+      "step": 7576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9162229487764744,
+      "learning_rate": 7.556028226478925e-06,
+      "loss": 0.6764,
+      "step": 7577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0960516309549795,
+      "learning_rate": 7.555449044985282e-06,
+      "loss": 0.5347,
+      "step": 7578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6816243033142406,
+      "learning_rate": 7.5548698204620875e-06,
+      "loss": 0.5271,
+      "step": 7579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8342777945296704,
+      "learning_rate": 7.5542905529214705e-06,
+      "loss": 0.6661,
+      "step": 7580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.149042039415481,
+      "learning_rate": 7.553711242375558e-06,
+      "loss": 0.453,
+      "step": 7581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.134499847182831,
+      "learning_rate": 7.553131888836475e-06,
+      "loss": 0.6796,
+      "step": 7582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7719928917988643,
+      "learning_rate": 7.552552492316352e-06,
+      "loss": 0.4947,
+      "step": 7583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5016855987088245,
+      "learning_rate": 7.551973052827315e-06,
+      "loss": 0.4309,
+      "step": 7584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2813307859016376,
+      "learning_rate": 7.5513935703815e-06,
+      "loss": 0.6408,
+      "step": 7585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9169860550251077,
+      "learning_rate": 7.550814044991033e-06,
+      "loss": 0.6271,
+      "step": 7586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7862073169081647,
+      "learning_rate": 7.550234476668047e-06,
+      "loss": 0.6277,
+      "step": 7587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0025817542876534,
+      "learning_rate": 7.5496548654246786e-06,
+      "loss": 0.6448,
+      "step": 7588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6681285597520175,
+      "learning_rate": 7.549075211273055e-06,
+      "loss": 0.6014,
+      "step": 7589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8842812250343441,
+      "learning_rate": 7.548495514225317e-06,
+      "loss": 0.6665,
+      "step": 7590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8815740727192423,
+      "learning_rate": 7.547915774293597e-06,
+      "loss": 0.4838,
+      "step": 7591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8163466563024264,
+      "learning_rate": 7.547335991490033e-06,
+      "loss": 0.643,
+      "step": 7592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5061103368688378,
+      "learning_rate": 7.5467561658267605e-06,
+      "loss": 0.3987,
+      "step": 7593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7154858137559257,
+      "learning_rate": 7.54617629731592e-06,
+      "loss": 0.5484,
+      "step": 7594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7741429337606671,
+      "learning_rate": 7.545596385969648e-06,
+      "loss": 0.5504,
+      "step": 7595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5985267168797463,
+      "learning_rate": 7.545016431800088e-06,
+      "loss": 0.5072,
+      "step": 7596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5952595267784253,
+      "learning_rate": 7.544436434819378e-06,
+      "loss": 0.437,
+      "step": 7597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3501785049409014,
+      "learning_rate": 7.543856395039662e-06,
+      "loss": 0.3951,
+      "step": 7598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.849683046493784,
+      "learning_rate": 7.543276312473081e-06,
+      "loss": 0.6697,
+      "step": 7599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9468657194249888,
+      "learning_rate": 7.54269618713178e-06,
+      "loss": 0.4895,
+      "step": 7600
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5470448136329651,
+      "eval_runtime": 68.0236,
+      "eval_samples_per_second": 50.13,
+      "eval_steps_per_second": 1.573,
+      "step": 7600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9987676332446778,
+      "learning_rate": 7.542116019027904e-06,
+      "loss": 0.574,
+      "step": 7601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6080200245340053,
+      "learning_rate": 7.541535808173597e-06,
+      "loss": 0.4742,
+      "step": 7602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7731570936417225,
+      "learning_rate": 7.540955554581005e-06,
+      "loss": 0.5842,
+      "step": 7603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5636721521078223,
+      "learning_rate": 7.540375258262277e-06,
+      "loss": 0.51,
+      "step": 7604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8523009271968114,
+      "learning_rate": 7.5397949192295604e-06,
+      "loss": 0.5944,
+      "step": 7605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5971429129421781,
+      "learning_rate": 7.539214537495003e-06,
+      "loss": 0.4312,
+      "step": 7606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0242582177755555,
+      "learning_rate": 7.5386341130707566e-06,
+      "loss": 0.676,
+      "step": 7607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.869623158005537,
+      "learning_rate": 7.538053645968971e-06,
+      "loss": 0.6549,
+      "step": 7608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.39916085466353,
+      "learning_rate": 7.537473136201799e-06,
+      "loss": 0.3703,
+      "step": 7609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9840947870702217,
+      "learning_rate": 7.536892583781392e-06,
+      "loss": 0.5553,
+      "step": 7610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.324743069704983,
+      "learning_rate": 7.536311988719904e-06,
+      "loss": 0.6146,
+      "step": 7611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4860922013708546,
+      "learning_rate": 7.535731351029489e-06,
+      "loss": 0.4655,
+      "step": 7612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7915395396153715,
+      "learning_rate": 7.535150670722302e-06,
+      "loss": 0.4957,
+      "step": 7613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7035949681388567,
+      "learning_rate": 7.5345699478105014e-06,
+      "loss": 0.6007,
+      "step": 7614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.197678636269882,
+      "learning_rate": 7.5339891823062415e-06,
+      "loss": 0.5429,
+      "step": 7615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8025857719351022,
+      "learning_rate": 7.533408374221681e-06,
+      "loss": 0.5531,
+      "step": 7616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7628954901249356,
+      "learning_rate": 7.532827523568982e-06,
+      "loss": 0.4693,
+      "step": 7617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8774950941212738,
+      "learning_rate": 7.532246630360298e-06,
+      "loss": 0.5541,
+      "step": 7618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.770217032100057,
+      "learning_rate": 7.531665694607795e-06,
+      "loss": 0.5604,
+      "step": 7619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6420407313678698,
+      "learning_rate": 7.531084716323631e-06,
+      "loss": 0.419,
+      "step": 7620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5481492289298981,
+      "learning_rate": 7.530503695519972e-06,
+      "loss": 0.5817,
+      "step": 7621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1645630672350418,
+      "learning_rate": 7.529922632208977e-06,
+      "loss": 0.6071,
+      "step": 7622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8958247031289388,
+      "learning_rate": 7.529341526402814e-06,
+      "loss": 0.6581,
+      "step": 7623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5822991193100413,
+      "learning_rate": 7.528760378113646e-06,
+      "loss": 0.3376,
+      "step": 7624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.999525078390096,
+      "learning_rate": 7.5281791873536415e-06,
+      "loss": 0.6021,
+      "step": 7625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6623489935908864,
+      "learning_rate": 7.527597954134964e-06,
+      "loss": 0.4482,
+      "step": 7626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7349581216713807,
+      "learning_rate": 7.527016678469783e-06,
+      "loss": 0.4674,
+      "step": 7627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5684237125432927,
+      "learning_rate": 7.526435360370267e-06,
+      "loss": 0.449,
+      "step": 7628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8770957990460975,
+      "learning_rate": 7.525853999848584e-06,
+      "loss": 0.7246,
+      "step": 7629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1402108401118256,
+      "learning_rate": 7.525272596916907e-06,
+      "loss": 0.6423,
+      "step": 7630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7558857199571638,
+      "learning_rate": 7.524691151587406e-06,
+      "loss": 0.4776,
+      "step": 7631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7609792052763094,
+      "learning_rate": 7.524109663872253e-06,
+      "loss": 0.5041,
+      "step": 7632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.808119210520194,
+      "learning_rate": 7.523528133783622e-06,
+      "loss": 0.5163,
+      "step": 7633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8848149993750016,
+      "learning_rate": 7.522946561333686e-06,
+      "loss": 0.4338,
+      "step": 7634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5599580873148393,
+      "learning_rate": 7.522364946534621e-06,
+      "loss": 0.4341,
+      "step": 7635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.442555168143261,
+      "learning_rate": 7.521783289398602e-06,
+      "loss": 0.497,
+      "step": 7636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0383156174270645,
+      "learning_rate": 7.521201589937804e-06,
+      "loss": 0.5275,
+      "step": 7637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5101158727488815,
+      "learning_rate": 7.520619848164407e-06,
+      "loss": 0.5279,
+      "step": 7638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.686201767654814,
+      "learning_rate": 7.520038064090589e-06,
+      "loss": 0.705,
+      "step": 7639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.083396505061956,
+      "learning_rate": 7.519456237728527e-06,
+      "loss": 0.7105,
+      "step": 7640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1454143280205877,
+      "learning_rate": 7.518874369090405e-06,
+      "loss": 0.6599,
+      "step": 7641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6395306100564708,
+      "learning_rate": 7.518292458188401e-06,
+      "loss": 0.4959,
+      "step": 7642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7808596826312166,
+      "learning_rate": 7.5177105050347e-06,
+      "loss": 0.5699,
+      "step": 7643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0305951484122864,
+      "learning_rate": 7.5171285096414805e-06,
+      "loss": 0.6202,
+      "step": 7644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1794608678671596,
+      "learning_rate": 7.516546472020929e-06,
+      "loss": 0.5969,
+      "step": 7645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9234548545263754,
+      "learning_rate": 7.51596439218523e-06,
+      "loss": 0.6542,
+      "step": 7646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7947400769124782,
+      "learning_rate": 7.515382270146569e-06,
+      "loss": 0.5941,
+      "step": 7647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8575542995896253,
+      "learning_rate": 7.51480010591713e-06,
+      "loss": 0.5443,
+      "step": 7648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.75565515435291,
+      "learning_rate": 7.514217899509104e-06,
+      "loss": 0.402,
+      "step": 7649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.567968975513903,
+      "learning_rate": 7.513635650934678e-06,
+      "loss": 0.5501,
+      "step": 7650
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5443139672279358,
+      "eval_runtime": 68.0067,
+      "eval_samples_per_second": 50.142,
+      "eval_steps_per_second": 1.573,
+      "step": 7650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7237988205868608,
+      "learning_rate": 7.513053360206039e-06,
+      "loss": 0.4186,
+      "step": 7651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7443630293572427,
+      "learning_rate": 7.512471027335378e-06,
+      "loss": 0.6175,
+      "step": 7652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.603888117192093,
+      "learning_rate": 7.511888652334888e-06,
+      "loss": 0.5242,
+      "step": 7653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.953519280640265,
+      "learning_rate": 7.511306235216758e-06,
+      "loss": 0.6396,
+      "step": 7654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0212700766171974,
+      "learning_rate": 7.5107237759931795e-06,
+      "loss": 0.5745,
+      "step": 7655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5210320213744173,
+      "learning_rate": 7.510141274676349e-06,
+      "loss": 0.5367,
+      "step": 7656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6952074002241917,
+      "learning_rate": 7.50955873127846e-06,
+      "loss": 0.4906,
+      "step": 7657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8312218831157112,
+      "learning_rate": 7.508976145811706e-06,
+      "loss": 0.5192,
+      "step": 7658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3934451377080992,
+      "learning_rate": 7.508393518288285e-06,
+      "loss": 0.4105,
+      "step": 7659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4754570132835954,
+      "learning_rate": 7.507810848720393e-06,
+      "loss": 0.4631,
+      "step": 7660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5304887670003868,
+      "learning_rate": 7.507228137120227e-06,
+      "loss": 0.3462,
+      "step": 7661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8150498015363774,
+      "learning_rate": 7.506645383499988e-06,
+      "loss": 0.5065,
+      "step": 7662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4672726325087555,
+      "learning_rate": 7.506062587871874e-06,
+      "loss": 0.7506,
+      "step": 7663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8111680575493476,
+      "learning_rate": 7.505479750248085e-06,
+      "loss": 0.5628,
+      "step": 7664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.964507510940575,
+      "learning_rate": 7.504896870640824e-06,
+      "loss": 0.5499,
+      "step": 7665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2855254940028,
+      "learning_rate": 7.504313949062291e-06,
+      "loss": 0.7964,
+      "step": 7666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8255335272401612,
+      "learning_rate": 7.503730985524692e-06,
+      "loss": 0.5512,
+      "step": 7667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9113645454924033,
+      "learning_rate": 7.503147980040228e-06,
+      "loss": 0.8142,
+      "step": 7668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.845376760649547,
+      "learning_rate": 7.502564932621106e-06,
+      "loss": 0.4934,
+      "step": 7669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6734181949551423,
+      "learning_rate": 7.50198184327953e-06,
+      "loss": 0.4829,
+      "step": 7670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6517048083723493,
+      "learning_rate": 7.50139871202771e-06,
+      "loss": 0.4046,
+      "step": 7671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.107865264032404,
+      "learning_rate": 7.500815538877848e-06,
+      "loss": 0.7046,
+      "step": 7672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5846939161263547,
+      "learning_rate": 7.5002323238421585e-06,
+      "loss": 0.569,
+      "step": 7673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.978623865651484,
+      "learning_rate": 7.499649066932846e-06,
+      "loss": 0.5183,
+      "step": 7674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0528432085445054,
+      "learning_rate": 7.4990657681621215e-06,
+      "loss": 0.5692,
+      "step": 7675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.739832880214109,
+      "learning_rate": 7.498482427542199e-06,
+      "loss": 0.6347,
+      "step": 7676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.989279124514874,
+      "learning_rate": 7.497899045085285e-06,
+      "loss": 0.5401,
+      "step": 7677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.178311565929239,
+      "learning_rate": 7.497315620803598e-06,
+      "loss": 0.4826,
+      "step": 7678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7592713945087004,
+      "learning_rate": 7.496732154709349e-06,
+      "loss": 0.6356,
+      "step": 7679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0515127004499356,
+      "learning_rate": 7.4961486468147514e-06,
+      "loss": 0.6558,
+      "step": 7680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8712749987979305,
+      "learning_rate": 7.4955650971320226e-06,
+      "loss": 0.4521,
+      "step": 7681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6470599093213898,
+      "learning_rate": 7.494981505673378e-06,
+      "loss": 0.3873,
+      "step": 7682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7451344936939277,
+      "learning_rate": 7.494397872451034e-06,
+      "loss": 0.4224,
+      "step": 7683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4678041463617024,
+      "learning_rate": 7.493814197477211e-06,
+      "loss": 0.6642,
+      "step": 7684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0344914265241547,
+      "learning_rate": 7.493230480764123e-06,
+      "loss": 0.564,
+      "step": 7685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.226197764521574,
+      "learning_rate": 7.492646722323995e-06,
+      "loss": 0.7048,
+      "step": 7686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7450496980097612,
+      "learning_rate": 7.492062922169046e-06,
+      "loss": 0.6414,
+      "step": 7687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8168427375515308,
+      "learning_rate": 7.491479080311497e-06,
+      "loss": 0.513,
+      "step": 7688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8316494583177736,
+      "learning_rate": 7.490895196763569e-06,
+      "loss": 0.4756,
+      "step": 7689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.722841410239442,
+      "learning_rate": 7.490311271537487e-06,
+      "loss": 0.4538,
+      "step": 7690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7373452226270976,
+      "learning_rate": 7.489727304645477e-06,
+      "loss": 0.5233,
+      "step": 7691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7887097821377949,
+      "learning_rate": 7.489143296099759e-06,
+      "loss": 0.6047,
+      "step": 7692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5333219052718001,
+      "learning_rate": 7.488559245912564e-06,
+      "loss": 0.4752,
+      "step": 7693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5864102933320317,
+      "learning_rate": 7.487975154096113e-06,
+      "loss": 0.6271,
+      "step": 7694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1964469108654567,
+      "learning_rate": 7.487391020662639e-06,
+      "loss": 0.6122,
+      "step": 7695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4735880451387822,
+      "learning_rate": 7.4868068456243685e-06,
+      "loss": 0.463,
+      "step": 7696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0711015247324784,
+      "learning_rate": 7.486222628993531e-06,
+      "loss": 0.5094,
+      "step": 7697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.774823903780623,
+      "learning_rate": 7.485638370782354e-06,
+      "loss": 0.5049,
+      "step": 7698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0191810343359675,
+      "learning_rate": 7.4850540710030735e-06,
+      "loss": 0.7757,
+      "step": 7699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5334343692616157,
+      "learning_rate": 7.484469729667919e-06,
+      "loss": 0.4267,
+      "step": 7700
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5463433265686035,
+      "eval_runtime": 67.944,
+      "eval_samples_per_second": 50.188,
+      "eval_steps_per_second": 1.575,
+      "step": 7700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8008102907610568,
+      "learning_rate": 7.483885346789123e-06,
+      "loss": 0.5669,
+      "step": 7701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.70202998566005,
+      "learning_rate": 7.483300922378919e-06,
+      "loss": 0.5574,
+      "step": 7702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7105976462449444,
+      "learning_rate": 7.482716456449543e-06,
+      "loss": 0.5633,
+      "step": 7703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8922038149057938,
+      "learning_rate": 7.482131949013229e-06,
+      "loss": 0.4328,
+      "step": 7704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9505735156795752,
+      "learning_rate": 7.481547400082213e-06,
+      "loss": 0.5805,
+      "step": 7705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7941046284974365,
+      "learning_rate": 7.4809628096687335e-06,
+      "loss": 0.603,
+      "step": 7706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.038513242057382,
+      "learning_rate": 7.480378177785028e-06,
+      "loss": 0.6222,
+      "step": 7707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7592080994244828,
+      "learning_rate": 7.479793504443337e-06,
+      "loss": 0.6139,
+      "step": 7708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6760482701243238,
+      "learning_rate": 7.479208789655895e-06,
+      "loss": 0.5105,
+      "step": 7709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0373322398732854,
+      "learning_rate": 7.478624033434949e-06,
+      "loss": 0.5571,
+      "step": 7710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7805010658241407,
+      "learning_rate": 7.478039235792737e-06,
+      "loss": 0.5174,
+      "step": 7711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1249051924176667,
+      "learning_rate": 7.477454396741502e-06,
+      "loss": 0.631,
+      "step": 7712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.125265988239902,
+      "learning_rate": 7.476869516293489e-06,
+      "loss": 0.6438,
+      "step": 7713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.84128924329599,
+      "learning_rate": 7.47628459446094e-06,
+      "loss": 0.5521,
+      "step": 7714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1266928698333345,
+      "learning_rate": 7.475699631256099e-06,
+      "loss": 0.8227,
+      "step": 7715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.659948813708367,
+      "learning_rate": 7.4751146266912154e-06,
+      "loss": 0.4582,
+      "step": 7716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.830257810509578,
+      "learning_rate": 7.474529580778534e-06,
+      "loss": 0.6157,
+      "step": 7717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7638195653990827,
+      "learning_rate": 7.473944493530299e-06,
+      "loss": 0.5502,
+      "step": 7718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.752263355732796,
+      "learning_rate": 7.473359364958766e-06,
+      "loss": 0.4816,
+      "step": 7719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8979850500467783,
+      "learning_rate": 7.4727741950761766e-06,
+      "loss": 0.5359,
+      "step": 7720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7959453285403917,
+      "learning_rate": 7.472188983894787e-06,
+      "loss": 0.5399,
+      "step": 7721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4858784311281357,
+      "learning_rate": 7.471603731426845e-06,
+      "loss": 0.5013,
+      "step": 7722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8417534971515015,
+      "learning_rate": 7.471018437684605e-06,
+      "loss": 0.6179,
+      "step": 7723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.857140459126684,
+      "learning_rate": 7.470433102680318e-06,
+      "loss": 0.5715,
+      "step": 7724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5654998973886078,
+      "learning_rate": 7.469847726426237e-06,
+      "loss": 0.4986,
+      "step": 7725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9167854594874558,
+      "learning_rate": 7.4692623089346186e-06,
+      "loss": 0.5943,
+      "step": 7726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1725009705453213,
+      "learning_rate": 7.468676850217716e-06,
+      "loss": 0.636,
+      "step": 7727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6699813056212935,
+      "learning_rate": 7.468091350287786e-06,
+      "loss": 0.4647,
+      "step": 7728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.251516889242804,
+      "learning_rate": 7.467505809157086e-06,
+      "loss": 0.5447,
+      "step": 7729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.957931099402218,
+      "learning_rate": 7.466920226837874e-06,
+      "loss": 0.5624,
+      "step": 7730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5551054748024158,
+      "learning_rate": 7.46633460334241e-06,
+      "loss": 0.4929,
+      "step": 7731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5060005009637263,
+      "learning_rate": 7.4657489386829515e-06,
+      "loss": 0.3445,
+      "step": 7732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7201733481639196,
+      "learning_rate": 7.465163232871759e-06,
+      "loss": 0.522,
+      "step": 7733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8614497311506264,
+      "learning_rate": 7.4645774859210975e-06,
+      "loss": 0.5332,
+      "step": 7734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.793560576411564,
+      "learning_rate": 7.463991697843224e-06,
+      "loss": 0.5332,
+      "step": 7735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.208159332651028,
+      "learning_rate": 7.463405868650405e-06,
+      "loss": 0.721,
+      "step": 7736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7093673561145841,
+      "learning_rate": 7.462819998354904e-06,
+      "loss": 0.4826,
+      "step": 7737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5725069877160778,
+      "learning_rate": 7.462234086968985e-06,
+      "loss": 0.4858,
+      "step": 7738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5439328984177232,
+      "learning_rate": 7.461648134504914e-06,
+      "loss": 0.3973,
+      "step": 7739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6234220643021284,
+      "learning_rate": 7.461062140974959e-06,
+      "loss": 0.4082,
+      "step": 7740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.063773700091296,
+      "learning_rate": 7.460476106391385e-06,
+      "loss": 0.4925,
+      "step": 7741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0100211511643007,
+      "learning_rate": 7.459890030766462e-06,
+      "loss": 0.6287,
+      "step": 7742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7725502401817836,
+      "learning_rate": 7.4593039141124575e-06,
+      "loss": 0.4589,
+      "step": 7743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.65823959448978,
+      "learning_rate": 7.458717756441642e-06,
+      "loss": 0.5343,
+      "step": 7744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6272308639889135,
+      "learning_rate": 7.458131557766289e-06,
+      "loss": 0.4613,
+      "step": 7745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8056933357038014,
+      "learning_rate": 7.457545318098667e-06,
+      "loss": 0.4467,
+      "step": 7746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6187811516733783,
+      "learning_rate": 7.4569590374510504e-06,
+      "loss": 0.4976,
+      "step": 7747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.248989177135954,
+      "learning_rate": 7.456372715835711e-06,
+      "loss": 0.5651,
+      "step": 7748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6265124788961416,
+      "learning_rate": 7.455786353264923e-06,
+      "loss": 0.4553,
+      "step": 7749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9812567515409294,
+      "learning_rate": 7.455199949750965e-06,
+      "loss": 0.4747,
+      "step": 7750
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5465705990791321,
+      "eval_runtime": 68.041,
+      "eval_samples_per_second": 50.117,
+      "eval_steps_per_second": 1.573,
+      "step": 7750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.024059038140696,
+      "learning_rate": 7.45461350530611e-06,
+      "loss": 0.4704,
+      "step": 7751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2649944699200772,
+      "learning_rate": 7.454027019942633e-06,
+      "loss": 0.8372,
+      "step": 7752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6916950222583251,
+      "learning_rate": 7.453440493672817e-06,
+      "loss": 0.3985,
+      "step": 7753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.162958145989186,
+      "learning_rate": 7.452853926508935e-06,
+      "loss": 0.6858,
+      "step": 7754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8925418060613897,
+      "learning_rate": 7.452267318463271e-06,
+      "loss": 0.621,
+      "step": 7755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9739795195354293,
+      "learning_rate": 7.4516806695481026e-06,
+      "loss": 0.8386,
+      "step": 7756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5130296795904759,
+      "learning_rate": 7.451093979775711e-06,
+      "loss": 0.4506,
+      "step": 7757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6365227011965746,
+      "learning_rate": 7.4505072491583795e-06,
+      "loss": 0.6796,
+      "step": 7758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9875670039661129,
+      "learning_rate": 7.449920477708389e-06,
+      "loss": 0.5149,
+      "step": 7759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8645599553873067,
+      "learning_rate": 7.449333665438027e-06,
+      "loss": 0.6785,
+      "step": 7760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8726082151415258,
+      "learning_rate": 7.448746812359574e-06,
+      "loss": 0.5854,
+      "step": 7761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.850306976797289,
+      "learning_rate": 7.448159918485316e-06,
+      "loss": 0.5477,
+      "step": 7762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7029393156905228,
+      "learning_rate": 7.4475729838275414e-06,
+      "loss": 0.4582,
+      "step": 7763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0152858681388435,
+      "learning_rate": 7.446986008398534e-06,
+      "loss": 0.5536,
+      "step": 7764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7854392399456707,
+      "learning_rate": 7.446398992210586e-06,
+      "loss": 0.4414,
+      "step": 7765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9486301603728164,
+      "learning_rate": 7.445811935275985e-06,
+      "loss": 0.5191,
+      "step": 7766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8833711553304502,
+      "learning_rate": 7.4452248376070165e-06,
+      "loss": 0.646,
+      "step": 7767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6905503815238094,
+      "learning_rate": 7.444637699215976e-06,
+      "loss": 0.5503,
+      "step": 7768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.026514577709958,
+      "learning_rate": 7.444050520115154e-06,
+      "loss": 0.6718,
+      "step": 7769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6483142527052281,
+      "learning_rate": 7.44346330031684e-06,
+      "loss": 0.4143,
+      "step": 7770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.133079270891955,
+      "learning_rate": 7.44287603983333e-06,
+      "loss": 0.4714,
+      "step": 7771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6031408529392404,
+      "learning_rate": 7.442288738676917e-06,
+      "loss": 0.4754,
+      "step": 7772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6276754540914529,
+      "learning_rate": 7.441701396859895e-06,
+      "loss": 0.5426,
+      "step": 7773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2453353538637444,
+      "learning_rate": 7.4411140143945615e-06,
+      "loss": 0.6441,
+      "step": 7774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5324480229003399,
+      "learning_rate": 7.440526591293211e-06,
+      "loss": 0.435,
+      "step": 7775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6988593833143566,
+      "learning_rate": 7.439939127568141e-06,
+      "loss": 0.4177,
+      "step": 7776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786089013650956,
+      "learning_rate": 7.439351623231651e-06,
+      "loss": 0.5653,
+      "step": 7777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5003011367362928,
+      "learning_rate": 7.438764078296039e-06,
+      "loss": 0.4022,
+      "step": 7778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8322967925196254,
+      "learning_rate": 7.438176492773605e-06,
+      "loss": 0.4656,
+      "step": 7779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.640306674792717,
+      "learning_rate": 7.437588866676651e-06,
+      "loss": 0.7222,
+      "step": 7780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7682875497733967,
+      "learning_rate": 7.437001200017477e-06,
+      "loss": 0.5231,
+      "step": 7781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.030695719119281,
+      "learning_rate": 7.436413492808388e-06,
+      "loss": 0.4445,
+      "step": 7782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7091150018485473,
+      "learning_rate": 7.4358257450616815e-06,
+      "loss": 0.5596,
+      "step": 7783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1479994547630885,
+      "learning_rate": 7.435237956789668e-06,
+      "loss": 0.737,
+      "step": 7784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8786580699237636,
+      "learning_rate": 7.434650128004649e-06,
+      "loss": 0.5241,
+      "step": 7785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6580450399229363,
+      "learning_rate": 7.434062258718932e-06,
+      "loss": 0.4183,
+      "step": 7786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9658511340960931,
+      "learning_rate": 7.433474348944821e-06,
+      "loss": 0.555,
+      "step": 7787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4727535002436625,
+      "learning_rate": 7.432886398694627e-06,
+      "loss": 0.4172,
+      "step": 7788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1064481845459735,
+      "learning_rate": 7.432298407980656e-06,
+      "loss": 0.6173,
+      "step": 7789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.610008582333273,
+      "learning_rate": 7.431710376815219e-06,
+      "loss": 0.5024,
+      "step": 7790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9332512885840822,
+      "learning_rate": 7.431122305210624e-06,
+      "loss": 0.5745,
+      "step": 7791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8683341908075968,
+      "learning_rate": 7.430534193179184e-06,
+      "loss": 0.5032,
+      "step": 7792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.164969595648172,
+      "learning_rate": 7.42994604073321e-06,
+      "loss": 0.645,
+      "step": 7793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.983990533975628,
+      "learning_rate": 7.429357847885013e-06,
+      "loss": 0.5989,
+      "step": 7794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0524931199841903,
+      "learning_rate": 7.428769614646908e-06,
+      "loss": 0.4208,
+      "step": 7795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8895943124486942,
+      "learning_rate": 7.428181341031211e-06,
+      "loss": 0.4335,
+      "step": 7796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8286729988416988,
+      "learning_rate": 7.4275930270502325e-06,
+      "loss": 0.5655,
+      "step": 7797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4541848236138881,
+      "learning_rate": 7.427004672716293e-06,
+      "loss": 0.4315,
+      "step": 7798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9226916602147117,
+      "learning_rate": 7.426416278041707e-06,
+      "loss": 0.517,
+      "step": 7799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4368973434569545,
+      "learning_rate": 7.425827843038793e-06,
+      "loss": 0.7417,
+      "step": 7800
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5444765686988831,
+      "eval_runtime": 67.9721,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 7800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5908518379643901,
+      "learning_rate": 7.425239367719869e-06,
+      "loss": 0.4276,
+      "step": 7801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.038657138578111,
+      "learning_rate": 7.424650852097255e-06,
+      "loss": 0.5982,
+      "step": 7802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9167704092737745,
+      "learning_rate": 7.424062296183271e-06,
+      "loss": 0.5792,
+      "step": 7803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.997511643538155,
+      "learning_rate": 7.423473699990238e-06,
+      "loss": 0.5724,
+      "step": 7804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.929222272879269,
+      "learning_rate": 7.4228850635304775e-06,
+      "loss": 0.4787,
+      "step": 7805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7256040707263876,
+      "learning_rate": 7.422296386816313e-06,
+      "loss": 0.3577,
+      "step": 7806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0912349108025667,
+      "learning_rate": 7.421707669860067e-06,
+      "loss": 0.6309,
+      "step": 7807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6642899619960592,
+      "learning_rate": 7.4211189126740646e-06,
+      "loss": 0.5908,
+      "step": 7808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3811130408129229,
+      "learning_rate": 7.420530115270631e-06,
+      "loss": 0.3684,
+      "step": 7809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9148504509080941,
+      "learning_rate": 7.419941277662093e-06,
+      "loss": 0.6204,
+      "step": 7810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.006856395445758,
+      "learning_rate": 7.419352399860776e-06,
+      "loss": 0.4674,
+      "step": 7811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6629868151299894,
+      "learning_rate": 7.4187634818790084e-06,
+      "loss": 0.4652,
+      "step": 7812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.755227892813993,
+      "learning_rate": 7.4181745237291204e-06,
+      "loss": 0.5505,
+      "step": 7813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1950727532900984,
+      "learning_rate": 7.41758552542344e-06,
+      "loss": 0.5427,
+      "step": 7814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.97715467148467,
+      "learning_rate": 7.416996486974298e-06,
+      "loss": 0.6983,
+      "step": 7815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.873699399245189,
+      "learning_rate": 7.416407408394024e-06,
+      "loss": 0.6577,
+      "step": 7816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.881136820201882,
+      "learning_rate": 7.415818289694951e-06,
+      "loss": 0.5861,
+      "step": 7817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.725890310249564,
+      "learning_rate": 7.415229130889412e-06,
+      "loss": 0.5382,
+      "step": 7818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.011530361501101,
+      "learning_rate": 7.414639931989741e-06,
+      "loss": 0.4541,
+      "step": 7819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.275183752595961,
+      "learning_rate": 7.4140506930082735e-06,
+      "loss": 0.4321,
+      "step": 7820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.124091182459048,
+      "learning_rate": 7.413461413957343e-06,
+      "loss": 0.5576,
+      "step": 7821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.819190462740879,
+      "learning_rate": 7.4128720948492865e-06,
+      "loss": 0.5656,
+      "step": 7822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8292607397752108,
+      "learning_rate": 7.41228273569644e-06,
+      "loss": 0.4634,
+      "step": 7823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3143354685157753,
+      "learning_rate": 7.411693336511142e-06,
+      "loss": 0.7361,
+      "step": 7824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7103771611383258,
+      "learning_rate": 7.4111038973057324e-06,
+      "loss": 0.4577,
+      "step": 7825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8074963381382567,
+      "learning_rate": 7.4105144180925485e-06,
+      "loss": 0.5658,
+      "step": 7826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.390686189748748,
+      "learning_rate": 7.409924898883933e-06,
+      "loss": 0.3518,
+      "step": 7827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8301857398779524,
+      "learning_rate": 7.409335339692225e-06,
+      "loss": 0.5606,
+      "step": 7828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9307486466723751,
+      "learning_rate": 7.408745740529769e-06,
+      "loss": 0.6223,
+      "step": 7829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3080796112603172,
+      "learning_rate": 7.408156101408906e-06,
+      "loss": 0.5725,
+      "step": 7830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5161891821222422,
+      "learning_rate": 7.40756642234198e-06,
+      "loss": 0.4362,
+      "step": 7831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.209942914787588,
+      "learning_rate": 7.406976703341337e-06,
+      "loss": 0.7244,
+      "step": 7832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9632794923909538,
+      "learning_rate": 7.4063869444193194e-06,
+      "loss": 0.6918,
+      "step": 7833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8277706117537402,
+      "learning_rate": 7.405797145588276e-06,
+      "loss": 0.6267,
+      "step": 7834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.763866374892802,
+      "learning_rate": 7.405207306860552e-06,
+      "loss": 0.6149,
+      "step": 7835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7645697906051017,
+      "learning_rate": 7.404617428248499e-06,
+      "loss": 0.4456,
+      "step": 7836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.370582168867314,
+      "learning_rate": 7.404027509764461e-06,
+      "loss": 0.7017,
+      "step": 7837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8894039496624935,
+      "learning_rate": 7.40343755142079e-06,
+      "loss": 0.5464,
+      "step": 7838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2878056454253457,
+      "learning_rate": 7.402847553229836e-06,
+      "loss": 0.7498,
+      "step": 7839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7335560794667393,
+      "learning_rate": 7.402257515203952e-06,
+      "loss": 0.4846,
+      "step": 7840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9420729925619622,
+      "learning_rate": 7.401667437355486e-06,
+      "loss": 0.7295,
+      "step": 7841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7414311004401162,
+      "learning_rate": 7.401077319696794e-06,
+      "loss": 0.4474,
+      "step": 7842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1677372096051357,
+      "learning_rate": 7.40048716224023e-06,
+      "loss": 0.526,
+      "step": 7843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8059014985105217,
+      "learning_rate": 7.3998969649981475e-06,
+      "loss": 0.5294,
+      "step": 7844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0505379190659285,
+      "learning_rate": 7.399306727982902e-06,
+      "loss": 0.595,
+      "step": 7845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5302633647836814,
+      "learning_rate": 7.39871645120685e-06,
+      "loss": 0.4302,
+      "step": 7846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.073648696239884,
+      "learning_rate": 7.398126134682347e-06,
+      "loss": 0.5808,
+      "step": 7847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7058263659116095,
+      "learning_rate": 7.397535778421755e-06,
+      "loss": 0.5353,
+      "step": 7848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6927345421552837,
+      "learning_rate": 7.396945382437428e-06,
+      "loss": 0.4861,
+      "step": 7849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.676378308848748,
+      "learning_rate": 7.396354946741728e-06,
+      "loss": 0.5712,
+      "step": 7850
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5439610481262207,
+      "eval_runtime": 68.0192,
+      "eval_samples_per_second": 50.133,
+      "eval_steps_per_second": 1.573,
+      "step": 7850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7449633195764687,
+      "learning_rate": 7.395764471347016e-06,
+      "loss": 0.4459,
+      "step": 7851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5808124250944617,
+      "learning_rate": 7.3951739562656515e-06,
+      "loss": 0.388,
+      "step": 7852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5100142895226825,
+      "learning_rate": 7.394583401509998e-06,
+      "loss": 0.4022,
+      "step": 7853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8588036945360575,
+      "learning_rate": 7.393992807092417e-06,
+      "loss": 0.5757,
+      "step": 7854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6582736315335949,
+      "learning_rate": 7.393402173025273e-06,
+      "loss": 0.404,
+      "step": 7855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.105635461131989,
+      "learning_rate": 7.392811499320931e-06,
+      "loss": 0.591,
+      "step": 7856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6753916663675412,
+      "learning_rate": 7.392220785991755e-06,
+      "loss": 0.5639,
+      "step": 7857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7342536335000023,
+      "learning_rate": 7.3916300330501135e-06,
+      "loss": 0.5826,
+      "step": 7858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8611690060640145,
+      "learning_rate": 7.391039240508372e-06,
+      "loss": 0.4045,
+      "step": 7859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.626841830065207,
+      "learning_rate": 7.390448408378899e-06,
+      "loss": 0.5318,
+      "step": 7860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.609737280082455,
+      "learning_rate": 7.389857536674061e-06,
+      "loss": 0.5064,
+      "step": 7861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7212789728820075,
+      "learning_rate": 7.38926662540623e-06,
+      "loss": 0.4709,
+      "step": 7862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3378119510290345,
+      "learning_rate": 7.388675674587776e-06,
+      "loss": 0.6869,
+      "step": 7863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0749916638981047,
+      "learning_rate": 7.38808468423107e-06,
+      "loss": 0.6326,
+      "step": 7864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4057443297180887,
+      "learning_rate": 7.387493654348483e-06,
+      "loss": 0.6113,
+      "step": 7865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8946937589942616,
+      "learning_rate": 7.386902584952389e-06,
+      "loss": 0.4636,
+      "step": 7866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4944431586121436,
+      "learning_rate": 7.386311476055163e-06,
+      "loss": 0.4399,
+      "step": 7867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.906440596061823,
+      "learning_rate": 7.385720327669175e-06,
+      "loss": 0.539,
+      "step": 7868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.792445863793417,
+      "learning_rate": 7.385129139806806e-06,
+      "loss": 0.5881,
+      "step": 7869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5610953970823138,
+      "learning_rate": 7.384537912480428e-06,
+      "loss": 0.3629,
+      "step": 7870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8218209811416712,
+      "learning_rate": 7.383946645702419e-06,
+      "loss": 0.4969,
+      "step": 7871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0756644930810433,
+      "learning_rate": 7.383355339485158e-06,
+      "loss": 0.5995,
+      "step": 7872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6730753891378642,
+      "learning_rate": 7.3827639938410216e-06,
+      "loss": 0.4005,
+      "step": 7873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4529904450061983,
+      "learning_rate": 7.382172608782392e-06,
+      "loss": 0.4569,
+      "step": 7874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6014636949547163,
+      "learning_rate": 7.3815811843216465e-06,
+      "loss": 0.4149,
+      "step": 7875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0046050500428105,
+      "learning_rate": 7.380989720471168e-06,
+      "loss": 0.5471,
+      "step": 7876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5028712065314944,
+      "learning_rate": 7.3803982172433385e-06,
+      "loss": 0.4279,
+      "step": 7877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5870490580564942,
+      "learning_rate": 7.37980667465054e-06,
+      "loss": 0.4273,
+      "step": 7878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.033223592415201,
+      "learning_rate": 7.379215092705157e-06,
+      "loss": 0.803,
+      "step": 7879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1375570363392007,
+      "learning_rate": 7.378623471419573e-06,
+      "loss": 0.5469,
+      "step": 7880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.057302269181127,
+      "learning_rate": 7.378031810806173e-06,
+      "loss": 0.5154,
+      "step": 7881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.019262629238968,
+      "learning_rate": 7.377440110877343e-06,
+      "loss": 0.5389,
+      "step": 7882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.626010503723284,
+      "learning_rate": 7.376848371645472e-06,
+      "loss": 0.3973,
+      "step": 7883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.497984320308863,
+      "learning_rate": 7.376256593122944e-06,
+      "loss": 0.5286,
+      "step": 7884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.878864479895243,
+      "learning_rate": 7.375664775322152e-06,
+      "loss": 0.5234,
+      "step": 7885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.142889428900326,
+      "learning_rate": 7.3750729182554805e-06,
+      "loss": 0.6299,
+      "step": 7886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6039897262268867,
+      "learning_rate": 7.374481021935323e-06,
+      "loss": 0.4613,
+      "step": 7887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9041506265480155,
+      "learning_rate": 7.373889086374071e-06,
+      "loss": 0.4484,
+      "step": 7888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0882168774503977,
+      "learning_rate": 7.3732971115841135e-06,
+      "loss": 0.7617,
+      "step": 7889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2562838713925717,
+      "learning_rate": 7.372705097577845e-06,
+      "loss": 0.55,
+      "step": 7890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3027151270628614,
+      "learning_rate": 7.372113044367657e-06,
+      "loss": 0.2771,
+      "step": 7891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6216588622784682,
+      "learning_rate": 7.371520951965946e-06,
+      "loss": 0.4966,
+      "step": 7892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8515594897025753,
+      "learning_rate": 7.370928820385107e-06,
+      "loss": 0.6041,
+      "step": 7893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7640570637512687,
+      "learning_rate": 7.370336649637536e-06,
+      "loss": 0.5761,
+      "step": 7894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5608690622530406,
+      "learning_rate": 7.369744439735628e-06,
+      "loss": 0.407,
+      "step": 7895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9642468153124455,
+      "learning_rate": 7.369152190691782e-06,
+      "loss": 0.6323,
+      "step": 7896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6031370813498065,
+      "learning_rate": 7.368559902518394e-06,
+      "loss": 0.584,
+      "step": 7897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7591770681506123,
+      "learning_rate": 7.3679675752278665e-06,
+      "loss": 0.6067,
+      "step": 7898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.966606015089269,
+      "learning_rate": 7.367375208832599e-06,
+      "loss": 0.6409,
+      "step": 7899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.270973701127489,
+      "learning_rate": 7.36678280334499e-06,
+      "loss": 0.5315,
+      "step": 7900
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5458500981330872,
+      "eval_runtime": 68.0003,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 7900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0147366179626838,
+      "learning_rate": 7.366190358777443e-06,
+      "loss": 0.5472,
+      "step": 7901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6430867012896253,
+      "learning_rate": 7.365597875142359e-06,
+      "loss": 0.5787,
+      "step": 7902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7282965751185086,
+      "learning_rate": 7.3650053524521425e-06,
+      "loss": 0.5629,
+      "step": 7903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.828580278904919,
+      "learning_rate": 7.364412790719199e-06,
+      "loss": 0.6524,
+      "step": 7904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5966414618212927,
+      "learning_rate": 7.363820189955929e-06,
+      "loss": 0.4768,
+      "step": 7905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8784811013066796,
+      "learning_rate": 7.363227550174744e-06,
+      "loss": 0.4335,
+      "step": 7906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6728881220525365,
+      "learning_rate": 7.3626348713880465e-06,
+      "loss": 0.4507,
+      "step": 7907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.917233669764073,
+      "learning_rate": 7.362042153608243e-06,
+      "loss": 0.6593,
+      "step": 7908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4719225362852517,
+      "learning_rate": 7.361449396847747e-06,
+      "loss": 0.3753,
+      "step": 7909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9419492706140034,
+      "learning_rate": 7.360856601118961e-06,
+      "loss": 0.6527,
+      "step": 7910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8635587023836007,
+      "learning_rate": 7.3602637664343e-06,
+      "loss": 0.5051,
+      "step": 7911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9442117722448704,
+      "learning_rate": 7.359670892806171e-06,
+      "loss": 0.5672,
+      "step": 7912
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9333811363840925,
+      "learning_rate": 7.359077980246988e-06,
+      "loss": 0.5698,
+      "step": 7913
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6561364760581965,
+      "learning_rate": 7.358485028769161e-06,
+      "loss": 0.6007,
+      "step": 7914
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8824409764302525,
+      "learning_rate": 7.357892038385105e-06,
+      "loss": 0.5368,
+      "step": 7915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.022463679945833,
+      "learning_rate": 7.357299009107232e-06,
+      "loss": 0.6713,
+      "step": 7916
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7905672526253484,
+      "learning_rate": 7.356705940947959e-06,
+      "loss": 0.4592,
+      "step": 7917
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6407214250061042,
+      "learning_rate": 7.356112833919702e-06,
+      "loss": 0.5166,
+      "step": 7918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5679981956235207,
+      "learning_rate": 7.355519688034874e-06,
+      "loss": 0.4994,
+      "step": 7919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6613844915907237,
+      "learning_rate": 7.354926503305892e-06,
+      "loss": 0.5445,
+      "step": 7920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5341371602738867,
+      "learning_rate": 7.354333279745178e-06,
+      "loss": 0.4718,
+      "step": 7921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4923647701077378,
+      "learning_rate": 7.3537400173651485e-06,
+      "loss": 0.4548,
+      "step": 7922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5027229300721803,
+      "learning_rate": 7.353146716178225e-06,
+      "loss": 0.4752,
+      "step": 7923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9165036758530105,
+      "learning_rate": 7.3525533761968225e-06,
+      "loss": 0.5893,
+      "step": 7924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6454830206840436,
+      "learning_rate": 7.351959997433366e-06,
+      "loss": 0.4998,
+      "step": 7925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8162334967933125,
+      "learning_rate": 7.35136657990028e-06,
+      "loss": 0.4474,
+      "step": 7926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4722936547304748,
+      "learning_rate": 7.350773123609982e-06,
+      "loss": 0.3684,
+      "step": 7927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6215211368570666,
+      "learning_rate": 7.3501796285749005e-06,
+      "loss": 0.7512,
+      "step": 7928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9420609207078392,
+      "learning_rate": 7.349586094807455e-06,
+      "loss": 0.5734,
+      "step": 7929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6887249544785607,
+      "learning_rate": 7.348992522320075e-06,
+      "loss": 0.5139,
+      "step": 7930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7593734877766392,
+      "learning_rate": 7.348398911125185e-06,
+      "loss": 0.398,
+      "step": 7931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0362697313161933,
+      "learning_rate": 7.347805261235211e-06,
+      "loss": 0.6545,
+      "step": 7932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6207936722008092,
+      "learning_rate": 7.347211572662583e-06,
+      "loss": 0.3837,
+      "step": 7933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9571550615622966,
+      "learning_rate": 7.346617845419727e-06,
+      "loss": 0.6072,
+      "step": 7934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9895185680435985,
+      "learning_rate": 7.346024079519074e-06,
+      "loss": 0.5149,
+      "step": 7935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5451591994513774,
+      "learning_rate": 7.345430274973053e-06,
+      "loss": 0.4572,
+      "step": 7936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.674756463981675,
+      "learning_rate": 7.344836431794096e-06,
+      "loss": 0.461,
+      "step": 7937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7307939454913206,
+      "learning_rate": 7.3442425499946335e-06,
+      "loss": 0.4379,
+      "step": 7938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9781318047748704,
+      "learning_rate": 7.343648629587099e-06,
+      "loss": 0.5257,
+      "step": 7939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5888600708253438,
+      "learning_rate": 7.343054670583924e-06,
+      "loss": 0.4599,
+      "step": 7940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6756098766908443,
+      "learning_rate": 7.3424606729975465e-06,
+      "loss": 0.5077,
+      "step": 7941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0505720467454753,
+      "learning_rate": 7.341866636840399e-06,
+      "loss": 0.5212,
+      "step": 7942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7268356307431791,
+      "learning_rate": 7.341272562124917e-06,
+      "loss": 0.4406,
+      "step": 7943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.463489408517857,
+      "learning_rate": 7.340678448863537e-06,
+      "loss": 0.4099,
+      "step": 7944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.595504623229051,
+      "learning_rate": 7.340084297068698e-06,
+      "loss": 0.3985,
+      "step": 7945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9149775208633306,
+      "learning_rate": 7.339490106752837e-06,
+      "loss": 0.4863,
+      "step": 7946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8281001584623076,
+      "learning_rate": 7.338895877928393e-06,
+      "loss": 0.5476,
+      "step": 7947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.684043825278913,
+      "learning_rate": 7.338301610607804e-06,
+      "loss": 0.5655,
+      "step": 7948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8616125910114907,
+      "learning_rate": 7.337707304803515e-06,
+      "loss": 0.5972,
+      "step": 7949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1357976073013796,
+      "learning_rate": 7.337112960527964e-06,
+      "loss": 0.5737,
+      "step": 7950
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5453633069992065,
+      "eval_runtime": 67.9981,
+      "eval_samples_per_second": 50.148,
+      "eval_steps_per_second": 1.574,
+      "step": 7950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.716096525837795,
+      "learning_rate": 7.336518577793595e-06,
+      "loss": 0.4575,
+      "step": 7951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6251342065486245,
+      "learning_rate": 7.3359241566128515e-06,
+      "loss": 0.6226,
+      "step": 7952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.744409257853174,
+      "learning_rate": 7.335329696998173e-06,
+      "loss": 0.5153,
+      "step": 7953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0111929771773402,
+      "learning_rate": 7.3347351989620095e-06,
+      "loss": 0.5273,
+      "step": 7954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3147387262906167,
+      "learning_rate": 7.334140662516803e-06,
+      "loss": 0.6486,
+      "step": 7955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2901176654398268,
+      "learning_rate": 7.3335460876750016e-06,
+      "loss": 0.4074,
+      "step": 7956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5567459332385347,
+      "learning_rate": 7.332951474449052e-06,
+      "loss": 0.4311,
+      "step": 7957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2655859914351124,
+      "learning_rate": 7.332356822851401e-06,
+      "loss": 0.5059,
+      "step": 7958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5163933846870294,
+      "learning_rate": 7.3317621328945e-06,
+      "loss": 0.4245,
+      "step": 7959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6655579509254475,
+      "learning_rate": 7.331167404590795e-06,
+      "loss": 0.4865,
+      "step": 7960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6908600010585568,
+      "learning_rate": 7.330572637952736e-06,
+      "loss": 0.5523,
+      "step": 7961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1314633231313027,
+      "learning_rate": 7.329977832992779e-06,
+      "loss": 0.5529,
+      "step": 7962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9317815496086301,
+      "learning_rate": 7.329382989723369e-06,
+      "loss": 0.426,
+      "step": 7963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7311145070652232,
+      "learning_rate": 7.328788108156964e-06,
+      "loss": 0.5208,
+      "step": 7964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.133957761441365,
+      "learning_rate": 7.3281931883060184e-06,
+      "loss": 0.5144,
+      "step": 7965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6595442582441136,
+      "learning_rate": 7.327598230182981e-06,
+      "loss": 0.4061,
+      "step": 7966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8466175121846475,
+      "learning_rate": 7.32700323380031e-06,
+      "loss": 0.4877,
+      "step": 7967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9495476114873582,
+      "learning_rate": 7.3264081991704625e-06,
+      "loss": 0.5359,
+      "step": 7968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9209421668010997,
+      "learning_rate": 7.325813126305894e-06,
+      "loss": 0.4841,
+      "step": 7969
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.815567536851963,
+      "learning_rate": 7.325218015219061e-06,
+      "loss": 0.5448,
+      "step": 7970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8467708167924166,
+      "learning_rate": 7.324622865922423e-06,
+      "loss": 0.6338,
+      "step": 7971
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5944376143080496,
+      "learning_rate": 7.324027678428437e-06,
+      "loss": 0.4671,
+      "step": 7972
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.074553985749331,
+      "learning_rate": 7.323432452749566e-06,
+      "loss": 0.5917,
+      "step": 7973
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4233442134968004,
+      "learning_rate": 7.3228371888982685e-06,
+      "loss": 0.5081,
+      "step": 7974
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2127672056301053,
+      "learning_rate": 7.3222418868870075e-06,
+      "loss": 0.735,
+      "step": 7975
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7831065235248518,
+      "learning_rate": 7.321646546728244e-06,
+      "loss": 0.4104,
+      "step": 7976
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8360082699146583,
+      "learning_rate": 7.32105116843444e-06,
+      "loss": 0.6156,
+      "step": 7977
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.483010497613506,
+      "learning_rate": 7.320455752018063e-06,
+      "loss": 0.7487,
+      "step": 7978
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8621744436749563,
+      "learning_rate": 7.319860297491573e-06,
+      "loss": 0.5091,
+      "step": 7979
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5222655676385572,
+      "learning_rate": 7.319264804867441e-06,
+      "loss": 0.4383,
+      "step": 7980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5664206585451546,
+      "learning_rate": 7.31866927415813e-06,
+      "loss": 0.4449,
+      "step": 7981
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.719325131341008,
+      "learning_rate": 7.318073705376106e-06,
+      "loss": 0.5335,
+      "step": 7982
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8025970186992397,
+      "learning_rate": 7.317478098533839e-06,
+      "loss": 0.4465,
+      "step": 7983
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4875920763052335,
+      "learning_rate": 7.316882453643797e-06,
+      "loss": 0.4035,
+      "step": 7984
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.744490649527107,
+      "learning_rate": 7.316286770718449e-06,
+      "loss": 0.5284,
+      "step": 7985
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6009347819684852,
+      "learning_rate": 7.315691049770266e-06,
+      "loss": 0.5003,
+      "step": 7986
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7385896180135179,
+      "learning_rate": 7.315095290811717e-06,
+      "loss": 0.5297,
+      "step": 7987
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9943301661180202,
+      "learning_rate": 7.314499493855277e-06,
+      "loss": 0.5477,
+      "step": 7988
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7961332589528842,
+      "learning_rate": 7.313903658913416e-06,
+      "loss": 0.5439,
+      "step": 7989
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.503287614193437,
+      "learning_rate": 7.313307785998609e-06,
+      "loss": 0.3667,
+      "step": 7990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.844915398044933,
+      "learning_rate": 7.312711875123331e-06,
+      "loss": 0.6382,
+      "step": 7991
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7261230557769416,
+      "learning_rate": 7.312115926300054e-06,
+      "loss": 0.5449,
+      "step": 7992
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8092031467744047,
+      "learning_rate": 7.311519939541256e-06,
+      "loss": 0.667,
+      "step": 7993
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7695663964973334,
+      "learning_rate": 7.310923914859415e-06,
+      "loss": 0.4732,
+      "step": 7994
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9787695414735902,
+      "learning_rate": 7.310327852267004e-06,
+      "loss": 0.5722,
+      "step": 7995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9019948940900546,
+      "learning_rate": 7.309731751776506e-06,
+      "loss": 0.5626,
+      "step": 7996
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1207204590931723,
+      "learning_rate": 7.309135613400398e-06,
+      "loss": 0.4832,
+      "step": 7997
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.216664432589838,
+      "learning_rate": 7.308539437151158e-06,
+      "loss": 0.604,
+      "step": 7998
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9327433714513043,
+      "learning_rate": 7.30794322304127e-06,
+      "loss": 0.7309,
+      "step": 7999
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.090025412691573,
+      "learning_rate": 7.307346971083212e-06,
+      "loss": 0.5849,
+      "step": 8000
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5460476875305176,
+      "eval_runtime": 67.9734,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 8000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7484068689373902,
+      "learning_rate": 7.306750681289469e-06,
+      "loss": 0.6451,
+      "step": 8001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8077213163234178,
+      "learning_rate": 7.306154353672521e-06,
+      "loss": 0.4795,
+      "step": 8002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7005604987176974,
+      "learning_rate": 7.305557988244855e-06,
+      "loss": 0.4702,
+      "step": 8003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6873422375389302,
+      "learning_rate": 7.304961585018953e-06,
+      "loss": 0.5137,
+      "step": 8004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.922826368342636,
+      "learning_rate": 7.304365144007303e-06,
+      "loss": 0.5176,
+      "step": 8005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6124850896627874,
+      "learning_rate": 7.303768665222386e-06,
+      "loss": 0.4267,
+      "step": 8006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.115453139625372,
+      "learning_rate": 7.303172148676696e-06,
+      "loss": 0.59,
+      "step": 8007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6522447279463628,
+      "learning_rate": 7.3025755943827144e-06,
+      "loss": 0.6073,
+      "step": 8008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7477256882578736,
+      "learning_rate": 7.3019790023529345e-06,
+      "loss": 0.5285,
+      "step": 8009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8597078328826022,
+      "learning_rate": 7.3013823725998426e-06,
+      "loss": 0.4939,
+      "step": 8010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9249902972381774,
+      "learning_rate": 7.300785705135929e-06,
+      "loss": 0.5245,
+      "step": 8011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7810910387790813,
+      "learning_rate": 7.300188999973686e-06,
+      "loss": 0.5305,
+      "step": 8012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7156778740631131,
+      "learning_rate": 7.299592257125605e-06,
+      "loss": 0.4837,
+      "step": 8013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.435222426830195,
+      "learning_rate": 7.298995476604176e-06,
+      "loss": 0.4708,
+      "step": 8014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9133311897895962,
+      "learning_rate": 7.298398658421895e-06,
+      "loss": 0.4903,
+      "step": 8015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6411600591291715,
+      "learning_rate": 7.2978018025912555e-06,
+      "loss": 0.4757,
+      "step": 8016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7994161345247839,
+      "learning_rate": 7.297204909124751e-06,
+      "loss": 0.5564,
+      "step": 8017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9629377714382676,
+      "learning_rate": 7.296607978034879e-06,
+      "loss": 0.7223,
+      "step": 8018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8537443975768093,
+      "learning_rate": 7.296011009334134e-06,
+      "loss": 0.4804,
+      "step": 8019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8990542573843927,
+      "learning_rate": 7.295414003035014e-06,
+      "loss": 0.4947,
+      "step": 8020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6100289813519588,
+      "learning_rate": 7.294816959150018e-06,
+      "loss": 0.4695,
+      "step": 8021
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.413057972876278,
+      "learning_rate": 7.294219877691644e-06,
+      "loss": 0.4178,
+      "step": 8022
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8420205467965147,
+      "learning_rate": 7.2936227586723905e-06,
+      "loss": 0.6244,
+      "step": 8023
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6311105603990215,
+      "learning_rate": 7.29302560210476e-06,
+      "loss": 0.5096,
+      "step": 8024
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6318482796169191,
+      "learning_rate": 7.29242840800125e-06,
+      "loss": 0.4473,
+      "step": 8025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8404985883910392,
+      "learning_rate": 7.2918311763743655e-06,
+      "loss": 0.5209,
+      "step": 8026
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2814766007709273,
+      "learning_rate": 7.291233907236608e-06,
+      "loss": 0.5731,
+      "step": 8027
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6207619050860358,
+      "learning_rate": 7.290636600600481e-06,
+      "loss": 0.4291,
+      "step": 8028
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6423898645749937,
+      "learning_rate": 7.29003925647849e-06,
+      "loss": 0.5212,
+      "step": 8029
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6848626361908314,
+      "learning_rate": 7.289441874883139e-06,
+      "loss": 0.4081,
+      "step": 8030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.961583373794314,
+      "learning_rate": 7.288844455826933e-06,
+      "loss": 0.5038,
+      "step": 8031
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9684256733167718,
+      "learning_rate": 7.288246999322379e-06,
+      "loss": 0.4181,
+      "step": 8032
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1734756281189456,
+      "learning_rate": 7.287649505381986e-06,
+      "loss": 0.612,
+      "step": 8033
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9227147618993803,
+      "learning_rate": 7.287051974018262e-06,
+      "loss": 0.4256,
+      "step": 8034
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.601213584190589,
+      "learning_rate": 7.286454405243713e-06,
+      "loss": 0.4821,
+      "step": 8035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9556089897337894,
+      "learning_rate": 7.285856799070851e-06,
+      "loss": 0.5028,
+      "step": 8036
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8021006445925705,
+      "learning_rate": 7.285259155512185e-06,
+      "loss": 0.6055,
+      "step": 8037
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8391770012342847,
+      "learning_rate": 7.284661474580228e-06,
+      "loss": 0.4262,
+      "step": 8038
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.975211811446324,
+      "learning_rate": 7.284063756287493e-06,
+      "loss": 0.5901,
+      "step": 8039
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.328891371633461,
+      "learning_rate": 7.283466000646492e-06,
+      "loss": 0.3938,
+      "step": 8040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.803274024478346,
+      "learning_rate": 7.282868207669735e-06,
+      "loss": 0.6286,
+      "step": 8041
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7936155555154152,
+      "learning_rate": 7.282270377369741e-06,
+      "loss": 0.7687,
+      "step": 8042
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.059101604828371,
+      "learning_rate": 7.281672509759024e-06,
+      "loss": 0.804,
+      "step": 8043
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.613802583215946,
+      "learning_rate": 7.2810746048500995e-06,
+      "loss": 0.5122,
+      "step": 8044
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7666506495722745,
+      "learning_rate": 7.280476662655484e-06,
+      "loss": 0.5736,
+      "step": 8045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.131008096400983,
+      "learning_rate": 7.279878683187695e-06,
+      "loss": 0.6483,
+      "step": 8046
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8057604577777568,
+      "learning_rate": 7.279280666459253e-06,
+      "loss": 0.6764,
+      "step": 8047
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6129916543992353,
+      "learning_rate": 7.278682612482674e-06,
+      "loss": 0.4265,
+      "step": 8048
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8138310751258904,
+      "learning_rate": 7.278084521270481e-06,
+      "loss": 0.6251,
+      "step": 8049
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.210876040673632,
+      "learning_rate": 7.2774863928351905e-06,
+      "loss": 0.599,
+      "step": 8050
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5449053049087524,
+      "eval_runtime": 67.9421,
+      "eval_samples_per_second": 50.19,
+      "eval_steps_per_second": 1.575,
+      "step": 8050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6650070745063883,
+      "learning_rate": 7.2768882271893274e-06,
+      "loss": 0.6079,
+      "step": 8051
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8844037057640128,
+      "learning_rate": 7.276290024345414e-06,
+      "loss": 0.5887,
+      "step": 8052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.479904063610998,
+      "learning_rate": 7.27569178431597e-06,
+      "loss": 0.4997,
+      "step": 8053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.862376168886467,
+      "learning_rate": 7.275093507113523e-06,
+      "loss": 0.6298,
+      "step": 8054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1534628269579117,
+      "learning_rate": 7.274495192750597e-06,
+      "loss": 0.6085,
+      "step": 8055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.986185950355823,
+      "learning_rate": 7.2738968412397135e-06,
+      "loss": 0.576,
+      "step": 8056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2397525909716083,
+      "learning_rate": 7.273298452593404e-06,
+      "loss": 0.5425,
+      "step": 8057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7400114532972755,
+      "learning_rate": 7.2727000268241924e-06,
+      "loss": 0.5084,
+      "step": 8058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8160678835369433,
+      "learning_rate": 7.2721015639446044e-06,
+      "loss": 0.4715,
+      "step": 8059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5331311949213071,
+      "learning_rate": 7.271503063967174e-06,
+      "loss": 0.4681,
+      "step": 8060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6741770320151563,
+      "learning_rate": 7.270904526904424e-06,
+      "loss": 0.5508,
+      "step": 8061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.205923176693688,
+      "learning_rate": 7.270305952768892e-06,
+      "loss": 0.4912,
+      "step": 8062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9426422908208032,
+      "learning_rate": 7.269707341573101e-06,
+      "loss": 0.5148,
+      "step": 8063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8833580896014115,
+      "learning_rate": 7.269108693329586e-06,
+      "loss": 0.5231,
+      "step": 8064
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.369250800862709,
+      "learning_rate": 7.268510008050881e-06,
+      "loss": 0.4725,
+      "step": 8065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7194400317123215,
+      "learning_rate": 7.267911285749517e-06,
+      "loss": 0.4952,
+      "step": 8066
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9316999875671703,
+      "learning_rate": 7.267312526438028e-06,
+      "loss": 0.6344,
+      "step": 8067
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0260191134107335,
+      "learning_rate": 7.26671373012895e-06,
+      "loss": 0.5837,
+      "step": 8068
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.862730731189686,
+      "learning_rate": 7.266114896834816e-06,
+      "loss": 0.4904,
+      "step": 8069
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.397528106002344,
+      "learning_rate": 7.265516026568163e-06,
+      "loss": 0.4235,
+      "step": 8070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0031573417872317,
+      "learning_rate": 7.26491711934153e-06,
+      "loss": 0.7588,
+      "step": 8071
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.210583589268288,
+      "learning_rate": 7.264318175167453e-06,
+      "loss": 0.5919,
+      "step": 8072
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5484472979764583,
+      "learning_rate": 7.263719194058472e-06,
+      "loss": 0.4571,
+      "step": 8073
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5680266717677185,
+      "learning_rate": 7.263120176027124e-06,
+      "loss": 0.4297,
+      "step": 8074
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1557278833165494,
+      "learning_rate": 7.26252112108595e-06,
+      "loss": 0.7236,
+      "step": 8075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9228067904546593,
+      "learning_rate": 7.261922029247492e-06,
+      "loss": 0.56,
+      "step": 8076
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.326127693463268,
+      "learning_rate": 7.26132290052429e-06,
+      "loss": 0.7105,
+      "step": 8077
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3162658914005316,
+      "learning_rate": 7.260723734928887e-06,
+      "loss": 0.6341,
+      "step": 8078
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5834126578503829,
+      "learning_rate": 7.260124532473827e-06,
+      "loss": 0.4715,
+      "step": 8079
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.069715904978176,
+      "learning_rate": 7.259525293171652e-06,
+      "loss": 0.4842,
+      "step": 8080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7679263221323749,
+      "learning_rate": 7.2589260170349095e-06,
+      "loss": 0.6367,
+      "step": 8081
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8277920787471902,
+      "learning_rate": 7.258326704076141e-06,
+      "loss": 0.5685,
+      "step": 8082
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5142657496273444,
+      "learning_rate": 7.257727354307896e-06,
+      "loss": 0.4309,
+      "step": 8083
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.276005975317174,
+      "learning_rate": 7.257127967742723e-06,
+      "loss": 0.6612,
+      "step": 8084
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7669781405023377,
+      "learning_rate": 7.256528544393164e-06,
+      "loss": 0.5354,
+      "step": 8085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3321028667809136,
+      "learning_rate": 7.255929084271773e-06,
+      "loss": 0.7524,
+      "step": 8086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.209587144685329,
+      "learning_rate": 7.255329587391097e-06,
+      "loss": 0.5482,
+      "step": 8087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8830395992574696,
+      "learning_rate": 7.254730053763683e-06,
+      "loss": 0.5078,
+      "step": 8088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.499148675117834,
+      "learning_rate": 7.25413048340209e-06,
+      "loss": 0.4675,
+      "step": 8089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0890739598257393,
+      "learning_rate": 7.253530876318863e-06,
+      "loss": 0.6836,
+      "step": 8090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8687307953692536,
+      "learning_rate": 7.252931232526557e-06,
+      "loss": 0.5182,
+      "step": 8091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6618008931065256,
+      "learning_rate": 7.252331552037723e-06,
+      "loss": 0.48,
+      "step": 8092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6744802412802688,
+      "learning_rate": 7.251731834864918e-06,
+      "loss": 0.4552,
+      "step": 8093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8109001883934412,
+      "learning_rate": 7.251132081020693e-06,
+      "loss": 0.4762,
+      "step": 8094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8301111910528496,
+      "learning_rate": 7.250532290517608e-06,
+      "loss": 0.4235,
+      "step": 8095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5500388311818885,
+      "learning_rate": 7.249932463368214e-06,
+      "loss": 0.4077,
+      "step": 8096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7250728550271028,
+      "learning_rate": 7.249332599585073e-06,
+      "loss": 0.5219,
+      "step": 8097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.147791360894073,
+      "learning_rate": 7.24873269918074e-06,
+      "loss": 0.597,
+      "step": 8098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8229226059438348,
+      "learning_rate": 7.248132762167774e-06,
+      "loss": 0.5899,
+      "step": 8099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.918336314369844,
+      "learning_rate": 7.247532788558735e-06,
+      "loss": 0.6313,
+      "step": 8100
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5442965626716614,
+      "eval_runtime": 68.0164,
+      "eval_samples_per_second": 50.135,
+      "eval_steps_per_second": 1.573,
+      "step": 8100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.84098008702934,
+      "learning_rate": 7.246932778366182e-06,
+      "loss": 0.5444,
+      "step": 8101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9478964464089223,
+      "learning_rate": 7.246332731602676e-06,
+      "loss": 0.4228,
+      "step": 8102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8860365346968384,
+      "learning_rate": 7.2457326482807785e-06,
+      "loss": 0.4351,
+      "step": 8103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3844016557199799,
+      "learning_rate": 7.245132528413054e-06,
+      "loss": 0.4849,
+      "step": 8104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6263573984227433,
+      "learning_rate": 7.244532372012062e-06,
+      "loss": 0.4704,
+      "step": 8105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.44965605348867,
+      "learning_rate": 7.24393217909037e-06,
+      "loss": 0.4934,
+      "step": 8106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8648546745490215,
+      "learning_rate": 7.2433319496605395e-06,
+      "loss": 0.5721,
+      "step": 8107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.472864196054038,
+      "learning_rate": 7.24273168373514e-06,
+      "loss": 0.4635,
+      "step": 8108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8402313591363753,
+      "learning_rate": 7.242131381326733e-06,
+      "loss": 0.6296,
+      "step": 8109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.607545523006393,
+      "learning_rate": 7.241531042447888e-06,
+      "loss": 0.4706,
+      "step": 8110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.642232632187804,
+      "learning_rate": 7.240930667111173e-06,
+      "loss": 0.4986,
+      "step": 8111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7603430775957327,
+      "learning_rate": 7.240330255329153e-06,
+      "loss": 0.5674,
+      "step": 8112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.843304842686943,
+      "learning_rate": 7.239729807114404e-06,
+      "loss": 0.5286,
+      "step": 8113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6656108467159483,
+      "learning_rate": 7.23912932247949e-06,
+      "loss": 0.5051,
+      "step": 8114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.635129737913381,
+      "learning_rate": 7.238528801436985e-06,
+      "loss": 0.4606,
+      "step": 8115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8365495203496756,
+      "learning_rate": 7.237928243999458e-06,
+      "loss": 0.7443,
+      "step": 8116
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8398053703448958,
+      "learning_rate": 7.237327650179483e-06,
+      "loss": 0.5906,
+      "step": 8117
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3080040548746834,
+      "learning_rate": 7.236727019989633e-06,
+      "loss": 0.6681,
+      "step": 8118
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1143581657183654,
+      "learning_rate": 7.236126353442482e-06,
+      "loss": 0.7135,
+      "step": 8119
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.750070171594016,
+      "learning_rate": 7.235525650550604e-06,
+      "loss": 0.546,
+      "step": 8120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8309601284579375,
+      "learning_rate": 7.234924911326574e-06,
+      "loss": 0.5647,
+      "step": 8121
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9482262999889863,
+      "learning_rate": 7.23432413578297e-06,
+      "loss": 0.6871,
+      "step": 8122
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8999613969883478,
+      "learning_rate": 7.233723323932365e-06,
+      "loss": 0.6254,
+      "step": 8123
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6859541390476862,
+      "learning_rate": 7.2331224757873406e-06,
+      "loss": 0.5592,
+      "step": 8124
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4893575766185279,
+      "learning_rate": 7.2325215913604725e-06,
+      "loss": 0.4902,
+      "step": 8125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6081817173032102,
+      "learning_rate": 7.231920670664343e-06,
+      "loss": 0.4496,
+      "step": 8126
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8403012065486748,
+      "learning_rate": 7.231319713711528e-06,
+      "loss": 0.5417,
+      "step": 8127
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1889454306780025,
+      "learning_rate": 7.230718720514611e-06,
+      "loss": 0.4922,
+      "step": 8128
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.463767484847048,
+      "learning_rate": 7.230117691086173e-06,
+      "loss": 0.759,
+      "step": 8129
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.169617303631373,
+      "learning_rate": 7.229516625438796e-06,
+      "loss": 0.6566,
+      "step": 8130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7051761323867367,
+      "learning_rate": 7.228915523585062e-06,
+      "loss": 0.4328,
+      "step": 8131
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.114990487335082,
+      "learning_rate": 7.228314385537555e-06,
+      "loss": 0.6387,
+      "step": 8132
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5951654534365274,
+      "learning_rate": 7.227713211308861e-06,
+      "loss": 0.4533,
+      "step": 8133
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1197584838373227,
+      "learning_rate": 7.227112000911563e-06,
+      "loss": 0.7991,
+      "step": 8134
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6274496802102536,
+      "learning_rate": 7.226510754358249e-06,
+      "loss": 0.4386,
+      "step": 8135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.892564387932698,
+      "learning_rate": 7.225909471661503e-06,
+      "loss": 0.5918,
+      "step": 8136
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.173155422636181,
+      "learning_rate": 7.225308152833915e-06,
+      "loss": 0.6345,
+      "step": 8137
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9070007192607927,
+      "learning_rate": 7.224706797888072e-06,
+      "loss": 0.6017,
+      "step": 8138
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6445228323783356,
+      "learning_rate": 7.224105406836565e-06,
+      "loss": 0.5405,
+      "step": 8139
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2200860721398725,
+      "learning_rate": 7.2235039796919815e-06,
+      "loss": 0.6591,
+      "step": 8140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4330831222726244,
+      "learning_rate": 7.222902516466912e-06,
+      "loss": 0.3951,
+      "step": 8141
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3646670313761344,
+      "learning_rate": 7.222301017173949e-06,
+      "loss": 0.293,
+      "step": 8142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.229113064080224,
+      "learning_rate": 7.221699481825686e-06,
+      "loss": 0.5677,
+      "step": 8143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0985252275124635,
+      "learning_rate": 7.2210979104347105e-06,
+      "loss": 0.7113,
+      "step": 8144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7031780844502193,
+      "learning_rate": 7.22049630301362e-06,
+      "loss": 0.5659,
+      "step": 8145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.805678847407213,
+      "learning_rate": 7.219894659575009e-06,
+      "loss": 0.5162,
+      "step": 8146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1082511612360753,
+      "learning_rate": 7.2192929801314705e-06,
+      "loss": 0.5167,
+      "step": 8147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2249888094431856,
+      "learning_rate": 7.218691264695602e-06,
+      "loss": 0.687,
+      "step": 8148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0555798065696953,
+      "learning_rate": 7.2180895132799985e-06,
+      "loss": 0.6383,
+      "step": 8149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5448378846542337,
+      "learning_rate": 7.21748772589726e-06,
+      "loss": 0.4171,
+      "step": 8150
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5422620177268982,
+      "eval_runtime": 67.9844,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 8150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1575059150484606,
+      "learning_rate": 7.216885902559982e-06,
+      "loss": 0.7106,
+      "step": 8151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.877764178870167,
+      "learning_rate": 7.216284043280763e-06,
+      "loss": 0.5527,
+      "step": 8152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8768627440862888,
+      "learning_rate": 7.215682148072205e-06,
+      "loss": 0.6127,
+      "step": 8153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8074729492678068,
+      "learning_rate": 7.215080216946906e-06,
+      "loss": 0.4093,
+      "step": 8154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7736614066561565,
+      "learning_rate": 7.214478249917468e-06,
+      "loss": 0.5944,
+      "step": 8155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.947077833604283,
+      "learning_rate": 7.213876246996494e-06,
+      "loss": 0.6092,
+      "step": 8156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6343210022226822,
+      "learning_rate": 7.213274208196583e-06,
+      "loss": 0.4587,
+      "step": 8157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9066539694225229,
+      "learning_rate": 7.212672133530342e-06,
+      "loss": 0.5776,
+      "step": 8158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5982094346912565,
+      "learning_rate": 7.212070023010375e-06,
+      "loss": 0.4279,
+      "step": 8159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7027440473711692,
+      "learning_rate": 7.211467876649283e-06,
+      "loss": 0.4016,
+      "step": 8160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6272318388142395,
+      "learning_rate": 7.210865694459675e-06,
+      "loss": 0.491,
+      "step": 8161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.372219604871397,
+      "learning_rate": 7.2102634764541565e-06,
+      "loss": 0.3806,
+      "step": 8162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4073135340028364,
+      "learning_rate": 7.209661222645333e-06,
+      "loss": 0.6441,
+      "step": 8163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5808184308036721,
+      "learning_rate": 7.209058933045815e-06,
+      "loss": 0.5869,
+      "step": 8164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7304221776406872,
+      "learning_rate": 7.208456607668209e-06,
+      "loss": 0.5485,
+      "step": 8165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0350451094015485,
+      "learning_rate": 7.207854246525127e-06,
+      "loss": 0.5434,
+      "step": 8166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9087121482921747,
+      "learning_rate": 7.207251849629176e-06,
+      "loss": 0.6523,
+      "step": 8167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6377251195148663,
+      "learning_rate": 7.206649416992968e-06,
+      "loss": 0.4719,
+      "step": 8168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1099884149729506,
+      "learning_rate": 7.206046948629114e-06,
+      "loss": 0.6514,
+      "step": 8169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.060629616177886,
+      "learning_rate": 7.205444444550226e-06,
+      "loss": 0.7915,
+      "step": 8170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7624891335366337,
+      "learning_rate": 7.20484190476892e-06,
+      "loss": 0.4617,
+      "step": 8171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6054405823465352,
+      "learning_rate": 7.204239329297806e-06,
+      "loss": 0.395,
+      "step": 8172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5092006467996648,
+      "learning_rate": 7.203636718149502e-06,
+      "loss": 0.4263,
+      "step": 8173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7721966342776045,
+      "learning_rate": 7.20303407133662e-06,
+      "loss": 0.4764,
+      "step": 8174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9231549395743284,
+      "learning_rate": 7.202431388871778e-06,
+      "loss": 0.5125,
+      "step": 8175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5785952434148067,
+      "learning_rate": 7.201828670767591e-06,
+      "loss": 0.4487,
+      "step": 8176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0711513824645094,
+      "learning_rate": 7.20122591703668e-06,
+      "loss": 0.7144,
+      "step": 8177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1568599009803777,
+      "learning_rate": 7.200623127691659e-06,
+      "loss": 0.5598,
+      "step": 8178
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0440252343244123,
+      "learning_rate": 7.200020302745151e-06,
+      "loss": 0.5152,
+      "step": 8179
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9103253563547595,
+      "learning_rate": 7.199417442209773e-06,
+      "loss": 0.6144,
+      "step": 8180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.599426683658917,
+      "learning_rate": 7.198814546098146e-06,
+      "loss": 0.4611,
+      "step": 8181
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0262728738910907,
+      "learning_rate": 7.198211614422892e-06,
+      "loss": 0.6945,
+      "step": 8182
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9045658305071793,
+      "learning_rate": 7.197608647196632e-06,
+      "loss": 0.6052,
+      "step": 8183
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8254161018435158,
+      "learning_rate": 7.197005644431991e-06,
+      "loss": 0.4526,
+      "step": 8184
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7658784018658091,
+      "learning_rate": 7.196402606141591e-06,
+      "loss": 0.4706,
+      "step": 8185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7589271030971447,
+      "learning_rate": 7.195799532338054e-06,
+      "loss": 0.5678,
+      "step": 8186
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8713794471351297,
+      "learning_rate": 7.195196423034008e-06,
+      "loss": 0.5896,
+      "step": 8187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8460355380995803,
+      "learning_rate": 7.194593278242078e-06,
+      "loss": 0.7409,
+      "step": 8188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.798389577382997,
+      "learning_rate": 7.193990097974889e-06,
+      "loss": 0.448,
+      "step": 8189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.179477625466614,
+      "learning_rate": 7.193386882245071e-06,
+      "loss": 0.5128,
+      "step": 8190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7143719419209444,
+      "learning_rate": 7.1927836310652525e-06,
+      "loss": 0.4512,
+      "step": 8191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9736756566096345,
+      "learning_rate": 7.192180344448056e-06,
+      "loss": 0.6046,
+      "step": 8192
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9791235225524761,
+      "learning_rate": 7.191577022406117e-06,
+      "loss": 0.6281,
+      "step": 8193
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8052062823473027,
+      "learning_rate": 7.190973664952064e-06,
+      "loss": 0.6327,
+      "step": 8194
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.78485300079244,
+      "learning_rate": 7.190370272098527e-06,
+      "loss": 0.4798,
+      "step": 8195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0286750330900487,
+      "learning_rate": 7.189766843858138e-06,
+      "loss": 0.5086,
+      "step": 8196
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6682460136419064,
+      "learning_rate": 7.18916338024353e-06,
+      "loss": 0.4715,
+      "step": 8197
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9359374752947605,
+      "learning_rate": 7.188559881267336e-06,
+      "loss": 0.5231,
+      "step": 8198
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7437363527313712,
+      "learning_rate": 7.187956346942191e-06,
+      "loss": 0.5463,
+      "step": 8199
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4213831403273887,
+      "learning_rate": 7.187352777280726e-06,
+      "loss": 0.4029,
+      "step": 8200
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5437533259391785,
+      "eval_runtime": 68.0222,
+      "eval_samples_per_second": 50.131,
+      "eval_steps_per_second": 1.573,
+      "step": 8200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6044613991517676,
+      "learning_rate": 7.186749172295581e-06,
+      "loss": 0.3895,
+      "step": 8201
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3960669616286079,
+      "learning_rate": 7.186145531999387e-06,
+      "loss": 0.3821,
+      "step": 8202
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.683047065426744,
+      "learning_rate": 7.1855418564047855e-06,
+      "loss": 0.3441,
+      "step": 8203
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7280337313677099,
+      "learning_rate": 7.184938145524412e-06,
+      "loss": 0.6662,
+      "step": 8204
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1141767800080893,
+      "learning_rate": 7.184334399370904e-06,
+      "loss": 0.432,
+      "step": 8205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5179797891655886,
+      "learning_rate": 7.183730617956902e-06,
+      "loss": 0.5831,
+      "step": 8206
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5166275721252747,
+      "learning_rate": 7.183126801295045e-06,
+      "loss": 0.3897,
+      "step": 8207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7037195619240064,
+      "learning_rate": 7.182522949397975e-06,
+      "loss": 0.5458,
+      "step": 8208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4588386211557491,
+      "learning_rate": 7.181919062278332e-06,
+      "loss": 0.3955,
+      "step": 8209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7506826525375854,
+      "learning_rate": 7.181315139948757e-06,
+      "loss": 0.4831,
+      "step": 8210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.088076290009511,
+      "learning_rate": 7.180711182421893e-06,
+      "loss": 0.5115,
+      "step": 8211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8369977769502972,
+      "learning_rate": 7.180107189710386e-06,
+      "loss": 0.5074,
+      "step": 8212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6411895754729666,
+      "learning_rate": 7.179503161826878e-06,
+      "loss": 0.625,
+      "step": 8213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0269629486845,
+      "learning_rate": 7.1788990987840144e-06,
+      "loss": 0.6281,
+      "step": 8214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9568938206622606,
+      "learning_rate": 7.178295000594443e-06,
+      "loss": 0.6415,
+      "step": 8215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9172199579943878,
+      "learning_rate": 7.177690867270806e-06,
+      "loss": 0.5285,
+      "step": 8216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8734449300804832,
+      "learning_rate": 7.177086698825752e-06,
+      "loss": 0.6242,
+      "step": 8217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5381160841221753,
+      "learning_rate": 7.176482495271932e-06,
+      "loss": 0.4215,
+      "step": 8218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9621260453304283,
+      "learning_rate": 7.175878256621991e-06,
+      "loss": 0.5977,
+      "step": 8219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0458961236220046,
+      "learning_rate": 7.1752739828885795e-06,
+      "loss": 0.5966,
+      "step": 8220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8007636969832082,
+      "learning_rate": 7.174669674084347e-06,
+      "loss": 0.4867,
+      "step": 8221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5139244172717734,
+      "learning_rate": 7.174065330221946e-06,
+      "loss": 0.3081,
+      "step": 8222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6475660276190223,
+      "learning_rate": 7.173460951314028e-06,
+      "loss": 0.5076,
+      "step": 8223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6366613021174974,
+      "learning_rate": 7.172856537373243e-06,
+      "loss": 0.5547,
+      "step": 8224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6522462517932526,
+      "learning_rate": 7.172252088412247e-06,
+      "loss": 0.5197,
+      "step": 8225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8996682210341553,
+      "learning_rate": 7.17164760444369e-06,
+      "loss": 0.5963,
+      "step": 8226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8176930465018875,
+      "learning_rate": 7.17104308548023e-06,
+      "loss": 0.6065,
+      "step": 8227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.950226130338244,
+      "learning_rate": 7.170438531534521e-06,
+      "loss": 0.4475,
+      "step": 8228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.807022695942534,
+      "learning_rate": 7.169833942619218e-06,
+      "loss": 0.4664,
+      "step": 8229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7717129161656926,
+      "learning_rate": 7.1692293187469795e-06,
+      "loss": 0.494,
+      "step": 8230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.326550331300188,
+      "learning_rate": 7.168624659930462e-06,
+      "loss": 0.5558,
+      "step": 8231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9616410659063583,
+      "learning_rate": 7.168019966182324e-06,
+      "loss": 0.5127,
+      "step": 8232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9632252068610065,
+      "learning_rate": 7.1674152375152225e-06,
+      "loss": 0.5058,
+      "step": 8233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.639737427341864,
+      "learning_rate": 7.16681047394182e-06,
+      "loss": 0.4581,
+      "step": 8234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8614185231457712,
+      "learning_rate": 7.166205675474775e-06,
+      "loss": 0.5696,
+      "step": 8235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6554875477520896,
+      "learning_rate": 7.16560084212675e-06,
+      "loss": 0.4944,
+      "step": 8236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.656442858620652,
+      "learning_rate": 7.164995973910403e-06,
+      "loss": 0.6151,
+      "step": 8237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.588035254603232,
+      "learning_rate": 7.164391070838404e-06,
+      "loss": 0.4093,
+      "step": 8238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8306996434927736,
+      "learning_rate": 7.163786132923408e-06,
+      "loss": 0.6315,
+      "step": 8239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7664538130307517,
+      "learning_rate": 7.163181160178084e-06,
+      "loss": 0.3796,
+      "step": 8240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.558314646087074,
+      "learning_rate": 7.162576152615096e-06,
+      "loss": 0.4267,
+      "step": 8241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7126852532807264,
+      "learning_rate": 7.161971110247108e-06,
+      "loss": 0.5803,
+      "step": 8242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6610788627214608,
+      "learning_rate": 7.1613660330867875e-06,
+      "loss": 0.6064,
+      "step": 8243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8100367851524333,
+      "learning_rate": 7.160760921146802e-06,
+      "loss": 0.6508,
+      "step": 8244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.505760144768377,
+      "learning_rate": 7.160155774439816e-06,
+      "loss": 0.4349,
+      "step": 8245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7105971724883162,
+      "learning_rate": 7.159550592978501e-06,
+      "loss": 0.4695,
+      "step": 8246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7768960396655606,
+      "learning_rate": 7.158945376775526e-06,
+      "loss": 0.6088,
+      "step": 8247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7021921044896051,
+      "learning_rate": 7.1583401258435604e-06,
+      "loss": 0.5687,
+      "step": 8248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7588607778286336,
+      "learning_rate": 7.157734840195273e-06,
+      "loss": 0.4233,
+      "step": 8249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9608409598174399,
+      "learning_rate": 7.157129519843335e-06,
+      "loss": 0.5715,
+      "step": 8250
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5437836050987244,
+      "eval_runtime": 67.9739,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 8250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8896342251125084,
+      "learning_rate": 7.156524164800422e-06,
+      "loss": 0.5328,
+      "step": 8251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9870406218175172,
+      "learning_rate": 7.155918775079204e-06,
+      "loss": 0.5527,
+      "step": 8252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.610475352047403,
+      "learning_rate": 7.155313350692356e-06,
+      "loss": 0.6828,
+      "step": 8253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8650059328025408,
+      "learning_rate": 7.154707891652549e-06,
+      "loss": 0.4959,
+      "step": 8254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7281967991093323,
+      "learning_rate": 7.154102397972461e-06,
+      "loss": 0.6514,
+      "step": 8255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4516622956370426,
+      "learning_rate": 7.1534968696647665e-06,
+      "loss": 0.4018,
+      "step": 8256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6566042652592472,
+      "learning_rate": 7.152891306742143e-06,
+      "loss": 0.3605,
+      "step": 8257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6737989580896522,
+      "learning_rate": 7.152285709217266e-06,
+      "loss": 0.5899,
+      "step": 8258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6876890499414994,
+      "learning_rate": 7.151680077102814e-06,
+      "loss": 0.5186,
+      "step": 8259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.601368473173683,
+      "learning_rate": 7.151074410411465e-06,
+      "loss": 0.4942,
+      "step": 8260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0827103157814877,
+      "learning_rate": 7.1504687091558976e-06,
+      "loss": 0.527,
+      "step": 8261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.807763793922266,
+      "learning_rate": 7.149862973348796e-06,
+      "loss": 0.4405,
+      "step": 8262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9836016536088876,
+      "learning_rate": 7.149257203002836e-06,
+      "loss": 0.4068,
+      "step": 8263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.172563218514065,
+      "learning_rate": 7.148651398130701e-06,
+      "loss": 0.5957,
+      "step": 8264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7070828547360473,
+      "learning_rate": 7.1480455587450725e-06,
+      "loss": 0.5173,
+      "step": 8265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5228219215939987,
+      "learning_rate": 7.147439684858635e-06,
+      "loss": 0.462,
+      "step": 8266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0534078148549946,
+      "learning_rate": 7.1468337764840705e-06,
+      "loss": 0.5927,
+      "step": 8267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8832329653066058,
+      "learning_rate": 7.146227833634064e-06,
+      "loss": 0.4454,
+      "step": 8268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.064543096870778,
+      "learning_rate": 7.145621856321299e-06,
+      "loss": 0.7249,
+      "step": 8269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7604095418564492,
+      "learning_rate": 7.1450158445584646e-06,
+      "loss": 0.5649,
+      "step": 8270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.027101537808236,
+      "learning_rate": 7.144409798358244e-06,
+      "loss": 0.5878,
+      "step": 8271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6359333231831041,
+      "learning_rate": 7.143803717733326e-06,
+      "loss": 0.5216,
+      "step": 8272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4427826886208854,
+      "learning_rate": 7.143197602696399e-06,
+      "loss": 0.4747,
+      "step": 8273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7180344687483777,
+      "learning_rate": 7.142591453260148e-06,
+      "loss": 0.4302,
+      "step": 8274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0459280159687387,
+      "learning_rate": 7.141985269437268e-06,
+      "loss": 0.707,
+      "step": 8275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7123388886073392,
+      "learning_rate": 7.141379051240446e-06,
+      "loss": 0.5654,
+      "step": 8276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9264102020275167,
+      "learning_rate": 7.1407727986823715e-06,
+      "loss": 0.5295,
+      "step": 8277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9369994332544056,
+      "learning_rate": 7.140166511775739e-06,
+      "loss": 0.5243,
+      "step": 8278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6752376066966157,
+      "learning_rate": 7.139560190533238e-06,
+      "loss": 0.4656,
+      "step": 8279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7626635885899944,
+      "learning_rate": 7.138953834967564e-06,
+      "loss": 0.5057,
+      "step": 8280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.419344228877794,
+      "learning_rate": 7.138347445091409e-06,
+      "loss": 0.4679,
+      "step": 8281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.620123585043562,
+      "learning_rate": 7.137741020917468e-06,
+      "loss": 0.6687,
+      "step": 8282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8391203946833912,
+      "learning_rate": 7.137134562458437e-06,
+      "loss": 0.5182,
+      "step": 8283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9470485465166694,
+      "learning_rate": 7.136528069727009e-06,
+      "loss": 0.5403,
+      "step": 8284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.869795642740288,
+      "learning_rate": 7.135921542735882e-06,
+      "loss": 0.7651,
+      "step": 8285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9131938280154268,
+      "learning_rate": 7.135314981497755e-06,
+      "loss": 0.4783,
+      "step": 8286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6242346089492012,
+      "learning_rate": 7.134708386025323e-06,
+      "loss": 0.5307,
+      "step": 8287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9426343246791475,
+      "learning_rate": 7.134101756331289e-06,
+      "loss": 0.6256,
+      "step": 8288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6855978587006444,
+      "learning_rate": 7.133495092428347e-06,
+      "loss": 0.516,
+      "step": 8289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5925079860054714,
+      "learning_rate": 7.132888394329202e-06,
+      "loss": 0.529,
+      "step": 8290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.858044875273145,
+      "learning_rate": 7.1322816620465505e-06,
+      "loss": 0.5306,
+      "step": 8291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7867340324328909,
+      "learning_rate": 7.131674895593097e-06,
+      "loss": 0.6277,
+      "step": 8292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8891557373739094,
+      "learning_rate": 7.131068094981545e-06,
+      "loss": 0.4858,
+      "step": 8293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6673953707818896,
+      "learning_rate": 7.1304612602245936e-06,
+      "loss": 0.4237,
+      "step": 8294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6200290636232921,
+      "learning_rate": 7.129854391334949e-06,
+      "loss": 0.3903,
+      "step": 8295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6608954322023122,
+      "learning_rate": 7.129247488325315e-06,
+      "loss": 0.4695,
+      "step": 8296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.888435686969756,
+      "learning_rate": 7.128640551208398e-06,
+      "loss": 0.4725,
+      "step": 8297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6387114608008337,
+      "learning_rate": 7.1280335799969e-06,
+      "loss": 0.5086,
+      "step": 8298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.905490696052942,
+      "learning_rate": 7.127426574703532e-06,
+      "loss": 0.5624,
+      "step": 8299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.00191629888257,
+      "learning_rate": 7.126819535340999e-06,
+      "loss": 0.6135,
+      "step": 8300
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5441112518310547,
+      "eval_runtime": 67.9513,
+      "eval_samples_per_second": 50.183,
+      "eval_steps_per_second": 1.575,
+      "step": 8300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9946338631405292,
+      "learning_rate": 7.12621246192201e-06,
+      "loss": 0.7562,
+      "step": 8301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.12189396791582,
+      "learning_rate": 7.125605354459273e-06,
+      "loss": 0.4172,
+      "step": 8302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5101562235696846,
+      "learning_rate": 7.124998212965496e-06,
+      "loss": 0.3363,
+      "step": 8303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8840681860837003,
+      "learning_rate": 7.124391037453393e-06,
+      "loss": 0.554,
+      "step": 8304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7988127923772539,
+      "learning_rate": 7.123783827935671e-06,
+      "loss": 0.479,
+      "step": 8305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8565886424560976,
+      "learning_rate": 7.123176584425044e-06,
+      "loss": 0.5228,
+      "step": 8306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.695245688223734,
+      "learning_rate": 7.122569306934224e-06,
+      "loss": 0.5193,
+      "step": 8307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6673929587500853,
+      "learning_rate": 7.121961995475921e-06,
+      "loss": 0.4536,
+      "step": 8308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7063095970115867,
+      "learning_rate": 7.121354650062853e-06,
+      "loss": 0.44,
+      "step": 8309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8173018261845826,
+      "learning_rate": 7.120747270707733e-06,
+      "loss": 0.4612,
+      "step": 8310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8856935986710655,
+      "learning_rate": 7.120139857423274e-06,
+      "loss": 0.649,
+      "step": 8311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5174365032779156,
+      "learning_rate": 7.119532410222194e-06,
+      "loss": 0.3651,
+      "step": 8312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8340427114928226,
+      "learning_rate": 7.118924929117209e-06,
+      "loss": 0.6845,
+      "step": 8313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5075488597099966,
+      "learning_rate": 7.118317414121037e-06,
+      "loss": 0.4084,
+      "step": 8314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.088811210244777,
+      "learning_rate": 7.1177098652463936e-06,
+      "loss": 0.6717,
+      "step": 8315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8078122912542833,
+      "learning_rate": 7.1171022825059985e-06,
+      "loss": 0.5604,
+      "step": 8316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9995193910248945,
+      "learning_rate": 7.116494665912573e-06,
+      "loss": 0.6038,
+      "step": 8317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7203130066054821,
+      "learning_rate": 7.115887015478837e-06,
+      "loss": 0.4805,
+      "step": 8318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.123359674877253,
+      "learning_rate": 7.115279331217506e-06,
+      "loss": 0.5825,
+      "step": 8319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.456926767434663,
+      "learning_rate": 7.11467161314131e-06,
+      "loss": 0.4076,
+      "step": 8320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7981030474596051,
+      "learning_rate": 7.114063861262964e-06,
+      "loss": 0.5306,
+      "step": 8321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9683031491239535,
+      "learning_rate": 7.1134560755951946e-06,
+      "loss": 0.5476,
+      "step": 8322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.903919667317964,
+      "learning_rate": 7.112848256150723e-06,
+      "loss": 0.4668,
+      "step": 8323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4698495615488913,
+      "learning_rate": 7.112240402942275e-06,
+      "loss": 0.5036,
+      "step": 8324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2438488301653803,
+      "learning_rate": 7.111632515982577e-06,
+      "loss": 0.6441,
+      "step": 8325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.866707183265216,
+      "learning_rate": 7.111024595284354e-06,
+      "loss": 0.4655,
+      "step": 8326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8089552110063198,
+      "learning_rate": 7.11041664086033e-06,
+      "loss": 0.4644,
+      "step": 8327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6529536975663124,
+      "learning_rate": 7.109808652723234e-06,
+      "loss": 0.4925,
+      "step": 8328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3167726372992095,
+      "learning_rate": 7.1092006308857945e-06,
+      "loss": 0.3809,
+      "step": 8329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8326735978268693,
+      "learning_rate": 7.1085925753607385e-06,
+      "loss": 0.5869,
+      "step": 8330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0635264473241843,
+      "learning_rate": 7.1079844861607976e-06,
+      "loss": 0.6771,
+      "step": 8331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5860821972160488,
+      "learning_rate": 7.107376363298698e-06,
+      "loss": 0.5397,
+      "step": 8332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3113736146621195,
+      "learning_rate": 7.1067682067871735e-06,
+      "loss": 0.4138,
+      "step": 8333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6992626892918958,
+      "learning_rate": 7.1061600166389555e-06,
+      "loss": 0.4742,
+      "step": 8334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2702330367525545,
+      "learning_rate": 7.105551792866773e-06,
+      "loss": 0.5977,
+      "step": 8335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9194228176424009,
+      "learning_rate": 7.104943535483363e-06,
+      "loss": 0.6064,
+      "step": 8336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.044759162309934,
+      "learning_rate": 7.1043352445014565e-06,
+      "loss": 0.6745,
+      "step": 8337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6235170799951162,
+      "learning_rate": 7.103726919933788e-06,
+      "loss": 0.6402,
+      "step": 8338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8532139084605794,
+      "learning_rate": 7.103118561793093e-06,
+      "loss": 0.4843,
+      "step": 8339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7795942506109963,
+      "learning_rate": 7.102510170092106e-06,
+      "loss": 0.5365,
+      "step": 8340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6658384590041049,
+      "learning_rate": 7.101901744843564e-06,
+      "loss": 0.5698,
+      "step": 8341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6876929101667357,
+      "learning_rate": 7.101293286060206e-06,
+      "loss": 0.5878,
+      "step": 8342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.745749671879053,
+      "learning_rate": 7.100684793754764e-06,
+      "loss": 0.3994,
+      "step": 8343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7732456094503046,
+      "learning_rate": 7.1000762679399825e-06,
+      "loss": 0.4842,
+      "step": 8344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7964733188752555,
+      "learning_rate": 7.099467708628596e-06,
+      "loss": 0.4541,
+      "step": 8345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.608030453820506,
+      "learning_rate": 7.09885911583335e-06,
+      "loss": 0.4967,
+      "step": 8346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2862142837411052,
+      "learning_rate": 7.098250489566979e-06,
+      "loss": 0.3225,
+      "step": 8347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8119319188453988,
+      "learning_rate": 7.0976418298422265e-06,
+      "loss": 0.4895,
+      "step": 8348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7123249401276182,
+      "learning_rate": 7.097033136671836e-06,
+      "loss": 0.6007,
+      "step": 8349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6376481315833493,
+      "learning_rate": 7.0964244100685484e-06,
+      "loss": 0.425,
+      "step": 8350
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5428789258003235,
+      "eval_runtime": 68.0047,
+      "eval_samples_per_second": 50.144,
+      "eval_steps_per_second": 1.573,
+      "step": 8350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9030874703083984,
+      "learning_rate": 7.095815650045107e-06,
+      "loss": 0.6284,
+      "step": 8351
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5155458713457566,
+      "learning_rate": 7.095206856614256e-06,
+      "loss": 0.4027,
+      "step": 8352
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7275782017836805,
+      "learning_rate": 7.09459802978874e-06,
+      "loss": 0.5803,
+      "step": 8353
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5531290440312728,
+      "learning_rate": 7.093989169581305e-06,
+      "loss": 0.5108,
+      "step": 8354
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7784562865188955,
+      "learning_rate": 7.093380276004699e-06,
+      "loss": 0.4969,
+      "step": 8355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8831188712108167,
+      "learning_rate": 7.0927713490716635e-06,
+      "loss": 0.671,
+      "step": 8356
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8745005627955016,
+      "learning_rate": 7.09216238879495e-06,
+      "loss": 0.5449,
+      "step": 8357
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5067594321202462,
+      "learning_rate": 7.0915533951873065e-06,
+      "loss": 0.508,
+      "step": 8358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.354595703485477,
+      "learning_rate": 7.090944368261478e-06,
+      "loss": 0.5651,
+      "step": 8359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.886575527388395,
+      "learning_rate": 7.090335308030222e-06,
+      "loss": 0.7391,
+      "step": 8360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5999234597196883,
+      "learning_rate": 7.089726214506281e-06,
+      "loss": 0.5554,
+      "step": 8361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3958104531079565,
+      "learning_rate": 7.089117087702411e-06,
+      "loss": 0.4582,
+      "step": 8362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7560628780454628,
+      "learning_rate": 7.088507927631361e-06,
+      "loss": 0.5214,
+      "step": 8363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5935864223651655,
+      "learning_rate": 7.087898734305883e-06,
+      "loss": 0.5439,
+      "step": 8364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6664025590272673,
+      "learning_rate": 7.087289507738734e-06,
+      "loss": 0.3938,
+      "step": 8365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4077139297740437,
+      "learning_rate": 7.086680247942664e-06,
+      "loss": 0.4636,
+      "step": 8366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.803801559256751,
+      "learning_rate": 7.0860709549304285e-06,
+      "loss": 0.4787,
+      "step": 8367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9205020222400249,
+      "learning_rate": 7.085461628714783e-06,
+      "loss": 0.689,
+      "step": 8368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.053735504090645,
+      "learning_rate": 7.084852269308481e-06,
+      "loss": 0.8046,
+      "step": 8369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6534205353758928,
+      "learning_rate": 7.084242876724282e-06,
+      "loss": 0.4151,
+      "step": 8370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6507353299652359,
+      "learning_rate": 7.083633450974944e-06,
+      "loss": 0.5077,
+      "step": 8371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8483018659545434,
+      "learning_rate": 7.083023992073221e-06,
+      "loss": 0.621,
+      "step": 8372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2911643539869029,
+      "learning_rate": 7.082414500031877e-06,
+      "loss": 0.3803,
+      "step": 8373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7932694188886065,
+      "learning_rate": 7.081804974863666e-06,
+      "loss": 0.5283,
+      "step": 8374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1162227505987565,
+      "learning_rate": 7.081195416581351e-06,
+      "loss": 0.6402,
+      "step": 8375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3328688691190353,
+      "learning_rate": 7.0805858251976926e-06,
+      "loss": 0.4333,
+      "step": 8376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5166940964278692,
+      "learning_rate": 7.07997620072545e-06,
+      "loss": 0.4384,
+      "step": 8377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2593443631541756,
+      "learning_rate": 7.079366543177388e-06,
+      "loss": 0.6915,
+      "step": 8378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.019209561567957,
+      "learning_rate": 7.07875685256627e-06,
+      "loss": 0.6524,
+      "step": 8379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8774821995444144,
+      "learning_rate": 7.0781471289048556e-06,
+      "loss": 0.5019,
+      "step": 8380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6954831703715978,
+      "learning_rate": 7.077537372205913e-06,
+      "loss": 0.4608,
+      "step": 8381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.018809993698181,
+      "learning_rate": 7.0769275824822026e-06,
+      "loss": 0.6405,
+      "step": 8382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8411947047446966,
+      "learning_rate": 7.076317759746495e-06,
+      "loss": 0.5249,
+      "step": 8383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1252055174338484,
+      "learning_rate": 7.075707904011554e-06,
+      "loss": 0.6079,
+      "step": 8384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.195690266236481,
+      "learning_rate": 7.075098015290145e-06,
+      "loss": 0.5442,
+      "step": 8385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1954085745303735,
+      "learning_rate": 7.07448809359504e-06,
+      "loss": 0.5604,
+      "step": 8386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.827542288031462,
+      "learning_rate": 7.073878138939003e-06,
+      "loss": 0.4505,
+      "step": 8387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9352304069307307,
+      "learning_rate": 7.073268151334805e-06,
+      "loss": 0.4362,
+      "step": 8388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.704913375327249,
+      "learning_rate": 7.072658130795217e-06,
+      "loss": 0.6085,
+      "step": 8389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.93487252814261,
+      "learning_rate": 7.072048077333006e-06,
+      "loss": 0.5478,
+      "step": 8390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5489824698762253,
+      "learning_rate": 7.071437990960945e-06,
+      "loss": 0.4538,
+      "step": 8391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1583218708512266,
+      "learning_rate": 7.070827871691807e-06,
+      "loss": 0.5664,
+      "step": 8392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.535122578174503,
+      "learning_rate": 7.070217719538362e-06,
+      "loss": 0.4369,
+      "step": 8393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0080167407702185,
+      "learning_rate": 7.0696075345133855e-06,
+      "loss": 0.5714,
+      "step": 8394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9896313307369569,
+      "learning_rate": 7.06899731662965e-06,
+      "loss": 0.4836,
+      "step": 8395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.586917165695312,
+      "learning_rate": 7.068387065899928e-06,
+      "loss": 0.4315,
+      "step": 8396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.69946112082285,
+      "learning_rate": 7.067776782336999e-06,
+      "loss": 0.4637,
+      "step": 8397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.148465278774325,
+      "learning_rate": 7.067166465953637e-06,
+      "loss": 0.7798,
+      "step": 8398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.09770162578845,
+      "learning_rate": 7.066556116762618e-06,
+      "loss": 0.5912,
+      "step": 8399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3454733032111619,
+      "learning_rate": 7.065945734776719e-06,
+      "loss": 0.3569,
+      "step": 8400
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.5422218441963196,
+      "eval_runtime": 67.9345,
+      "eval_samples_per_second": 50.195,
+      "eval_steps_per_second": 1.575,
+      "step": 8400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3533081766193307,
+      "learning_rate": 7.0653353200087195e-06,
+      "loss": 0.4825,
+      "step": 8401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6476406466750737,
+      "learning_rate": 7.064724872471397e-06,
+      "loss": 0.4275,
+      "step": 8402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.894422449972467,
+      "learning_rate": 7.064114392177531e-06,
+      "loss": 0.5425,
+      "step": 8403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0624484990429734,
+      "learning_rate": 7.063503879139902e-06,
+      "loss": 0.5189,
+      "step": 8404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6774739726920593,
+      "learning_rate": 7.06289333337129e-06,
+      "loss": 0.5727,
+      "step": 8405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8215634675737016,
+      "learning_rate": 7.062282754884479e-06,
+      "loss": 0.5721,
+      "step": 8406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3469562269016597,
+      "learning_rate": 7.061672143692246e-06,
+      "loss": 0.4534,
+      "step": 8407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4597827952060833,
+      "learning_rate": 7.061061499807378e-06,
+      "loss": 0.4075,
+      "step": 8408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4262840021963328,
+      "learning_rate": 7.060450823242656e-06,
+      "loss": 0.3804,
+      "step": 8409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4190142158924621,
+      "learning_rate": 7.059840114010866e-06,
+      "loss": 0.3769,
+      "step": 8410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8194605976985314,
+      "learning_rate": 7.059229372124793e-06,
+      "loss": 0.648,
+      "step": 8411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8022460014639667,
+      "learning_rate": 7.05861859759722e-06,
+      "loss": 0.4524,
+      "step": 8412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.790036050069774,
+      "learning_rate": 7.058007790440936e-06,
+      "loss": 0.5558,
+      "step": 8413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7040653202561418,
+      "learning_rate": 7.057396950668727e-06,
+      "loss": 0.5342,
+      "step": 8414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8475323391703722,
+      "learning_rate": 7.056786078293378e-06,
+      "loss": 0.4539,
+      "step": 8415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1602172762015344,
+      "learning_rate": 7.056175173327681e-06,
+      "loss": 0.609,
+      "step": 8416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4676265514758486,
+      "learning_rate": 7.055564235784423e-06,
+      "loss": 0.4441,
+      "step": 8417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6939427228912558,
+      "learning_rate": 7.054953265676395e-06,
+      "loss": 0.4092,
+      "step": 8418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7823666264526348,
+      "learning_rate": 7.054342263016385e-06,
+      "loss": 0.5062,
+      "step": 8419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9540666737679546,
+      "learning_rate": 7.053731227817186e-06,
+      "loss": 0.5141,
+      "step": 8420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.616708973903898,
+      "learning_rate": 7.053120160091588e-06,
+      "loss": 0.5319,
+      "step": 8421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.924640921062328,
+      "learning_rate": 7.052509059852383e-06,
+      "loss": 0.6496,
+      "step": 8422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9813381623844524,
+      "learning_rate": 7.0518979271123665e-06,
+      "loss": 0.4354,
+      "step": 8423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9295264572599313,
+      "learning_rate": 7.0512867618843315e-06,
+      "loss": 0.5761,
+      "step": 8424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6748404348366275,
+      "learning_rate": 7.05067556418107e-06,
+      "loss": 0.6127,
+      "step": 8425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.867756098477568,
+      "learning_rate": 7.05006433401538e-06,
+      "loss": 0.5756,
+      "step": 8426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.218421692191706,
+      "learning_rate": 7.049453071400055e-06,
+      "loss": 0.5945,
+      "step": 8427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9737470379516033,
+      "learning_rate": 7.048841776347894e-06,
+      "loss": 0.5207,
+      "step": 8428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6629375107190487,
+      "learning_rate": 7.048230448871691e-06,
+      "loss": 0.5394,
+      "step": 8429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9341770095488338,
+      "learning_rate": 7.047619088984245e-06,
+      "loss": 0.6639,
+      "step": 8430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1625926015615966,
+      "learning_rate": 7.047007696698356e-06,
+      "loss": 0.6494,
+      "step": 8431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5318240330190842,
+      "learning_rate": 7.04639627202682e-06,
+      "loss": 0.5182,
+      "step": 8432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.84008279872367,
+      "learning_rate": 7.045784814982439e-06,
+      "loss": 0.4191,
+      "step": 8433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6983280177952347,
+      "learning_rate": 7.045173325578014e-06,
+      "loss": 0.4238,
+      "step": 8434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9024206328520494,
+      "learning_rate": 7.044561803826345e-06,
+      "loss": 0.6775,
+      "step": 8435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0793791468976024,
+      "learning_rate": 7.043950249740233e-06,
+      "loss": 0.5665,
+      "step": 8436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7610540662039396,
+      "learning_rate": 7.043338663332483e-06,
+      "loss": 0.4588,
+      "step": 8437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6976486905388235,
+      "learning_rate": 7.042727044615896e-06,
+      "loss": 0.4576,
+      "step": 8438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6046783832683356,
+      "learning_rate": 7.042115393603276e-06,
+      "loss": 0.4923,
+      "step": 8439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7904939075589166,
+      "learning_rate": 7.041503710307429e-06,
+      "loss": 0.6669,
+      "step": 8440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.319613144119945,
+      "learning_rate": 7.040891994741158e-06,
+      "loss": 0.7103,
+      "step": 8441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1724397744085477,
+      "learning_rate": 7.040280246917271e-06,
+      "loss": 0.6609,
+      "step": 8442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3915096010977992,
+      "learning_rate": 7.039668466848575e-06,
+      "loss": 0.3721,
+      "step": 8443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7872961466290784,
+      "learning_rate": 7.039056654547874e-06,
+      "loss": 0.4569,
+      "step": 8444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.061700700752379,
+      "learning_rate": 7.038444810027979e-06,
+      "loss": 0.5502,
+      "step": 8445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8160645299957958,
+      "learning_rate": 7.037832933301697e-06,
+      "loss": 0.5602,
+      "step": 8446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6660571282314969,
+      "learning_rate": 7.0372210243818376e-06,
+      "loss": 0.4544,
+      "step": 8447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7391019255274227,
+      "learning_rate": 7.036609083281212e-06,
+      "loss": 0.4533,
+      "step": 8448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8644706432641276,
+      "learning_rate": 7.0359971100126275e-06,
+      "loss": 0.7324,
+      "step": 8449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.977753193317594,
+      "learning_rate": 7.0353851045888986e-06,
+      "loss": 0.6031,
+      "step": 8450
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.541221022605896,
+      "eval_runtime": 67.9684,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 8450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9913406175340909,
+      "learning_rate": 7.034773067022835e-06,
+      "loss": 0.462,
+      "step": 8451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8215867393288854,
+      "learning_rate": 7.034160997327253e-06,
+      "loss": 0.4855,
+      "step": 8452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.171486942045073,
+      "learning_rate": 7.033548895514962e-06,
+      "loss": 0.5727,
+      "step": 8453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7185733949516249,
+      "learning_rate": 7.032936761598778e-06,
+      "loss": 0.6008,
+      "step": 8454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0406319463589133,
+      "learning_rate": 7.032324595591514e-06,
+      "loss": 0.4535,
+      "step": 8455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5964669982771909,
+      "learning_rate": 7.031712397505989e-06,
+      "loss": 0.4325,
+      "step": 8456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.627012710478444,
+      "learning_rate": 7.031100167355012e-06,
+      "loss": 0.6197,
+      "step": 8457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.970581157557746,
+      "learning_rate": 7.030487905151407e-06,
+      "loss": 0.5691,
+      "step": 8458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8561421001722491,
+      "learning_rate": 7.029875610907988e-06,
+      "loss": 0.5631,
+      "step": 8459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.06643748675406,
+      "learning_rate": 7.029263284637573e-06,
+      "loss": 0.5296,
+      "step": 8460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.78181582025657,
+      "learning_rate": 7.028650926352981e-06,
+      "loss": 0.4646,
+      "step": 8461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3405872745745055,
+      "learning_rate": 7.02803853606703e-06,
+      "loss": 0.6236,
+      "step": 8462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8510246386894342,
+      "learning_rate": 7.027426113792543e-06,
+      "loss": 0.3553,
+      "step": 8463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8889993358076778,
+      "learning_rate": 7.02681365954234e-06,
+      "loss": 0.5517,
+      "step": 8464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9099293052333173,
+      "learning_rate": 7.0262011733292405e-06,
+      "loss": 0.4699,
+      "step": 8465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.462469523657964,
+      "learning_rate": 7.025588655166067e-06,
+      "loss": 0.4171,
+      "step": 8466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9661484787209602,
+      "learning_rate": 7.024976105065641e-06,
+      "loss": 0.7101,
+      "step": 8467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.231387076526439,
+      "learning_rate": 7.0243635230407905e-06,
+      "loss": 0.3638,
+      "step": 8468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9970213272844954,
+      "learning_rate": 7.023750909104334e-06,
+      "loss": 0.5006,
+      "step": 8469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.984389524715462,
+      "learning_rate": 7.0231382632691e-06,
+      "loss": 0.5668,
+      "step": 8470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9979901687297612,
+      "learning_rate": 7.0225255855479135e-06,
+      "loss": 0.4983,
+      "step": 8471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5049425616678347,
+      "learning_rate": 7.021912875953598e-06,
+      "loss": 0.547,
+      "step": 8472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4889389830887643,
+      "learning_rate": 7.021300134498981e-06,
+      "loss": 0.4672,
+      "step": 8473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4048291465915006,
+      "learning_rate": 7.020687361196892e-06,
+      "loss": 0.4435,
+      "step": 8474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8849605264397113,
+      "learning_rate": 7.020074556060156e-06,
+      "loss": 0.7486,
+      "step": 8475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.929570955424169,
+      "learning_rate": 7.019461719101606e-06,
+      "loss": 0.5857,
+      "step": 8476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.689942895067235,
+      "learning_rate": 7.018848850334067e-06,
+      "loss": 0.4864,
+      "step": 8477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6894030662256199,
+      "learning_rate": 7.01823594977037e-06,
+      "loss": 0.4679,
+      "step": 8478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8733443419721707,
+      "learning_rate": 7.017623017423348e-06,
+      "loss": 0.5785,
+      "step": 8479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5451374899403962,
+      "learning_rate": 7.0170100533058296e-06,
+      "loss": 0.4188,
+      "step": 8480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.723956384563685,
+      "learning_rate": 7.016397057430647e-06,
+      "loss": 0.4509,
+      "step": 8481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7974179046871543,
+      "learning_rate": 7.015784029810636e-06,
+      "loss": 0.4508,
+      "step": 8482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.894331400836234,
+      "learning_rate": 7.015170970458628e-06,
+      "loss": 0.5224,
+      "step": 8483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8477887618176072,
+      "learning_rate": 7.014557879387455e-06,
+      "loss": 0.4162,
+      "step": 8484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6716268017876874,
+      "learning_rate": 7.013944756609955e-06,
+      "loss": 0.5902,
+      "step": 8485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7947397063013475,
+      "learning_rate": 7.013331602138961e-06,
+      "loss": 0.5287,
+      "step": 8486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3932628719436608,
+      "learning_rate": 7.012718415987311e-06,
+      "loss": 0.3779,
+      "step": 8487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.042361447980263,
+      "learning_rate": 7.01210519816784e-06,
+      "loss": 0.5677,
+      "step": 8488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7315521300596126,
+      "learning_rate": 7.011491948693385e-06,
+      "loss": 0.5277,
+      "step": 8489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7992782719524303,
+      "learning_rate": 7.010878667576786e-06,
+      "loss": 0.7297,
+      "step": 8490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.923702262133363,
+      "learning_rate": 7.01026535483088e-06,
+      "loss": 0.6187,
+      "step": 8491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0018525873340063,
+      "learning_rate": 7.009652010468506e-06,
+      "loss": 0.5549,
+      "step": 8492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6187817044869857,
+      "learning_rate": 7.009038634502506e-06,
+      "loss": 0.4778,
+      "step": 8493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.068858968136066,
+      "learning_rate": 7.0084252269457186e-06,
+      "loss": 0.6073,
+      "step": 8494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.010934282522212,
+      "learning_rate": 7.007811787810987e-06,
+      "loss": 0.5365,
+      "step": 8495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.772994448365756,
+      "learning_rate": 7.007198317111154e-06,
+      "loss": 0.4056,
+      "step": 8496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.942493234814343,
+      "learning_rate": 7.006584814859057e-06,
+      "loss": 0.508,
+      "step": 8497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6014983779496146,
+      "learning_rate": 7.005971281067543e-06,
+      "loss": 0.4781,
+      "step": 8498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5257145594231565,
+      "learning_rate": 7.005357715749457e-06,
+      "loss": 0.4438,
+      "step": 8499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5539726658244997,
+      "learning_rate": 7.004744118917641e-06,
+      "loss": 0.4813,
+      "step": 8500
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.5410424470901489,
+      "eval_runtime": 67.9494,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 8500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2596028463322697,
+      "learning_rate": 7.004130490584943e-06,
+      "loss": 0.7552,
+      "step": 8501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6108202174190753,
+      "learning_rate": 7.003516830764207e-06,
+      "loss": 0.4211,
+      "step": 8502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6928923549455677,
+      "learning_rate": 7.0029031394682804e-06,
+      "loss": 0.5942,
+      "step": 8503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.732127901583906,
+      "learning_rate": 7.002289416710007e-06,
+      "loss": 0.4369,
+      "step": 8504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.089605895774753,
+      "learning_rate": 7.001675662502241e-06,
+      "loss": 0.6826,
+      "step": 8505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2486942882588803,
+      "learning_rate": 7.001061876857827e-06,
+      "loss": 0.6483,
+      "step": 8506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6104271026631611,
+      "learning_rate": 7.000448059789614e-06,
+      "loss": 0.4986,
+      "step": 8507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9628662347811408,
+      "learning_rate": 6.9998342113104545e-06,
+      "loss": 0.4903,
+      "step": 8508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9936037023118252,
+      "learning_rate": 6.999220331433196e-06,
+      "loss": 0.7982,
+      "step": 8509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7611142367703199,
+      "learning_rate": 6.998606420170692e-06,
+      "loss": 0.5559,
+      "step": 8510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9451555530400995,
+      "learning_rate": 6.997992477535792e-06,
+      "loss": 0.5321,
+      "step": 8511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6389975881043484,
+      "learning_rate": 6.997378503541352e-06,
+      "loss": 0.5056,
+      "step": 8512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7285532269733443,
+      "learning_rate": 6.996764498200221e-06,
+      "loss": 0.5501,
+      "step": 8513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8576813915477617,
+      "learning_rate": 6.996150461525256e-06,
+      "loss": 0.4517,
+      "step": 8514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6479746185346662,
+      "learning_rate": 6.99553639352931e-06,
+      "loss": 0.5451,
+      "step": 8515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0169563512326927,
+      "learning_rate": 6.994922294225239e-06,
+      "loss": 0.5772,
+      "step": 8516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0455896387219332,
+      "learning_rate": 6.994308163625899e-06,
+      "loss": 0.4871,
+      "step": 8517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8346990955293052,
+      "learning_rate": 6.993694001744143e-06,
+      "loss": 0.5366,
+      "step": 8518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0002573711591007,
+      "learning_rate": 6.993079808592835e-06,
+      "loss": 0.6649,
+      "step": 8519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.075634503852235,
+      "learning_rate": 6.992465584184824e-06,
+      "loss": 0.5231,
+      "step": 8520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0964305965044403,
+      "learning_rate": 6.9918513285329755e-06,
+      "loss": 0.6128,
+      "step": 8521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0396136881974636,
+      "learning_rate": 6.991237041650146e-06,
+      "loss": 0.5591,
+      "step": 8522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.461078139492272,
+      "learning_rate": 6.990622723549197e-06,
+      "loss": 0.4986,
+      "step": 8523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7374663218144208,
+      "learning_rate": 6.990008374242983e-06,
+      "loss": 0.6122,
+      "step": 8524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1003610638664556,
+      "learning_rate": 6.9893939937443735e-06,
+      "loss": 0.4389,
+      "step": 8525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.681046821503429,
+      "learning_rate": 6.988779582066222e-06,
+      "loss": 0.3537,
+      "step": 8526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.143018566062631,
+      "learning_rate": 6.988165139221397e-06,
+      "loss": 0.5292,
+      "step": 8527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0616209389718465,
+      "learning_rate": 6.9875506652227585e-06,
+      "loss": 0.5055,
+      "step": 8528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6548880194922488,
+      "learning_rate": 6.986936160083171e-06,
+      "loss": 0.4592,
+      "step": 8529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5625602399243546,
+      "learning_rate": 6.9863216238154985e-06,
+      "loss": 0.5972,
+      "step": 8530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.690261840115271,
+      "learning_rate": 6.985707056432604e-06,
+      "loss": 0.4511,
+      "step": 8531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1604531261102617,
+      "learning_rate": 6.9850924579473575e-06,
+      "loss": 0.6803,
+      "step": 8532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7337345778868403,
+      "learning_rate": 6.984477828372622e-06,
+      "loss": 0.5057,
+      "step": 8533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.660091163751602,
+      "learning_rate": 6.9838631677212656e-06,
+      "loss": 0.4275,
+      "step": 8534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.303437852649926,
+      "learning_rate": 6.983248476006156e-06,
+      "loss": 0.3484,
+      "step": 8535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4590182126186693,
+      "learning_rate": 6.98263375324016e-06,
+      "loss": 0.5111,
+      "step": 8536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.163977369971323,
+      "learning_rate": 6.982018999436148e-06,
+      "loss": 0.5516,
+      "step": 8537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.130711969504231,
+      "learning_rate": 6.981404214606987e-06,
+      "loss": 0.6066,
+      "step": 8538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9704513945240405,
+      "learning_rate": 6.980789398765549e-06,
+      "loss": 0.5471,
+      "step": 8539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1102846828800366,
+      "learning_rate": 6.980174551924708e-06,
+      "loss": 0.6135,
+      "step": 8540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.617826313471493,
+      "learning_rate": 6.97955967409733e-06,
+      "loss": 0.5179,
+      "step": 8541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5963710674369636,
+      "learning_rate": 6.978944765296289e-06,
+      "loss": 0.43,
+      "step": 8542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5632829632733696,
+      "learning_rate": 6.9783298255344564e-06,
+      "loss": 0.4655,
+      "step": 8543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.745725092791536,
+      "learning_rate": 6.977714854824709e-06,
+      "loss": 0.4998,
+      "step": 8544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.678203892187683,
+      "learning_rate": 6.97709985317992e-06,
+      "loss": 0.4985,
+      "step": 8545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8505737141603407,
+      "learning_rate": 6.976484820612962e-06,
+      "loss": 0.4965,
+      "step": 8546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5286523098351785,
+      "learning_rate": 6.975869757136711e-06,
+      "loss": 0.3224,
+      "step": 8547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6682930382732577,
+      "learning_rate": 6.975254662764042e-06,
+      "loss": 0.5317,
+      "step": 8548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6515176332252008,
+      "learning_rate": 6.974639537507836e-06,
+      "loss": 0.3524,
+      "step": 8549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8684742758979664,
+      "learning_rate": 6.974024381380965e-06,
+      "loss": 0.5315,
+      "step": 8550
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.5405775308609009,
+      "eval_runtime": 67.9995,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 8550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8236004012513012,
+      "learning_rate": 6.97340919439631e-06,
+      "loss": 0.5092,
+      "step": 8551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7363155995448067,
+      "learning_rate": 6.972793976566748e-06,
+      "loss": 0.5252,
+      "step": 8552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7477783965194655,
+      "learning_rate": 6.972178727905159e-06,
+      "loss": 0.518,
+      "step": 8553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.946813167190017,
+      "learning_rate": 6.971563448424424e-06,
+      "loss": 0.4907,
+      "step": 8554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4015061227634948,
+      "learning_rate": 6.97094813813742e-06,
+      "loss": 0.4953,
+      "step": 8555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3371563067534966,
+      "learning_rate": 6.970332797057031e-06,
+      "loss": 0.6242,
+      "step": 8556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8034279277654046,
+      "learning_rate": 6.969717425196141e-06,
+      "loss": 0.4014,
+      "step": 8557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9041694035192693,
+      "learning_rate": 6.969102022567626e-06,
+      "loss": 0.5821,
+      "step": 8558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7389555539653354,
+      "learning_rate": 6.968486589184374e-06,
+      "loss": 0.5782,
+      "step": 8559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9971689802570598,
+      "learning_rate": 6.967871125059267e-06,
+      "loss": 0.484,
+      "step": 8560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5069492445608192,
+      "learning_rate": 6.967255630205191e-06,
+      "loss": 0.5476,
+      "step": 8561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.995239670676187,
+      "learning_rate": 6.966640104635029e-06,
+      "loss": 0.7441,
+      "step": 8562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8087257460180588,
+      "learning_rate": 6.9660245483616665e-06,
+      "loss": 0.5925,
+      "step": 8563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.123267981583754,
+      "learning_rate": 6.965408961397993e-06,
+      "loss": 0.4679,
+      "step": 8564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.659637650761147,
+      "learning_rate": 6.964793343756891e-06,
+      "loss": 0.6825,
+      "step": 8565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.138931091299346,
+      "learning_rate": 6.9641776954512494e-06,
+      "loss": 0.6415,
+      "step": 8566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7619084937080411,
+      "learning_rate": 6.963562016493959e-06,
+      "loss": 0.4542,
+      "step": 8567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8026808404320813,
+      "learning_rate": 6.9629463068979065e-06,
+      "loss": 0.5071,
+      "step": 8568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0106602146464603,
+      "learning_rate": 6.962330566675981e-06,
+      "loss": 0.7273,
+      "step": 8569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7002481716275166,
+      "learning_rate": 6.961714795841075e-06,
+      "loss": 0.5361,
+      "step": 8570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.713804071528051,
+      "learning_rate": 6.961098994406076e-06,
+      "loss": 0.6184,
+      "step": 8571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0975285572352456,
+      "learning_rate": 6.960483162383876e-06,
+      "loss": 0.5571,
+      "step": 8572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7939986629730391,
+      "learning_rate": 6.959867299787371e-06,
+      "loss": 0.5117,
+      "step": 8573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8009160143196339,
+      "learning_rate": 6.95925140662945e-06,
+      "loss": 0.5741,
+      "step": 8574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7136319401399267,
+      "learning_rate": 6.958635482923006e-06,
+      "loss": 0.494,
+      "step": 8575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6257761672332645,
+      "learning_rate": 6.958019528680935e-06,
+      "loss": 0.457,
+      "step": 8576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6569085629663567,
+      "learning_rate": 6.957403543916132e-06,
+      "loss": 0.4759,
+      "step": 8577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.945134881016213,
+      "learning_rate": 6.956787528641489e-06,
+      "loss": 0.6103,
+      "step": 8578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8122567419417397,
+      "learning_rate": 6.956171482869905e-06,
+      "loss": 0.4798,
+      "step": 8579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5602445084854029,
+      "learning_rate": 6.9555554066142736e-06,
+      "loss": 0.501,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.928794380855918,
+      "learning_rate": 6.954939299887495e-06,
+      "loss": 0.736,
+      "step": 8581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8922945444590202,
+      "learning_rate": 6.954323162702467e-06,
+      "loss": 0.6793,
+      "step": 8582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0450574198506293,
+      "learning_rate": 6.953706995072086e-06,
+      "loss": 0.5307,
+      "step": 8583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7935173797729298,
+      "learning_rate": 6.953090797009252e-06,
+      "loss": 0.5779,
+      "step": 8584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.567578157077902,
+      "learning_rate": 6.952474568526865e-06,
+      "loss": 0.4264,
+      "step": 8585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4957254652615133,
+      "learning_rate": 6.951858309637826e-06,
+      "loss": 0.5338,
+      "step": 8586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.806702367728035,
+      "learning_rate": 6.951242020355032e-06,
+      "loss": 0.5106,
+      "step": 8587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.721313185724387,
+      "learning_rate": 6.95062570069139e-06,
+      "loss": 0.5043,
+      "step": 8588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6592004444177202,
+      "learning_rate": 6.9500093506597995e-06,
+      "loss": 0.488,
+      "step": 8589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7906978065510615,
+      "learning_rate": 6.949392970273164e-06,
+      "loss": 0.4929,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.042755876413758,
+      "learning_rate": 6.948776559544388e-06,
+      "loss": 0.9157,
+      "step": 8591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1297766518795953,
+      "learning_rate": 6.948160118486373e-06,
+      "loss": 0.7001,
+      "step": 8592
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.563501424139489,
+      "learning_rate": 6.9475436471120265e-06,
+      "loss": 0.5677,
+      "step": 8593
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.866378882837601,
+      "learning_rate": 6.946927145434253e-06,
+      "loss": 0.5586,
+      "step": 8594
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.904452462456913,
+      "learning_rate": 6.946310613465956e-06,
+      "loss": 0.4928,
+      "step": 8595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.03601493658665,
+      "learning_rate": 6.945694051220046e-06,
+      "loss": 0.5513,
+      "step": 8596
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1920988084353814,
+      "learning_rate": 6.945077458709431e-06,
+      "loss": 0.5509,
+      "step": 8597
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9080189877593168,
+      "learning_rate": 6.944460835947015e-06,
+      "loss": 0.5606,
+      "step": 8598
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8262652210242045,
+      "learning_rate": 6.94384418294571e-06,
+      "loss": 0.504,
+      "step": 8599
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6649227874057362,
+      "learning_rate": 6.943227499718421e-06,
+      "loss": 0.3962,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5406767129898071,
+      "eval_runtime": 67.9931,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.659493282061497,
+      "learning_rate": 6.9426107862780625e-06,
+      "loss": 0.4795,
+      "step": 8601
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0309269528603453,
+      "learning_rate": 6.941994042637542e-06,
+      "loss": 0.6867,
+      "step": 8602
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7555517341597475,
+      "learning_rate": 6.941377268809775e-06,
+      "loss": 0.5678,
+      "step": 8603
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1283011896998554,
+      "learning_rate": 6.940760464807667e-06,
+      "loss": 0.7764,
+      "step": 8604
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9499757391761965,
+      "learning_rate": 6.940143630644135e-06,
+      "loss": 0.4752,
+      "step": 8605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1528641061197225,
+      "learning_rate": 6.9395267663320895e-06,
+      "loss": 0.5527,
+      "step": 8606
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9958054176140942,
+      "learning_rate": 6.938909871884446e-06,
+      "loss": 0.4702,
+      "step": 8607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6011556787551604,
+      "learning_rate": 6.938292947314118e-06,
+      "loss": 0.536,
+      "step": 8608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6419818149374321,
+      "learning_rate": 6.937675992634022e-06,
+      "loss": 0.5819,
+      "step": 8609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.674463538496651,
+      "learning_rate": 6.937059007857071e-06,
+      "loss": 0.4798,
+      "step": 8610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9129504881884802,
+      "learning_rate": 6.936441992996181e-06,
+      "loss": 0.5357,
+      "step": 8611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9503050284527985,
+      "learning_rate": 6.935824948064271e-06,
+      "loss": 0.6567,
+      "step": 8612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9095230535542513,
+      "learning_rate": 6.935207873074258e-06,
+      "loss": 0.5151,
+      "step": 8613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.07057194299277,
+      "learning_rate": 6.93459076803906e-06,
+      "loss": 0.6029,
+      "step": 8614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5496479377432033,
+      "learning_rate": 6.933973632971594e-06,
+      "loss": 0.4747,
+      "step": 8615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.054977392210811,
+      "learning_rate": 6.933356467884781e-06,
+      "loss": 0.5724,
+      "step": 8616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8271603422213047,
+      "learning_rate": 6.932739272791542e-06,
+      "loss": 0.5461,
+      "step": 8617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.503651310220456,
+      "learning_rate": 6.932122047704794e-06,
+      "loss": 0.6865,
+      "step": 8618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8568896455985586,
+      "learning_rate": 6.931504792637463e-06,
+      "loss": 0.5582,
+      "step": 8619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.587003548216015,
+      "learning_rate": 6.930887507602466e-06,
+      "loss": 0.5054,
+      "step": 8620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0829151668208685,
+      "learning_rate": 6.930270192612729e-06,
+      "loss": 0.5697,
+      "step": 8621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.886415053476499,
+      "learning_rate": 6.929652847681174e-06,
+      "loss": 0.5109,
+      "step": 8622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.032822757198007,
+      "learning_rate": 6.929035472820724e-06,
+      "loss": 0.6105,
+      "step": 8623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.227042389021828,
+      "learning_rate": 6.9284180680443036e-06,
+      "loss": 0.6591,
+      "step": 8624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8223236229215765,
+      "learning_rate": 6.927800633364839e-06,
+      "loss": 0.5996,
+      "step": 8625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8187459190502473,
+      "learning_rate": 6.927183168795254e-06,
+      "loss": 0.465,
+      "step": 8626
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6746123331793044,
+      "learning_rate": 6.926565674348477e-06,
+      "loss": 0.4474,
+      "step": 8627
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8323358153168516,
+      "learning_rate": 6.925948150037433e-06,
+      "loss": 0.5588,
+      "step": 8628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0316082313329953,
+      "learning_rate": 6.925330595875049e-06,
+      "loss": 0.5521,
+      "step": 8629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3245503415941557,
+      "learning_rate": 6.924713011874256e-06,
+      "loss": 0.3606,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6713125941769844,
+      "learning_rate": 6.924095398047982e-06,
+      "loss": 0.4626,
+      "step": 8631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7890312204686367,
+      "learning_rate": 6.923477754409154e-06,
+      "loss": 0.552,
+      "step": 8632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9197913983871446,
+      "learning_rate": 6.922860080970703e-06,
+      "loss": 0.551,
+      "step": 8633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1379937154229327,
+      "learning_rate": 6.92224237774556e-06,
+      "loss": 0.6804,
+      "step": 8634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.746557191278718,
+      "learning_rate": 6.921624644746656e-06,
+      "loss": 0.4582,
+      "step": 8635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.148263146517975,
+      "learning_rate": 6.921006881986926e-06,
+      "loss": 0.5185,
+      "step": 8636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8617644299337186,
+      "learning_rate": 6.9203890894792965e-06,
+      "loss": 0.5553,
+      "step": 8637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6140185996721055,
+      "learning_rate": 6.9197712672367045e-06,
+      "loss": 0.5107,
+      "step": 8638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.003759674269103,
+      "learning_rate": 6.919153415272082e-06,
+      "loss": 0.5752,
+      "step": 8639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.031689936067538,
+      "learning_rate": 6.918535533598365e-06,
+      "loss": 0.5823,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.994035961259673,
+      "learning_rate": 6.917917622228488e-06,
+      "loss": 0.5343,
+      "step": 8641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7425676056526707,
+      "learning_rate": 6.9172996811753865e-06,
+      "loss": 0.5999,
+      "step": 8642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9087462779446773,
+      "learning_rate": 6.916681710451995e-06,
+      "loss": 0.5157,
+      "step": 8643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6491356336649836,
+      "learning_rate": 6.916063710071254e-06,
+      "loss": 0.6303,
+      "step": 8644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.922086132203609,
+      "learning_rate": 6.915445680046097e-06,
+      "loss": 0.7041,
+      "step": 8645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6473840343444897,
+      "learning_rate": 6.914827620389464e-06,
+      "loss": 0.3643,
+      "step": 8646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8651469623359704,
+      "learning_rate": 6.914209531114294e-06,
+      "loss": 0.6392,
+      "step": 8647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9921609696161717,
+      "learning_rate": 6.913591412233526e-06,
+      "loss": 0.4758,
+      "step": 8648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5400797404834066,
+      "learning_rate": 6.9129732637601e-06,
+      "loss": 0.5072,
+      "step": 8649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6568040917136655,
+      "learning_rate": 6.912355085706955e-06,
+      "loss": 0.5136,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5412369966506958,
+      "eval_runtime": 67.99,
+      "eval_samples_per_second": 50.154,
+      "eval_steps_per_second": 1.574,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.674396248432447,
+      "learning_rate": 6.911736878087036e-06,
+      "loss": 0.6946,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7939921702425774,
+      "learning_rate": 6.91111864091328e-06,
+      "loss": 0.3707,
+      "step": 8652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7192348632360637,
+      "learning_rate": 6.910500374198632e-06,
+      "loss": 0.513,
+      "step": 8653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7907259226437944,
+      "learning_rate": 6.909882077956036e-06,
+      "loss": 0.6409,
+      "step": 8654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8235399127480378,
+      "learning_rate": 6.909263752198433e-06,
+      "loss": 0.5837,
+      "step": 8655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4534385952594249,
+      "learning_rate": 6.908645396938769e-06,
+      "loss": 0.4569,
+      "step": 8656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6851612024341698,
+      "learning_rate": 6.908027012189989e-06,
+      "loss": 0.4976,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7870145023417632,
+      "learning_rate": 6.907408597965038e-06,
+      "loss": 0.3861,
+      "step": 8658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.824931589389376,
+      "learning_rate": 6.906790154276862e-06,
+      "loss": 0.545,
+      "step": 8659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.906808703365079,
+      "learning_rate": 6.906171681138409e-06,
+      "loss": 0.504,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9766977187586592,
+      "learning_rate": 6.905553178562624e-06,
+      "loss": 0.6521,
+      "step": 8661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6228412812561477,
+      "learning_rate": 6.904934646562456e-06,
+      "loss": 0.5693,
+      "step": 8662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.164847295961308,
+      "learning_rate": 6.904316085150855e-06,
+      "loss": 0.8892,
+      "step": 8663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7890362361013101,
+      "learning_rate": 6.903697494340767e-06,
+      "loss": 0.6031,
+      "step": 8664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8539653443823934,
+      "learning_rate": 6.903078874145146e-06,
+      "loss": 0.4428,
+      "step": 8665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.777440209010927,
+      "learning_rate": 6.9024602245769405e-06,
+      "loss": 0.5551,
+      "step": 8666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.725960234818395,
+      "learning_rate": 6.901841545649101e-06,
+      "loss": 0.5328,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.014531037721224,
+      "learning_rate": 6.901222837374578e-06,
+      "loss": 0.7224,
+      "step": 8668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8664729003115645,
+      "learning_rate": 6.900604099766326e-06,
+      "loss": 0.5479,
+      "step": 8669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8199558839474692,
+      "learning_rate": 6.899985332837297e-06,
+      "loss": 0.5343,
+      "step": 8670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.145972245304769,
+      "learning_rate": 6.899366536600445e-06,
+      "loss": 0.6156,
+      "step": 8671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.417631122754757,
+      "learning_rate": 6.898747711068723e-06,
+      "loss": 0.4254,
+      "step": 8672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7426679306948194,
+      "learning_rate": 6.898128856255087e-06,
+      "loss": 0.429,
+      "step": 8673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9924862095756584,
+      "learning_rate": 6.897509972172491e-06,
+      "loss": 0.6197,
+      "step": 8674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9963798632740566,
+      "learning_rate": 6.896891058833893e-06,
+      "loss": 0.5577,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8304148790120993,
+      "learning_rate": 6.8962721162522485e-06,
+      "loss": 0.5652,
+      "step": 8676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5215643707124364,
+      "learning_rate": 6.895653144440513e-06,
+      "loss": 0.5822,
+      "step": 8677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5671615464405564,
+      "learning_rate": 6.895034143411645e-06,
+      "loss": 0.4112,
+      "step": 8678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7132438870723037,
+      "learning_rate": 6.894415113178607e-06,
+      "loss": 0.5002,
+      "step": 8679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7563907171016053,
+      "learning_rate": 6.893796053754352e-06,
+      "loss": 0.4694,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1170350644625433,
+      "learning_rate": 6.893176965151844e-06,
+      "loss": 0.4486,
+      "step": 8681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.257918374031791,
+      "learning_rate": 6.8925578473840405e-06,
+      "loss": 0.592,
+      "step": 8682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.716828201066442,
+      "learning_rate": 6.891938700463904e-06,
+      "loss": 0.5144,
+      "step": 8683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6255156911251611,
+      "learning_rate": 6.891319524404396e-06,
+      "loss": 0.5088,
+      "step": 8684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0685701416146225,
+      "learning_rate": 6.890700319218476e-06,
+      "loss": 0.469,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7285989329296754,
+      "learning_rate": 6.89008108491911e-06,
+      "loss": 0.6053,
+      "step": 8686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5700263905029825,
+      "learning_rate": 6.88946182151926e-06,
+      "loss": 0.4863,
+      "step": 8687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7787216503392693,
+      "learning_rate": 6.8888425290318895e-06,
+      "loss": 0.4683,
+      "step": 8688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6352663563286771,
+      "learning_rate": 6.888223207469963e-06,
+      "loss": 0.4787,
+      "step": 8689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8641144193616133,
+      "learning_rate": 6.8876038568464455e-06,
+      "loss": 0.5919,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8413998543639023,
+      "learning_rate": 6.886984477174305e-06,
+      "loss": 0.5608,
+      "step": 8691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5351407684332685,
+      "learning_rate": 6.886365068466505e-06,
+      "loss": 0.4921,
+      "step": 8692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7564098510187391,
+      "learning_rate": 6.885745630736011e-06,
+      "loss": 0.5293,
+      "step": 8693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1129871100746374,
+      "learning_rate": 6.885126163995795e-06,
+      "loss": 0.5974,
+      "step": 8694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.844353147267103,
+      "learning_rate": 6.8845066682588236e-06,
+      "loss": 0.5551,
+      "step": 8695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.217728587567656,
+      "learning_rate": 6.883887143538065e-06,
+      "loss": 0.4967,
+      "step": 8696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7147420431394984,
+      "learning_rate": 6.88326758984649e-06,
+      "loss": 0.4621,
+      "step": 8697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0828991790187565,
+      "learning_rate": 6.882648007197063e-06,
+      "loss": 0.5992,
+      "step": 8698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3604899798411507,
+      "learning_rate": 6.882028395602762e-06,
+      "loss": 0.693,
+      "step": 8699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.771873397110332,
+      "learning_rate": 6.881408755076554e-06,
+      "loss": 0.6861,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5400500893592834,
+      "eval_runtime": 67.9872,
+      "eval_samples_per_second": 50.156,
+      "eval_steps_per_second": 1.574,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.435109947983848,
+      "learning_rate": 6.880789085631412e-06,
+      "loss": 0.4342,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3251236714389765,
+      "learning_rate": 6.880169387280308e-06,
+      "loss": 0.5522,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.098782698378118,
+      "learning_rate": 6.879549660036214e-06,
+      "loss": 0.6757,
+      "step": 8703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6837728120188966,
+      "learning_rate": 6.878929903912106e-06,
+      "loss": 0.4269,
+      "step": 8704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9220416237991225,
+      "learning_rate": 6.878310118920958e-06,
+      "loss": 0.6438,
+      "step": 8705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9099358290537933,
+      "learning_rate": 6.87769030507574e-06,
+      "loss": 0.554,
+      "step": 8706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.903948297332591,
+      "learning_rate": 6.877070462389435e-06,
+      "loss": 0.5773,
+      "step": 8707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2498648316620917,
+      "learning_rate": 6.8764505908750145e-06,
+      "loss": 0.3079,
+      "step": 8708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6110831521256146,
+      "learning_rate": 6.875830690545455e-06,
+      "loss": 0.3841,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9934097376305646,
+      "learning_rate": 6.875210761413735e-06,
+      "loss": 0.5958,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7204811054012514,
+      "learning_rate": 6.8745908034928326e-06,
+      "loss": 0.5094,
+      "step": 8711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0084250557584022,
+      "learning_rate": 6.8739708167957245e-06,
+      "loss": 0.6063,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9085568846817658,
+      "learning_rate": 6.873350801335391e-06,
+      "loss": 0.657,
+      "step": 8713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.67650740692771,
+      "learning_rate": 6.872730757124809e-06,
+      "loss": 0.5873,
+      "step": 8714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6253065577512373,
+      "learning_rate": 6.872110684176965e-06,
+      "loss": 0.4696,
+      "step": 8715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.663053951362995,
+      "learning_rate": 6.871490582504835e-06,
+      "loss": 0.4234,
+      "step": 8716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6222195109946498,
+      "learning_rate": 6.870870452121401e-06,
+      "loss": 0.5148,
+      "step": 8717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.555175316143455,
+      "learning_rate": 6.870250293039646e-06,
+      "loss": 0.3983,
+      "step": 8718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7358041589348825,
+      "learning_rate": 6.869630105272551e-06,
+      "loss": 0.4165,
+      "step": 8719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3988863460010625,
+      "learning_rate": 6.869009888833101e-06,
+      "loss": 0.3921,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0612545213984013,
+      "learning_rate": 6.868389643734281e-06,
+      "loss": 0.5448,
+      "step": 8721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.100749252693489,
+      "learning_rate": 6.867769369989071e-06,
+      "loss": 0.5779,
+      "step": 8722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5738347251043667,
+      "learning_rate": 6.867149067610461e-06,
+      "loss": 0.5566,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.753997234018586,
+      "learning_rate": 6.866528736611433e-06,
+      "loss": 0.5049,
+      "step": 8724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5840980209297906,
+      "learning_rate": 6.865908377004976e-06,
+      "loss": 0.4095,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6418285787034401,
+      "learning_rate": 6.865287988804076e-06,
+      "loss": 0.4495,
+      "step": 8726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6894552125281814,
+      "learning_rate": 6.864667572021717e-06,
+      "loss": 0.4432,
+      "step": 8727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5563794389106171,
+      "learning_rate": 6.8640471266708915e-06,
+      "loss": 0.5513,
+      "step": 8728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6370403246895378,
+      "learning_rate": 6.863426652764588e-06,
+      "loss": 0.4962,
+      "step": 8729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8296881606500062,
+      "learning_rate": 6.862806150315793e-06,
+      "loss": 0.5109,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8871755259095382,
+      "learning_rate": 6.862185619337499e-06,
+      "loss": 0.5493,
+      "step": 8731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7640075581099701,
+      "learning_rate": 6.8615650598426945e-06,
+      "loss": 0.3821,
+      "step": 8732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8698844378496675,
+      "learning_rate": 6.860944471844371e-06,
+      "loss": 0.5515,
+      "step": 8733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7229661259943314,
+      "learning_rate": 6.860323855355521e-06,
+      "loss": 0.4485,
+      "step": 8734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8848286680022484,
+      "learning_rate": 6.859703210389134e-06,
+      "loss": 0.5585,
+      "step": 8735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7806167612714285,
+      "learning_rate": 6.859082536958206e-06,
+      "loss": 0.4769,
+      "step": 8736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.701963747282994,
+      "learning_rate": 6.858461835075729e-06,
+      "loss": 0.498,
+      "step": 8737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7857614265493618,
+      "learning_rate": 6.857841104754697e-06,
+      "loss": 0.5697,
+      "step": 8738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.213796324077444,
+      "learning_rate": 6.857220346008106e-06,
+      "loss": 0.6289,
+      "step": 8739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0245856809377307,
+      "learning_rate": 6.85659955884895e-06,
+      "loss": 0.499,
+      "step": 8740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.217070568981951,
+      "learning_rate": 6.855978743290224e-06,
+      "loss": 0.5159,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6550540383593861,
+      "learning_rate": 6.855357899344926e-06,
+      "loss": 0.6656,
+      "step": 8742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7064283696339617,
+      "learning_rate": 6.854737027026051e-06,
+      "loss": 0.4622,
+      "step": 8743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5602910313532357,
+      "learning_rate": 6.854116126346599e-06,
+      "loss": 0.3893,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5975733170885607,
+      "learning_rate": 6.853495197319568e-06,
+      "loss": 0.5902,
+      "step": 8745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2356906478137173,
+      "learning_rate": 6.852874239957953e-06,
+      "loss": 0.6112,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7068072845722058,
+      "learning_rate": 6.852253254274757e-06,
+      "loss": 0.5068,
+      "step": 8747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.633217892072193,
+      "learning_rate": 6.851632240282979e-06,
+      "loss": 0.4555,
+      "step": 8748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7904778185707917,
+      "learning_rate": 6.85101119799562e-06,
+      "loss": 0.594,
+      "step": 8749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4369037916018987,
+      "learning_rate": 6.850390127425682e-06,
+      "loss": 0.417,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5395739674568176,
+      "eval_runtime": 67.9579,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.575,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.184126945700475,
+      "learning_rate": 6.849769028586163e-06,
+      "loss": 0.5502,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8032016192045042,
+      "learning_rate": 6.849147901490069e-06,
+      "loss": 0.6229,
+      "step": 8752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.757485175517596,
+      "learning_rate": 6.848526746150401e-06,
+      "loss": 0.5842,
+      "step": 8753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8783448853920508,
+      "learning_rate": 6.847905562580164e-06,
+      "loss": 0.6502,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2118522478722147,
+      "learning_rate": 6.847284350792361e-06,
+      "loss": 0.6452,
+      "step": 8755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6848112429311366,
+      "learning_rate": 6.846663110799997e-06,
+      "loss": 0.4682,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7008992825567646,
+      "learning_rate": 6.846041842616078e-06,
+      "loss": 0.4709,
+      "step": 8757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.436351093415407,
+      "learning_rate": 6.845420546253608e-06,
+      "loss": 0.4324,
+      "step": 8758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8573861211457292,
+      "learning_rate": 6.8447992217255935e-06,
+      "loss": 0.5336,
+      "step": 8759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9581483615286472,
+      "learning_rate": 6.844177869045046e-06,
+      "loss": 0.5345,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5285100696864087,
+      "learning_rate": 6.8435564882249674e-06,
+      "loss": 0.475,
+      "step": 8761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.510253101526285,
+      "learning_rate": 6.842935079278368e-06,
+      "loss": 0.3452,
+      "step": 8762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.827806574032183,
+      "learning_rate": 6.8423136422182565e-06,
+      "loss": 0.7455,
+      "step": 8763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.654270830989414,
+      "learning_rate": 6.841692177057644e-06,
+      "loss": 0.5817,
+      "step": 8764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5938687381376482,
+      "learning_rate": 6.841070683809538e-06,
+      "loss": 0.4133,
+      "step": 8765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.644379613401214,
+      "learning_rate": 6.840449162486951e-06,
+      "loss": 0.5025,
+      "step": 8766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5038377000191743,
+      "learning_rate": 6.839827613102893e-06,
+      "loss": 0.4675,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.781819500518713,
+      "learning_rate": 6.839206035670376e-06,
+      "loss": 0.4516,
+      "step": 8768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.957534433906976,
+      "learning_rate": 6.838584430202414e-06,
+      "loss": 0.6154,
+      "step": 8769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8393915811471702,
+      "learning_rate": 6.837962796712016e-06,
+      "loss": 0.5677,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.563330827089301,
+      "learning_rate": 6.837341135212199e-06,
+      "loss": 0.5325,
+      "step": 8771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5638268287205288,
+      "learning_rate": 6.836719445715976e-06,
+      "loss": 0.4326,
+      "step": 8772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9700366121092043,
+      "learning_rate": 6.836097728236362e-06,
+      "loss": 0.5432,
+      "step": 8773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6226506926110913,
+      "learning_rate": 6.835475982786371e-06,
+      "loss": 0.5644,
+      "step": 8774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8025818793739758,
+      "learning_rate": 6.834854209379019e-06,
+      "loss": 0.5441,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6288614432388238,
+      "learning_rate": 6.834232408027326e-06,
+      "loss": 0.5051,
+      "step": 8776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6993560632783153,
+      "learning_rate": 6.833610578744305e-06,
+      "loss": 0.5224,
+      "step": 8777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8178540332285782,
+      "learning_rate": 6.832988721542973e-06,
+      "loss": 0.5283,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8219563085561703,
+      "learning_rate": 6.832366836436353e-06,
+      "loss": 0.3834,
+      "step": 8779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9945680213617614,
+      "learning_rate": 6.83174492343746e-06,
+      "loss": 0.9028,
+      "step": 8780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0407199636747793,
+      "learning_rate": 6.831122982559314e-06,
+      "loss": 0.4726,
+      "step": 8781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.006417612707001,
+      "learning_rate": 6.830501013814934e-06,
+      "loss": 0.6237,
+      "step": 8782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5256519208064971,
+      "learning_rate": 6.829879017217343e-06,
+      "loss": 0.3656,
+      "step": 8783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7655753716588456,
+      "learning_rate": 6.829256992779561e-06,
+      "loss": 0.5405,
+      "step": 8784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1592892045113077,
+      "learning_rate": 6.828634940514609e-06,
+      "loss": 0.7662,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8545174133252424,
+      "learning_rate": 6.828012860435508e-06,
+      "loss": 0.6206,
+      "step": 8786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4035554880941234,
+      "learning_rate": 6.8273907525552855e-06,
+      "loss": 0.3722,
+      "step": 8787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7764657580555214,
+      "learning_rate": 6.8267686168869605e-06,
+      "loss": 0.5284,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.007667947901432,
+      "learning_rate": 6.8261464534435586e-06,
+      "loss": 0.5138,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7295512023172985,
+      "learning_rate": 6.8255242622381055e-06,
+      "loss": 0.5225,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6536369531752908,
+      "learning_rate": 6.8249020432836235e-06,
+      "loss": 0.4295,
+      "step": 8791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7688060733747415,
+      "learning_rate": 6.824279796593142e-06,
+      "loss": 0.6362,
+      "step": 8792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6392218274029655,
+      "learning_rate": 6.823657522179686e-06,
+      "loss": 0.493,
+      "step": 8793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5317812545043468,
+      "learning_rate": 6.8230352200562786e-06,
+      "loss": 0.3921,
+      "step": 8794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7029987939812896,
+      "learning_rate": 6.822412890235953e-06,
+      "loss": 0.4571,
+      "step": 8795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0787045769255554,
+      "learning_rate": 6.821790532731734e-06,
+      "loss": 0.5809,
+      "step": 8796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7273089070024301,
+      "learning_rate": 6.821168147556653e-06,
+      "loss": 0.4645,
+      "step": 8797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.061209934090238,
+      "learning_rate": 6.8205457347237366e-06,
+      "loss": 0.4547,
+      "step": 8798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3716254590361476,
+      "learning_rate": 6.8199232942460145e-06,
+      "loss": 0.5079,
+      "step": 8799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.044963240677361,
+      "learning_rate": 6.81930082613652e-06,
+      "loss": 0.6537,
+      "step": 8800
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5390601754188538,
+      "eval_runtime": 67.9714,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 8800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.055362158823467,
+      "learning_rate": 6.81867833040828e-06,
+      "loss": 0.4998,
+      "step": 8801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7654683022198536,
+      "learning_rate": 6.8180558070743274e-06,
+      "loss": 0.5057,
+      "step": 8802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.041560780738976,
+      "learning_rate": 6.817433256147698e-06,
+      "loss": 0.6044,
+      "step": 8803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5305102191914113,
+      "learning_rate": 6.816810677641419e-06,
+      "loss": 0.5084,
+      "step": 8804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0752406836914505,
+      "learning_rate": 6.816188071568528e-06,
+      "loss": 0.6946,
+      "step": 8805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4917319188130593,
+      "learning_rate": 6.8155654379420585e-06,
+      "loss": 0.4776,
+      "step": 8806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7224491737075558,
+      "learning_rate": 6.814942776775042e-06,
+      "loss": 0.5748,
+      "step": 8807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9056824519887827,
+      "learning_rate": 6.814320088080515e-06,
+      "loss": 0.6346,
+      "step": 8808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.91290927886194,
+      "learning_rate": 6.8136973718715134e-06,
+      "loss": 0.6163,
+      "step": 8809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5284755520571787,
+      "learning_rate": 6.813074628161073e-06,
+      "loss": 0.5035,
+      "step": 8810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5071655556017693,
+      "learning_rate": 6.812451856962232e-06,
+      "loss": 0.3625,
+      "step": 8811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7407974851839347,
+      "learning_rate": 6.811829058288027e-06,
+      "loss": 0.4496,
+      "step": 8812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7411519790710503,
+      "learning_rate": 6.811206232151493e-06,
+      "loss": 0.4576,
+      "step": 8813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0063489961793533,
+      "learning_rate": 6.810583378565675e-06,
+      "loss": 0.4838,
+      "step": 8814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5745196097819871,
+      "learning_rate": 6.809960497543604e-06,
+      "loss": 0.4073,
+      "step": 8815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7108646338887714,
+      "learning_rate": 6.8093375890983245e-06,
+      "loss": 0.4416,
+      "step": 8816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5589471536084636,
+      "learning_rate": 6.808714653242877e-06,
+      "loss": 0.5311,
+      "step": 8817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8509478751859312,
+      "learning_rate": 6.808091689990301e-06,
+      "loss": 0.6157,
+      "step": 8818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0229823613891544,
+      "learning_rate": 6.807468699353638e-06,
+      "loss": 0.5497,
+      "step": 8819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7028532656984108,
+      "learning_rate": 6.806845681345928e-06,
+      "loss": 0.5464,
+      "step": 8820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.73018509875592,
+      "learning_rate": 6.806222635980217e-06,
+      "loss": 0.4295,
+      "step": 8821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9020987899662785,
+      "learning_rate": 6.805599563269547e-06,
+      "loss": 0.7738,
+      "step": 8822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5128428482723595,
+      "learning_rate": 6.80497646322696e-06,
+      "loss": 0.4109,
+      "step": 8823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.794288547809859,
+      "learning_rate": 6.804353335865502e-06,
+      "loss": 0.433,
+      "step": 8824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7620331754838914,
+      "learning_rate": 6.803730181198216e-06,
+      "loss": 0.4771,
+      "step": 8825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9488884661093497,
+      "learning_rate": 6.803106999238148e-06,
+      "loss": 0.5536,
+      "step": 8826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3077984982603867,
+      "learning_rate": 6.802483789998346e-06,
+      "loss": 0.5772,
+      "step": 8827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.857934093539009,
+      "learning_rate": 6.801860553491855e-06,
+      "loss": 0.4705,
+      "step": 8828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7811406611336362,
+      "learning_rate": 6.80123728973172e-06,
+      "loss": 0.676,
+      "step": 8829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8047855247995683,
+      "learning_rate": 6.800613998730993e-06,
+      "loss": 0.7726,
+      "step": 8830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8830309030429073,
+      "learning_rate": 6.7999906805027185e-06,
+      "loss": 0.4541,
+      "step": 8831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7751024247141092,
+      "learning_rate": 6.7993673350599465e-06,
+      "loss": 0.5367,
+      "step": 8832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.239181117535282,
+      "learning_rate": 6.798743962415727e-06,
+      "loss": 0.5788,
+      "step": 8833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5506960014663058,
+      "learning_rate": 6.7981205625831085e-06,
+      "loss": 0.458,
+      "step": 8834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5500378166255953,
+      "learning_rate": 6.797497135575142e-06,
+      "loss": 0.5145,
+      "step": 8835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5279473789736864,
+      "learning_rate": 6.79687368140488e-06,
+      "loss": 0.4271,
+      "step": 8836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.030314074948522,
+      "learning_rate": 6.796250200085373e-06,
+      "loss": 0.5678,
+      "step": 8837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.878217638668873,
+      "learning_rate": 6.795626691629672e-06,
+      "loss": 0.6609,
+      "step": 8838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.407773487185688,
+      "learning_rate": 6.795003156050832e-06,
+      "loss": 0.4296,
+      "step": 8839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9069751349600597,
+      "learning_rate": 6.794379593361905e-06,
+      "loss": 0.3375,
+      "step": 8840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.304301596517241,
+      "learning_rate": 6.793756003575945e-06,
+      "loss": 0.6141,
+      "step": 8841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8736669687883254,
+      "learning_rate": 6.793132386706006e-06,
+      "loss": 0.5733,
+      "step": 8842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.435784816440325,
+      "learning_rate": 6.7925087427651455e-06,
+      "loss": 0.772,
+      "step": 8843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.790357339546339,
+      "learning_rate": 6.791885071766416e-06,
+      "loss": 0.5547,
+      "step": 8844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6404644946513274,
+      "learning_rate": 6.791261373722876e-06,
+      "loss": 0.491,
+      "step": 8845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9854803890117918,
+      "learning_rate": 6.7906376486475804e-06,
+      "loss": 0.4853,
+      "step": 8846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8337870762786286,
+      "learning_rate": 6.790013896553587e-06,
+      "loss": 0.4731,
+      "step": 8847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5405961562397674,
+      "learning_rate": 6.789390117453953e-06,
+      "loss": 0.4576,
+      "step": 8848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4225307592365533,
+      "learning_rate": 6.788766311361739e-06,
+      "loss": 0.4559,
+      "step": 8849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.671172231420834,
+      "learning_rate": 6.788142478290003e-06,
+      "loss": 0.5367,
+      "step": 8850
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5382565855979919,
+      "eval_runtime": 67.9654,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 8850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1399691654517503,
+      "learning_rate": 6.787518618251804e-06,
+      "loss": 0.6167,
+      "step": 8851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.08806137294351,
+      "learning_rate": 6.7868947312602016e-06,
+      "loss": 0.606,
+      "step": 8852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5519138740438845,
+      "learning_rate": 6.786270817328258e-06,
+      "loss": 0.558,
+      "step": 8853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9802734222071152,
+      "learning_rate": 6.785646876469035e-06,
+      "loss": 0.5334,
+      "step": 8854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5179945685950662,
+      "learning_rate": 6.785022908695592e-06,
+      "loss": 0.5293,
+      "step": 8855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6740969065897002,
+      "learning_rate": 6.784398914020993e-06,
+      "loss": 0.4811,
+      "step": 8856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6370849320159846,
+      "learning_rate": 6.7837748924583e-06,
+      "loss": 0.6257,
+      "step": 8857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7852891139733926,
+      "learning_rate": 6.7831508440205775e-06,
+      "loss": 0.5529,
+      "step": 8858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8891582223593206,
+      "learning_rate": 6.78252676872089e-06,
+      "loss": 0.599,
+      "step": 8859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6512094669565496,
+      "learning_rate": 6.781902666572299e-06,
+      "loss": 0.5807,
+      "step": 8860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7126665737517812,
+      "learning_rate": 6.781278537587873e-06,
+      "loss": 0.5057,
+      "step": 8861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8688756106809608,
+      "learning_rate": 6.780654381780676e-06,
+      "loss": 0.5936,
+      "step": 8862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8796599966197394,
+      "learning_rate": 6.780030199163775e-06,
+      "loss": 0.7204,
+      "step": 8863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5500751487430802,
+      "learning_rate": 6.779405989750238e-06,
+      "loss": 0.4083,
+      "step": 8864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6866565508304494,
+      "learning_rate": 6.77878175355313e-06,
+      "loss": 0.43,
+      "step": 8865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8151964543910213,
+      "learning_rate": 6.778157490585522e-06,
+      "loss": 0.5332,
+      "step": 8866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.011472822941833,
+      "learning_rate": 6.777533200860479e-06,
+      "loss": 0.6813,
+      "step": 8867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.880433773465686,
+      "learning_rate": 6.7769088843910725e-06,
+      "loss": 0.4857,
+      "step": 8868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9739345520974314,
+      "learning_rate": 6.776284541190372e-06,
+      "loss": 0.5044,
+      "step": 8869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.024751490257435,
+      "learning_rate": 6.775660171271447e-06,
+      "loss": 0.6366,
+      "step": 8870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5833926505927431,
+      "learning_rate": 6.7750357746473685e-06,
+      "loss": 0.4426,
+      "step": 8871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8333089914146343,
+      "learning_rate": 6.7744113513312095e-06,
+      "loss": 0.4637,
+      "step": 8872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.087066410453923,
+      "learning_rate": 6.773786901336039e-06,
+      "loss": 0.5776,
+      "step": 8873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8488113408568911,
+      "learning_rate": 6.773162424674931e-06,
+      "loss": 0.4117,
+      "step": 8874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7097607237252828,
+      "learning_rate": 6.772537921360957e-06,
+      "loss": 0.5932,
+      "step": 8875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8701963977924647,
+      "learning_rate": 6.771913391407194e-06,
+      "loss": 0.4264,
+      "step": 8876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5569663495583852,
+      "learning_rate": 6.771288834826713e-06,
+      "loss": 0.4478,
+      "step": 8877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9287522969503534,
+      "learning_rate": 6.770664251632592e-06,
+      "loss": 0.4477,
+      "step": 8878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.661806185487496,
+      "learning_rate": 6.770039641837901e-06,
+      "loss": 0.5173,
+      "step": 8879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5602253397359445,
+      "learning_rate": 6.7694150054557216e-06,
+      "loss": 0.4927,
+      "step": 8880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.825925702507115,
+      "learning_rate": 6.768790342499126e-06,
+      "loss": 0.4534,
+      "step": 8881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5706940902766016,
+      "learning_rate": 6.768165652981195e-06,
+      "loss": 0.4513,
+      "step": 8882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.585232592961546,
+      "learning_rate": 6.767540936915e-06,
+      "loss": 0.4854,
+      "step": 8883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8077509051865241,
+      "learning_rate": 6.7669161943136254e-06,
+      "loss": 0.3985,
+      "step": 8884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5525162763808806,
+      "learning_rate": 6.766291425190147e-06,
+      "loss": 0.4229,
+      "step": 8885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0593950137261663,
+      "learning_rate": 6.7656666295576436e-06,
+      "loss": 0.5442,
+      "step": 8886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9553450653485704,
+      "learning_rate": 6.765041807429194e-06,
+      "loss": 0.4947,
+      "step": 8887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6140036262822282,
+      "learning_rate": 6.764416958817883e-06,
+      "loss": 0.4963,
+      "step": 8888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9349395446128377,
+      "learning_rate": 6.763792083736786e-06,
+      "loss": 0.6199,
+      "step": 8889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.633195881207857,
+      "learning_rate": 6.763167182198987e-06,
+      "loss": 0.4477,
+      "step": 8890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8686132472011099,
+      "learning_rate": 6.76254225421757e-06,
+      "loss": 0.6005,
+      "step": 8891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6874627252815442,
+      "learning_rate": 6.761917299805613e-06,
+      "loss": 0.4022,
+      "step": 8892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.024357845226284,
+      "learning_rate": 6.7612923189762015e-06,
+      "loss": 0.5084,
+      "step": 8893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9729381758693882,
+      "learning_rate": 6.760667311742417e-06,
+      "loss": 0.5878,
+      "step": 8894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.510487385863476,
+      "learning_rate": 6.760042278117348e-06,
+      "loss": 0.5353,
+      "step": 8895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.00257856996558,
+      "learning_rate": 6.759417218114076e-06,
+      "loss": 0.6018,
+      "step": 8896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7479710829173924,
+      "learning_rate": 6.758792131745686e-06,
+      "loss": 0.5037,
+      "step": 8897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.436873493171192,
+      "learning_rate": 6.758167019025264e-06,
+      "loss": 0.5157,
+      "step": 8898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5759564538233832,
+      "learning_rate": 6.757541879965899e-06,
+      "loss": 0.5314,
+      "step": 8899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.299321941784006,
+      "learning_rate": 6.756916714580672e-06,
+      "loss": 0.7096,
+      "step": 8900
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5392040610313416,
+      "eval_runtime": 67.9338,
+      "eval_samples_per_second": 50.196,
+      "eval_steps_per_second": 1.575,
+      "step": 8900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7059885786604394,
+      "learning_rate": 6.756291522882679e-06,
+      "loss": 0.5433,
+      "step": 8901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8249159054102893,
+      "learning_rate": 6.755666304885001e-06,
+      "loss": 0.6126,
+      "step": 8902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8368678666186775,
+      "learning_rate": 6.7550410606007285e-06,
+      "loss": 0.5239,
+      "step": 8903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.087317255627488,
+      "learning_rate": 6.754415790042952e-06,
+      "loss": 0.6782,
+      "step": 8904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591954006254952,
+      "learning_rate": 6.753790493224757e-06,
+      "loss": 0.5519,
+      "step": 8905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7015075777754307,
+      "learning_rate": 6.753165170159242e-06,
+      "loss": 0.4906,
+      "step": 8906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8599481033068184,
+      "learning_rate": 6.75253982085949e-06,
+      "loss": 0.5359,
+      "step": 8907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8700066271620626,
+      "learning_rate": 6.751914445338594e-06,
+      "loss": 0.5044,
+      "step": 8908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9258028018404179,
+      "learning_rate": 6.751289043609648e-06,
+      "loss": 0.56,
+      "step": 8909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6509045032370622,
+      "learning_rate": 6.750663615685741e-06,
+      "loss": 0.5405,
+      "step": 8910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8022929341409062,
+      "learning_rate": 6.750038161579971e-06,
+      "loss": 0.4524,
+      "step": 8911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7767161995135974,
+      "learning_rate": 6.749412681305426e-06,
+      "loss": 0.5821,
+      "step": 8912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.96890827493042,
+      "learning_rate": 6.7487871748752035e-06,
+      "loss": 0.6716,
+      "step": 8913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.997686398411329,
+      "learning_rate": 6.748161642302396e-06,
+      "loss": 0.588,
+      "step": 8914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5473979339992088,
+      "learning_rate": 6.747536083600102e-06,
+      "loss": 0.3982,
+      "step": 8915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9076490549846232,
+      "learning_rate": 6.746910498781414e-06,
+      "loss": 0.4981,
+      "step": 8916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3550363817555273,
+      "learning_rate": 6.746284887859429e-06,
+      "loss": 0.5804,
+      "step": 8917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7509492348297078,
+      "learning_rate": 6.745659250847244e-06,
+      "loss": 0.6461,
+      "step": 8918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6249963666580294,
+      "learning_rate": 6.745033587757956e-06,
+      "loss": 0.4785,
+      "step": 8919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8261468554234233,
+      "learning_rate": 6.744407898604664e-06,
+      "loss": 0.4763,
+      "step": 8920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.246616252125342,
+      "learning_rate": 6.7437821834004645e-06,
+      "loss": 0.5248,
+      "step": 8921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7908965298914912,
+      "learning_rate": 6.743156442158458e-06,
+      "loss": 0.4657,
+      "step": 8922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5893588972330968,
+      "learning_rate": 6.742530674891744e-06,
+      "loss": 0.3858,
+      "step": 8923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5532272032438093,
+      "learning_rate": 6.74190488161342e-06,
+      "loss": 0.4628,
+      "step": 8924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9248307323231324,
+      "learning_rate": 6.741279062336592e-06,
+      "loss": 0.6417,
+      "step": 8925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8346260552348836,
+      "learning_rate": 6.740653217074356e-06,
+      "loss": 0.4318,
+      "step": 8926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7763069987404414,
+      "learning_rate": 6.740027345839815e-06,
+      "loss": 0.6765,
+      "step": 8927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8643053102831728,
+      "learning_rate": 6.739401448646073e-06,
+      "loss": 0.5347,
+      "step": 8928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8367184484322885,
+      "learning_rate": 6.738775525506232e-06,
+      "loss": 0.487,
+      "step": 8929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.943276432071611,
+      "learning_rate": 6.738149576433392e-06,
+      "loss": 0.5649,
+      "step": 8930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8204556930780291,
+      "learning_rate": 6.737523601440663e-06,
+      "loss": 0.6198,
+      "step": 8931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8383584763369167,
+      "learning_rate": 6.7368976005411435e-06,
+      "loss": 0.4929,
+      "step": 8932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7895360883246474,
+      "learning_rate": 6.736271573747942e-06,
+      "loss": 0.6631,
+      "step": 8933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.654215085551337,
+      "learning_rate": 6.735645521074163e-06,
+      "loss": 0.4399,
+      "step": 8934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.520151710880392,
+      "learning_rate": 6.7350194425329115e-06,
+      "loss": 0.4456,
+      "step": 8935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9441726525697418,
+      "learning_rate": 6.734393338137297e-06,
+      "loss": 0.4973,
+      "step": 8936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.737878399382875,
+      "learning_rate": 6.733767207900425e-06,
+      "loss": 0.5272,
+      "step": 8937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8547295521156044,
+      "learning_rate": 6.7331410518354005e-06,
+      "loss": 0.5621,
+      "step": 8938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6333342997448328,
+      "learning_rate": 6.7325148699553345e-06,
+      "loss": 0.5124,
+      "step": 8939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4564356010517918,
+      "learning_rate": 6.731888662273336e-06,
+      "loss": 0.2819,
+      "step": 8940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7160215208318323,
+      "learning_rate": 6.7312624288025145e-06,
+      "loss": 0.4979,
+      "step": 8941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9168888766384933,
+      "learning_rate": 6.730636169555976e-06,
+      "loss": 0.7005,
+      "step": 8942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.57760252256519,
+      "learning_rate": 6.730009884546837e-06,
+      "loss": 0.4005,
+      "step": 8943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7893725091969412,
+      "learning_rate": 6.729383573788202e-06,
+      "loss": 0.5429,
+      "step": 8944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7045107551918588,
+      "learning_rate": 6.728757237293187e-06,
+      "loss": 0.4891,
+      "step": 8945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8349362951137393,
+      "learning_rate": 6.7281308750749034e-06,
+      "loss": 0.5212,
+      "step": 8946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.696111986920655,
+      "learning_rate": 6.72750448714646e-06,
+      "loss": 0.5298,
+      "step": 8947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.037354620222616,
+      "learning_rate": 6.726878073520974e-06,
+      "loss": 0.4902,
+      "step": 8948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6924233522733905,
+      "learning_rate": 6.726251634211558e-06,
+      "loss": 0.552,
+      "step": 8949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.819296015305697,
+      "learning_rate": 6.725625169231325e-06,
+      "loss": 0.4843,
+      "step": 8950
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5404635071754456,
+      "eval_runtime": 67.9506,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 8950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.544572486584579,
+      "learning_rate": 6.724998678593391e-06,
+      "loss": 0.554,
+      "step": 8951
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6611166837635891,
+      "learning_rate": 6.72437216231087e-06,
+      "loss": 0.5042,
+      "step": 8952
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.355497523268971,
+      "learning_rate": 6.7237456203968786e-06,
+      "loss": 0.3678,
+      "step": 8953
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6469343401608514,
+      "learning_rate": 6.723119052864534e-06,
+      "loss": 0.5092,
+      "step": 8954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1731218053326242,
+      "learning_rate": 6.72249245972695e-06,
+      "loss": 0.4867,
+      "step": 8955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.797962420295741,
+      "learning_rate": 6.721865840997246e-06,
+      "loss": 0.4849,
+      "step": 8956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.244818413312252,
+      "learning_rate": 6.721239196688541e-06,
+      "loss": 0.6034,
+      "step": 8957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7210923357364987,
+      "learning_rate": 6.72061252681395e-06,
+      "loss": 0.5223,
+      "step": 8958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2884791349898033,
+      "learning_rate": 6.719985831386594e-06,
+      "loss": 0.4021,
+      "step": 8959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6956390533855354,
+      "learning_rate": 6.719359110419594e-06,
+      "loss": 0.5316,
+      "step": 8960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5972054701838752,
+      "learning_rate": 6.718732363926066e-06,
+      "loss": 0.5431,
+      "step": 8961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.787306233646309,
+      "learning_rate": 6.7181055919191355e-06,
+      "loss": 0.5134,
+      "step": 8962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6006965750750501,
+      "learning_rate": 6.717478794411918e-06,
+      "loss": 0.4352,
+      "step": 8963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9676080781494232,
+      "learning_rate": 6.716851971417541e-06,
+      "loss": 0.6096,
+      "step": 8964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7154674216128791,
+      "learning_rate": 6.716225122949123e-06,
+      "loss": 0.4851,
+      "step": 8965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7366716340847421,
+      "learning_rate": 6.715598249019787e-06,
+      "loss": 0.507,
+      "step": 8966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.530005739046869,
+      "learning_rate": 6.714971349642657e-06,
+      "loss": 0.7925,
+      "step": 8967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.727326588334344,
+      "learning_rate": 6.714344424830857e-06,
+      "loss": 0.4557,
+      "step": 8968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8198551659135722,
+      "learning_rate": 6.713717474597508e-06,
+      "loss": 0.5551,
+      "step": 8969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1202809755436367,
+      "learning_rate": 6.713090498955741e-06,
+      "loss": 0.6006,
+      "step": 8970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.108569146064458,
+      "learning_rate": 6.7124634979186756e-06,
+      "loss": 0.552,
+      "step": 8971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7100604567427078,
+      "learning_rate": 6.7118364714994425e-06,
+      "loss": 0.5533,
+      "step": 8972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.752816598785541,
+      "learning_rate": 6.7112094197111635e-06,
+      "loss": 0.5842,
+      "step": 8973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9962386208672631,
+      "learning_rate": 6.710582342566966e-06,
+      "loss": 0.4938,
+      "step": 8974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4176548637582593,
+      "learning_rate": 6.7099552400799815e-06,
+      "loss": 0.4575,
+      "step": 8975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7204349442409612,
+      "learning_rate": 6.709328112263337e-06,
+      "loss": 0.5509,
+      "step": 8976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.323263985217948,
+      "learning_rate": 6.708700959130156e-06,
+      "loss": 0.367,
+      "step": 8977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7065565203660822,
+      "learning_rate": 6.708073780693574e-06,
+      "loss": 0.6138,
+      "step": 8978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5968334494190441,
+      "learning_rate": 6.707446576966716e-06,
+      "loss": 0.4117,
+      "step": 8979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6883959581170505,
+      "learning_rate": 6.7068193479627155e-06,
+      "loss": 0.5367,
+      "step": 8980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.728665064037161,
+      "learning_rate": 6.7061920936947e-06,
+      "loss": 0.5389,
+      "step": 8981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.729773208852611,
+      "learning_rate": 6.705564814175802e-06,
+      "loss": 0.493,
+      "step": 8982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6561041524968854,
+      "learning_rate": 6.704937509419155e-06,
+      "loss": 0.5601,
+      "step": 8983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.022505209979809,
+      "learning_rate": 6.70431017943789e-06,
+      "loss": 0.7122,
+      "step": 8984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8051660838118004,
+      "learning_rate": 6.703682824245137e-06,
+      "loss": 0.6378,
+      "step": 8985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7010987844128667,
+      "learning_rate": 6.703055443854034e-06,
+      "loss": 0.5733,
+      "step": 8986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8693834076013531,
+      "learning_rate": 6.702428038277711e-06,
+      "loss": 0.5708,
+      "step": 8987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.064096237329891,
+      "learning_rate": 6.701800607529305e-06,
+      "loss": 0.5814,
+      "step": 8988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.350746391371499,
+      "learning_rate": 6.701173151621949e-06,
+      "loss": 0.5721,
+      "step": 8989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6520195077075908,
+      "learning_rate": 6.700545670568779e-06,
+      "loss": 0.5297,
+      "step": 8990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7665334517770772,
+      "learning_rate": 6.699918164382931e-06,
+      "loss": 0.5797,
+      "step": 8991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7444968143828805,
+      "learning_rate": 6.699290633077542e-06,
+      "loss": 0.5046,
+      "step": 8992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6167564322948862,
+      "learning_rate": 6.69866307666575e-06,
+      "loss": 0.5661,
+      "step": 8993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7133491959730849,
+      "learning_rate": 6.698035495160689e-06,
+      "loss": 0.5036,
+      "step": 8994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7495198515706876,
+      "learning_rate": 6.697407888575499e-06,
+      "loss": 0.4961,
+      "step": 8995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3008308566296338,
+      "learning_rate": 6.69678025692332e-06,
+      "loss": 0.4609,
+      "step": 8996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5764842280057865,
+      "learning_rate": 6.696152600217288e-06,
+      "loss": 0.4196,
+      "step": 8997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8054178505900385,
+      "learning_rate": 6.695524918470544e-06,
+      "loss": 0.3935,
+      "step": 8998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.607969016712663,
+      "learning_rate": 6.69489721169623e-06,
+      "loss": 0.3909,
+      "step": 8999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.610365146842044,
+      "learning_rate": 6.694269479907484e-06,
+      "loss": 0.3402,
+      "step": 9000
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5383872389793396,
+      "eval_runtime": 68.013,
+      "eval_samples_per_second": 50.138,
+      "eval_steps_per_second": 1.573,
+      "step": 9000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5718524005738181,
+      "learning_rate": 6.693641723117449e-06,
+      "loss": 0.4978,
+      "step": 9001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5270971449392179,
+      "learning_rate": 6.693013941339266e-06,
+      "loss": 0.4437,
+      "step": 9002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.60084920082863,
+      "learning_rate": 6.692386134586076e-06,
+      "loss": 0.5394,
+      "step": 9003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8622958133198106,
+      "learning_rate": 6.691758302871026e-06,
+      "loss": 0.5075,
+      "step": 9004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0028939081146753,
+      "learning_rate": 6.691130446207253e-06,
+      "loss": 0.656,
+      "step": 9005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.100050087608707,
+      "learning_rate": 6.690502564607904e-06,
+      "loss": 0.6651,
+      "step": 9006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.718817136167922,
+      "learning_rate": 6.689874658086125e-06,
+      "loss": 0.5702,
+      "step": 9007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7323801150742093,
+      "learning_rate": 6.689246726655058e-06,
+      "loss": 0.5929,
+      "step": 9008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8286656908609507,
+      "learning_rate": 6.688618770327848e-06,
+      "loss": 0.4886,
+      "step": 9009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7844147462405306,
+      "learning_rate": 6.687990789117644e-06,
+      "loss": 0.5221,
+      "step": 9010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5824219156920627,
+      "learning_rate": 6.687362783037591e-06,
+      "loss": 0.5152,
+      "step": 9011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0040534061885347,
+      "learning_rate": 6.6867347521008364e-06,
+      "loss": 0.5768,
+      "step": 9012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6333719672476992,
+      "learning_rate": 6.686106696320526e-06,
+      "loss": 0.4502,
+      "step": 9013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9006908478178384,
+      "learning_rate": 6.685478615709806e-06,
+      "loss": 0.7109,
+      "step": 9014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.203130405375062,
+      "learning_rate": 6.684850510281832e-06,
+      "loss": 0.6453,
+      "step": 9015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8417002679799674,
+      "learning_rate": 6.684222380049745e-06,
+      "loss": 0.4453,
+      "step": 9016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6568603481204756,
+      "learning_rate": 6.6835942250267006e-06,
+      "loss": 0.4992,
+      "step": 9017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5845141540359895,
+      "learning_rate": 6.682966045225845e-06,
+      "loss": 0.4626,
+      "step": 9018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.696405413880873,
+      "learning_rate": 6.682337840660329e-06,
+      "loss": 0.4275,
+      "step": 9019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.001597330369564,
+      "learning_rate": 6.681709611343307e-06,
+      "loss": 0.494,
+      "step": 9020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7135656110212658,
+      "learning_rate": 6.681081357287926e-06,
+      "loss": 0.5738,
+      "step": 9021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1937420369562233,
+      "learning_rate": 6.680453078507341e-06,
+      "loss": 0.9644,
+      "step": 9022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7651560752382671,
+      "learning_rate": 6.679824775014705e-06,
+      "loss": 0.5257,
+      "step": 9023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8080890513050978,
+      "learning_rate": 6.67919644682317e-06,
+      "loss": 0.5727,
+      "step": 9024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5864617334734459,
+      "learning_rate": 6.6785680939458886e-06,
+      "loss": 0.4476,
+      "step": 9025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.756075620123738,
+      "learning_rate": 6.677939716396017e-06,
+      "loss": 0.4854,
+      "step": 9026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7955087590571441,
+      "learning_rate": 6.677311314186709e-06,
+      "loss": 0.5617,
+      "step": 9027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8403461337622367,
+      "learning_rate": 6.67668288733112e-06,
+      "loss": 0.4461,
+      "step": 9028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.623034412877647,
+      "learning_rate": 6.676054435842405e-06,
+      "loss": 0.4914,
+      "step": 9029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7609702366836686,
+      "learning_rate": 6.6754259597337214e-06,
+      "loss": 0.6198,
+      "step": 9030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.159483759163162,
+      "learning_rate": 6.674797459018227e-06,
+      "loss": 0.479,
+      "step": 9031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.780484297953878,
+      "learning_rate": 6.674168933709074e-06,
+      "loss": 0.4985,
+      "step": 9032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.595473155330733,
+      "learning_rate": 6.673540383819427e-06,
+      "loss": 0.4484,
+      "step": 9033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5254591023031572,
+      "learning_rate": 6.672911809362439e-06,
+      "loss": 0.4665,
+      "step": 9034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8154313233879158,
+      "learning_rate": 6.672283210351272e-06,
+      "loss": 0.469,
+      "step": 9035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.635974273261574,
+      "learning_rate": 6.6716545867990835e-06,
+      "loss": 0.4691,
+      "step": 9036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.173855278204101,
+      "learning_rate": 6.671025938719034e-06,
+      "loss": 0.599,
+      "step": 9037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6120020236509187,
+      "learning_rate": 6.670397266124283e-06,
+      "loss": 0.3984,
+      "step": 9038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5761512559426134,
+      "learning_rate": 6.6697685690279935e-06,
+      "loss": 0.3742,
+      "step": 9039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8605302326296194,
+      "learning_rate": 6.669139847443325e-06,
+      "loss": 0.4732,
+      "step": 9040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9107458084755404,
+      "learning_rate": 6.66851110138344e-06,
+      "loss": 0.4768,
+      "step": 9041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1929060367135724,
+      "learning_rate": 6.667882330861501e-06,
+      "loss": 0.5874,
+      "step": 9042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0519208694635123,
+      "learning_rate": 6.66725353589067e-06,
+      "loss": 0.5393,
+      "step": 9043
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7684821223624871,
+      "learning_rate": 6.6666247164841125e-06,
+      "loss": 0.4394,
+      "step": 9044
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.768248072559843,
+      "learning_rate": 6.66599587265499e-06,
+      "loss": 0.5486,
+      "step": 9045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6323266677011994,
+      "learning_rate": 6.665367004416469e-06,
+      "loss": 0.4129,
+      "step": 9046
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.888277390032012,
+      "learning_rate": 6.664738111781715e-06,
+      "loss": 0.4919,
+      "step": 9047
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6251457430919365,
+      "learning_rate": 6.6641091947638894e-06,
+      "loss": 0.402,
+      "step": 9048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9892159320817857,
+      "learning_rate": 6.6634802533761615e-06,
+      "loss": 0.5909,
+      "step": 9049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0807405435973436,
+      "learning_rate": 6.662851287631696e-06,
+      "loss": 0.4746,
+      "step": 9050
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5392112135887146,
+      "eval_runtime": 67.9889,
+      "eval_samples_per_second": 50.155,
+      "eval_steps_per_second": 1.574,
+      "step": 9050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5092691730729442,
+      "learning_rate": 6.662222297543663e-06,
+      "loss": 0.3985,
+      "step": 9051
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1335296805796116,
+      "learning_rate": 6.661593283125228e-06,
+      "loss": 0.5355,
+      "step": 9052
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2176061324554297,
+      "learning_rate": 6.66096424438956e-06,
+      "loss": 0.7208,
+      "step": 9053
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0063511218560195,
+      "learning_rate": 6.660335181349824e-06,
+      "loss": 0.5188,
+      "step": 9054
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0034803222110185,
+      "learning_rate": 6.659706094019194e-06,
+      "loss": 0.4296,
+      "step": 9055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9782600594033546,
+      "learning_rate": 6.659076982410836e-06,
+      "loss": 0.4515,
+      "step": 9056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7314969339871786,
+      "learning_rate": 6.6584478465379224e-06,
+      "loss": 0.4907,
+      "step": 9057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5761337714682242,
+      "learning_rate": 6.657818686413622e-06,
+      "loss": 0.3998,
+      "step": 9058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8299813673630867,
+      "learning_rate": 6.6571895020511055e-06,
+      "loss": 0.5934,
+      "step": 9059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.805933662606076,
+      "learning_rate": 6.656560293463546e-06,
+      "loss": 0.5204,
+      "step": 9060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.671564241640023,
+      "learning_rate": 6.655931060664116e-06,
+      "loss": 0.4557,
+      "step": 9061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8471532425652515,
+      "learning_rate": 6.655301803665988e-06,
+      "loss": 0.5224,
+      "step": 9062
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6925626004806495,
+      "learning_rate": 6.654672522482333e-06,
+      "loss": 0.5288,
+      "step": 9063
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6440032227537167,
+      "learning_rate": 6.654043217126326e-06,
+      "loss": 0.5031,
+      "step": 9064
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7109137918412247,
+      "learning_rate": 6.653413887611142e-06,
+      "loss": 0.4994,
+      "step": 9065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5923310043312575,
+      "learning_rate": 6.652784533949954e-06,
+      "loss": 0.3871,
+      "step": 9066
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7996202267282055,
+      "learning_rate": 6.652155156155937e-06,
+      "loss": 0.5932,
+      "step": 9067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.960977543682451,
+      "learning_rate": 6.651525754242268e-06,
+      "loss": 0.5225,
+      "step": 9068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6728733941950902,
+      "learning_rate": 6.650896328222123e-06,
+      "loss": 0.5632,
+      "step": 9069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2049391126766715,
+      "learning_rate": 6.650266878108678e-06,
+      "loss": 0.5918,
+      "step": 9070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7358221332836652,
+      "learning_rate": 6.64963740391511e-06,
+      "loss": 0.484,
+      "step": 9071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4926002112595083,
+      "learning_rate": 6.649007905654595e-06,
+      "loss": 0.402,
+      "step": 9072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9613470722567579,
+      "learning_rate": 6.648378383340317e-06,
+      "loss": 0.6747,
+      "step": 9073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5788530207572191,
+      "learning_rate": 6.64774883698545e-06,
+      "loss": 0.4174,
+      "step": 9074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.073289306068199,
+      "learning_rate": 6.647119266603173e-06,
+      "loss": 0.6311,
+      "step": 9075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3711142120187398,
+      "learning_rate": 6.646489672206667e-06,
+      "loss": 0.4168,
+      "step": 9076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0020063479597323,
+      "learning_rate": 6.645860053809111e-06,
+      "loss": 0.4532,
+      "step": 9077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8127786536922699,
+      "learning_rate": 6.645230411423687e-06,
+      "loss": 0.6208,
+      "step": 9078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.732420915998382,
+      "learning_rate": 6.644600745063576e-06,
+      "loss": 0.5076,
+      "step": 9079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0271977770338623,
+      "learning_rate": 6.64397105474196e-06,
+      "loss": 0.6401,
+      "step": 9080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8564123475539522,
+      "learning_rate": 6.64334134047202e-06,
+      "loss": 0.6569,
+      "step": 9081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8924296317904432,
+      "learning_rate": 6.642711602266939e-06,
+      "loss": 0.5664,
+      "step": 9082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.12222674476115,
+      "learning_rate": 6.6420818401399e-06,
+      "loss": 0.5244,
+      "step": 9083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.554675154693887,
+      "learning_rate": 6.6414520541040874e-06,
+      "loss": 0.4765,
+      "step": 9084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7076751700962416,
+      "learning_rate": 6.640822244172684e-06,
+      "loss": 0.3767,
+      "step": 9085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4938425321396407,
+      "learning_rate": 6.6401924103588774e-06,
+      "loss": 0.4901,
+      "step": 9086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5521497356907745,
+      "learning_rate": 6.6395625526758514e-06,
+      "loss": 0.4092,
+      "step": 9087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.654946003075199,
+      "learning_rate": 6.638932671136787e-06,
+      "loss": 0.4893,
+      "step": 9088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.603611937333904,
+      "learning_rate": 6.638302765754879e-06,
+      "loss": 0.4139,
+      "step": 9089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8142429029439997,
+      "learning_rate": 6.637672836543307e-06,
+      "loss": 0.4714,
+      "step": 9090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4875544999423855,
+      "learning_rate": 6.637042883515262e-06,
+      "loss": 0.5851,
+      "step": 9091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7710091214157468,
+      "learning_rate": 6.636412906683931e-06,
+      "loss": 0.4345,
+      "step": 9092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6723683263543492,
+      "learning_rate": 6.6357829060625e-06,
+      "loss": 0.4619,
+      "step": 9093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7241869683667026,
+      "learning_rate": 6.635152881664161e-06,
+      "loss": 0.5529,
+      "step": 9094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.892229829873811,
+      "learning_rate": 6.634522833502101e-06,
+      "loss": 0.5481,
+      "step": 9095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6635614239466725,
+      "learning_rate": 6.63389276158951e-06,
+      "loss": 0.4766,
+      "step": 9096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.680677910952465,
+      "learning_rate": 6.633262665939577e-06,
+      "loss": 0.5358,
+      "step": 9097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9479986262369384,
+      "learning_rate": 6.632632546565495e-06,
+      "loss": 0.5626,
+      "step": 9098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0068251435416657,
+      "learning_rate": 6.632002403480455e-06,
+      "loss": 0.6206,
+      "step": 9099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.886748979921657,
+      "learning_rate": 6.631372236697646e-06,
+      "loss": 0.5147,
+      "step": 9100
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5403112769126892,
+      "eval_runtime": 67.9585,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.574,
+      "step": 9100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0610543673339747,
+      "learning_rate": 6.630742046230263e-06,
+      "loss": 0.5166,
+      "step": 9101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5368916028709845,
+      "learning_rate": 6.630111832091498e-06,
+      "loss": 0.5418,
+      "step": 9102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.819516089769834,
+      "learning_rate": 6.6294815942945444e-06,
+      "loss": 0.4049,
+      "step": 9103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6401464837078648,
+      "learning_rate": 6.628851332852593e-06,
+      "loss": 0.6153,
+      "step": 9104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7380726965923707,
+      "learning_rate": 6.628221047778841e-06,
+      "loss": 0.4544,
+      "step": 9105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1299396568400906,
+      "learning_rate": 6.627590739086483e-06,
+      "loss": 0.7077,
+      "step": 9106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.814413851168235,
+      "learning_rate": 6.626960406788711e-06,
+      "loss": 0.4618,
+      "step": 9107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.064549344911494,
+      "learning_rate": 6.6263300508987265e-06,
+      "loss": 0.5784,
+      "step": 9108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6807783335058581,
+      "learning_rate": 6.625699671429719e-06,
+      "loss": 0.5636,
+      "step": 9109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.499522543540604,
+      "learning_rate": 6.6250692683948895e-06,
+      "loss": 0.4015,
+      "step": 9110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7446554003838952,
+      "learning_rate": 6.6244388418074335e-06,
+      "loss": 0.6023,
+      "step": 9111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5878734614003114,
+      "learning_rate": 6.623808391680548e-06,
+      "loss": 0.3593,
+      "step": 9112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4019377703411382,
+      "learning_rate": 6.623177918027434e-06,
+      "loss": 0.4219,
+      "step": 9113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7960516893342549,
+      "learning_rate": 6.622547420861285e-06,
+      "loss": 0.6086,
+      "step": 9114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3364248481083076,
+      "learning_rate": 6.621916900195305e-06,
+      "loss": 0.7553,
+      "step": 9115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.928084719492276,
+      "learning_rate": 6.62128635604269e-06,
+      "loss": 0.6184,
+      "step": 9116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8568175772240574,
+      "learning_rate": 6.620655788416642e-06,
+      "loss": 0.6046,
+      "step": 9117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9517646472183303,
+      "learning_rate": 6.620025197330362e-06,
+      "loss": 0.6934,
+      "step": 9118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9471887389584104,
+      "learning_rate": 6.619394582797049e-06,
+      "loss": 0.5966,
+      "step": 9119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6877752154990426,
+      "learning_rate": 6.6187639448299065e-06,
+      "loss": 0.6363,
+      "step": 9120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7418948323526704,
+      "learning_rate": 6.618133283442135e-06,
+      "loss": 0.5923,
+      "step": 9121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.116803436323607,
+      "learning_rate": 6.617502598646938e-06,
+      "loss": 0.6887,
+      "step": 9122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8707191744299752,
+      "learning_rate": 6.616871890457519e-06,
+      "loss": 0.5339,
+      "step": 9123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8028637848238416,
+      "learning_rate": 6.61624115888708e-06,
+      "loss": 0.6191,
+      "step": 9124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.640498416817811,
+      "learning_rate": 6.6156104039488255e-06,
+      "loss": 0.5758,
+      "step": 9125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.457450097751306,
+      "learning_rate": 6.61497962565596e-06,
+      "loss": 0.5936,
+      "step": 9126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.305421791690329,
+      "learning_rate": 6.61434882402169e-06,
+      "loss": 0.5916,
+      "step": 9127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7888136030734136,
+      "learning_rate": 6.613717999059218e-06,
+      "loss": 0.4535,
+      "step": 9128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.489934811424703,
+      "learning_rate": 6.613087150781752e-06,
+      "loss": 0.4637,
+      "step": 9129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8086234445903135,
+      "learning_rate": 6.612456279202499e-06,
+      "loss": 0.5903,
+      "step": 9130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6415129154841281,
+      "learning_rate": 6.6118253843346624e-06,
+      "loss": 0.4988,
+      "step": 9131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6084095911092928,
+      "learning_rate": 6.611194466191454e-06,
+      "loss": 0.4186,
+      "step": 9132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.572364274869224,
+      "learning_rate": 6.610563524786078e-06,
+      "loss": 0.4013,
+      "step": 9133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3715317519781807,
+      "learning_rate": 6.609932560131745e-06,
+      "loss": 0.6148,
+      "step": 9134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.783871915883729,
+      "learning_rate": 6.609301572241664e-06,
+      "loss": 0.4673,
+      "step": 9135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8448072886919804,
+      "learning_rate": 6.608670561129042e-06,
+      "loss": 0.577,
+      "step": 9136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6518476883448885,
+      "learning_rate": 6.608039526807093e-06,
+      "loss": 0.4603,
+      "step": 9137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9242042506839045,
+      "learning_rate": 6.607408469289022e-06,
+      "loss": 0.6427,
+      "step": 9138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7697366685923557,
+      "learning_rate": 6.606777388588045e-06,
+      "loss": 0.6284,
+      "step": 9139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8413452212360006,
+      "learning_rate": 6.60614628471737e-06,
+      "loss": 0.622,
+      "step": 9140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7994662500346814,
+      "learning_rate": 6.6055151576902085e-06,
+      "loss": 0.4849,
+      "step": 9141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5932210299515144,
+      "learning_rate": 6.604884007519775e-06,
+      "loss": 0.4908,
+      "step": 9142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8413300954494594,
+      "learning_rate": 6.60425283421928e-06,
+      "loss": 0.4659,
+      "step": 9143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7789226269796001,
+      "learning_rate": 6.603621637801938e-06,
+      "loss": 0.6671,
+      "step": 9144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1179466878166826,
+      "learning_rate": 6.602990418280965e-06,
+      "loss": 0.5547,
+      "step": 9145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.149445986923636,
+      "learning_rate": 6.602359175669568e-06,
+      "loss": 0.5396,
+      "step": 9146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8597909831728416,
+      "learning_rate": 6.601727909980969e-06,
+      "loss": 0.496,
+      "step": 9147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.954769365696497,
+      "learning_rate": 6.60109662122838e-06,
+      "loss": 0.4489,
+      "step": 9148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6433206489774277,
+      "learning_rate": 6.600465309425017e-06,
+      "loss": 0.4887,
+      "step": 9149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.413726328163659,
+      "learning_rate": 6.5998339745840975e-06,
+      "loss": 0.4127,
+      "step": 9150
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5386484861373901,
+      "eval_runtime": 67.9932,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 9150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.20703130529047,
+      "learning_rate": 6.5992026167188375e-06,
+      "loss": 0.6577,
+      "step": 9151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.762309290752133,
+      "learning_rate": 6.59857123584245e-06,
+      "loss": 0.588,
+      "step": 9152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.373598807333473,
+      "learning_rate": 6.59793983196816e-06,
+      "loss": 0.4122,
+      "step": 9153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8895245498961784,
+      "learning_rate": 6.5973084051091785e-06,
+      "loss": 0.5248,
+      "step": 9154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.15854515844515,
+      "learning_rate": 6.596676955278729e-06,
+      "loss": 0.562,
+      "step": 9155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9170414744490962,
+      "learning_rate": 6.5960454824900285e-06,
+      "loss": 0.7163,
+      "step": 9156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9880030456174003,
+      "learning_rate": 6.595413986756296e-06,
+      "loss": 0.4999,
+      "step": 9157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8422923435234142,
+      "learning_rate": 6.594782468090752e-06,
+      "loss": 0.4747,
+      "step": 9158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0119778068757106,
+      "learning_rate": 6.594150926506618e-06,
+      "loss": 0.4242,
+      "step": 9159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.710708365115933,
+      "learning_rate": 6.593519362017111e-06,
+      "loss": 0.5899,
+      "step": 9160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7607771923246935,
+      "learning_rate": 6.592887774635459e-06,
+      "loss": 0.4875,
+      "step": 9161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7362704919075227,
+      "learning_rate": 6.592256164374878e-06,
+      "loss": 0.5696,
+      "step": 9162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.539027093749468,
+      "learning_rate": 6.591624531248594e-06,
+      "loss": 0.4233,
+      "step": 9163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7467296680633821,
+      "learning_rate": 6.590992875269828e-06,
+      "loss": 0.5127,
+      "step": 9164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9008188652797557,
+      "learning_rate": 6.590361196451803e-06,
+      "loss": 0.5861,
+      "step": 9165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.175807856155807,
+      "learning_rate": 6.589729494807745e-06,
+      "loss": 0.6033,
+      "step": 9166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8379896332268881,
+      "learning_rate": 6.589097770350876e-06,
+      "loss": 0.4791,
+      "step": 9167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1323862994860447,
+      "learning_rate": 6.588466023094421e-06,
+      "loss": 0.5833,
+      "step": 9168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8529692700860847,
+      "learning_rate": 6.587834253051608e-06,
+      "loss": 0.4543,
+      "step": 9169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9168123988116152,
+      "learning_rate": 6.5872024602356575e-06,
+      "loss": 0.5154,
+      "step": 9170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6403158741014703,
+      "learning_rate": 6.586570644659801e-06,
+      "loss": 0.5426,
+      "step": 9171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0602039099860963,
+      "learning_rate": 6.585938806337263e-06,
+      "loss": 0.7142,
+      "step": 9172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6812991988292267,
+      "learning_rate": 6.585306945281269e-06,
+      "loss": 0.4637,
+      "step": 9173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7640152674696319,
+      "learning_rate": 6.584675061505049e-06,
+      "loss": 0.4836,
+      "step": 9174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.593684038101005,
+      "learning_rate": 6.584043155021832e-06,
+      "loss": 0.4675,
+      "step": 9175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5666368435425948,
+      "learning_rate": 6.583411225844842e-06,
+      "loss": 0.3929,
+      "step": 9176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4654244267255367,
+      "learning_rate": 6.582779273987314e-06,
+      "loss": 0.3718,
+      "step": 9177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9394547030930962,
+      "learning_rate": 6.582147299462472e-06,
+      "loss": 0.6277,
+      "step": 9178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8678301121532437,
+      "learning_rate": 6.58151530228355e-06,
+      "loss": 0.627,
+      "step": 9179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4589955516487345,
+      "learning_rate": 6.580883282463776e-06,
+      "loss": 0.4204,
+      "step": 9180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.62670351932754,
+      "learning_rate": 6.580251240016381e-06,
+      "loss": 0.4092,
+      "step": 9181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7197514761042434,
+      "learning_rate": 6.579619174954599e-06,
+      "loss": 0.5268,
+      "step": 9182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6306986192206454,
+      "learning_rate": 6.578987087291658e-06,
+      "loss": 0.5446,
+      "step": 9183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2351264364876973,
+      "learning_rate": 6.578354977040795e-06,
+      "loss": 0.5741,
+      "step": 9184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6960435673726817,
+      "learning_rate": 6.5777228442152395e-06,
+      "loss": 0.5402,
+      "step": 9185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7622684686615615,
+      "learning_rate": 6.577090688828225e-06,
+      "loss": 0.6494,
+      "step": 9186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5011103133187134,
+      "learning_rate": 6.576458510892987e-06,
+      "loss": 0.4641,
+      "step": 9187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.958560350582808,
+      "learning_rate": 6.575826310422758e-06,
+      "loss": 0.6092,
+      "step": 9188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.004521903016179,
+      "learning_rate": 6.575194087430773e-06,
+      "loss": 0.5486,
+      "step": 9189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2280984907384047,
+      "learning_rate": 6.574561841930268e-06,
+      "loss": 0.5055,
+      "step": 9190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5797381382484392,
+      "learning_rate": 6.5739295739344775e-06,
+      "loss": 0.4595,
+      "step": 9191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7977294276461127,
+      "learning_rate": 6.57329728345664e-06,
+      "loss": 0.6157,
+      "step": 9192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1547689267218053,
+      "learning_rate": 6.57266497050999e-06,
+      "loss": 0.4921,
+      "step": 9193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0112030745297154,
+      "learning_rate": 6.5720326351077644e-06,
+      "loss": 0.6193,
+      "step": 9194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7159298819332096,
+      "learning_rate": 6.5714002772632014e-06,
+      "loss": 0.4775,
+      "step": 9195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6525547063608461,
+      "learning_rate": 6.570767896989541e-06,
+      "loss": 0.5963,
+      "step": 9196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5505685471427444,
+      "learning_rate": 6.570135494300016e-06,
+      "loss": 0.5594,
+      "step": 9197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9224907644356417,
+      "learning_rate": 6.5695030692078724e-06,
+      "loss": 0.4555,
+      "step": 9198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5954217206008676,
+      "learning_rate": 6.568870621726345e-06,
+      "loss": 0.486,
+      "step": 9199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7584276381984147,
+      "learning_rate": 6.5682381518686756e-06,
+      "loss": 0.6317,
+      "step": 9200
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5390273928642273,
+      "eval_runtime": 67.9959,
+      "eval_samples_per_second": 50.15,
+      "eval_steps_per_second": 1.574,
+      "step": 9200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6582396140657236,
+      "learning_rate": 6.567605659648104e-06,
+      "loss": 0.4575,
+      "step": 9201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9628926754971117,
+      "learning_rate": 6.5669731450778706e-06,
+      "loss": 0.6022,
+      "step": 9202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0746198637545863,
+      "learning_rate": 6.566340608171218e-06,
+      "loss": 0.6277,
+      "step": 9203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4746785202716501,
+      "learning_rate": 6.565708048941386e-06,
+      "loss": 0.462,
+      "step": 9204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6812148314325752,
+      "learning_rate": 6.565075467401618e-06,
+      "loss": 0.484,
+      "step": 9205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5073747329804985,
+      "learning_rate": 6.564442863565157e-06,
+      "loss": 0.4426,
+      "step": 9206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9068799943837687,
+      "learning_rate": 6.563810237445246e-06,
+      "loss": 0.5603,
+      "step": 9207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7341143395010827,
+      "learning_rate": 6.563177589055128e-06,
+      "loss": 0.4673,
+      "step": 9208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6449651365165217,
+      "learning_rate": 6.56254491840805e-06,
+      "loss": 0.3951,
+      "step": 9209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6395061069158092,
+      "learning_rate": 6.56191222551725e-06,
+      "loss": 0.4193,
+      "step": 9210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1387179753611156,
+      "learning_rate": 6.561279510395979e-06,
+      "loss": 0.6555,
+      "step": 9211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9029397657895954,
+      "learning_rate": 6.560646773057482e-06,
+      "loss": 0.5379,
+      "step": 9212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.742269168976007,
+      "learning_rate": 6.560014013515e-06,
+      "loss": 0.7394,
+      "step": 9213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0347906618984517,
+      "learning_rate": 6.559381231781787e-06,
+      "loss": 0.7587,
+      "step": 9214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9049639938036578,
+      "learning_rate": 6.558748427871084e-06,
+      "loss": 0.4507,
+      "step": 9215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9545321448532793,
+      "learning_rate": 6.558115601796141e-06,
+      "loss": 0.6653,
+      "step": 9216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0822111293063257,
+      "learning_rate": 6.557482753570205e-06,
+      "loss": 0.5465,
+      "step": 9217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8073598203115895,
+      "learning_rate": 6.556849883206523e-06,
+      "loss": 0.553,
+      "step": 9218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8433540746277512,
+      "learning_rate": 6.556216990718347e-06,
+      "loss": 0.6348,
+      "step": 9219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.186235599490943,
+      "learning_rate": 6.5555840761189215e-06,
+      "loss": 0.5907,
+      "step": 9220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.112329698559991,
+      "learning_rate": 6.554951139421501e-06,
+      "loss": 0.5847,
+      "step": 9221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6137166660942264,
+      "learning_rate": 6.554318180639334e-06,
+      "loss": 0.4419,
+      "step": 9222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7739241487253188,
+      "learning_rate": 6.5536851997856695e-06,
+      "loss": 0.5381,
+      "step": 9223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.801819513477907,
+      "learning_rate": 6.553052196873762e-06,
+      "loss": 0.5947,
+      "step": 9224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.029285499818263,
+      "learning_rate": 6.552419171916859e-06,
+      "loss": 0.5937,
+      "step": 9225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5534275368608115,
+      "learning_rate": 6.551786124928216e-06,
+      "loss": 0.4674,
+      "step": 9226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0374017736778627,
+      "learning_rate": 6.551153055921081e-06,
+      "loss": 0.5315,
+      "step": 9227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.564264227274358,
+      "learning_rate": 6.550519964908712e-06,
+      "loss": 0.4529,
+      "step": 9228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.94537380491218,
+      "learning_rate": 6.549886851904359e-06,
+      "loss": 0.6363,
+      "step": 9229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.165536687761575,
+      "learning_rate": 6.5492537169212774e-06,
+      "loss": 0.5991,
+      "step": 9230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9813981404807905,
+      "learning_rate": 6.548620559972721e-06,
+      "loss": 0.467,
+      "step": 9231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9476899483039092,
+      "learning_rate": 6.547987381071943e-06,
+      "loss": 0.6425,
+      "step": 9232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0271680825207334,
+      "learning_rate": 6.547354180232202e-06,
+      "loss": 0.5185,
+      "step": 9233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0837674797864083,
+      "learning_rate": 6.546720957466751e-06,
+      "loss": 0.6286,
+      "step": 9234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5198859725647118,
+      "learning_rate": 6.546087712788847e-06,
+      "loss": 0.3934,
+      "step": 9235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.597294392270965,
+      "learning_rate": 6.545454446211745e-06,
+      "loss": 0.4722,
+      "step": 9236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6166237674802506,
+      "learning_rate": 6.544821157748707e-06,
+      "loss": 0.503,
+      "step": 9237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.143881529660195,
+      "learning_rate": 6.544187847412985e-06,
+      "loss": 0.5352,
+      "step": 9238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8516708215968982,
+      "learning_rate": 6.543554515217838e-06,
+      "loss": 0.5881,
+      "step": 9239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.096581771976498,
+      "learning_rate": 6.542921161176527e-06,
+      "loss": 0.5386,
+      "step": 9240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4826580359211157,
+      "learning_rate": 6.5422877853023095e-06,
+      "loss": 0.4185,
+      "step": 9241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7060769641923081,
+      "learning_rate": 6.541654387608443e-06,
+      "loss": 0.5177,
+      "step": 9242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8441214863154551,
+      "learning_rate": 6.541020968108189e-06,
+      "loss": 0.527,
+      "step": 9243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.020070035473151,
+      "learning_rate": 6.5403875268148075e-06,
+      "loss": 0.5891,
+      "step": 9244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.177405416142187,
+      "learning_rate": 6.539754063741562e-06,
+      "loss": 0.5881,
+      "step": 9245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7785739554589426,
+      "learning_rate": 6.539120578901708e-06,
+      "loss": 0.4075,
+      "step": 9246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7993476027422397,
+      "learning_rate": 6.538487072308511e-06,
+      "loss": 0.5574,
+      "step": 9247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.037036178638035,
+      "learning_rate": 6.5378535439752325e-06,
+      "loss": 0.6448,
+      "step": 9248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6855592174350424,
+      "learning_rate": 6.537219993915135e-06,
+      "loss": 0.4314,
+      "step": 9249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.876819749277815,
+      "learning_rate": 6.53658642214148e-06,
+      "loss": 0.423,
+      "step": 9250
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.5385669469833374,
+      "eval_runtime": 67.917,
+      "eval_samples_per_second": 50.208,
+      "eval_steps_per_second": 1.575,
+      "step": 9250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7103808106698972,
+      "learning_rate": 6.535952828667533e-06,
+      "loss": 0.5401,
+      "step": 9251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4905220857440602,
+      "learning_rate": 6.535319213506558e-06,
+      "loss": 0.4731,
+      "step": 9252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5964645664734007,
+      "learning_rate": 6.534685576671816e-06,
+      "loss": 0.4291,
+      "step": 9253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8034246622737102,
+      "learning_rate": 6.534051918176576e-06,
+      "loss": 0.3907,
+      "step": 9254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6429363640290546,
+      "learning_rate": 6.533418238034099e-06,
+      "loss": 0.5335,
+      "step": 9255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6215848573345868,
+      "learning_rate": 6.532784536257656e-06,
+      "loss": 0.5202,
+      "step": 9256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5627236882666828,
+      "learning_rate": 6.532150812860509e-06,
+      "loss": 0.3973,
+      "step": 9257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8980917983660968,
+      "learning_rate": 6.531517067855926e-06,
+      "loss": 0.5714,
+      "step": 9258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.657514485222864,
+      "learning_rate": 6.530883301257173e-06,
+      "loss": 0.4558,
+      "step": 9259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.075259106666264,
+      "learning_rate": 6.530249513077519e-06,
+      "loss": 0.6606,
+      "step": 9260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6924281373656225,
+      "learning_rate": 6.529615703330232e-06,
+      "loss": 0.5151,
+      "step": 9261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9464332572830791,
+      "learning_rate": 6.528981872028579e-06,
+      "loss": 0.5674,
+      "step": 9262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8070127589517744,
+      "learning_rate": 6.528348019185829e-06,
+      "loss": 0.5119,
+      "step": 9263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5645757194306458,
+      "learning_rate": 6.527714144815254e-06,
+      "loss": 0.5272,
+      "step": 9264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8960392040526228,
+      "learning_rate": 6.527080248930121e-06,
+      "loss": 0.6726,
+      "step": 9265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.602859601233465,
+      "learning_rate": 6.5264463315437e-06,
+      "loss": 0.3801,
+      "step": 9266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1677266386941705,
+      "learning_rate": 6.525812392669263e-06,
+      "loss": 0.6685,
+      "step": 9267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5046770134265888,
+      "learning_rate": 6.5251784323200805e-06,
+      "loss": 0.4402,
+      "step": 9268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.227605783003439,
+      "learning_rate": 6.524544450509423e-06,
+      "loss": 0.8111,
+      "step": 9269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6905918106524744,
+      "learning_rate": 6.5239104472505665e-06,
+      "loss": 0.4745,
+      "step": 9270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6558372274547295,
+      "learning_rate": 6.523276422556778e-06,
+      "loss": 0.5649,
+      "step": 9271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.341040210604756,
+      "learning_rate": 6.522642376441335e-06,
+      "loss": 0.4611,
+      "step": 9272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1976709244505366,
+      "learning_rate": 6.522008308917509e-06,
+      "loss": 0.682,
+      "step": 9273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.788902252718071,
+      "learning_rate": 6.5213742199985745e-06,
+      "loss": 0.6139,
+      "step": 9274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9861365992961153,
+      "learning_rate": 6.520740109697804e-06,
+      "loss": 0.546,
+      "step": 9275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8161489566637525,
+      "learning_rate": 6.520105978028473e-06,
+      "loss": 0.5466,
+      "step": 9276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8425806041839663,
+      "learning_rate": 6.519471825003858e-06,
+      "loss": 0.5755,
+      "step": 9277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.06131956660973,
+      "learning_rate": 6.518837650637233e-06,
+      "loss": 0.5139,
+      "step": 9278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3745884577385814,
+      "learning_rate": 6.518203454941874e-06,
+      "loss": 0.5762,
+      "step": 9279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7347133708885711,
+      "learning_rate": 6.51756923793106e-06,
+      "loss": 0.6709,
+      "step": 9280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7181991135485775,
+      "learning_rate": 6.5169349996180645e-06,
+      "loss": 0.5307,
+      "step": 9281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.815724987002305,
+      "learning_rate": 6.516300740016166e-06,
+      "loss": 0.5061,
+      "step": 9282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1030595026568473,
+      "learning_rate": 6.515666459138643e-06,
+      "loss": 0.5675,
+      "step": 9283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8985372532166103,
+      "learning_rate": 6.515032156998774e-06,
+      "loss": 0.6028,
+      "step": 9284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5836313431999824,
+      "learning_rate": 6.514397833609838e-06,
+      "loss": 0.5029,
+      "step": 9285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6189383464331633,
+      "learning_rate": 6.513763488985113e-06,
+      "loss": 0.4571,
+      "step": 9286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7475946712553652,
+      "learning_rate": 6.513129123137878e-06,
+      "loss": 0.551,
+      "step": 9287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.827753798192451,
+      "learning_rate": 6.512494736081415e-06,
+      "loss": 0.491,
+      "step": 9288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9689270163138104,
+      "learning_rate": 6.511860327829003e-06,
+      "loss": 0.5846,
+      "step": 9289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6634871611667228,
+      "learning_rate": 6.511225898393924e-06,
+      "loss": 0.4179,
+      "step": 9290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6637243777287483,
+      "learning_rate": 6.510591447789459e-06,
+      "loss": 0.4387,
+      "step": 9291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7110055934239439,
+      "learning_rate": 6.509956976028889e-06,
+      "loss": 0.6078,
+      "step": 9292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7989219531411897,
+      "learning_rate": 6.509322483125499e-06,
+      "loss": 0.5959,
+      "step": 9293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8189716568942702,
+      "learning_rate": 6.508687969092566e-06,
+      "loss": 0.5555,
+      "step": 9294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8166518139342012,
+      "learning_rate": 6.50805343394338e-06,
+      "loss": 0.6114,
+      "step": 9295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6716037032651825,
+      "learning_rate": 6.507418877691221e-06,
+      "loss": 0.5273,
+      "step": 9296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6161012859462542,
+      "learning_rate": 6.506784300349373e-06,
+      "loss": 0.5206,
+      "step": 9297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0969171131920077,
+      "learning_rate": 6.5061497019311205e-06,
+      "loss": 0.6081,
+      "step": 9298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6976403106762448,
+      "learning_rate": 6.50551508244975e-06,
+      "loss": 0.5088,
+      "step": 9299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.840538599260247,
+      "learning_rate": 6.504880441918545e-06,
+      "loss": 0.4663,
+      "step": 9300
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.5372392535209656,
+      "eval_runtime": 67.9393,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 9300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8417277286224387,
+      "learning_rate": 6.504245780350793e-06,
+      "loss": 0.5942,
+      "step": 9301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8547194680868186,
+      "learning_rate": 6.50361109775978e-06,
+      "loss": 0.4959,
+      "step": 9302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7033602216150108,
+      "learning_rate": 6.50297639415879e-06,
+      "loss": 0.4613,
+      "step": 9303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6180928416423965,
+      "learning_rate": 6.502341669561114e-06,
+      "loss": 0.4282,
+      "step": 9304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9149820024269923,
+      "learning_rate": 6.501706923980037e-06,
+      "loss": 0.556,
+      "step": 9305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3064142924230224,
+      "learning_rate": 6.501072157428846e-06,
+      "loss": 0.6082,
+      "step": 9306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.816297971860027,
+      "learning_rate": 6.500437369920833e-06,
+      "loss": 0.5951,
+      "step": 9307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.913142564867728,
+      "learning_rate": 6.499802561469284e-06,
+      "loss": 0.6087,
+      "step": 9308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8279037670772478,
+      "learning_rate": 6.49916773208749e-06,
+      "loss": 0.4116,
+      "step": 9309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6452104675009538,
+      "learning_rate": 6.498532881788742e-06,
+      "loss": 0.4158,
+      "step": 9310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4489758048393413,
+      "learning_rate": 6.497898010586326e-06,
+      "loss": 0.4057,
+      "step": 9311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6690208707250376,
+      "learning_rate": 6.4972631184935355e-06,
+      "loss": 0.5353,
+      "step": 9312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8217752013433433,
+      "learning_rate": 6.496628205523661e-06,
+      "loss": 0.3854,
+      "step": 9313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7165844925529452,
+      "learning_rate": 6.495993271689994e-06,
+      "loss": 0.5378,
+      "step": 9314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8384738519101436,
+      "learning_rate": 6.4953583170058276e-06,
+      "loss": 0.4876,
+      "step": 9315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8292222235281972,
+      "learning_rate": 6.494723341484451e-06,
+      "loss": 0.5394,
+      "step": 9316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8857538613996794,
+      "learning_rate": 6.494088345139161e-06,
+      "loss": 0.5344,
+      "step": 9317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2309610803783064,
+      "learning_rate": 6.493453327983246e-06,
+      "loss": 0.6594,
+      "step": 9318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1050085802315275,
+      "learning_rate": 6.492818290030005e-06,
+      "loss": 0.6865,
+      "step": 9319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8468913280865993,
+      "learning_rate": 6.4921832312927305e-06,
+      "loss": 0.5253,
+      "step": 9320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9968948510035138,
+      "learning_rate": 6.491548151784715e-06,
+      "loss": 0.5222,
+      "step": 9321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5925512465765899,
+      "learning_rate": 6.490913051519255e-06,
+      "loss": 0.4615,
+      "step": 9322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6729191947435813,
+      "learning_rate": 6.490277930509647e-06,
+      "loss": 0.5382,
+      "step": 9323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9925513229645404,
+      "learning_rate": 6.489642788769184e-06,
+      "loss": 0.5189,
+      "step": 9324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7887294745489049,
+      "learning_rate": 6.489007626311165e-06,
+      "loss": 0.4503,
+      "step": 9325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7457798462178422,
+      "learning_rate": 6.488372443148885e-06,
+      "loss": 0.5607,
+      "step": 9326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9680733069423995,
+      "learning_rate": 6.4877372392956414e-06,
+      "loss": 0.4895,
+      "step": 9327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5512645145128479,
+      "learning_rate": 6.487102014764733e-06,
+      "loss": 0.4895,
+      "step": 9328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8990621981408096,
+      "learning_rate": 6.486466769569456e-06,
+      "loss": 0.813,
+      "step": 9329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8421935731330592,
+      "learning_rate": 6.4858315037231114e-06,
+      "loss": 0.448,
+      "step": 9330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5381722094593075,
+      "learning_rate": 6.485196217238996e-06,
+      "loss": 0.4807,
+      "step": 9331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8386644127925653,
+      "learning_rate": 6.484560910130409e-06,
+      "loss": 0.5784,
+      "step": 9332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6978084754305507,
+      "learning_rate": 6.4839255824106514e-06,
+      "loss": 0.4342,
+      "step": 9333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6988517289307201,
+      "learning_rate": 6.483290234093024e-06,
+      "loss": 0.5407,
+      "step": 9334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.585438321863583,
+      "learning_rate": 6.482654865190824e-06,
+      "loss": 0.4735,
+      "step": 9335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7101758894174868,
+      "learning_rate": 6.482019475717356e-06,
+      "loss": 0.5027,
+      "step": 9336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.499542242537594,
+      "learning_rate": 6.481384065685919e-06,
+      "loss": 0.4256,
+      "step": 9337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.802787894958348,
+      "learning_rate": 6.4807486351098195e-06,
+      "loss": 0.443,
+      "step": 9338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7830403209868724,
+      "learning_rate": 6.4801131840023535e-06,
+      "loss": 0.6665,
+      "step": 9339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6524567499833895,
+      "learning_rate": 6.479477712376827e-06,
+      "loss": 0.3949,
+      "step": 9340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8505866662566604,
+      "learning_rate": 6.478842220246544e-06,
+      "loss": 0.4619,
+      "step": 9341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.98835688106641,
+      "learning_rate": 6.478206707624805e-06,
+      "loss": 0.5738,
+      "step": 9342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.319847666799132,
+      "learning_rate": 6.4775711745249185e-06,
+      "loss": 0.716,
+      "step": 9343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7593093868689071,
+      "learning_rate": 6.476935620960186e-06,
+      "loss": 0.5691,
+      "step": 9344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.742517846711716,
+      "learning_rate": 6.476300046943913e-06,
+      "loss": 0.5274,
+      "step": 9345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7254233974538533,
+      "learning_rate": 6.4756644524894045e-06,
+      "loss": 0.4687,
+      "step": 9346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.075335172567298,
+      "learning_rate": 6.475028837609968e-06,
+      "loss": 0.519,
+      "step": 9347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7837049998135075,
+      "learning_rate": 6.474393202318907e-06,
+      "loss": 0.4775,
+      "step": 9348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8842996159931187,
+      "learning_rate": 6.47375754662953e-06,
+      "loss": 0.5955,
+      "step": 9349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6381214909451085,
+      "learning_rate": 6.473121870555144e-06,
+      "loss": 0.4088,
+      "step": 9350
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.5364348292350769,
+      "eval_runtime": 68.0075,
+      "eval_samples_per_second": 50.142,
+      "eval_steps_per_second": 1.573,
+      "step": 9350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.965963364171843,
+      "learning_rate": 6.472486174109055e-06,
+      "loss": 0.722,
+      "step": 9351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5216342865294155,
+      "learning_rate": 6.4718504573045736e-06,
+      "loss": 0.4389,
+      "step": 9352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.577553265129469,
+      "learning_rate": 6.4712147201550045e-06,
+      "loss": 0.4806,
+      "step": 9353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0274335447816734,
+      "learning_rate": 6.470578962673661e-06,
+      "loss": 0.5605,
+      "step": 9354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.65419191115954,
+      "learning_rate": 6.469943184873848e-06,
+      "loss": 0.4605,
+      "step": 9355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7569454395717279,
+      "learning_rate": 6.469307386768876e-06,
+      "loss": 0.4593,
+      "step": 9356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5873669153993828,
+      "learning_rate": 6.468671568372058e-06,
+      "loss": 0.4915,
+      "step": 9357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7149778602427441,
+      "learning_rate": 6.468035729696701e-06,
+      "loss": 0.6158,
+      "step": 9358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5085573048002043,
+      "learning_rate": 6.467399870756118e-06,
+      "loss": 0.6303,
+      "step": 9359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6018462370499313,
+      "learning_rate": 6.466763991563619e-06,
+      "loss": 0.4399,
+      "step": 9360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7777781942022493,
+      "learning_rate": 6.466128092132516e-06,
+      "loss": 0.4815,
+      "step": 9361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.156152813532522,
+      "learning_rate": 6.465492172476123e-06,
+      "loss": 0.6,
+      "step": 9362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9020422352852335,
+      "learning_rate": 6.464856232607751e-06,
+      "loss": 0.4605,
+      "step": 9363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6201538012060646,
+      "learning_rate": 6.4642202725407125e-06,
+      "loss": 0.4529,
+      "step": 9364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7801383040937182,
+      "learning_rate": 6.463584292288321e-06,
+      "loss": 0.4278,
+      "step": 9365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7994843202097173,
+      "learning_rate": 6.462948291863891e-06,
+      "loss": 0.4826,
+      "step": 9366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9845163572196167,
+      "learning_rate": 6.4623122712807365e-06,
+      "loss": 0.4958,
+      "step": 9367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7510628927530558,
+      "learning_rate": 6.461676230552172e-06,
+      "loss": 0.5518,
+      "step": 9368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7273012052773187,
+      "learning_rate": 6.4610401696915134e-06,
+      "loss": 0.5109,
+      "step": 9369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5184546686054805,
+      "learning_rate": 6.4604040887120765e-06,
+      "loss": 0.3972,
+      "step": 9370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.155495402853003,
+      "learning_rate": 6.459767987627176e-06,
+      "loss": 0.7038,
+      "step": 9371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6988331874533273,
+      "learning_rate": 6.459131866450128e-06,
+      "loss": 0.5188,
+      "step": 9372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.953608169895991,
+      "learning_rate": 6.4584957251942515e-06,
+      "loss": 0.6283,
+      "step": 9373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.033234048576611,
+      "learning_rate": 6.457859563872861e-06,
+      "loss": 0.5696,
+      "step": 9374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.72168632018129,
+      "learning_rate": 6.457223382499276e-06,
+      "loss": 0.4282,
+      "step": 9375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.73880408012025,
+      "learning_rate": 6.456587181086814e-06,
+      "loss": 0.4412,
+      "step": 9376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5423179925118817,
+      "learning_rate": 6.4559509596487935e-06,
+      "loss": 0.5588,
+      "step": 9377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8830911053468888,
+      "learning_rate": 6.4553147181985344e-06,
+      "loss": 0.5336,
+      "step": 9378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.052229819707831,
+      "learning_rate": 6.454678456749355e-06,
+      "loss": 0.5916,
+      "step": 9379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6406651593654824,
+      "learning_rate": 6.454042175314573e-06,
+      "loss": 0.517,
+      "step": 9380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5877261108331973,
+      "learning_rate": 6.453405873907512e-06,
+      "loss": 0.7639,
+      "step": 9381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.04876448466355,
+      "learning_rate": 6.452769552541491e-06,
+      "loss": 0.5908,
+      "step": 9382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2929897324637056,
+      "learning_rate": 6.452133211229832e-06,
+      "loss": 0.5958,
+      "step": 9383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5777187406539848,
+      "learning_rate": 6.451496849985856e-06,
+      "loss": 0.5254,
+      "step": 9384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9692629799435106,
+      "learning_rate": 6.450860468822883e-06,
+      "loss": 0.4467,
+      "step": 9385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8933904734809823,
+      "learning_rate": 6.450224067754237e-06,
+      "loss": 0.6601,
+      "step": 9386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8689578073838515,
+      "learning_rate": 6.449587646793241e-06,
+      "loss": 0.542,
+      "step": 9387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0616611720720477,
+      "learning_rate": 6.448951205953216e-06,
+      "loss": 0.5282,
+      "step": 9388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4649840580967914,
+      "learning_rate": 6.448314745247488e-06,
+      "loss": 0.4148,
+      "step": 9389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.138748529288699,
+      "learning_rate": 6.447678264689379e-06,
+      "loss": 0.526,
+      "step": 9390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8822173179304158,
+      "learning_rate": 6.447041764292214e-06,
+      "loss": 0.571,
+      "step": 9391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8347584493906182,
+      "learning_rate": 6.446405244069319e-06,
+      "loss": 0.6097,
+      "step": 9392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3089859791010636,
+      "learning_rate": 6.445768704034018e-06,
+      "loss": 0.5875,
+      "step": 9393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.989302264245777,
+      "learning_rate": 6.445132144199636e-06,
+      "loss": 0.7705,
+      "step": 9394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.342478320502983,
+      "learning_rate": 6.4444955645795004e-06,
+      "loss": 0.4272,
+      "step": 9395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.852794222515833,
+      "learning_rate": 6.4438589651869355e-06,
+      "loss": 0.5286,
+      "step": 9396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6968885184979503,
+      "learning_rate": 6.44322234603527e-06,
+      "loss": 0.5232,
+      "step": 9397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9313934500224057,
+      "learning_rate": 6.442585707137832e-06,
+      "loss": 0.3635,
+      "step": 9398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4705696293079993,
+      "learning_rate": 6.441949048507946e-06,
+      "loss": 0.3853,
+      "step": 9399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1174586912468616,
+      "learning_rate": 6.4413123701589434e-06,
+      "loss": 0.5929,
+      "step": 9400
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.5346692800521851,
+      "eval_runtime": 68.0167,
+      "eval_samples_per_second": 50.135,
+      "eval_steps_per_second": 1.573,
+      "step": 9400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.116958071296811,
+      "learning_rate": 6.4406756721041485e-06,
+      "loss": 0.6975,
+      "step": 9401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5545215729429873,
+      "learning_rate": 6.440038954356895e-06,
+      "loss": 0.4646,
+      "step": 9402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.684406447572235,
+      "learning_rate": 6.439402216930509e-06,
+      "loss": 0.6163,
+      "step": 9403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8716726357736684,
+      "learning_rate": 6.438765459838321e-06,
+      "loss": 0.4918,
+      "step": 9404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8960200719889624,
+      "learning_rate": 6.438128683093662e-06,
+      "loss": 0.5258,
+      "step": 9405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9799683683389324,
+      "learning_rate": 6.437491886709859e-06,
+      "loss": 0.4963,
+      "step": 9406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0257050934218928,
+      "learning_rate": 6.436855070700249e-06,
+      "loss": 0.7074,
+      "step": 9407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9494468764733512,
+      "learning_rate": 6.436218235078158e-06,
+      "loss": 0.554,
+      "step": 9408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6096397593269085,
+      "learning_rate": 6.435581379856919e-06,
+      "loss": 0.3856,
+      "step": 9409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0612244356254394,
+      "learning_rate": 6.434944505049867e-06,
+      "loss": 0.5371,
+      "step": 9410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8594955971656004,
+      "learning_rate": 6.434307610670331e-06,
+      "loss": 0.4447,
+      "step": 9411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3997162926693532,
+      "learning_rate": 6.433670696731648e-06,
+      "loss": 0.4509,
+      "step": 9412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6945101208251037,
+      "learning_rate": 6.433033763247146e-06,
+      "loss": 0.4828,
+      "step": 9413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8380808917606204,
+      "learning_rate": 6.432396810230162e-06,
+      "loss": 0.5669,
+      "step": 9414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6616062631502901,
+      "learning_rate": 6.4317598376940316e-06,
+      "loss": 0.5497,
+      "step": 9415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0261595687356353,
+      "learning_rate": 6.4311228456520865e-06,
+      "loss": 0.6333,
+      "step": 9416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8097675080116555,
+      "learning_rate": 6.430485834117664e-06,
+      "loss": 0.5615,
+      "step": 9417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6892244230691744,
+      "learning_rate": 6.429848803104096e-06,
+      "loss": 0.4608,
+      "step": 9418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6206275085098403,
+      "learning_rate": 6.4292117526247246e-06,
+      "loss": 0.4711,
+      "step": 9419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6785217875950418,
+      "learning_rate": 6.428574682692879e-06,
+      "loss": 0.5426,
+      "step": 9420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8451038241712405,
+      "learning_rate": 6.4279375933219e-06,
+      "loss": 0.5537,
+      "step": 9421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0449473887214062,
+      "learning_rate": 6.427300484525126e-06,
+      "loss": 0.5531,
+      "step": 9422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7356803487170234,
+      "learning_rate": 6.426663356315889e-06,
+      "loss": 0.5183,
+      "step": 9423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8564709172105764,
+      "learning_rate": 6.426026208707533e-06,
+      "loss": 0.4772,
+      "step": 9424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.876524545116582,
+      "learning_rate": 6.425389041713391e-06,
+      "loss": 0.5874,
+      "step": 9425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4331892938471749,
+      "learning_rate": 6.424751855346807e-06,
+      "loss": 0.4155,
+      "step": 9426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9503549211114757,
+      "learning_rate": 6.424114649621116e-06,
+      "loss": 0.465,
+      "step": 9427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.077052413348374,
+      "learning_rate": 6.423477424549658e-06,
+      "loss": 0.4782,
+      "step": 9428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.521656544202639,
+      "learning_rate": 6.4228401801457754e-06,
+      "loss": 0.4172,
+      "step": 9429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6237968412187738,
+      "learning_rate": 6.422202916422805e-06,
+      "loss": 0.5089,
+      "step": 9430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9187632676770188,
+      "learning_rate": 6.421565633394091e-06,
+      "loss": 0.5214,
+      "step": 9431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4009141944930792,
+      "learning_rate": 6.420928331072972e-06,
+      "loss": 0.3698,
+      "step": 9432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6899628561998457,
+      "learning_rate": 6.42029100947279e-06,
+      "loss": 0.4544,
+      "step": 9433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6670204223272165,
+      "learning_rate": 6.419653668606889e-06,
+      "loss": 0.4281,
+      "step": 9434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5024632848271229,
+      "learning_rate": 6.419016308488607e-06,
+      "loss": 0.3376,
+      "step": 9435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4440571825424364,
+      "learning_rate": 6.418378929131291e-06,
+      "loss": 0.5074,
+      "step": 9436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0793116435058456,
+      "learning_rate": 6.417741530548284e-06,
+      "loss": 0.5547,
+      "step": 9437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4193214729497277,
+      "learning_rate": 6.4171041127529255e-06,
+      "loss": 0.4445,
+      "step": 9438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8309813978240705,
+      "learning_rate": 6.4164666757585636e-06,
+      "loss": 0.4585,
+      "step": 9439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.578860970985913,
+      "learning_rate": 6.41582921957854e-06,
+      "loss": 0.4898,
+      "step": 9440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0821968932985175,
+      "learning_rate": 6.4151917442262e-06,
+      "loss": 0.4993,
+      "step": 9441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5769999781944493,
+      "learning_rate": 6.414554249714891e-06,
+      "loss": 0.478,
+      "step": 9442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7308513021012821,
+      "learning_rate": 6.413916736057956e-06,
+      "loss": 0.3903,
+      "step": 9443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7451473879990491,
+      "learning_rate": 6.413279203268742e-06,
+      "loss": 0.3883,
+      "step": 9444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.015988454375038,
+      "learning_rate": 6.412641651360595e-06,
+      "loss": 0.518,
+      "step": 9445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8562967809372635,
+      "learning_rate": 6.412004080346861e-06,
+      "loss": 0.5079,
+      "step": 9446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6841506758604088,
+      "learning_rate": 6.411366490240889e-06,
+      "loss": 0.5262,
+      "step": 9447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9389553207697372,
+      "learning_rate": 6.4107288810560265e-06,
+      "loss": 0.6717,
+      "step": 9448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.938679745473197,
+      "learning_rate": 6.410091252805619e-06,
+      "loss": 0.5292,
+      "step": 9449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.129506487801159,
+      "learning_rate": 6.409453605503018e-06,
+      "loss": 0.4721,
+      "step": 9450
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5352295637130737,
+      "eval_runtime": 67.9476,
+      "eval_samples_per_second": 50.186,
+      "eval_steps_per_second": 1.575,
+      "step": 9450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5589704323825329,
+      "learning_rate": 6.408815939161569e-06,
+      "loss": 0.4516,
+      "step": 9451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2674911951045726,
+      "learning_rate": 6.408178253794623e-06,
+      "loss": 0.6943,
+      "step": 9452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.57302738621643,
+      "learning_rate": 6.40754054941553e-06,
+      "loss": 0.4168,
+      "step": 9453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9671467347091982,
+      "learning_rate": 6.40690282603764e-06,
+      "loss": 0.5567,
+      "step": 9454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7037958050753463,
+      "learning_rate": 6.406265083674302e-06,
+      "loss": 0.5057,
+      "step": 9455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2439949382269704,
+      "learning_rate": 6.405627322338867e-06,
+      "loss": 0.3149,
+      "step": 9456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9064347708654925,
+      "learning_rate": 6.404989542044686e-06,
+      "loss": 0.4423,
+      "step": 9457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.594694034201175,
+      "learning_rate": 6.404351742805111e-06,
+      "loss": 0.4229,
+      "step": 9458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7701894890318315,
+      "learning_rate": 6.403713924633495e-06,
+      "loss": 0.5266,
+      "step": 9459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5069580963127227,
+      "learning_rate": 6.4030760875431894e-06,
+      "loss": 0.4386,
+      "step": 9460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7132795841970747,
+      "learning_rate": 6.402438231547548e-06,
+      "loss": 0.5831,
+      "step": 9461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.686515519580792,
+      "learning_rate": 6.4018003566599194e-06,
+      "loss": 0.6112,
+      "step": 9462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3330798964860535,
+      "learning_rate": 6.401162462893663e-06,
+      "loss": 0.6512,
+      "step": 9463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7321315685483814,
+      "learning_rate": 6.400524550262128e-06,
+      "loss": 0.4099,
+      "step": 9464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.655342884558333,
+      "learning_rate": 6.399886618778674e-06,
+      "loss": 0.4069,
+      "step": 9465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6546803856105265,
+      "learning_rate": 6.39924866845665e-06,
+      "loss": 0.4589,
+      "step": 9466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6091423492614991,
+      "learning_rate": 6.398610699309414e-06,
+      "loss": 0.54,
+      "step": 9467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5876639616984465,
+      "learning_rate": 6.397972711350322e-06,
+      "loss": 0.517,
+      "step": 9468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6264678217047706,
+      "learning_rate": 6.397334704592729e-06,
+      "loss": 0.4658,
+      "step": 9469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.015781990306186,
+      "learning_rate": 6.396696679049991e-06,
+      "loss": 0.5945,
+      "step": 9470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.156758235734967,
+      "learning_rate": 6.3960586347354644e-06,
+      "loss": 0.6026,
+      "step": 9471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5685588276796225,
+      "learning_rate": 6.3954205716625085e-06,
+      "loss": 0.4186,
+      "step": 9472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9862846663550284,
+      "learning_rate": 6.394782489844478e-06,
+      "loss": 0.4935,
+      "step": 9473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7540071619686703,
+      "learning_rate": 6.394144389294732e-06,
+      "loss": 0.5371,
+      "step": 9474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.875243413666982,
+      "learning_rate": 6.393506270026627e-06,
+      "loss": 0.4729,
+      "step": 9475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9649627243634071,
+      "learning_rate": 6.392868132053526e-06,
+      "loss": 0.4657,
+      "step": 9476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7325332856972855,
+      "learning_rate": 6.392229975388784e-06,
+      "loss": 0.5318,
+      "step": 9477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4890431677950242,
+      "learning_rate": 6.391591800045761e-06,
+      "loss": 0.3681,
+      "step": 9478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7273396579671125,
+      "learning_rate": 6.390953606037818e-06,
+      "loss": 0.4613,
+      "step": 9479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4892606623484306,
+      "learning_rate": 6.3903153933783144e-06,
+      "loss": 0.3707,
+      "step": 9480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5520222386027778,
+      "learning_rate": 6.389677162080612e-06,
+      "loss": 0.3904,
+      "step": 9481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.387851290110352,
+      "learning_rate": 6.389038912158071e-06,
+      "loss": 0.3868,
+      "step": 9482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6249379112980429,
+      "learning_rate": 6.388400643624051e-06,
+      "loss": 0.3546,
+      "step": 9483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.661922692897283,
+      "learning_rate": 6.387762356491916e-06,
+      "loss": 0.5373,
+      "step": 9484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9085652068223866,
+      "learning_rate": 6.387124050775028e-06,
+      "loss": 0.6006,
+      "step": 9485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.243638679906707,
+      "learning_rate": 6.386485726486748e-06,
+      "loss": 0.6133,
+      "step": 9486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.879088788288381,
+      "learning_rate": 6.385847383640441e-06,
+      "loss": 0.5034,
+      "step": 9487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7373925898471758,
+      "learning_rate": 6.3852090222494675e-06,
+      "loss": 0.4596,
+      "step": 9488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0049699141252373,
+      "learning_rate": 6.384570642327196e-06,
+      "loss": 0.5404,
+      "step": 9489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8438779010237505,
+      "learning_rate": 6.383932243886985e-06,
+      "loss": 0.6031,
+      "step": 9490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7298058382778687,
+      "learning_rate": 6.3832938269422e-06,
+      "loss": 0.442,
+      "step": 9491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8606864128723466,
+      "learning_rate": 6.382655391506211e-06,
+      "loss": 0.5274,
+      "step": 9492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.979985839986682,
+      "learning_rate": 6.382016937592379e-06,
+      "loss": 0.5154,
+      "step": 9493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0088928344942243,
+      "learning_rate": 6.381378465214069e-06,
+      "loss": 0.5069,
+      "step": 9494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.864869481658132,
+      "learning_rate": 6.380739974384648e-06,
+      "loss": 0.6164,
+      "step": 9495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8405343696188556,
+      "learning_rate": 6.380101465117484e-06,
+      "loss": 0.5406,
+      "step": 9496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7357604980222123,
+      "learning_rate": 6.379462937425941e-06,
+      "loss": 0.4139,
+      "step": 9497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.410858259820211,
+      "learning_rate": 6.3788243913233894e-06,
+      "loss": 0.6446,
+      "step": 9498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6384291869622103,
+      "learning_rate": 6.378185826823194e-06,
+      "loss": 0.4959,
+      "step": 9499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0477835992739757,
+      "learning_rate": 6.377547243938724e-06,
+      "loss": 0.569,
+      "step": 9500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5352896451950073,
+      "eval_runtime": 67.929,
+      "eval_samples_per_second": 50.199,
+      "eval_steps_per_second": 1.575,
+      "step": 9500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1792269766206167,
+      "learning_rate": 6.376908642683349e-06,
+      "loss": 0.6244,
+      "step": 9501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.677908415783788,
+      "learning_rate": 6.376270023070433e-06,
+      "loss": 0.4786,
+      "step": 9502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.048981315635748,
+      "learning_rate": 6.37563138511335e-06,
+      "loss": 0.52,
+      "step": 9503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7851820692190825,
+      "learning_rate": 6.374992728825467e-06,
+      "loss": 0.4665,
+      "step": 9504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1023857710027816,
+      "learning_rate": 6.3743540542201565e-06,
+      "loss": 0.5459,
+      "step": 9505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7976052130247726,
+      "learning_rate": 6.373715361310786e-06,
+      "loss": 0.522,
+      "step": 9506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5734951701819009,
+      "learning_rate": 6.3730766501107265e-06,
+      "loss": 0.4265,
+      "step": 9507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5488909744335657,
+      "learning_rate": 6.372437920633351e-06,
+      "loss": 0.4637,
+      "step": 9508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.872722879710208,
+      "learning_rate": 6.3717991728920295e-06,
+      "loss": 0.5734,
+      "step": 9509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.946325376397586,
+      "learning_rate": 6.371160406900132e-06,
+      "loss": 0.541,
+      "step": 9510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7561738441369026,
+      "learning_rate": 6.370521622671034e-06,
+      "loss": 0.4574,
+      "step": 9511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.546127661958916,
+      "learning_rate": 6.3698828202181065e-06,
+      "loss": 0.5999,
+      "step": 9512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2206790528741607,
+      "learning_rate": 6.369243999554722e-06,
+      "loss": 0.6363,
+      "step": 9513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.89868737124122,
+      "learning_rate": 6.368605160694255e-06,
+      "loss": 0.5474,
+      "step": 9514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6895390213524584,
+      "learning_rate": 6.367966303650075e-06,
+      "loss": 0.4247,
+      "step": 9515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.975066538090186,
+      "learning_rate": 6.367327428435564e-06,
+      "loss": 0.6089,
+      "step": 9516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5117333036282572,
+      "learning_rate": 6.366688535064091e-06,
+      "loss": 0.4154,
+      "step": 9517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7852106831732237,
+      "learning_rate": 6.366049623549029e-06,
+      "loss": 0.5088,
+      "step": 9518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7526016268199516,
+      "learning_rate": 6.3654106939037585e-06,
+      "loss": 0.5119,
+      "step": 9519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6716370546266204,
+      "learning_rate": 6.364771746141652e-06,
+      "loss": 0.4357,
+      "step": 9520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.729470622591884,
+      "learning_rate": 6.3641327802760875e-06,
+      "loss": 0.615,
+      "step": 9521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0797076660989218,
+      "learning_rate": 6.3634937963204394e-06,
+      "loss": 0.669,
+      "step": 9522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.665450539973602,
+      "learning_rate": 6.362854794288085e-06,
+      "loss": 0.5034,
+      "step": 9523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6041123859916684,
+      "learning_rate": 6.3622157741924e-06,
+      "loss": 0.3751,
+      "step": 9524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6166540770102122,
+      "learning_rate": 6.361576736046765e-06,
+      "loss": 0.5049,
+      "step": 9525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7978375661611001,
+      "learning_rate": 6.360937679864554e-06,
+      "loss": 0.4794,
+      "step": 9526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5665306130099805,
+      "learning_rate": 6.36029860565915e-06,
+      "loss": 0.4198,
+      "step": 9527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3910269439801621,
+      "learning_rate": 6.359659513443928e-06,
+      "loss": 0.3908,
+      "step": 9528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.457282475157226,
+      "learning_rate": 6.3590204032322675e-06,
+      "loss": 0.5023,
+      "step": 9529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4671675015417331,
+      "learning_rate": 6.35838127503755e-06,
+      "loss": 0.4579,
+      "step": 9530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.736634056692047,
+      "learning_rate": 6.3577421288731525e-06,
+      "loss": 0.5784,
+      "step": 9531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5420178900293013,
+      "learning_rate": 6.357102964752455e-06,
+      "loss": 0.4501,
+      "step": 9532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.124264961928176,
+      "learning_rate": 6.356463782688841e-06,
+      "loss": 0.5038,
+      "step": 9533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8300639406008146,
+      "learning_rate": 6.35582458269569e-06,
+      "loss": 0.5554,
+      "step": 9534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0346784293299427,
+      "learning_rate": 6.3551853647863824e-06,
+      "loss": 0.532,
+      "step": 9535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9651694974065819,
+      "learning_rate": 6.354546128974299e-06,
+      "loss": 0.6779,
+      "step": 9536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6631423267515477,
+      "learning_rate": 6.353906875272825e-06,
+      "loss": 0.4686,
+      "step": 9537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8268719527752202,
+      "learning_rate": 6.353267603695341e-06,
+      "loss": 0.5072,
+      "step": 9538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5306724123965867,
+      "learning_rate": 6.352628314255229e-06,
+      "loss": 0.4279,
+      "step": 9539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8389915166010184,
+      "learning_rate": 6.351989006965872e-06,
+      "loss": 0.5425,
+      "step": 9540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.007848050715354,
+      "learning_rate": 6.351349681840655e-06,
+      "loss": 0.5884,
+      "step": 9541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7433557260868426,
+      "learning_rate": 6.350710338892962e-06,
+      "loss": 0.6453,
+      "step": 9542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7855078559279611,
+      "learning_rate": 6.350070978136176e-06,
+      "loss": 0.6307,
+      "step": 9543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0933415547212677,
+      "learning_rate": 6.349431599583681e-06,
+      "loss": 0.7288,
+      "step": 9544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2727447746245937,
+      "learning_rate": 6.348792203248863e-06,
+      "loss": 0.7473,
+      "step": 9545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7847761053280873,
+      "learning_rate": 6.34815278914511e-06,
+      "loss": 0.5586,
+      "step": 9546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8926551632029998,
+      "learning_rate": 6.347513357285804e-06,
+      "loss": 0.4894,
+      "step": 9547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.486466572433058,
+      "learning_rate": 6.346873907684331e-06,
+      "loss": 0.4583,
+      "step": 9548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0987747728493247,
+      "learning_rate": 6.34623444035408e-06,
+      "loss": 0.6896,
+      "step": 9549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8678898558474906,
+      "learning_rate": 6.345594955308436e-06,
+      "loss": 0.711,
+      "step": 9550
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5353037714958191,
+      "eval_runtime": 67.9401,
+      "eval_samples_per_second": 50.191,
+      "eval_steps_per_second": 1.575,
+      "step": 9550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7331317193183098,
+      "learning_rate": 6.344955452560787e-06,
+      "loss": 0.4718,
+      "step": 9551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7435930850048,
+      "learning_rate": 6.344315932124521e-06,
+      "loss": 0.5976,
+      "step": 9552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6644568026379525,
+      "learning_rate": 6.343676394013025e-06,
+      "loss": 0.4133,
+      "step": 9553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.625630728137034,
+      "learning_rate": 6.343036838239687e-06,
+      "loss": 0.3678,
+      "step": 9554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6466045714027695,
+      "learning_rate": 6.3423972648178966e-06,
+      "loss": 0.4132,
+      "step": 9555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.26912923334431,
+      "learning_rate": 6.341757673761044e-06,
+      "loss": 0.677,
+      "step": 9556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5959561108203075,
+      "learning_rate": 6.341118065082515e-06,
+      "loss": 0.5049,
+      "step": 9557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7646774119170527,
+      "learning_rate": 6.340478438795704e-06,
+      "loss": 0.5411,
+      "step": 9558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5599481793469538,
+      "learning_rate": 6.339838794913998e-06,
+      "loss": 0.3972,
+      "step": 9559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9610617314435377,
+      "learning_rate": 6.33919913345079e-06,
+      "loss": 0.4476,
+      "step": 9560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2829071646413372,
+      "learning_rate": 6.338559454419467e-06,
+      "loss": 0.6388,
+      "step": 9561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.016790610118335,
+      "learning_rate": 6.337919757833424e-06,
+      "loss": 0.554,
+      "step": 9562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7261890459623825,
+      "learning_rate": 6.337280043706051e-06,
+      "loss": 0.5816,
+      "step": 9563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0222407371481075,
+      "learning_rate": 6.33664031205074e-06,
+      "loss": 0.5192,
+      "step": 9564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2530442457106723,
+      "learning_rate": 6.336000562880884e-06,
+      "loss": 0.6274,
+      "step": 9565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7239041001283486,
+      "learning_rate": 6.335360796209874e-06,
+      "loss": 0.4454,
+      "step": 9566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8630940812209693,
+      "learning_rate": 6.3347210120511075e-06,
+      "loss": 0.6305,
+      "step": 9567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7533862306775747,
+      "learning_rate": 6.334081210417973e-06,
+      "loss": 0.5442,
+      "step": 9568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4955553906692605,
+      "learning_rate": 6.3334413913238666e-06,
+      "loss": 0.4913,
+      "step": 9569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7770338115953388,
+      "learning_rate": 6.332801554782184e-06,
+      "loss": 0.4777,
+      "step": 9570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5395541970827022,
+      "learning_rate": 6.332161700806317e-06,
+      "loss": 0.473,
+      "step": 9571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6085607518610359,
+      "learning_rate": 6.331521829409661e-06,
+      "loss": 0.6612,
+      "step": 9572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.732493258977984,
+      "learning_rate": 6.330881940605612e-06,
+      "loss": 0.433,
+      "step": 9573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5268105081908172,
+      "learning_rate": 6.3302420344075675e-06,
+      "loss": 0.4728,
+      "step": 9574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9959709000086103,
+      "learning_rate": 6.329602110828921e-06,
+      "loss": 0.5521,
+      "step": 9575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.157700658637154,
+      "learning_rate": 6.328962169883068e-06,
+      "loss": 0.5988,
+      "step": 9576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.038417942146651,
+      "learning_rate": 6.328322211583408e-06,
+      "loss": 0.5308,
+      "step": 9577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7238194900448907,
+      "learning_rate": 6.327682235943338e-06,
+      "loss": 0.5092,
+      "step": 9578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7670299930605187,
+      "learning_rate": 6.3270422429762515e-06,
+      "loss": 0.4805,
+      "step": 9579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.019757305810726,
+      "learning_rate": 6.326402232695552e-06,
+      "loss": 0.5339,
+      "step": 9580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6120982345199149,
+      "learning_rate": 6.325762205114634e-06,
+      "loss": 0.4598,
+      "step": 9581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.872635418083162,
+      "learning_rate": 6.325122160246898e-06,
+      "loss": 0.5203,
+      "step": 9582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9913623694220444,
+      "learning_rate": 6.3244820981057405e-06,
+      "loss": 0.5363,
+      "step": 9583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.086144646528429,
+      "learning_rate": 6.323842018704564e-06,
+      "loss": 0.5976,
+      "step": 9584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0309298066140724,
+      "learning_rate": 6.323201922056766e-06,
+      "loss": 0.6001,
+      "step": 9585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3024206172856787,
+      "learning_rate": 6.322561808175746e-06,
+      "loss": 0.5311,
+      "step": 9586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0198341615983098,
+      "learning_rate": 6.321921677074907e-06,
+      "loss": 0.6119,
+      "step": 9587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5565133829690043,
+      "learning_rate": 6.321281528767649e-06,
+      "loss": 0.3967,
+      "step": 9588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7470010522113146,
+      "learning_rate": 6.3206413632673705e-06,
+      "loss": 0.5844,
+      "step": 9589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9539539770482786,
+      "learning_rate": 6.320001180587475e-06,
+      "loss": 0.4582,
+      "step": 9590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9371532061778602,
+      "learning_rate": 6.319360980741363e-06,
+      "loss": 0.6222,
+      "step": 9591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9668288964154246,
+      "learning_rate": 6.31872076374244e-06,
+      "loss": 0.4864,
+      "step": 9592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6308614139792665,
+      "learning_rate": 6.3180805296041044e-06,
+      "loss": 0.4548,
+      "step": 9593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3433897116064815,
+      "learning_rate": 6.317440278339764e-06,
+      "loss": 0.6882,
+      "step": 9594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.34133423565748,
+      "learning_rate": 6.316800009962817e-06,
+      "loss": 0.6802,
+      "step": 9595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6297997349881659,
+      "learning_rate": 6.316159724486668e-06,
+      "loss": 0.378,
+      "step": 9596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6473294909600695,
+      "learning_rate": 6.315519421924724e-06,
+      "loss": 0.4813,
+      "step": 9597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8535461091415932,
+      "learning_rate": 6.314879102290387e-06,
+      "loss": 0.4916,
+      "step": 9598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6624150090445464,
+      "learning_rate": 6.314238765597062e-06,
+      "loss": 0.4458,
+      "step": 9599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5954393440595331,
+      "learning_rate": 6.313598411858154e-06,
+      "loss": 0.4407,
+      "step": 9600
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5342313051223755,
+      "eval_runtime": 67.9164,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 9600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7241010309004852,
+      "learning_rate": 6.312958041087068e-06,
+      "loss": 0.3983,
+      "step": 9601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.150919655122565,
+      "learning_rate": 6.312317653297212e-06,
+      "loss": 0.5272,
+      "step": 9602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5423478047815327,
+      "learning_rate": 6.3116772485019895e-06,
+      "loss": 0.4343,
+      "step": 9603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7867253910676748,
+      "learning_rate": 6.311036826714808e-06,
+      "loss": 0.5597,
+      "step": 9604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9422401017554958,
+      "learning_rate": 6.3103963879490745e-06,
+      "loss": 0.6367,
+      "step": 9605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1015517962430827,
+      "learning_rate": 6.3097559322181965e-06,
+      "loss": 0.4799,
+      "step": 9606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7328579081968245,
+      "learning_rate": 6.309115459535581e-06,
+      "loss": 0.5659,
+      "step": 9607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6587411413888298,
+      "learning_rate": 6.308474969914634e-06,
+      "loss": 0.6132,
+      "step": 9608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4954863379793504,
+      "learning_rate": 6.307834463368767e-06,
+      "loss": 0.4825,
+      "step": 9609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4124503139551388,
+      "learning_rate": 6.307193939911387e-06,
+      "loss": 0.4648,
+      "step": 9610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.79958266792939,
+      "learning_rate": 6.306553399555905e-06,
+      "loss": 0.4729,
+      "step": 9611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.183650275951084,
+      "learning_rate": 6.3059128423157254e-06,
+      "loss": 0.6112,
+      "step": 9612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.734762563925726,
+      "learning_rate": 6.305272268204262e-06,
+      "loss": 0.4776,
+      "step": 9613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.96874582938953,
+      "learning_rate": 6.304631677234925e-06,
+      "loss": 0.6315,
+      "step": 9614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0279599183991723,
+      "learning_rate": 6.303991069421123e-06,
+      "loss": 0.5914,
+      "step": 9615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5854750574319765,
+      "learning_rate": 6.3033504447762675e-06,
+      "loss": 0.4023,
+      "step": 9616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6214542558226195,
+      "learning_rate": 6.302709803313769e-06,
+      "loss": 0.685,
+      "step": 9617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5648752161843473,
+      "learning_rate": 6.30206914504704e-06,
+      "loss": 0.3659,
+      "step": 9618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8002874276711491,
+      "learning_rate": 6.301428469989491e-06,
+      "loss": 0.4596,
+      "step": 9619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0845822087856316,
+      "learning_rate": 6.3007877781545355e-06,
+      "loss": 0.6374,
+      "step": 9620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4930600839547623,
+      "learning_rate": 6.300147069555585e-06,
+      "loss": 0.4224,
+      "step": 9621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.466554467655762,
+      "learning_rate": 6.299506344206052e-06,
+      "loss": 0.6105,
+      "step": 9622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9330264290717074,
+      "learning_rate": 6.2988656021193515e-06,
+      "loss": 0.541,
+      "step": 9623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6456203262323446,
+      "learning_rate": 6.298224843308894e-06,
+      "loss": 0.615,
+      "step": 9624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.100575456799597,
+      "learning_rate": 6.297584067788098e-06,
+      "loss": 0.7464,
+      "step": 9625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.998173131938254,
+      "learning_rate": 6.296943275570372e-06,
+      "loss": 0.5991,
+      "step": 9626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7982614868249305,
+      "learning_rate": 6.2963024666691344e-06,
+      "loss": 0.5114,
+      "step": 9627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2768601267399733,
+      "learning_rate": 6.2956616410978e-06,
+      "loss": 0.5815,
+      "step": 9628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6253160790822327,
+      "learning_rate": 6.295020798869782e-06,
+      "loss": 0.4423,
+      "step": 9629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.665765266389084,
+      "learning_rate": 6.2943799399985e-06,
+      "loss": 0.5849,
+      "step": 9630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.870853883079246,
+      "learning_rate": 6.293739064497365e-06,
+      "loss": 0.523,
+      "step": 9631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6053029909237466,
+      "learning_rate": 6.293098172379796e-06,
+      "loss": 0.4482,
+      "step": 9632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.280238613155019,
+      "learning_rate": 6.292457263659211e-06,
+      "loss": 0.5003,
+      "step": 9633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.45239885610985,
+      "learning_rate": 6.291816338349023e-06,
+      "loss": 0.355,
+      "step": 9634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8702026052114258,
+      "learning_rate": 6.291175396462652e-06,
+      "loss": 0.3972,
+      "step": 9635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7540789906516976,
+      "learning_rate": 6.290534438013517e-06,
+      "loss": 0.5702,
+      "step": 9636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6708556641730716,
+      "learning_rate": 6.289893463015034e-06,
+      "loss": 0.5063,
+      "step": 9637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7216956515650585,
+      "learning_rate": 6.289252471480621e-06,
+      "loss": 0.625,
+      "step": 9638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9549276287723825,
+      "learning_rate": 6.2886114634236994e-06,
+      "loss": 0.6179,
+      "step": 9639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7318326270638593,
+      "learning_rate": 6.287970438857685e-06,
+      "loss": 0.5904,
+      "step": 9640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7762093797785112,
+      "learning_rate": 6.287329397796001e-06,
+      "loss": 0.5832,
+      "step": 9641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.096345857451086,
+      "learning_rate": 6.2866883402520625e-06,
+      "loss": 0.5978,
+      "step": 9642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4762044810361872,
+      "learning_rate": 6.286047266239293e-06,
+      "loss": 0.4707,
+      "step": 9643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3235382289086823,
+      "learning_rate": 6.285406175771115e-06,
+      "loss": 0.3393,
+      "step": 9644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9440380556529513,
+      "learning_rate": 6.284765068860943e-06,
+      "loss": 0.4805,
+      "step": 9645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.040752042787232,
+      "learning_rate": 6.284123945522204e-06,
+      "loss": 0.4818,
+      "step": 9646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6023913756706,
+      "learning_rate": 6.283482805768316e-06,
+      "loss": 0.4933,
+      "step": 9647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.616874335477105,
+      "learning_rate": 6.282841649612702e-06,
+      "loss": 0.5519,
+      "step": 9648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7810295150635078,
+      "learning_rate": 6.282200477068784e-06,
+      "loss": 0.6039,
+      "step": 9649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7499729924707763,
+      "learning_rate": 6.281559288149986e-06,
+      "loss": 0.564,
+      "step": 9650
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5325720906257629,
+      "eval_runtime": 67.8922,
+      "eval_samples_per_second": 50.227,
+      "eval_steps_per_second": 1.576,
+      "step": 9650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6698413572845958,
+      "learning_rate": 6.280918082869729e-06,
+      "loss": 0.4842,
+      "step": 9651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6489013733005748,
+      "learning_rate": 6.280276861241438e-06,
+      "loss": 0.5044,
+      "step": 9652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7694340161042017,
+      "learning_rate": 6.279635623278534e-06,
+      "loss": 0.5577,
+      "step": 9653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8086531185208072,
+      "learning_rate": 6.278994368994444e-06,
+      "loss": 0.495,
+      "step": 9654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.880789698896061,
+      "learning_rate": 6.278353098402589e-06,
+      "loss": 0.5,
+      "step": 9655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4486894597171878,
+      "learning_rate": 6.277711811516397e-06,
+      "loss": 0.3902,
+      "step": 9656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.828216568371255,
+      "learning_rate": 6.277070508349291e-06,
+      "loss": 0.5198,
+      "step": 9657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7811739222460976,
+      "learning_rate": 6.276429188914697e-06,
+      "loss": 0.5434,
+      "step": 9658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7605624991667546,
+      "learning_rate": 6.275787853226038e-06,
+      "loss": 0.4618,
+      "step": 9659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4160747243848464,
+      "learning_rate": 6.275146501296745e-06,
+      "loss": 0.4179,
+      "step": 9660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7396735537307884,
+      "learning_rate": 6.2745051331402395e-06,
+      "loss": 0.4852,
+      "step": 9661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0456176730030196,
+      "learning_rate": 6.273863748769951e-06,
+      "loss": 0.555,
+      "step": 9662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8392033940434631,
+      "learning_rate": 6.273222348199304e-06,
+      "loss": 0.5401,
+      "step": 9663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.887073042743415,
+      "learning_rate": 6.272580931441727e-06,
+      "loss": 0.4647,
+      "step": 9664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7939621307208924,
+      "learning_rate": 6.2719394985106516e-06,
+      "loss": 0.4467,
+      "step": 9665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5825430598642147,
+      "learning_rate": 6.2712980494194994e-06,
+      "loss": 0.5079,
+      "step": 9666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9972184478797559,
+      "learning_rate": 6.270656584181704e-06,
+      "loss": 0.524,
+      "step": 9667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8275710005228303,
+      "learning_rate": 6.270015102810689e-06,
+      "loss": 0.4556,
+      "step": 9668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6485664033710792,
+      "learning_rate": 6.269373605319887e-06,
+      "loss": 0.6089,
+      "step": 9669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7455656831611728,
+      "learning_rate": 6.268732091722728e-06,
+      "loss": 0.5076,
+      "step": 9670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6681912926551241,
+      "learning_rate": 6.268090562032639e-06,
+      "loss": 0.3802,
+      "step": 9671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5001916191706937,
+      "learning_rate": 6.267449016263051e-06,
+      "loss": 0.5499,
+      "step": 9672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.968937774502472,
+      "learning_rate": 6.266807454427394e-06,
+      "loss": 0.6615,
+      "step": 9673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9231443877888237,
+      "learning_rate": 6.266165876539099e-06,
+      "loss": 0.4519,
+      "step": 9674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.722307918695121,
+      "learning_rate": 6.2655242826115995e-06,
+      "loss": 0.4132,
+      "step": 9675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9946550208457794,
+      "learning_rate": 6.264882672658323e-06,
+      "loss": 0.4931,
+      "step": 9676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.907261725370406,
+      "learning_rate": 6.264241046692703e-06,
+      "loss": 0.6245,
+      "step": 9677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7060582670353936,
+      "learning_rate": 6.263599404728171e-06,
+      "loss": 0.4971,
+      "step": 9678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2847263408324214,
+      "learning_rate": 6.26295774677816e-06,
+      "loss": 0.2979,
+      "step": 9679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8173502020401022,
+      "learning_rate": 6.262316072856103e-06,
+      "loss": 0.4333,
+      "step": 9680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.061755088844874,
+      "learning_rate": 6.261674382975433e-06,
+      "loss": 0.6115,
+      "step": 9681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8252329676498866,
+      "learning_rate": 6.2610326771495815e-06,
+      "loss": 0.6696,
+      "step": 9682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0154348852131205,
+      "learning_rate": 6.260390955391984e-06,
+      "loss": 0.6257,
+      "step": 9683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.600443378616619,
+      "learning_rate": 6.2597492177160746e-06,
+      "loss": 0.4833,
+      "step": 9684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9341213052638968,
+      "learning_rate": 6.259107464135286e-06,
+      "loss": 0.6447,
+      "step": 9685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9036823216927192,
+      "learning_rate": 6.2584656946630554e-06,
+      "loss": 0.656,
+      "step": 9686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.82247787469812,
+      "learning_rate": 6.257823909312816e-06,
+      "loss": 0.511,
+      "step": 9687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.646265480059428,
+      "learning_rate": 6.257182108098003e-06,
+      "loss": 0.4745,
+      "step": 9688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.806708732036495,
+      "learning_rate": 6.256540291032054e-06,
+      "loss": 0.4501,
+      "step": 9689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6310663471137037,
+      "learning_rate": 6.255898458128404e-06,
+      "loss": 0.4321,
+      "step": 9690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7788971322870866,
+      "learning_rate": 6.2552566094004896e-06,
+      "loss": 0.6072,
+      "step": 9691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7395863293630505,
+      "learning_rate": 6.254614744861747e-06,
+      "loss": 0.4312,
+      "step": 9692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.012918591822831,
+      "learning_rate": 6.253972864525612e-06,
+      "loss": 0.4546,
+      "step": 9693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9886084284107088,
+      "learning_rate": 6.253330968405525e-06,
+      "loss": 0.61,
+      "step": 9694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8066523887175503,
+      "learning_rate": 6.252689056514921e-06,
+      "loss": 0.4774,
+      "step": 9695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.633248290337768,
+      "learning_rate": 6.25204712886724e-06,
+      "loss": 0.4235,
+      "step": 9696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.200870389429612,
+      "learning_rate": 6.2514051854759195e-06,
+      "loss": 0.6579,
+      "step": 9697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6821665422642098,
+      "learning_rate": 6.2507632263543975e-06,
+      "loss": 0.4848,
+      "step": 9698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6723736999244012,
+      "learning_rate": 6.250121251516115e-06,
+      "loss": 0.4819,
+      "step": 9699
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6181152696475678,
+      "learning_rate": 6.249479260974508e-06,
+      "loss": 0.5512,
+      "step": 9700
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.5331730842590332,
+      "eval_runtime": 67.929,
+      "eval_samples_per_second": 50.199,
+      "eval_steps_per_second": 1.575,
+      "step": 9700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4488339643047974,
+      "learning_rate": 6.248837254743019e-06,
+      "loss": 0.444,
+      "step": 9701
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9896292947020378,
+      "learning_rate": 6.248195232835087e-06,
+      "loss": 0.609,
+      "step": 9702
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6660464233218966,
+      "learning_rate": 6.247553195264153e-06,
+      "loss": 0.4128,
+      "step": 9703
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8104292076594037,
+      "learning_rate": 6.246911142043657e-06,
+      "loss": 0.5129,
+      "step": 9704
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0467822713443313,
+      "learning_rate": 6.2462690731870416e-06,
+      "loss": 0.6234,
+      "step": 9705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4029466698965027,
+      "learning_rate": 6.2456269887077435e-06,
+      "loss": 0.8981,
+      "step": 9706
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1045262944818997,
+      "learning_rate": 6.24498488861921e-06,
+      "loss": 0.6142,
+      "step": 9707
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.076626225141291,
+      "learning_rate": 6.244342772934879e-06,
+      "loss": 0.5728,
+      "step": 9708
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1017527309113717,
+      "learning_rate": 6.243700641668196e-06,
+      "loss": 0.8424,
+      "step": 9709
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.81234394196888,
+      "learning_rate": 6.243058494832601e-06,
+      "loss": 0.4766,
+      "step": 9710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8965638178886612,
+      "learning_rate": 6.242416332441538e-06,
+      "loss": 0.5777,
+      "step": 9711
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7307570856509102,
+      "learning_rate": 6.2417741545084495e-06,
+      "loss": 0.5448,
+      "step": 9712
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7221376712415746,
+      "learning_rate": 6.241131961046781e-06,
+      "loss": 0.5719,
+      "step": 9713
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8476891002973246,
+      "learning_rate": 6.240489752069975e-06,
+      "loss": 0.6223,
+      "step": 9714
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5339300748768423,
+      "learning_rate": 6.2398475275914764e-06,
+      "loss": 0.4956,
+      "step": 9715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.817605344331235,
+      "learning_rate": 6.239205287624729e-06,
+      "loss": 0.6643,
+      "step": 9716
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.327591642152413,
+      "learning_rate": 6.238563032183177e-06,
+      "loss": 0.3694,
+      "step": 9717
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9541365800405281,
+      "learning_rate": 6.237920761280269e-06,
+      "loss": 0.5685,
+      "step": 9718
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3312049472245535,
+      "learning_rate": 6.237278474929447e-06,
+      "loss": 0.5524,
+      "step": 9719
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7121152142322134,
+      "learning_rate": 6.236636173144157e-06,
+      "loss": 0.5065,
+      "step": 9720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.761609320315244,
+      "learning_rate": 6.235993855937847e-06,
+      "loss": 0.4128,
+      "step": 9721
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6598386126192217,
+      "learning_rate": 6.235351523323962e-06,
+      "loss": 0.3577,
+      "step": 9722
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.783291336040703,
+      "learning_rate": 6.234709175315953e-06,
+      "loss": 0.5088,
+      "step": 9723
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.907220452349611,
+      "learning_rate": 6.234066811927261e-06,
+      "loss": 0.6838,
+      "step": 9724
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7631125371986838,
+      "learning_rate": 6.233424433171336e-06,
+      "loss": 0.4293,
+      "step": 9725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4795509957675603,
+      "learning_rate": 6.232782039061626e-06,
+      "loss": 0.4455,
+      "step": 9726
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.03887842092757,
+      "learning_rate": 6.23213962961158e-06,
+      "loss": 0.4886,
+      "step": 9727
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.750326104776596,
+      "learning_rate": 6.231497204834646e-06,
+      "loss": 0.4647,
+      "step": 9728
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9420901762606724,
+      "learning_rate": 6.2308547647442715e-06,
+      "loss": 0.6438,
+      "step": 9729
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7818232602612878,
+      "learning_rate": 6.230212309353905e-06,
+      "loss": 0.5332,
+      "step": 9730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8309486658670342,
+      "learning_rate": 6.229569838677e-06,
+      "loss": 0.5035,
+      "step": 9731
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6027071259895056,
+      "learning_rate": 6.228927352727002e-06,
+      "loss": 0.4411,
+      "step": 9732
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6723420168953924,
+      "learning_rate": 6.228284851517363e-06,
+      "loss": 0.5122,
+      "step": 9733
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7529878319789765,
+      "learning_rate": 6.227642335061535e-06,
+      "loss": 0.3842,
+      "step": 9734
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3528953431906832,
+      "learning_rate": 6.226999803372963e-06,
+      "loss": 0.335,
+      "step": 9735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7133815183466554,
+      "learning_rate": 6.226357256465104e-06,
+      "loss": 0.5296,
+      "step": 9736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.099661224835048,
+      "learning_rate": 6.225714694351406e-06,
+      "loss": 0.6599,
+      "step": 9737
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5766931017462016,
+      "learning_rate": 6.225072117045321e-06,
+      "loss": 0.4476,
+      "step": 9738
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9370767187184605,
+      "learning_rate": 6.224429524560304e-06,
+      "loss": 0.6141,
+      "step": 9739
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4589540593315222,
+      "learning_rate": 6.223786916909804e-06,
+      "loss": 0.4726,
+      "step": 9740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7949142255349322,
+      "learning_rate": 6.223144294107274e-06,
+      "loss": 0.389,
+      "step": 9741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7379037364671577,
+      "learning_rate": 6.222501656166167e-06,
+      "loss": 0.5513,
+      "step": 9742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.695287996070184,
+      "learning_rate": 6.221859003099937e-06,
+      "loss": 0.4657,
+      "step": 9743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6488116788720613,
+      "learning_rate": 6.221216334922037e-06,
+      "loss": 0.4604,
+      "step": 9744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.122249878146321,
+      "learning_rate": 6.2205736516459215e-06,
+      "loss": 0.5974,
+      "step": 9745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.034147122597805,
+      "learning_rate": 6.219930953285043e-06,
+      "loss": 0.4716,
+      "step": 9746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6046973207980606,
+      "learning_rate": 6.219288239852859e-06,
+      "loss": 0.5157,
+      "step": 9747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8586985372641267,
+      "learning_rate": 6.218645511362821e-06,
+      "loss": 0.6347,
+      "step": 9748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.002302106867676,
+      "learning_rate": 6.218002767828385e-06,
+      "loss": 0.3539,
+      "step": 9749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7119052475708836,
+      "learning_rate": 6.21736000926301e-06,
+      "loss": 0.5727,
+      "step": 9750
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.531967282295227,
+      "eval_runtime": 67.9187,
+      "eval_samples_per_second": 50.207,
+      "eval_steps_per_second": 1.575,
+      "step": 9750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5993693589585622,
+      "learning_rate": 6.216717235680144e-06,
+      "loss": 0.42,
+      "step": 9751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6960401222174917,
+      "learning_rate": 6.216074447093252e-06,
+      "loss": 0.5214,
+      "step": 9752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9345425864854495,
+      "learning_rate": 6.215431643515786e-06,
+      "loss": 0.5725,
+      "step": 9753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7861189596413556,
+      "learning_rate": 6.214788824961201e-06,
+      "loss": 0.5871,
+      "step": 9754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7389161280177925,
+      "learning_rate": 6.214145991442957e-06,
+      "loss": 0.571,
+      "step": 9755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7804825934266428,
+      "learning_rate": 6.213503142974509e-06,
+      "loss": 0.4839,
+      "step": 9756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3436312589465964,
+      "learning_rate": 6.212860279569317e-06,
+      "loss": 0.6849,
+      "step": 9757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4406923611855287,
+      "learning_rate": 6.212217401240839e-06,
+      "loss": 0.4847,
+      "step": 9758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2447994192239467,
+      "learning_rate": 6.2115745080025295e-06,
+      "loss": 0.5821,
+      "step": 9759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4982308207978168,
+      "learning_rate": 6.210931599867851e-06,
+      "loss": 0.3834,
+      "step": 9760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1595262384098226,
+      "learning_rate": 6.21028867685026e-06,
+      "loss": 0.6201,
+      "step": 9761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6002309960224885,
+      "learning_rate": 6.209645738963217e-06,
+      "loss": 0.3769,
+      "step": 9762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1190742500476505,
+      "learning_rate": 6.209002786220183e-06,
+      "loss": 0.4721,
+      "step": 9763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7582693445452569,
+      "learning_rate": 6.208359818634615e-06,
+      "loss": 0.4941,
+      "step": 9764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9814202639222285,
+      "learning_rate": 6.207716836219972e-06,
+      "loss": 0.7248,
+      "step": 9765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0968688873089434,
+      "learning_rate": 6.20707383898972e-06,
+      "loss": 0.4874,
+      "step": 9766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0476269142397037,
+      "learning_rate": 6.206430826957315e-06,
+      "loss": 0.513,
+      "step": 9767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.771359566704645,
+      "learning_rate": 6.2057878001362196e-06,
+      "loss": 0.6069,
+      "step": 9768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7950669516604063,
+      "learning_rate": 6.205144758539895e-06,
+      "loss": 0.6073,
+      "step": 9769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8698163223280935,
+      "learning_rate": 6.204501702181802e-06,
+      "loss": 0.506,
+      "step": 9770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4512100424440353,
+      "learning_rate": 6.203858631075404e-06,
+      "loss": 0.3385,
+      "step": 9771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8880547494318647,
+      "learning_rate": 6.203215545234162e-06,
+      "loss": 0.6943,
+      "step": 9772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8759063935403113,
+      "learning_rate": 6.202572444671542e-06,
+      "loss": 0.5249,
+      "step": 9773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9129752841021186,
+      "learning_rate": 6.201929329401002e-06,
+      "loss": 0.6112,
+      "step": 9774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.563956334690307,
+      "learning_rate": 6.2012861994360075e-06,
+      "loss": 0.4015,
+      "step": 9775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7748490405988706,
+      "learning_rate": 6.2006430547900235e-06,
+      "loss": 0.5511,
+      "step": 9776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8885850011713359,
+      "learning_rate": 6.19999989547651e-06,
+      "loss": 0.5028,
+      "step": 9777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.480809885819535,
+      "learning_rate": 6.199356721508936e-06,
+      "loss": 0.4275,
+      "step": 9778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6597017501654667,
+      "learning_rate": 6.198713532900762e-06,
+      "loss": 0.4019,
+      "step": 9779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.381508838901062,
+      "learning_rate": 6.198070329665454e-06,
+      "loss": 0.4617,
+      "step": 9780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.176729978739288,
+      "learning_rate": 6.197427111816478e-06,
+      "loss": 0.5985,
+      "step": 9781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9626646792136508,
+      "learning_rate": 6.196783879367298e-06,
+      "loss": 0.5936,
+      "step": 9782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.054535958119133,
+      "learning_rate": 6.196140632331381e-06,
+      "loss": 0.7552,
+      "step": 9783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8759811008647596,
+      "learning_rate": 6.195497370722192e-06,
+      "loss": 0.3562,
+      "step": 9784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1200872437193414,
+      "learning_rate": 6.194854094553197e-06,
+      "loss": 0.6463,
+      "step": 9785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0146625169810064,
+      "learning_rate": 6.194210803837862e-06,
+      "loss": 0.4472,
+      "step": 9786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7190096625684423,
+      "learning_rate": 6.193567498589657e-06,
+      "loss": 0.5805,
+      "step": 9787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.920783840857769,
+      "learning_rate": 6.192924178822046e-06,
+      "loss": 0.5668,
+      "step": 9788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3280227298736418,
+      "learning_rate": 6.192280844548498e-06,
+      "loss": 0.3432,
+      "step": 9789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.751531195501305,
+      "learning_rate": 6.191637495782482e-06,
+      "loss": 0.5653,
+      "step": 9790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.750718717002616,
+      "learning_rate": 6.1909941325374636e-06,
+      "loss": 0.4913,
+      "step": 9791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.553156537929278,
+      "learning_rate": 6.190350754826912e-06,
+      "loss": 0.5805,
+      "step": 9792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7249944282541358,
+      "learning_rate": 6.189707362664295e-06,
+      "loss": 0.5925,
+      "step": 9793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1885098818603663,
+      "learning_rate": 6.1890639560630835e-06,
+      "loss": 0.7953,
+      "step": 9794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8784033327680372,
+      "learning_rate": 6.188420535036748e-06,
+      "loss": 0.5032,
+      "step": 9795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.946324979702436,
+      "learning_rate": 6.187777099598754e-06,
+      "loss": 0.642,
+      "step": 9796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2118584526772236,
+      "learning_rate": 6.1871336497625744e-06,
+      "loss": 0.4728,
+      "step": 9797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4761609632076778,
+      "learning_rate": 6.1864901855416795e-06,
+      "loss": 0.4275,
+      "step": 9798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2223853461903023,
+      "learning_rate": 6.185846706949537e-06,
+      "loss": 0.5721,
+      "step": 9799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8090847980548674,
+      "learning_rate": 6.185203213999621e-06,
+      "loss": 0.5901,
+      "step": 9800
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.5324429273605347,
+      "eval_runtime": 67.967,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 9800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3639850521955965,
+      "learning_rate": 6.184559706705401e-06,
+      "loss": 0.3928,
+      "step": 9801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7924587497458218,
+      "learning_rate": 6.1839161850803495e-06,
+      "loss": 0.5976,
+      "step": 9802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9868453814523348,
+      "learning_rate": 6.183272649137936e-06,
+      "loss": 0.4202,
+      "step": 9803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7790624986176795,
+      "learning_rate": 6.182629098891634e-06,
+      "loss": 0.5205,
+      "step": 9804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3137921523894915,
+      "learning_rate": 6.181985534354918e-06,
+      "loss": 0.385,
+      "step": 9805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7524652581050806,
+      "learning_rate": 6.181341955541257e-06,
+      "loss": 0.4909,
+      "step": 9806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6547734900159912,
+      "learning_rate": 6.180698362464125e-06,
+      "loss": 0.4929,
+      "step": 9807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6864729576224904,
+      "learning_rate": 6.180054755136996e-06,
+      "loss": 0.6001,
+      "step": 9808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4565011887204236,
+      "learning_rate": 6.179411133573342e-06,
+      "loss": 0.4325,
+      "step": 9809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5003072703610307,
+      "learning_rate": 6.1787674977866405e-06,
+      "loss": 0.48,
+      "step": 9810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8313869090401889,
+      "learning_rate": 6.17812384779036e-06,
+      "loss": 0.5277,
+      "step": 9811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.604760359033317,
+      "learning_rate": 6.177480183597979e-06,
+      "loss": 0.4825,
+      "step": 9812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7071280403667075,
+      "learning_rate": 6.176836505222973e-06,
+      "loss": 0.4235,
+      "step": 9813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5632215161271972,
+      "learning_rate": 6.176192812678813e-06,
+      "loss": 0.4538,
+      "step": 9814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7195750986855964,
+      "learning_rate": 6.175549105978976e-06,
+      "loss": 0.5112,
+      "step": 9815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.745871639738924,
+      "learning_rate": 6.17490538513694e-06,
+      "loss": 0.4581,
+      "step": 9816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0260472829329013,
+      "learning_rate": 6.174261650166178e-06,
+      "loss": 0.4882,
+      "step": 9817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5843850548425917,
+      "learning_rate": 6.173617901080167e-06,
+      "loss": 0.4424,
+      "step": 9818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7747589293198183,
+      "learning_rate": 6.1729741378923845e-06,
+      "loss": 0.7845,
+      "step": 9819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.804871725224852,
+      "learning_rate": 6.172330360616306e-06,
+      "loss": 0.5982,
+      "step": 9820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8005875195226986,
+      "learning_rate": 6.17168656926541e-06,
+      "loss": 0.5393,
+      "step": 9821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.130439036577557,
+      "learning_rate": 6.1710427638531715e-06,
+      "loss": 0.6945,
+      "step": 9822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7971237338597492,
+      "learning_rate": 6.170398944393069e-06,
+      "loss": 0.5038,
+      "step": 9823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6990830497215614,
+      "learning_rate": 6.169755110898582e-06,
+      "loss": 0.584,
+      "step": 9824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8509278453072209,
+      "learning_rate": 6.169111263383187e-06,
+      "loss": 0.4852,
+      "step": 9825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4087342297475218,
+      "learning_rate": 6.168467401860365e-06,
+      "loss": 0.3213,
+      "step": 9826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7065423112004547,
+      "learning_rate": 6.167823526343593e-06,
+      "loss": 0.3989,
+      "step": 9827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6948532729238954,
+      "learning_rate": 6.167179636846348e-06,
+      "loss": 0.4307,
+      "step": 9828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.273449247685582,
+      "learning_rate": 6.166535733382113e-06,
+      "loss": 0.8362,
+      "step": 9829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9858204338092798,
+      "learning_rate": 6.165891815964367e-06,
+      "loss": 0.7835,
+      "step": 9830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.007395902761099,
+      "learning_rate": 6.165247884606588e-06,
+      "loss": 0.7503,
+      "step": 9831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.196046171250188,
+      "learning_rate": 6.164603939322259e-06,
+      "loss": 0.7168,
+      "step": 9832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7993810453862273,
+      "learning_rate": 6.1639599801248584e-06,
+      "loss": 0.5109,
+      "step": 9833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1499575312683836,
+      "learning_rate": 6.163316007027868e-06,
+      "loss": 0.5746,
+      "step": 9834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.15020208301116,
+      "learning_rate": 6.16267202004477e-06,
+      "loss": 0.5583,
+      "step": 9835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.548396325589164,
+      "learning_rate": 6.162028019189044e-06,
+      "loss": 0.4312,
+      "step": 9836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6620324114316174,
+      "learning_rate": 6.161384004474173e-06,
+      "loss": 0.4949,
+      "step": 9837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6296838433696892,
+      "learning_rate": 6.16073997591364e-06,
+      "loss": 0.528,
+      "step": 9838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8216674468267657,
+      "learning_rate": 6.160095933520923e-06,
+      "loss": 0.7162,
+      "step": 9839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8786256217889845,
+      "learning_rate": 6.15945187730951e-06,
+      "loss": 0.4146,
+      "step": 9840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.598934897674305,
+      "learning_rate": 6.1588078072928806e-06,
+      "loss": 0.4149,
+      "step": 9841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.612318882361504,
+      "learning_rate": 6.158163723484519e-06,
+      "loss": 0.4965,
+      "step": 9842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.840024418967838,
+      "learning_rate": 6.1575196258979086e-06,
+      "loss": 0.5557,
+      "step": 9843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.081831221204546,
+      "learning_rate": 6.156875514546533e-06,
+      "loss": 0.5045,
+      "step": 9844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0963429939704006,
+      "learning_rate": 6.156231389443877e-06,
+      "loss": 0.7191,
+      "step": 9845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8088842426782983,
+      "learning_rate": 6.155587250603426e-06,
+      "loss": 0.4471,
+      "step": 9846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8052103901099248,
+      "learning_rate": 6.1549430980386615e-06,
+      "loss": 0.528,
+      "step": 9847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5782905346033287,
+      "learning_rate": 6.154298931763071e-06,
+      "loss": 0.6537,
+      "step": 9848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.827420229569745,
+      "learning_rate": 6.153654751790137e-06,
+      "loss": 0.6824,
+      "step": 9849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.893235141017907,
+      "learning_rate": 6.153010558133347e-06,
+      "loss": 0.5139,
+      "step": 9850
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.5325884222984314,
+      "eval_runtime": 67.9535,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 9850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.010018125070688,
+      "learning_rate": 6.152366350806188e-06,
+      "loss": 0.7166,
+      "step": 9851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8181631748623253,
+      "learning_rate": 6.151722129822143e-06,
+      "loss": 0.6073,
+      "step": 9852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.317529260050339,
+      "learning_rate": 6.151077895194702e-06,
+      "loss": 0.4372,
+      "step": 9853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.180096722693597,
+      "learning_rate": 6.1504336469373486e-06,
+      "loss": 0.5644,
+      "step": 9854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9044185564148064,
+      "learning_rate": 6.1497893850635706e-06,
+      "loss": 0.6703,
+      "step": 9855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.82258905528487,
+      "learning_rate": 6.149145109586856e-06,
+      "loss": 0.5366,
+      "step": 9856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.810274351022971,
+      "learning_rate": 6.148500820520691e-06,
+      "loss": 0.575,
+      "step": 9857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8441582523972126,
+      "learning_rate": 6.147856517878564e-06,
+      "loss": 0.5377,
+      "step": 9858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.129323542326203,
+      "learning_rate": 6.1472122016739645e-06,
+      "loss": 0.2977,
+      "step": 9859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0995227482587757,
+      "learning_rate": 6.146567871920377e-06,
+      "loss": 0.6288,
+      "step": 9860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9038421905413372,
+      "learning_rate": 6.145923528631294e-06,
+      "loss": 0.5114,
+      "step": 9861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0512602540035365,
+      "learning_rate": 6.145279171820204e-06,
+      "loss": 0.6633,
+      "step": 9862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8649833317008828,
+      "learning_rate": 6.1446348015005956e-06,
+      "loss": 0.423,
+      "step": 9863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.106705650506377,
+      "learning_rate": 6.143990417685956e-06,
+      "loss": 0.6205,
+      "step": 9864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8227832885113764,
+      "learning_rate": 6.143346020389778e-06,
+      "loss": 0.6344,
+      "step": 9865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4803893781216824,
+      "learning_rate": 6.142701609625551e-06,
+      "loss": 0.4938,
+      "step": 9866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.53264143382188,
+      "learning_rate": 6.142057185406766e-06,
+      "loss": 0.5322,
+      "step": 9867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3129763771646816,
+      "learning_rate": 6.141412747746911e-06,
+      "loss": 0.3411,
+      "step": 9868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6152218541641903,
+      "learning_rate": 6.1407682966594794e-06,
+      "loss": 0.4129,
+      "step": 9869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6992292452828406,
+      "learning_rate": 6.1401238321579606e-06,
+      "loss": 0.4461,
+      "step": 9870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.72409023425377,
+      "learning_rate": 6.1394793542558485e-06,
+      "loss": 0.5794,
+      "step": 9871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8501096721702242,
+      "learning_rate": 6.138834862966633e-06,
+      "loss": 0.545,
+      "step": 9872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.450744353810793,
+      "learning_rate": 6.1381903583038035e-06,
+      "loss": 0.4017,
+      "step": 9873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7722295869620037,
+      "learning_rate": 6.1375458402808596e-06,
+      "loss": 0.4939,
+      "step": 9874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.007470174452993,
+      "learning_rate": 6.136901308911287e-06,
+      "loss": 0.6207,
+      "step": 9875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7832841653678604,
+      "learning_rate": 6.136256764208581e-06,
+      "loss": 0.5649,
+      "step": 9876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.040423207822207,
+      "learning_rate": 6.135612206186235e-06,
+      "loss": 0.5161,
+      "step": 9877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1611583939547665,
+      "learning_rate": 6.134967634857743e-06,
+      "loss": 0.6535,
+      "step": 9878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7784356577409544,
+      "learning_rate": 6.134323050236598e-06,
+      "loss": 0.4589,
+      "step": 9879
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.002903162270701,
+      "learning_rate": 6.133678452336294e-06,
+      "loss": 0.4412,
+      "step": 9880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6669054638284309,
+      "learning_rate": 6.133033841170324e-06,
+      "loss": 0.4479,
+      "step": 9881
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6703606610533883,
+      "learning_rate": 6.1323892167521845e-06,
+      "loss": 0.5086,
+      "step": 9882
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4308847748930285,
+      "learning_rate": 6.131744579095369e-06,
+      "loss": 0.3881,
+      "step": 9883
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7088811209813348,
+      "learning_rate": 6.131099928213375e-06,
+      "loss": 0.4344,
+      "step": 9884
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8258275943213191,
+      "learning_rate": 6.130455264119694e-06,
+      "loss": 0.4776,
+      "step": 9885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8818047468408055,
+      "learning_rate": 6.129810586827825e-06,
+      "loss": 0.5508,
+      "step": 9886
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7187796613485402,
+      "learning_rate": 6.129165896351261e-06,
+      "loss": 0.6145,
+      "step": 9887
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6906866616208278,
+      "learning_rate": 6.1285211927035e-06,
+      "loss": 0.4302,
+      "step": 9888
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9063248500878867,
+      "learning_rate": 6.127876475898039e-06,
+      "loss": 0.5251,
+      "step": 9889
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2175756195836422,
+      "learning_rate": 6.1272317459483735e-06,
+      "loss": 0.6463,
+      "step": 9890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7306668549993107,
+      "learning_rate": 6.126587002868002e-06,
+      "loss": 0.5035,
+      "step": 9891
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2789090986739422,
+      "learning_rate": 6.125942246670419e-06,
+      "loss": 0.3879,
+      "step": 9892
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5664643624292438,
+      "learning_rate": 6.125297477369126e-06,
+      "loss": 0.3481,
+      "step": 9893
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7382452650329618,
+      "learning_rate": 6.124652694977616e-06,
+      "loss": 0.5812,
+      "step": 9894
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.094167695403867,
+      "learning_rate": 6.124007899509392e-06,
+      "loss": 0.5415,
+      "step": 9895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4762941710244282,
+      "learning_rate": 6.12336309097795e-06,
+      "loss": 0.3692,
+      "step": 9896
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.120647840984015,
+      "learning_rate": 6.122718269396789e-06,
+      "loss": 0.5735,
+      "step": 9897
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.632223791682367,
+      "learning_rate": 6.122073434779406e-06,
+      "loss": 0.4331,
+      "step": 9898
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.896259968039248,
+      "learning_rate": 6.121428587139304e-06,
+      "loss": 0.5514,
+      "step": 9899
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7206253678297792,
+      "learning_rate": 6.120783726489979e-06,
+      "loss": 0.7262,
+      "step": 9900
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.5314356088638306,
+      "eval_runtime": 67.9424,
+      "eval_samples_per_second": 50.19,
+      "eval_steps_per_second": 1.575,
+      "step": 9900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8635604574891413,
+      "learning_rate": 6.120138852844935e-06,
+      "loss": 0.5081,
+      "step": 9901
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8246241781942611,
+      "learning_rate": 6.119493966217667e-06,
+      "loss": 0.5112,
+      "step": 9902
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8355069598636697,
+      "learning_rate": 6.118849066621677e-06,
+      "loss": 0.6193,
+      "step": 9903
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.582933777508484,
+      "learning_rate": 6.118204154070469e-06,
+      "loss": 0.4492,
+      "step": 9904
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.217783214417292,
+      "learning_rate": 6.117559228577538e-06,
+      "loss": 0.6028,
+      "step": 9905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7171510236454228,
+      "learning_rate": 6.116914290156389e-06,
+      "loss": 0.7049,
+      "step": 9906
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4262698777681273,
+      "learning_rate": 6.116269338820522e-06,
+      "loss": 0.4514,
+      "step": 9907
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8783209018538751,
+      "learning_rate": 6.11562437458344e-06,
+      "loss": 0.6651,
+      "step": 9908
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6283714230733384,
+      "learning_rate": 6.114979397458643e-06,
+      "loss": 0.4704,
+      "step": 9909
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6214307911596126,
+      "learning_rate": 6.114334407459636e-06,
+      "loss": 0.4669,
+      "step": 9910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.942234682317419,
+      "learning_rate": 6.113689404599919e-06,
+      "loss": 0.5879,
+      "step": 9911
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6825255587217836,
+      "learning_rate": 6.113044388892997e-06,
+      "loss": 0.5157,
+      "step": 9912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7201081133247487,
+      "learning_rate": 6.1123993603523694e-06,
+      "loss": 0.4197,
+      "step": 9913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1482485954704784,
+      "learning_rate": 6.111754318991543e-06,
+      "loss": 0.6156,
+      "step": 9914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.207647683423656,
+      "learning_rate": 6.111109264824019e-06,
+      "loss": 0.6445,
+      "step": 9915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6719130448821018,
+      "learning_rate": 6.110464197863302e-06,
+      "loss": 0.4913,
+      "step": 9916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1987087442102102,
+      "learning_rate": 6.1098191181228965e-06,
+      "loss": 0.5992,
+      "step": 9917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0115398732392564,
+      "learning_rate": 6.109174025616305e-06,
+      "loss": 0.5038,
+      "step": 9918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8811078495041855,
+      "learning_rate": 6.108528920357038e-06,
+      "loss": 0.6191,
+      "step": 9919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.742547612883504,
+      "learning_rate": 6.107883802358592e-06,
+      "loss": 0.4287,
+      "step": 9920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.691570942870138,
+      "learning_rate": 6.107238671634477e-06,
+      "loss": 0.468,
+      "step": 9921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6795253036853577,
+      "learning_rate": 6.1065935281981984e-06,
+      "loss": 0.4347,
+      "step": 9922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5052783584706244,
+      "learning_rate": 6.10594837206326e-06,
+      "loss": 0.3976,
+      "step": 9923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5655438749005826,
+      "learning_rate": 6.10530320324317e-06,
+      "loss": 0.4648,
+      "step": 9924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7879559210423532,
+      "learning_rate": 6.104658021751433e-06,
+      "loss": 0.4085,
+      "step": 9925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.469092670396477,
+      "learning_rate": 6.1040128276015556e-06,
+      "loss": 0.6113,
+      "step": 9926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7183129470025793,
+      "learning_rate": 6.103367620807045e-06,
+      "loss": 0.4758,
+      "step": 9927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.469671643017864,
+      "learning_rate": 6.102722401381408e-06,
+      "loss": 0.4965,
+      "step": 9928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0306931159167956,
+      "learning_rate": 6.102077169338151e-06,
+      "loss": 0.564,
+      "step": 9929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0721979439945866,
+      "learning_rate": 6.101431924690785e-06,
+      "loss": 0.6546,
+      "step": 9930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7766503140743013,
+      "learning_rate": 6.100786667452813e-06,
+      "loss": 0.5278,
+      "step": 9931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5502118002135896,
+      "learning_rate": 6.100141397637744e-06,
+      "loss": 0.3987,
+      "step": 9932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8666459927170607,
+      "learning_rate": 6.0994961152590915e-06,
+      "loss": 0.5058,
+      "step": 9933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5703701206746232,
+      "learning_rate": 6.0988508203303565e-06,
+      "loss": 0.5827,
+      "step": 9934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6588765034177995,
+      "learning_rate": 6.098205512865053e-06,
+      "loss": 0.639,
+      "step": 9935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7351147950686987,
+      "learning_rate": 6.0975601928766896e-06,
+      "loss": 0.5876,
+      "step": 9936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4753956306551335,
+      "learning_rate": 6.096914860378772e-06,
+      "loss": 0.3923,
+      "step": 9937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7293801331436844,
+      "learning_rate": 6.096269515384816e-06,
+      "loss": 0.4897,
+      "step": 9938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.405964961793398,
+      "learning_rate": 6.095624157908327e-06,
+      "loss": 0.3757,
+      "step": 9939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.193469257936965,
+      "learning_rate": 6.094978787962815e-06,
+      "loss": 0.4629,
+      "step": 9940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7361824197306879,
+      "learning_rate": 6.094333405561792e-06,
+      "loss": 0.4178,
+      "step": 9941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8496095138346205,
+      "learning_rate": 6.093688010718768e-06,
+      "loss": 0.5556,
+      "step": 9942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7252859954156352,
+      "learning_rate": 6.093042603447256e-06,
+      "loss": 0.5144,
+      "step": 9943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9045322830820821,
+      "learning_rate": 6.092397183760764e-06,
+      "loss": 0.7461,
+      "step": 9944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3891931150997372,
+      "learning_rate": 6.091751751672805e-06,
+      "loss": 0.3989,
+      "step": 9945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8716134385584235,
+      "learning_rate": 6.091106307196891e-06,
+      "loss": 0.5737,
+      "step": 9946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2391559311432054,
+      "learning_rate": 6.0904608503465335e-06,
+      "loss": 0.6437,
+      "step": 9947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9302539046864242,
+      "learning_rate": 6.089815381135244e-06,
+      "loss": 0.5677,
+      "step": 9948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.034376185863169,
+      "learning_rate": 6.089169899576538e-06,
+      "loss": 0.6341,
+      "step": 9949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7632631557950604,
+      "learning_rate": 6.088524405683923e-06,
+      "loss": 0.5224,
+      "step": 9950
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.5315147042274475,
+      "eval_runtime": 67.9554,
+      "eval_samples_per_second": 50.18,
+      "eval_steps_per_second": 1.575,
+      "step": 9950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9562879141757143,
+      "learning_rate": 6.0878788994709185e-06,
+      "loss": 0.6381,
+      "step": 9951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.113242098486831,
+      "learning_rate": 6.087233380951032e-06,
+      "loss": 0.7294,
+      "step": 9952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9957449816136805,
+      "learning_rate": 6.086587850137781e-06,
+      "loss": 0.6945,
+      "step": 9953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1769743961303827,
+      "learning_rate": 6.085942307044677e-06,
+      "loss": 0.6481,
+      "step": 9954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5963135085649547,
+      "learning_rate": 6.085296751685233e-06,
+      "loss": 0.5725,
+      "step": 9955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3167223287819634,
+      "learning_rate": 6.084651184072967e-06,
+      "loss": 0.6587,
+      "step": 9956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.789453250307438,
+      "learning_rate": 6.0840056042213925e-06,
+      "loss": 0.5645,
+      "step": 9957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1011770359813444,
+      "learning_rate": 6.08336001214402e-06,
+      "loss": 0.612,
+      "step": 9958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6361792925865677,
+      "learning_rate": 6.082714407854371e-06,
+      "loss": 0.4791,
+      "step": 9959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8186244467860633,
+      "learning_rate": 6.082068791365957e-06,
+      "loss": 0.413,
+      "step": 9960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.492429777697312,
+      "learning_rate": 6.0814231626922934e-06,
+      "loss": 0.4007,
+      "step": 9961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.118556904419105,
+      "learning_rate": 6.080777521846899e-06,
+      "loss": 0.5314,
+      "step": 9962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.090314511852874,
+      "learning_rate": 6.080131868843286e-06,
+      "loss": 0.4831,
+      "step": 9963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5783856394091857,
+      "learning_rate": 6.0794862036949745e-06,
+      "loss": 0.3969,
+      "step": 9964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4183581865005996,
+      "learning_rate": 6.078840526415479e-06,
+      "loss": 0.3812,
+      "step": 9965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1192007227974554,
+      "learning_rate": 6.078194837018314e-06,
+      "loss": 0.6633,
+      "step": 9966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.153321793706906,
+      "learning_rate": 6.0775491355170025e-06,
+      "loss": 0.679,
+      "step": 9967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6208710544272011,
+      "learning_rate": 6.076903421925056e-06,
+      "loss": 0.4096,
+      "step": 9968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6893100373451493,
+      "learning_rate": 6.0762576962559975e-06,
+      "loss": 0.5065,
+      "step": 9969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4776632783076098,
+      "learning_rate": 6.07561195852334e-06,
+      "loss": 0.4931,
+      "step": 9970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5559127012446308,
+      "learning_rate": 6.074966208740604e-06,
+      "loss": 0.3784,
+      "step": 9971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.832072587490964,
+      "learning_rate": 6.074320446921308e-06,
+      "loss": 0.663,
+      "step": 9972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3540539658095356,
+      "learning_rate": 6.07367467307897e-06,
+      "loss": 0.4777,
+      "step": 9973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.097439281458676,
+      "learning_rate": 6.073028887227108e-06,
+      "loss": 0.4748,
+      "step": 9974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7334448441467214,
+      "learning_rate": 6.072383089379243e-06,
+      "loss": 0.6393,
+      "step": 9975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5929171925528087,
+      "learning_rate": 6.0717372795488935e-06,
+      "loss": 0.5172,
+      "step": 9976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9988321574690868,
+      "learning_rate": 6.071091457749579e-06,
+      "loss": 0.7078,
+      "step": 9977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.998538950817682,
+      "learning_rate": 6.07044562399482e-06,
+      "loss": 0.7927,
+      "step": 9978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4284272414058115,
+      "learning_rate": 6.069799778298137e-06,
+      "loss": 0.3289,
+      "step": 9979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.07432509251289,
+      "learning_rate": 6.069153920673049e-06,
+      "loss": 0.5739,
+      "step": 9980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8219859250822759,
+      "learning_rate": 6.068508051133077e-06,
+      "loss": 0.4421,
+      "step": 9981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3074844792295894,
+      "learning_rate": 6.067862169691741e-06,
+      "loss": 0.7744,
+      "step": 9982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.779316972719202,
+      "learning_rate": 6.0672162763625634e-06,
+      "loss": 0.5403,
+      "step": 9983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8089072654915115,
+      "learning_rate": 6.066570371159065e-06,
+      "loss": 0.6095,
+      "step": 9984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8729435757911201,
+      "learning_rate": 6.065924454094769e-06,
+      "loss": 0.6665,
+      "step": 9985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7724890712872485,
+      "learning_rate": 6.0652785251831955e-06,
+      "loss": 0.4914,
+      "step": 9986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2966960796572655,
+      "learning_rate": 6.0646325844378686e-06,
+      "loss": 0.3517,
+      "step": 9987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5546168030012948,
+      "learning_rate": 6.063986631872307e-06,
+      "loss": 0.3925,
+      "step": 9988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9290569943116422,
+      "learning_rate": 6.063340667500037e-06,
+      "loss": 0.451,
+      "step": 9989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1054063656021293,
+      "learning_rate": 6.06269469133458e-06,
+      "loss": 0.5753,
+      "step": 9990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.439924990506133,
+      "learning_rate": 6.062048703389458e-06,
+      "loss": 0.4428,
+      "step": 9991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5174302493918004,
+      "learning_rate": 6.061402703678197e-06,
+      "loss": 0.3934,
+      "step": 9992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2323143317972955,
+      "learning_rate": 6.060756692214319e-06,
+      "loss": 0.4127,
+      "step": 9993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2204076999459468,
+      "learning_rate": 6.060110669011348e-06,
+      "loss": 0.4458,
+      "step": 9994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.152380541866223,
+      "learning_rate": 6.0594646340828055e-06,
+      "loss": 0.4621,
+      "step": 9995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2442885614920924,
+      "learning_rate": 6.058818587442221e-06,
+      "loss": 0.741,
+      "step": 9996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8734814545767529,
+      "learning_rate": 6.058172529103117e-06,
+      "loss": 0.7447,
+      "step": 9997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3206949751870076,
+      "learning_rate": 6.057526459079015e-06,
+      "loss": 0.6234,
+      "step": 9998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.023180516403649,
+      "learning_rate": 6.056880377383446e-06,
+      "loss": 0.5251,
+      "step": 9999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5898426614322339,
+      "learning_rate": 6.056234284029931e-06,
+      "loss": 0.3067,
+      "step": 10000
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.5313936471939087,
+      "eval_runtime": 67.935,
+      "eval_samples_per_second": 50.195,
+      "eval_steps_per_second": 1.575,
+      "step": 10000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.92303952169054,
+      "learning_rate": 6.055588179031996e-06,
+      "loss": 0.5878,
+      "step": 10001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7522372023708244,
+      "learning_rate": 6.054942062403169e-06,
+      "loss": 0.4332,
+      "step": 10002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8449516061113418,
+      "learning_rate": 6.054295934156973e-06,
+      "loss": 0.5761,
+      "step": 10003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5551784017225008,
+      "learning_rate": 6.053649794306936e-06,
+      "loss": 0.4373,
+      "step": 10004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1730663454395787,
+      "learning_rate": 6.053003642866585e-06,
+      "loss": 0.5264,
+      "step": 10005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.367447846943292,
+      "learning_rate": 6.052357479849446e-06,
+      "loss": 0.7237,
+      "step": 10006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8616878683051687,
+      "learning_rate": 6.051711305269047e-06,
+      "loss": 0.5026,
+      "step": 10007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.181698236007914,
+      "learning_rate": 6.051065119138914e-06,
+      "loss": 0.6304,
+      "step": 10008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6483312624838466,
+      "learning_rate": 6.0504189214725724e-06,
+      "loss": 0.4279,
+      "step": 10009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0732581968559405,
+      "learning_rate": 6.049772712283557e-06,
+      "loss": 0.6704,
+      "step": 10010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.735782302982471,
+      "learning_rate": 6.049126491585388e-06,
+      "loss": 0.4846,
+      "step": 10011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8492306091809911,
+      "learning_rate": 6.048480259391599e-06,
+      "loss": 0.6254,
+      "step": 10012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7501735554255737,
+      "learning_rate": 6.047834015715717e-06,
+      "loss": 0.4507,
+      "step": 10013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9759799623788188,
+      "learning_rate": 6.047187760571268e-06,
+      "loss": 0.5617,
+      "step": 10014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6786426992377514,
+      "learning_rate": 6.046541493971784e-06,
+      "loss": 0.4154,
+      "step": 10015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7262533579357753,
+      "learning_rate": 6.045895215930793e-06,
+      "loss": 0.4649,
+      "step": 10016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.617263815749601,
+      "learning_rate": 6.045248926461825e-06,
+      "loss": 0.4026,
+      "step": 10017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9090730052234428,
+      "learning_rate": 6.044602625578409e-06,
+      "loss": 0.4324,
+      "step": 10018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8639251081676647,
+      "learning_rate": 6.043956313294076e-06,
+      "loss": 0.4942,
+      "step": 10019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7541786925573883,
+      "learning_rate": 6.043309989622356e-06,
+      "loss": 0.4285,
+      "step": 10020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.146513042815706,
+      "learning_rate": 6.042663654576778e-06,
+      "loss": 0.397,
+      "step": 10021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5051246625373391,
+      "learning_rate": 6.042017308170873e-06,
+      "loss": 0.442,
+      "step": 10022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7843360819960083,
+      "learning_rate": 6.0413709504181736e-06,
+      "loss": 0.494,
+      "step": 10023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.559402451890376,
+      "learning_rate": 6.040724581332208e-06,
+      "loss": 0.4712,
+      "step": 10024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2001105301641855,
+      "learning_rate": 6.040078200926511e-06,
+      "loss": 0.693,
+      "step": 10025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9254797923014069,
+      "learning_rate": 6.039431809214611e-06,
+      "loss": 0.5497,
+      "step": 10026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9166643095859908,
+      "learning_rate": 6.038785406210042e-06,
+      "loss": 0.6699,
+      "step": 10027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2429886096747498,
+      "learning_rate": 6.038138991926334e-06,
+      "loss": 0.4052,
+      "step": 10028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8324047465596043,
+      "learning_rate": 6.03749256637702e-06,
+      "loss": 0.5385,
+      "step": 10029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6517165901565343,
+      "learning_rate": 6.0368461295756354e-06,
+      "loss": 0.421,
+      "step": 10030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8872069779928378,
+      "learning_rate": 6.036199681535709e-06,
+      "loss": 0.5938,
+      "step": 10031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7299426694697173,
+      "learning_rate": 6.035553222270774e-06,
+      "loss": 0.4193,
+      "step": 10032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4527282974473568,
+      "learning_rate": 6.034906751794367e-06,
+      "loss": 0.3555,
+      "step": 10033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9964123522782562,
+      "learning_rate": 6.034260270120018e-06,
+      "loss": 0.6077,
+      "step": 10034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.895949756340976,
+      "learning_rate": 6.0336137772612635e-06,
+      "loss": 0.4149,
+      "step": 10035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6209764886088063,
+      "learning_rate": 6.032967273231635e-06,
+      "loss": 0.4829,
+      "step": 10036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6080237018355055,
+      "learning_rate": 6.03232075804467e-06,
+      "loss": 0.5537,
+      "step": 10037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6670737275867074,
+      "learning_rate": 6.031674231713898e-06,
+      "loss": 0.5107,
+      "step": 10038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.918103555773594,
+      "learning_rate": 6.0310276942528565e-06,
+      "loss": 0.5006,
+      "step": 10039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6553159791676326,
+      "learning_rate": 6.03038114567508e-06,
+      "loss": 0.518,
+      "step": 10040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8579544274257895,
+      "learning_rate": 6.029734585994104e-06,
+      "loss": 0.5251,
+      "step": 10041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8016856370396523,
+      "learning_rate": 6.029088015223464e-06,
+      "loss": 0.3626,
+      "step": 10042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8584077010606606,
+      "learning_rate": 6.028441433376695e-06,
+      "loss": 0.5182,
+      "step": 10043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8460325730949896,
+      "learning_rate": 6.027794840467332e-06,
+      "loss": 0.6105,
+      "step": 10044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.569270935850642,
+      "learning_rate": 6.027148236508912e-06,
+      "loss": 0.5532,
+      "step": 10045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4786842958191004,
+      "learning_rate": 6.026501621514971e-06,
+      "loss": 0.4742,
+      "step": 10046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3874159302985372,
+      "learning_rate": 6.025854995499046e-06,
+      "loss": 0.3503,
+      "step": 10047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7918883639417689,
+      "learning_rate": 6.025208358474673e-06,
+      "loss": 0.5616,
+      "step": 10048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1094343244742397,
+      "learning_rate": 6.02456171045539e-06,
+      "loss": 0.7292,
+      "step": 10049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.734715366173689,
+      "learning_rate": 6.023915051454733e-06,
+      "loss": 0.4928,
+      "step": 10050
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.5302700996398926,
+      "eval_runtime": 67.9294,
+      "eval_samples_per_second": 50.199,
+      "eval_steps_per_second": 1.575,
+      "step": 10050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6807399409620492,
+      "learning_rate": 6.023268381486239e-06,
+      "loss": 0.4614,
+      "step": 10051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0371832444409486,
+      "learning_rate": 6.022621700563447e-06,
+      "loss": 0.5188,
+      "step": 10052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.423629810541582,
+      "learning_rate": 6.021975008699894e-06,
+      "loss": 0.628,
+      "step": 10053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.031723031144768,
+      "learning_rate": 6.021328305909119e-06,
+      "loss": 0.5037,
+      "step": 10054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.414836974329989,
+      "learning_rate": 6.02068159220466e-06,
+      "loss": 0.4845,
+      "step": 10055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1511530131129244,
+      "learning_rate": 6.0200348676000546e-06,
+      "loss": 0.5624,
+      "step": 10056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.196074501308284,
+      "learning_rate": 6.019388132108842e-06,
+      "loss": 0.7062,
+      "step": 10057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0070624596199873,
+      "learning_rate": 6.018741385744562e-06,
+      "loss": 0.5665,
+      "step": 10058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7678210180483143,
+      "learning_rate": 6.018094628520753e-06,
+      "loss": 0.4228,
+      "step": 10059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.512638303615254,
+      "learning_rate": 6.017447860450956e-06,
+      "loss": 0.415,
+      "step": 10060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4109119672011483,
+      "learning_rate": 6.016801081548708e-06,
+      "loss": 0.3646,
+      "step": 10061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0609320704954386,
+      "learning_rate": 6.01615429182755e-06,
+      "loss": 0.4217,
+      "step": 10062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9448928957306475,
+      "learning_rate": 6.015507491301023e-06,
+      "loss": 0.6825,
+      "step": 10063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6972549372551162,
+      "learning_rate": 6.014860679982667e-06,
+      "loss": 0.4345,
+      "step": 10064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7090685604787499,
+      "learning_rate": 6.0142138578860224e-06,
+      "loss": 0.5279,
+      "step": 10065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7102286326349525,
+      "learning_rate": 6.0135670250246305e-06,
+      "loss": 0.5279,
+      "step": 10066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3786404881673409,
+      "learning_rate": 6.012920181412032e-06,
+      "loss": 0.3753,
+      "step": 10067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0877420787206145,
+      "learning_rate": 6.012273327061767e-06,
+      "loss": 0.4918,
+      "step": 10068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9294816621072244,
+      "learning_rate": 6.011626461987379e-06,
+      "loss": 0.5757,
+      "step": 10069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.862836974240496,
+      "learning_rate": 6.010979586202409e-06,
+      "loss": 0.4714,
+      "step": 10070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0157524167659218,
+      "learning_rate": 6.010332699720398e-06,
+      "loss": 0.5751,
+      "step": 10071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.644285548708812,
+      "learning_rate": 6.00968580255489e-06,
+      "loss": 0.4243,
+      "step": 10072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8692562001397095,
+      "learning_rate": 6.009038894719424e-06,
+      "loss": 0.4375,
+      "step": 10073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4602182665977599,
+      "learning_rate": 6.008391976227547e-06,
+      "loss": 0.4544,
+      "step": 10074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.473303952399368,
+      "learning_rate": 6.0077450470928e-06,
+      "loss": 0.4217,
+      "step": 10075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.654782974774597,
+      "learning_rate": 6.007098107328726e-06,
+      "loss": 0.5463,
+      "step": 10076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8064341605899112,
+      "learning_rate": 6.006451156948866e-06,
+      "loss": 0.499,
+      "step": 10077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7298359584279956,
+      "learning_rate": 6.005804195966766e-06,
+      "loss": 0.5302,
+      "step": 10078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7820549921943916,
+      "learning_rate": 6.005157224395972e-06,
+      "loss": 0.429,
+      "step": 10079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8155123760814347,
+      "learning_rate": 6.0045102422500225e-06,
+      "loss": 0.4742,
+      "step": 10080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6112912627896676,
+      "learning_rate": 6.0038632495424655e-06,
+      "loss": 0.5646,
+      "step": 10081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8399598341199848,
+      "learning_rate": 6.003216246286844e-06,
+      "loss": 0.5197,
+      "step": 10082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7912171140940563,
+      "learning_rate": 6.002569232496701e-06,
+      "loss": 0.4945,
+      "step": 10083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.720877735050417,
+      "learning_rate": 6.001922208185587e-06,
+      "loss": 0.476,
+      "step": 10084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6327445659672255,
+      "learning_rate": 6.001275173367041e-06,
+      "loss": 0.4237,
+      "step": 10085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0890954346000767,
+      "learning_rate": 6.000628128054611e-06,
+      "loss": 0.4756,
+      "step": 10086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8301571430353076,
+      "learning_rate": 5.999981072261843e-06,
+      "loss": 0.5693,
+      "step": 10087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6567580469444136,
+      "learning_rate": 5.999334006002281e-06,
+      "loss": 0.4317,
+      "step": 10088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.539668805582771,
+      "learning_rate": 5.998686929289472e-06,
+      "loss": 0.4107,
+      "step": 10089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5297234091556964,
+      "learning_rate": 5.998039842136962e-06,
+      "loss": 0.5072,
+      "step": 10090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8703882607920226,
+      "learning_rate": 5.997392744558297e-06,
+      "loss": 0.5083,
+      "step": 10091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7266832626088067,
+      "learning_rate": 5.996745636567024e-06,
+      "loss": 0.4525,
+      "step": 10092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.130824073114249,
+      "learning_rate": 5.9960985181766885e-06,
+      "loss": 0.4981,
+      "step": 10093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5758878725282064,
+      "learning_rate": 5.995451389400841e-06,
+      "loss": 0.4415,
+      "step": 10094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.691099541978474,
+      "learning_rate": 5.9948042502530246e-06,
+      "loss": 0.5828,
+      "step": 10095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.903545636667632,
+      "learning_rate": 5.994157100746788e-06,
+      "loss": 0.6092,
+      "step": 10096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5646301744404618,
+      "learning_rate": 5.993509940895681e-06,
+      "loss": 0.4953,
+      "step": 10097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8389416647225598,
+      "learning_rate": 5.992862770713249e-06,
+      "loss": 0.537,
+      "step": 10098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.766723413871671,
+      "learning_rate": 5.992215590213041e-06,
+      "loss": 0.5291,
+      "step": 10099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6465546816440269,
+      "learning_rate": 5.991568399408607e-06,
+      "loss": 0.3959,
+      "step": 10100
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5287665128707886,
+      "eval_runtime": 67.9738,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 10100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8981635156852728,
+      "learning_rate": 5.990921198313492e-06,
+      "loss": 0.5559,
+      "step": 10101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.002576518528273,
+      "learning_rate": 5.990273986941247e-06,
+      "loss": 0.6662,
+      "step": 10102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4852117943012861,
+      "learning_rate": 5.989626765305422e-06,
+      "loss": 0.4595,
+      "step": 10103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.963044724614955,
+      "learning_rate": 5.9889795334195625e-06,
+      "loss": 0.561,
+      "step": 10104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.594209812883944,
+      "learning_rate": 5.988332291297222e-06,
+      "loss": 0.4612,
+      "step": 10105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6120703309053452,
+      "learning_rate": 5.987685038951947e-06,
+      "loss": 0.4258,
+      "step": 10106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.459336095385555,
+      "learning_rate": 5.9870377763972875e-06,
+      "loss": 0.4407,
+      "step": 10107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5458919913480156,
+      "learning_rate": 5.9863905036467985e-06,
+      "loss": 0.3714,
+      "step": 10108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4143481220259624,
+      "learning_rate": 5.985743220714023e-06,
+      "loss": 0.4343,
+      "step": 10109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.030675136539293,
+      "learning_rate": 5.9850959276125165e-06,
+      "loss": 0.7048,
+      "step": 10110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9202966287844212,
+      "learning_rate": 5.984448624355828e-06,
+      "loss": 0.6085,
+      "step": 10111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9312607112337827,
+      "learning_rate": 5.983801310957507e-06,
+      "loss": 0.5782,
+      "step": 10112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8410064787984803,
+      "learning_rate": 5.983153987431108e-06,
+      "loss": 0.5032,
+      "step": 10113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6639057786713791,
+      "learning_rate": 5.982506653790179e-06,
+      "loss": 0.4804,
+      "step": 10114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8401194320125567,
+      "learning_rate": 5.981859310048274e-06,
+      "loss": 0.5679,
+      "step": 10115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8923591198199874,
+      "learning_rate": 5.981211956218943e-06,
+      "loss": 0.6212,
+      "step": 10116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.789807998112438,
+      "learning_rate": 5.9805645923157394e-06,
+      "loss": 0.4803,
+      "step": 10117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9593080541884675,
+      "learning_rate": 5.979917218352214e-06,
+      "loss": 0.5898,
+      "step": 10118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8168963416079518,
+      "learning_rate": 5.979269834341919e-06,
+      "loss": 0.5007,
+      "step": 10119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6842068386880038,
+      "learning_rate": 5.97862244029841e-06,
+      "loss": 0.4789,
+      "step": 10120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6782799953684866,
+      "learning_rate": 5.977975036235235e-06,
+      "loss": 0.7767,
+      "step": 10121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.352968225849005,
+      "learning_rate": 5.977327622165949e-06,
+      "loss": 0.3643,
+      "step": 10122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5033240124082914,
+      "learning_rate": 5.976680198104108e-06,
+      "loss": 0.5466,
+      "step": 10123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3583904608670292,
+      "learning_rate": 5.976032764063262e-06,
+      "loss": 0.332,
+      "step": 10124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6863172092049352,
+      "learning_rate": 5.975385320056965e-06,
+      "loss": 0.4288,
+      "step": 10125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7132307834079223,
+      "learning_rate": 5.974737866098773e-06,
+      "loss": 0.4966,
+      "step": 10126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8016945798116533,
+      "learning_rate": 5.974090402202238e-06,
+      "loss": 0.6569,
+      "step": 10127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4172678620739216,
+      "learning_rate": 5.973442928380916e-06,
+      "loss": 0.4915,
+      "step": 10128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7837727678416766,
+      "learning_rate": 5.9727954446483595e-06,
+      "loss": 0.4628,
+      "step": 10129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9817898854446727,
+      "learning_rate": 5.972147951018124e-06,
+      "loss": 0.6265,
+      "step": 10130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8066094895014604,
+      "learning_rate": 5.971500447503763e-06,
+      "loss": 0.5064,
+      "step": 10131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9225065566319721,
+      "learning_rate": 5.970852934118835e-06,
+      "loss": 0.543,
+      "step": 10132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.026115884436435,
+      "learning_rate": 5.9702054108768925e-06,
+      "loss": 0.6406,
+      "step": 10133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7331932955257325,
+      "learning_rate": 5.969557877791491e-06,
+      "loss": 0.4982,
+      "step": 10134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5388715822299155,
+      "learning_rate": 5.968910334876189e-06,
+      "loss": 0.4182,
+      "step": 10135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1859579021608724,
+      "learning_rate": 5.9682627821445385e-06,
+      "loss": 0.7668,
+      "step": 10136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7442216663157106,
+      "learning_rate": 5.967615219610099e-06,
+      "loss": 0.4451,
+      "step": 10137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0962814948962696,
+      "learning_rate": 5.966967647286423e-06,
+      "loss": 0.5787,
+      "step": 10138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9040748476042606,
+      "learning_rate": 5.9663200651870725e-06,
+      "loss": 0.5666,
+      "step": 10139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9490845998939796,
+      "learning_rate": 5.9656724733255985e-06,
+      "loss": 0.4065,
+      "step": 10140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.793913131055735,
+      "learning_rate": 5.965024871715561e-06,
+      "loss": 0.5258,
+      "step": 10141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4375918951468887,
+      "learning_rate": 5.964377260370519e-06,
+      "loss": 0.4567,
+      "step": 10142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1672524837895275,
+      "learning_rate": 5.963729639304025e-06,
+      "loss": 0.7495,
+      "step": 10143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4819566614085553,
+      "learning_rate": 5.963082008529639e-06,
+      "loss": 0.6833,
+      "step": 10144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7939955906694756,
+      "learning_rate": 5.962434368060921e-06,
+      "loss": 0.404,
+      "step": 10145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7784414913491617,
+      "learning_rate": 5.961786717911424e-06,
+      "loss": 0.5729,
+      "step": 10146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.865748579726715,
+      "learning_rate": 5.9611390580947106e-06,
+      "loss": 0.4887,
+      "step": 10147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7232561522316663,
+      "learning_rate": 5.9604913886243374e-06,
+      "loss": 0.422,
+      "step": 10148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8837239109283737,
+      "learning_rate": 5.959843709513862e-06,
+      "loss": 0.7432,
+      "step": 10149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9177745177251864,
+      "learning_rate": 5.959196020776845e-06,
+      "loss": 0.6509,
+      "step": 10150
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5277344584465027,
+      "eval_runtime": 68.0316,
+      "eval_samples_per_second": 50.124,
+      "eval_steps_per_second": 1.573,
+      "step": 10150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5664265861422206,
+      "learning_rate": 5.958548322426844e-06,
+      "loss": 0.4525,
+      "step": 10151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9377528103705843,
+      "learning_rate": 5.957900614477418e-06,
+      "loss": 0.6122,
+      "step": 10152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2758065229298312,
+      "learning_rate": 5.957252896942129e-06,
+      "loss": 0.3959,
+      "step": 10153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.985029732245017,
+      "learning_rate": 5.956605169834533e-06,
+      "loss": 0.6102,
+      "step": 10154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5889441118425227,
+      "learning_rate": 5.955957433168192e-06,
+      "loss": 0.4148,
+      "step": 10155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4009152113188232,
+      "learning_rate": 5.9553096869566654e-06,
+      "loss": 0.4127,
+      "step": 10156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.19389347058946,
+      "learning_rate": 5.9546619312135124e-06,
+      "loss": 0.6797,
+      "step": 10157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.52435297124934,
+      "learning_rate": 5.954014165952295e-06,
+      "loss": 0.486,
+      "step": 10158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.142418883573149,
+      "learning_rate": 5.953366391186574e-06,
+      "loss": 0.8946,
+      "step": 10159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.877323796714911,
+      "learning_rate": 5.952718606929908e-06,
+      "loss": 0.3496,
+      "step": 10160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.64376437276972,
+      "learning_rate": 5.9520708131958595e-06,
+      "loss": 0.3944,
+      "step": 10161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7889734684336456,
+      "learning_rate": 5.95142300999799e-06,
+      "loss": 0.4547,
+      "step": 10162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8450227616742207,
+      "learning_rate": 5.950775197349858e-06,
+      "loss": 0.5068,
+      "step": 10163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7956306676852618,
+      "learning_rate": 5.950127375265029e-06,
+      "loss": 0.4405,
+      "step": 10164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8457512365876811,
+      "learning_rate": 5.949479543757063e-06,
+      "loss": 0.5027,
+      "step": 10165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5818313872221297,
+      "learning_rate": 5.948831702839521e-06,
+      "loss": 0.4443,
+      "step": 10166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3939705158199467,
+      "learning_rate": 5.948183852525967e-06,
+      "loss": 0.6156,
+      "step": 10167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5463936337406248,
+      "learning_rate": 5.947535992829962e-06,
+      "loss": 0.4434,
+      "step": 10168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.548640173847266,
+      "learning_rate": 5.94688812376507e-06,
+      "loss": 0.5058,
+      "step": 10169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4906182653308173,
+      "learning_rate": 5.946240245344851e-06,
+      "loss": 0.412,
+      "step": 10170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.585768941216899,
+      "learning_rate": 5.945592357582871e-06,
+      "loss": 0.5092,
+      "step": 10171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7279952768307218,
+      "learning_rate": 5.944944460492691e-06,
+      "loss": 0.5943,
+      "step": 10172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9534383357199943,
+      "learning_rate": 5.9442965540878746e-06,
+      "loss": 0.5734,
+      "step": 10173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4606302767909534,
+      "learning_rate": 5.943648638381987e-06,
+      "loss": 0.4273,
+      "step": 10174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7450472230532905,
+      "learning_rate": 5.94300071338859e-06,
+      "loss": 0.4368,
+      "step": 10175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5278381895533486,
+      "learning_rate": 5.9423527791212485e-06,
+      "loss": 0.499,
+      "step": 10176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7175832821811028,
+      "learning_rate": 5.941704835593527e-06,
+      "loss": 0.5609,
+      "step": 10177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.271473467550092,
+      "learning_rate": 5.941056882818988e-06,
+      "loss": 0.501,
+      "step": 10178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7128489531188873,
+      "learning_rate": 5.9404089208111985e-06,
+      "loss": 0.5136,
+      "step": 10179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.717780688829515,
+      "learning_rate": 5.939760949583719e-06,
+      "loss": 0.4847,
+      "step": 10180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8397469299588232,
+      "learning_rate": 5.939112969150118e-06,
+      "loss": 0.5353,
+      "step": 10181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0601905874133957,
+      "learning_rate": 5.938464979523961e-06,
+      "loss": 0.6411,
+      "step": 10182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.984195442964574,
+      "learning_rate": 5.937816980718813e-06,
+      "loss": 0.526,
+      "step": 10183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1218348280223984,
+      "learning_rate": 5.937168972748236e-06,
+      "loss": 0.5167,
+      "step": 10184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5545217962088171,
+      "learning_rate": 5.9365209556257995e-06,
+      "loss": 0.3664,
+      "step": 10185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6400972987876266,
+      "learning_rate": 5.9358729293650675e-06,
+      "loss": 0.4638,
+      "step": 10186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3335627271217585,
+      "learning_rate": 5.935224893979606e-06,
+      "loss": 0.3525,
+      "step": 10187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2684886200091126,
+      "learning_rate": 5.9345768494829825e-06,
+      "loss": 0.4993,
+      "step": 10188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.048191815730685,
+      "learning_rate": 5.933928795888762e-06,
+      "loss": 0.4891,
+      "step": 10189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2199732331416344,
+      "learning_rate": 5.933280733210512e-06,
+      "loss": 0.7154,
+      "step": 10190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5734417011302224,
+      "learning_rate": 5.932632661461799e-06,
+      "loss": 0.5541,
+      "step": 10191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4272157178096234,
+      "learning_rate": 5.93198458065619e-06,
+      "loss": 0.4398,
+      "step": 10192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.369791439551386,
+      "learning_rate": 5.931336490807253e-06,
+      "loss": 0.3007,
+      "step": 10193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.238676301353342,
+      "learning_rate": 5.930688391928553e-06,
+      "loss": 0.7141,
+      "step": 10194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.640193806452688,
+      "learning_rate": 5.93004028403366e-06,
+      "loss": 0.4651,
+      "step": 10195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5598428566116596,
+      "learning_rate": 5.929392167136141e-06,
+      "loss": 0.4797,
+      "step": 10196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6256530684324806,
+      "learning_rate": 5.928744041249561e-06,
+      "loss": 0.4645,
+      "step": 10197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6926932285053296,
+      "learning_rate": 5.928095906387494e-06,
+      "loss": 0.5098,
+      "step": 10198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.008727292635866,
+      "learning_rate": 5.927447762563503e-06,
+      "loss": 0.6106,
+      "step": 10199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5778038410248716,
+      "learning_rate": 5.92679960979116e-06,
+      "loss": 0.4964,
+      "step": 10200
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5267772078514099,
+      "eval_runtime": 67.9626,
+      "eval_samples_per_second": 50.175,
+      "eval_steps_per_second": 1.574,
+      "step": 10200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6439685826324475,
+      "learning_rate": 5.926151448084032e-06,
+      "loss": 0.4369,
+      "step": 10201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7542644183325893,
+      "learning_rate": 5.925503277455688e-06,
+      "loss": 0.4502,
+      "step": 10202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7792744653325567,
+      "learning_rate": 5.924855097919696e-06,
+      "loss": 0.4341,
+      "step": 10203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6682701517446172,
+      "learning_rate": 5.924206909489629e-06,
+      "loss": 0.5508,
+      "step": 10204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9794255070464175,
+      "learning_rate": 5.9235587121790506e-06,
+      "loss": 0.4206,
+      "step": 10205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8424263224753237,
+      "learning_rate": 5.922910506001536e-06,
+      "loss": 0.6206,
+      "step": 10206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6537629093212818,
+      "learning_rate": 5.922262290970652e-06,
+      "loss": 0.443,
+      "step": 10207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9762910072595723,
+      "learning_rate": 5.921614067099969e-06,
+      "loss": 0.6747,
+      "step": 10208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1620950671551578,
+      "learning_rate": 5.920965834403057e-06,
+      "loss": 0.676,
+      "step": 10209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.725869749881544,
+      "learning_rate": 5.920317592893488e-06,
+      "loss": 0.5102,
+      "step": 10210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1809453144019244,
+      "learning_rate": 5.91966934258483e-06,
+      "loss": 0.6491,
+      "step": 10211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.207900311124109,
+      "learning_rate": 5.919021083490656e-06,
+      "loss": 0.5474,
+      "step": 10212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8498800712650025,
+      "learning_rate": 5.918372815624535e-06,
+      "loss": 0.5903,
+      "step": 10213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3476363888403027,
+      "learning_rate": 5.917724539000039e-06,
+      "loss": 0.4836,
+      "step": 10214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9649141402702728,
+      "learning_rate": 5.9170762536307395e-06,
+      "loss": 0.6552,
+      "step": 10215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3366526882656806,
+      "learning_rate": 5.916427959530209e-06,
+      "loss": 0.4422,
+      "step": 10216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0182089421888785,
+      "learning_rate": 5.915779656712016e-06,
+      "loss": 0.6098,
+      "step": 10217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6613427580522522,
+      "learning_rate": 5.915131345189735e-06,
+      "loss": 0.4353,
+      "step": 10218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7813197829043281,
+      "learning_rate": 5.914483024976937e-06,
+      "loss": 0.586,
+      "step": 10219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6843085311869848,
+      "learning_rate": 5.913834696087193e-06,
+      "loss": 0.4235,
+      "step": 10220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6224081521144513,
+      "learning_rate": 5.913186358534079e-06,
+      "loss": 0.4724,
+      "step": 10221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.697683373398943,
+      "learning_rate": 5.912538012331164e-06,
+      "loss": 0.4569,
+      "step": 10222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1457480906872717,
+      "learning_rate": 5.911889657492021e-06,
+      "loss": 0.6358,
+      "step": 10223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8135717805956948,
+      "learning_rate": 5.911241294030226e-06,
+      "loss": 0.6442,
+      "step": 10224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1013004400026962,
+      "learning_rate": 5.910592921959349e-06,
+      "loss": 0.6221,
+      "step": 10225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8544180505609196,
+      "learning_rate": 5.909944541292965e-06,
+      "loss": 0.5421,
+      "step": 10226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4122729906373417,
+      "learning_rate": 5.9092961520446445e-06,
+      "loss": 0.4093,
+      "step": 10227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8916069208721793,
+      "learning_rate": 5.908647754227964e-06,
+      "loss": 0.509,
+      "step": 10228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.496501350755577,
+      "learning_rate": 5.907999347856497e-06,
+      "loss": 0.3745,
+      "step": 10229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.613357602840765,
+      "learning_rate": 5.907350932943817e-06,
+      "loss": 0.5179,
+      "step": 10230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.119649592360791,
+      "learning_rate": 5.906702509503498e-06,
+      "loss": 0.496,
+      "step": 10231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.241113120525062,
+      "learning_rate": 5.906054077549115e-06,
+      "loss": 0.6533,
+      "step": 10232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.932655177578485,
+      "learning_rate": 5.905405637094243e-06,
+      "loss": 0.5361,
+      "step": 10233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1259360801273313,
+      "learning_rate": 5.904757188152453e-06,
+      "loss": 0.5985,
+      "step": 10234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5308723293847828,
+      "learning_rate": 5.904108730737325e-06,
+      "loss": 0.4474,
+      "step": 10235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8722365568022339,
+      "learning_rate": 5.903460264862433e-06,
+      "loss": 0.5082,
+      "step": 10236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0400899792200575,
+      "learning_rate": 5.9028117905413475e-06,
+      "loss": 0.5617,
+      "step": 10237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3937364627965412,
+      "learning_rate": 5.902163307787651e-06,
+      "loss": 0.4096,
+      "step": 10238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3392310948023383,
+      "learning_rate": 5.901514816614912e-06,
+      "loss": 0.3917,
+      "step": 10239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4309960923094525,
+      "learning_rate": 5.900866317036712e-06,
+      "loss": 0.4604,
+      "step": 10240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5521178861119007,
+      "learning_rate": 5.900217809066624e-06,
+      "loss": 0.4699,
+      "step": 10241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9046974413453734,
+      "learning_rate": 5.899569292718225e-06,
+      "loss": 0.6211,
+      "step": 10242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4491366953614293,
+      "learning_rate": 5.898920768005089e-06,
+      "loss": 0.4211,
+      "step": 10243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9617811490133568,
+      "learning_rate": 5.898272234940796e-06,
+      "loss": 0.6601,
+      "step": 10244
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6570250010308962,
+      "learning_rate": 5.897623693538922e-06,
+      "loss": 0.4072,
+      "step": 10245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5131770140149783,
+      "learning_rate": 5.896975143813043e-06,
+      "loss": 0.4012,
+      "step": 10246
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6327372303703802,
+      "learning_rate": 5.896326585776733e-06,
+      "loss": 0.4463,
+      "step": 10247
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0611075360600197,
+      "learning_rate": 5.895678019443576e-06,
+      "loss": 0.47,
+      "step": 10248
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6629739625948405,
+      "learning_rate": 5.895029444827143e-06,
+      "loss": 0.3364,
+      "step": 10249
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.464925332630788,
+      "learning_rate": 5.894380861941014e-06,
+      "loss": 0.7262,
+      "step": 10250
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5259653329849243,
+      "eval_runtime": 67.9568,
+      "eval_samples_per_second": 50.179,
+      "eval_steps_per_second": 1.575,
+      "step": 10250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8400970144945734,
+      "learning_rate": 5.8937322707987685e-06,
+      "loss": 0.5816,
+      "step": 10251
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1047664235921504,
+      "learning_rate": 5.893083671413981e-06,
+      "loss": 0.5609,
+      "step": 10252
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5669200726268366,
+      "learning_rate": 5.892435063800231e-06,
+      "loss": 0.4346,
+      "step": 10253
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9801644120528608,
+      "learning_rate": 5.891786447971096e-06,
+      "loss": 0.5716,
+      "step": 10254
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.860968980054391,
+      "learning_rate": 5.8911378239401565e-06,
+      "loss": 0.7225,
+      "step": 10255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9390801007567438,
+      "learning_rate": 5.890489191720989e-06,
+      "loss": 0.588,
+      "step": 10256
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5863015555755389,
+      "learning_rate": 5.889840551327174e-06,
+      "loss": 0.4768,
+      "step": 10257
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6527769997156527,
+      "learning_rate": 5.8891919027722865e-06,
+      "loss": 0.5125,
+      "step": 10258
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4674090588019517,
+      "learning_rate": 5.888543246069911e-06,
+      "loss": 0.412,
+      "step": 10259
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.774617833795182,
+      "learning_rate": 5.887894581233621e-06,
+      "loss": 0.5917,
+      "step": 10260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8645791277905384,
+      "learning_rate": 5.887245908277002e-06,
+      "loss": 0.776,
+      "step": 10261
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8692448126710535,
+      "learning_rate": 5.886597227213628e-06,
+      "loss": 0.5726,
+      "step": 10262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0524450123055518,
+      "learning_rate": 5.885948538057083e-06,
+      "loss": 0.5763,
+      "step": 10263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5199772686934445,
+      "learning_rate": 5.885299840820945e-06,
+      "loss": 0.4398,
+      "step": 10264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.710762661329433,
+      "learning_rate": 5.884651135518794e-06,
+      "loss": 0.4755,
+      "step": 10265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.730437335738376,
+      "learning_rate": 5.884002422164209e-06,
+      "loss": 0.5199,
+      "step": 10266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7489328409036011,
+      "learning_rate": 5.883353700770773e-06,
+      "loss": 0.4561,
+      "step": 10267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7907196935273635,
+      "learning_rate": 5.882704971352063e-06,
+      "loss": 0.4439,
+      "step": 10268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8010346332573766,
+      "learning_rate": 5.882056233921666e-06,
+      "loss": 0.4781,
+      "step": 10269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8026186635776313,
+      "learning_rate": 5.881407488493158e-06,
+      "loss": 0.5646,
+      "step": 10270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4659738543716623,
+      "learning_rate": 5.880758735080119e-06,
+      "loss": 0.6175,
+      "step": 10271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8823575830272752,
+      "learning_rate": 5.880109973696134e-06,
+      "loss": 0.6605,
+      "step": 10272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8689132069799488,
+      "learning_rate": 5.879461204354783e-06,
+      "loss": 0.4852,
+      "step": 10273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7323265430533688,
+      "learning_rate": 5.878812427069647e-06,
+      "loss": 0.5169,
+      "step": 10274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6918388850233492,
+      "learning_rate": 5.8781636418543086e-06,
+      "loss": 0.5315,
+      "step": 10275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9631822832333914,
+      "learning_rate": 5.877514848722348e-06,
+      "loss": 0.5884,
+      "step": 10276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5805615196428608,
+      "learning_rate": 5.876866047687349e-06,
+      "loss": 0.4714,
+      "step": 10277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.505029719035521,
+      "learning_rate": 5.876217238762894e-06,
+      "loss": 0.34,
+      "step": 10278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8209322739204092,
+      "learning_rate": 5.875568421962563e-06,
+      "loss": 0.5867,
+      "step": 10279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1600368274208352,
+      "learning_rate": 5.874919597299943e-06,
+      "loss": 0.8214,
+      "step": 10280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4931716878982866,
+      "learning_rate": 5.874270764788613e-06,
+      "loss": 0.3751,
+      "step": 10281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6140655710188174,
+      "learning_rate": 5.873621924442156e-06,
+      "loss": 0.4393,
+      "step": 10282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4811535767211936,
+      "learning_rate": 5.872973076274158e-06,
+      "loss": 0.4609,
+      "step": 10283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8154592236133344,
+      "learning_rate": 5.872324220298198e-06,
+      "loss": 0.5471,
+      "step": 10284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6282368497494095,
+      "learning_rate": 5.8716753565278625e-06,
+      "loss": 0.5069,
+      "step": 10285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.716794409283064,
+      "learning_rate": 5.871026484976736e-06,
+      "loss": 0.543,
+      "step": 10286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8001947874938016,
+      "learning_rate": 5.870377605658398e-06,
+      "loss": 0.4822,
+      "step": 10287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4382262602652736,
+      "learning_rate": 5.869728718586436e-06,
+      "loss": 0.3634,
+      "step": 10288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.518474047583997,
+      "learning_rate": 5.8690798237744325e-06,
+      "loss": 0.427,
+      "step": 10289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5547825515126974,
+      "learning_rate": 5.8684309212359735e-06,
+      "loss": 0.4746,
+      "step": 10290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.752456708387909,
+      "learning_rate": 5.867782010984642e-06,
+      "loss": 0.6068,
+      "step": 10291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7161160628408805,
+      "learning_rate": 5.867133093034021e-06,
+      "loss": 0.4439,
+      "step": 10292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.043086340800578,
+      "learning_rate": 5.866484167397697e-06,
+      "loss": 0.6417,
+      "step": 10293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.484955581967785,
+      "learning_rate": 5.8658352340892565e-06,
+      "loss": 0.5334,
+      "step": 10294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5570893843585372,
+      "learning_rate": 5.8651862931222806e-06,
+      "loss": 0.4462,
+      "step": 10295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9708041140208525,
+      "learning_rate": 5.864537344510357e-06,
+      "loss": 0.5977,
+      "step": 10296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8804998901919554,
+      "learning_rate": 5.863888388267069e-06,
+      "loss": 0.5384,
+      "step": 10297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.060423749985293,
+      "learning_rate": 5.863239424406006e-06,
+      "loss": 0.7291,
+      "step": 10298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.009485874731229,
+      "learning_rate": 5.8625904529407505e-06,
+      "loss": 0.5614,
+      "step": 10299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.454585478468612,
+      "learning_rate": 5.861941473884889e-06,
+      "loss": 0.3511,
+      "step": 10300
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5270065665245056,
+      "eval_runtime": 67.9329,
+      "eval_samples_per_second": 50.197,
+      "eval_steps_per_second": 1.575,
+      "step": 10300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0714394224947115,
+      "learning_rate": 5.861292487252008e-06,
+      "loss": 0.6425,
+      "step": 10301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6965798857608105,
+      "learning_rate": 5.860643493055692e-06,
+      "loss": 0.482,
+      "step": 10302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.404651080533692,
+      "learning_rate": 5.859994491309529e-06,
+      "loss": 0.3734,
+      "step": 10303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.922212422163308,
+      "learning_rate": 5.859345482027105e-06,
+      "loss": 0.5958,
+      "step": 10304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0523903201583003,
+      "learning_rate": 5.858696465222007e-06,
+      "loss": 0.5021,
+      "step": 10305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.543363319502022,
+      "learning_rate": 5.85804744090782e-06,
+      "loss": 0.7383,
+      "step": 10306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6545421101638322,
+      "learning_rate": 5.857398409098134e-06,
+      "loss": 0.5609,
+      "step": 10307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7179403569757767,
+      "learning_rate": 5.856749369806533e-06,
+      "loss": 0.4299,
+      "step": 10308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7071550518049114,
+      "learning_rate": 5.8561003230466064e-06,
+      "loss": 0.504,
+      "step": 10309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8279908928996353,
+      "learning_rate": 5.855451268831941e-06,
+      "loss": 0.4427,
+      "step": 10310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7444747200953128,
+      "learning_rate": 5.854802207176121e-06,
+      "loss": 0.4825,
+      "step": 10311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.486591817270283,
+      "learning_rate": 5.854153138092741e-06,
+      "loss": 0.381,
+      "step": 10312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0195468216117494,
+      "learning_rate": 5.853504061595383e-06,
+      "loss": 0.5328,
+      "step": 10313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5746622565722583,
+      "learning_rate": 5.852854977697638e-06,
+      "loss": 0.4417,
+      "step": 10314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.607675473290184,
+      "learning_rate": 5.852205886413093e-06,
+      "loss": 0.6141,
+      "step": 10315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8742453282904292,
+      "learning_rate": 5.851556787755336e-06,
+      "loss": 0.555,
+      "step": 10316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4289880048560335,
+      "learning_rate": 5.8509076817379575e-06,
+      "loss": 0.3617,
+      "step": 10317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5588760843294776,
+      "learning_rate": 5.850258568374543e-06,
+      "loss": 0.3991,
+      "step": 10318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7597144198358883,
+      "learning_rate": 5.849609447678683e-06,
+      "loss": 0.5575,
+      "step": 10319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5390389444586428,
+      "learning_rate": 5.848960319663966e-06,
+      "loss": 0.375,
+      "step": 10320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6568940441339994,
+      "learning_rate": 5.848311184343982e-06,
+      "loss": 0.547,
+      "step": 10321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7229427322549675,
+      "learning_rate": 5.847662041732322e-06,
+      "loss": 0.5232,
+      "step": 10322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5107206705485676,
+      "learning_rate": 5.84701289184257e-06,
+      "loss": 0.4446,
+      "step": 10323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.676606179941347,
+      "learning_rate": 5.84636373468832e-06,
+      "loss": 0.5371,
+      "step": 10324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8106849886656808,
+      "learning_rate": 5.84571457028316e-06,
+      "loss": 0.5533,
+      "step": 10325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4139438193330331,
+      "learning_rate": 5.845065398640679e-06,
+      "loss": 0.3594,
+      "step": 10326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7415247657393187,
+      "learning_rate": 5.84441621977447e-06,
+      "loss": 0.4229,
+      "step": 10327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6149831043806362,
+      "learning_rate": 5.84376703369812e-06,
+      "loss": 0.4233,
+      "step": 10328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5411090921938693,
+      "learning_rate": 5.843117840425221e-06,
+      "loss": 0.414,
+      "step": 10329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.351416548064641,
+      "learning_rate": 5.842468639969364e-06,
+      "loss": 0.4602,
+      "step": 10330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6740113882953618,
+      "learning_rate": 5.841819432344138e-06,
+      "loss": 0.468,
+      "step": 10331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8199157765904312,
+      "learning_rate": 5.8411702175631345e-06,
+      "loss": 0.5662,
+      "step": 10332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.187020676658914,
+      "learning_rate": 5.8405209956399455e-06,
+      "loss": 0.6772,
+      "step": 10333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.874937973189193,
+      "learning_rate": 5.839871766588161e-06,
+      "loss": 0.4674,
+      "step": 10334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.027387830717575,
+      "learning_rate": 5.83922253042137e-06,
+      "loss": 0.6308,
+      "step": 10335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6449063570176137,
+      "learning_rate": 5.838573287153169e-06,
+      "loss": 0.4378,
+      "step": 10336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8938223654267163,
+      "learning_rate": 5.8379240367971435e-06,
+      "loss": 0.4847,
+      "step": 10337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4794061548432613,
+      "learning_rate": 5.837274779366891e-06,
+      "loss": 0.4066,
+      "step": 10338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8891179405880207,
+      "learning_rate": 5.836625514875999e-06,
+      "loss": 0.3982,
+      "step": 10339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5677534511823716,
+      "learning_rate": 5.835976243338059e-06,
+      "loss": 0.5247,
+      "step": 10340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7673645261870061,
+      "learning_rate": 5.835326964766669e-06,
+      "loss": 0.477,
+      "step": 10341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4649522030866289,
+      "learning_rate": 5.834677679175415e-06,
+      "loss": 0.4951,
+      "step": 10342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1618505312786387,
+      "learning_rate": 5.834028386577891e-06,
+      "loss": 0.7041,
+      "step": 10343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5250326689766585,
+      "learning_rate": 5.833379086987692e-06,
+      "loss": 0.3984,
+      "step": 10344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.614864611188482,
+      "learning_rate": 5.832729780418407e-06,
+      "loss": 0.41,
+      "step": 10345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3590307993311006,
+      "learning_rate": 5.832080466883633e-06,
+      "loss": 0.404,
+      "step": 10346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1658083802706316,
+      "learning_rate": 5.8314311463969585e-06,
+      "loss": 0.776,
+      "step": 10347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5899816529052946,
+      "learning_rate": 5.83078181897198e-06,
+      "loss": 0.3743,
+      "step": 10348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7873392447048393,
+      "learning_rate": 5.83013248462229e-06,
+      "loss": 0.413,
+      "step": 10349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7118416120193538,
+      "learning_rate": 5.82948314336148e-06,
+      "loss": 0.5163,
+      "step": 10350
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.5275726914405823,
+      "eval_runtime": 67.9673,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 10350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2092440404568476,
+      "learning_rate": 5.828833795203146e-06,
+      "loss": 0.628,
+      "step": 10351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6952295496005256,
+      "learning_rate": 5.82818444016088e-06,
+      "loss": 0.5472,
+      "step": 10352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9177483926282473,
+      "learning_rate": 5.8275350782482765e-06,
+      "loss": 0.5909,
+      "step": 10353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1013114926843666,
+      "learning_rate": 5.826885709478931e-06,
+      "loss": 0.5623,
+      "step": 10354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7946311619283115,
+      "learning_rate": 5.8262363338664366e-06,
+      "loss": 0.502,
+      "step": 10355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5208234090920043,
+      "learning_rate": 5.825586951424384e-06,
+      "loss": 0.4236,
+      "step": 10356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.935430143750103,
+      "learning_rate": 5.824937562166374e-06,
+      "loss": 0.5467,
+      "step": 10357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5780662458915975,
+      "learning_rate": 5.824288166105997e-06,
+      "loss": 0.4578,
+      "step": 10358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8291156795224213,
+      "learning_rate": 5.823638763256849e-06,
+      "loss": 0.489,
+      "step": 10359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9720327059122584,
+      "learning_rate": 5.822989353632523e-06,
+      "loss": 0.4535,
+      "step": 10360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5777820948080037,
+      "learning_rate": 5.822339937246616e-06,
+      "loss": 0.5576,
+      "step": 10361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8224652554547998,
+      "learning_rate": 5.821690514112725e-06,
+      "loss": 0.6402,
+      "step": 10362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8688478867762428,
+      "learning_rate": 5.821041084244441e-06,
+      "loss": 0.5468,
+      "step": 10363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2836134873824154,
+      "learning_rate": 5.8203916476553605e-06,
+      "loss": 0.4194,
+      "step": 10364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8417950150879774,
+      "learning_rate": 5.81974220435908e-06,
+      "loss": 0.673,
+      "step": 10365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.603544368529165,
+      "learning_rate": 5.819092754369196e-06,
+      "loss": 0.4409,
+      "step": 10366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.051906770752734,
+      "learning_rate": 5.8184432976993035e-06,
+      "loss": 0.6193,
+      "step": 10367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9205413520954266,
+      "learning_rate": 5.817793834362998e-06,
+      "loss": 0.53,
+      "step": 10368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8963461308933036,
+      "learning_rate": 5.8171443643738754e-06,
+      "loss": 0.4502,
+      "step": 10369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.704278010806306,
+      "learning_rate": 5.816494887745534e-06,
+      "loss": 0.4115,
+      "step": 10370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6239445086032283,
+      "learning_rate": 5.815845404491568e-06,
+      "loss": 0.4012,
+      "step": 10371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6280725498292004,
+      "learning_rate": 5.815195914625573e-06,
+      "loss": 0.5288,
+      "step": 10372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5427416288017042,
+      "learning_rate": 5.81454641816115e-06,
+      "loss": 0.4763,
+      "step": 10373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9411338727004397,
+      "learning_rate": 5.813896915111891e-06,
+      "loss": 0.4805,
+      "step": 10374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3811765816206083,
+      "learning_rate": 5.8132474054913964e-06,
+      "loss": 0.2906,
+      "step": 10375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4918494297811948,
+      "learning_rate": 5.812597889313262e-06,
+      "loss": 0.4662,
+      "step": 10376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2725332078010743,
+      "learning_rate": 5.811948366591082e-06,
+      "loss": 0.6353,
+      "step": 10377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8759211311204145,
+      "learning_rate": 5.81129883733846e-06,
+      "loss": 0.535,
+      "step": 10378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8726837600202455,
+      "learning_rate": 5.810649301568989e-06,
+      "loss": 0.5283,
+      "step": 10379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5784043726489623,
+      "learning_rate": 5.809999759296268e-06,
+      "loss": 0.4152,
+      "step": 10380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5704185115122875,
+      "learning_rate": 5.809350210533895e-06,
+      "loss": 0.4367,
+      "step": 10381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8324883311415934,
+      "learning_rate": 5.808700655295466e-06,
+      "loss": 0.4589,
+      "step": 10382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0929681404805596,
+      "learning_rate": 5.808051093594582e-06,
+      "loss": 0.5957,
+      "step": 10383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6459576965144027,
+      "learning_rate": 5.8074015254448405e-06,
+      "loss": 0.4428,
+      "step": 10384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6405685123823328,
+      "learning_rate": 5.806751950859838e-06,
+      "loss": 0.5151,
+      "step": 10385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9707431540367155,
+      "learning_rate": 5.806102369853174e-06,
+      "loss": 0.5132,
+      "step": 10386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.672912262705483,
+      "learning_rate": 5.805452782438447e-06,
+      "loss": 0.4215,
+      "step": 10387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.164802605155134,
+      "learning_rate": 5.804803188629254e-06,
+      "loss": 0.5334,
+      "step": 10388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.118392096675612,
+      "learning_rate": 5.804153588439198e-06,
+      "loss": 0.4888,
+      "step": 10389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2085711702764956,
+      "learning_rate": 5.803503981881874e-06,
+      "loss": 0.5565,
+      "step": 10390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6634242603987444,
+      "learning_rate": 5.802854368970884e-06,
+      "loss": 0.5009,
+      "step": 10391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0154184925035685,
+      "learning_rate": 5.802204749719825e-06,
+      "loss": 0.7297,
+      "step": 10392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8184008064400163,
+      "learning_rate": 5.801555124142297e-06,
+      "loss": 0.4657,
+      "step": 10393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9436057657116153,
+      "learning_rate": 5.800905492251901e-06,
+      "loss": 0.5042,
+      "step": 10394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.017038423906311,
+      "learning_rate": 5.800255854062234e-06,
+      "loss": 0.4509,
+      "step": 10395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5350764323168475,
+      "learning_rate": 5.799606209586898e-06,
+      "loss": 0.4331,
+      "step": 10396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.175048507225362,
+      "learning_rate": 5.798956558839493e-06,
+      "loss": 0.5335,
+      "step": 10397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.282908414564054,
+      "learning_rate": 5.7983069018336155e-06,
+      "loss": 0.4846,
+      "step": 10398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5460914174861684,
+      "learning_rate": 5.797657238582871e-06,
+      "loss": 0.459,
+      "step": 10399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.774363153404618,
+      "learning_rate": 5.797007569100857e-06,
+      "loss": 0.4439,
+      "step": 10400
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.5270489454269409,
+      "eval_runtime": 67.9651,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 10400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9561430219489915,
+      "learning_rate": 5.7963578934011715e-06,
+      "loss": 0.5846,
+      "step": 10401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9191424419879264,
+      "learning_rate": 5.795708211497418e-06,
+      "loss": 0.5561,
+      "step": 10402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8684246071047452,
+      "learning_rate": 5.795058523403199e-06,
+      "loss": 0.6545,
+      "step": 10403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2364526186499427,
+      "learning_rate": 5.794408829132111e-06,
+      "loss": 0.6415,
+      "step": 10404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.501895857468752,
+      "learning_rate": 5.793759128697758e-06,
+      "loss": 0.4356,
+      "step": 10405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.527924088460477,
+      "learning_rate": 5.79310942211374e-06,
+      "loss": 0.3492,
+      "step": 10406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2417418079931046,
+      "learning_rate": 5.792459709393659e-06,
+      "loss": 0.6215,
+      "step": 10407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4086579262935657,
+      "learning_rate": 5.7918099905511146e-06,
+      "loss": 0.4409,
+      "step": 10408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.659874346597104,
+      "learning_rate": 5.791160265599709e-06,
+      "loss": 0.4933,
+      "step": 10409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8547037403167685,
+      "learning_rate": 5.790510534553046e-06,
+      "loss": 0.5427,
+      "step": 10410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.003566695834514,
+      "learning_rate": 5.789860797424724e-06,
+      "loss": 0.5498,
+      "step": 10411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.139070858256411,
+      "learning_rate": 5.789211054228345e-06,
+      "loss": 0.6873,
+      "step": 10412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2386058939633764,
+      "learning_rate": 5.788561304977514e-06,
+      "loss": 0.6644,
+      "step": 10413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.172323083169788,
+      "learning_rate": 5.787911549685831e-06,
+      "loss": 0.6304,
+      "step": 10414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.946761061602331,
+      "learning_rate": 5.7872617883668995e-06,
+      "loss": 0.5135,
+      "step": 10415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.427472887583763,
+      "learning_rate": 5.78661202103432e-06,
+      "loss": 0.4789,
+      "step": 10416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3544287406833402,
+      "learning_rate": 5.785962247701695e-06,
+      "loss": 0.4188,
+      "step": 10417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1168414384833585,
+      "learning_rate": 5.785312468382629e-06,
+      "loss": 0.4505,
+      "step": 10418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7038149714493207,
+      "learning_rate": 5.7846626830907224e-06,
+      "loss": 0.5047,
+      "step": 10419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6363239662881182,
+      "learning_rate": 5.784012891839581e-06,
+      "loss": 0.4701,
+      "step": 10420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.809775687357784,
+      "learning_rate": 5.783363094642805e-06,
+      "loss": 0.66,
+      "step": 10421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4699664628830478,
+      "learning_rate": 5.782713291514e-06,
+      "loss": 0.4578,
+      "step": 10422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9296159551256178,
+      "learning_rate": 5.782063482466766e-06,
+      "loss": 0.5203,
+      "step": 10423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7282333618446648,
+      "learning_rate": 5.781413667514708e-06,
+      "loss": 0.6214,
+      "step": 10424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7915749131586536,
+      "learning_rate": 5.780763846671432e-06,
+      "loss": 0.4405,
+      "step": 10425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6414937602856654,
+      "learning_rate": 5.780114019950538e-06,
+      "loss": 0.4945,
+      "step": 10426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7561710637004202,
+      "learning_rate": 5.779464187365631e-06,
+      "loss": 0.4294,
+      "step": 10427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0581054341438283,
+      "learning_rate": 5.778814348930315e-06,
+      "loss": 0.7398,
+      "step": 10428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9578127561718652,
+      "learning_rate": 5.778164504658193e-06,
+      "loss": 0.5678,
+      "step": 10429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.228157753661959,
+      "learning_rate": 5.7775146545628705e-06,
+      "loss": 0.4778,
+      "step": 10430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1349802806336906,
+      "learning_rate": 5.776864798657951e-06,
+      "loss": 0.6815,
+      "step": 10431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.000820471426764,
+      "learning_rate": 5.776214936957039e-06,
+      "loss": 0.534,
+      "step": 10432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0906466597596096,
+      "learning_rate": 5.775565069473738e-06,
+      "loss": 0.708,
+      "step": 10433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6592222068478446,
+      "learning_rate": 5.774915196221655e-06,
+      "loss": 0.5483,
+      "step": 10434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6628565886948157,
+      "learning_rate": 5.774265317214391e-06,
+      "loss": 0.4156,
+      "step": 10435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4221961469377251,
+      "learning_rate": 5.773615432465555e-06,
+      "loss": 0.5454,
+      "step": 10436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.942626042429986,
+      "learning_rate": 5.77296554198875e-06,
+      "loss": 0.4971,
+      "step": 10437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7886580807551555,
+      "learning_rate": 5.7723156457975794e-06,
+      "loss": 0.4421,
+      "step": 10438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4653152299816323,
+      "learning_rate": 5.771665743905651e-06,
+      "loss": 0.4737,
+      "step": 10439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.550457299223102,
+      "learning_rate": 5.771015836326569e-06,
+      "loss": 0.3137,
+      "step": 10440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8637786453883396,
+      "learning_rate": 5.7703659230739375e-06,
+      "loss": 0.4633,
+      "step": 10441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.005661321365242,
+      "learning_rate": 5.769716004161366e-06,
+      "loss": 0.5772,
+      "step": 10442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6826713073405795,
+      "learning_rate": 5.7690660796024546e-06,
+      "loss": 0.4635,
+      "step": 10443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.64370690188876,
+      "learning_rate": 5.768416149410812e-06,
+      "loss": 0.3772,
+      "step": 10444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4774010873980985,
+      "learning_rate": 5.767766213600046e-06,
+      "loss": 0.3354,
+      "step": 10445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6874161189517027,
+      "learning_rate": 5.7671162721837595e-06,
+      "loss": 0.5221,
+      "step": 10446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.827219992976272,
+      "learning_rate": 5.766466325175561e-06,
+      "loss": 0.6122,
+      "step": 10447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7006167840387398,
+      "learning_rate": 5.7658163725890526e-06,
+      "loss": 0.4955,
+      "step": 10448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.055163758197648,
+      "learning_rate": 5.765166414437846e-06,
+      "loss": 0.5431,
+      "step": 10449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.272914856392133,
+      "learning_rate": 5.7645164507355446e-06,
+      "loss": 0.6226,
+      "step": 10450
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.5272147059440613,
+      "eval_runtime": 68.0028,
+      "eval_samples_per_second": 50.145,
+      "eval_steps_per_second": 1.573,
+      "step": 10450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7119217094919612,
+      "learning_rate": 5.763866481495755e-06,
+      "loss": 0.5172,
+      "step": 10451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.824364947046066,
+      "learning_rate": 5.763216506732084e-06,
+      "loss": 0.5737,
+      "step": 10452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8032039394288624,
+      "learning_rate": 5.762566526458141e-06,
+      "loss": 0.5654,
+      "step": 10453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7371943496602174,
+      "learning_rate": 5.76191654068753e-06,
+      "loss": 0.4758,
+      "step": 10454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9634211143440927,
+      "learning_rate": 5.7612665494338595e-06,
+      "loss": 0.624,
+      "step": 10455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.915970889251386,
+      "learning_rate": 5.760616552710735e-06,
+      "loss": 0.4858,
+      "step": 10456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.778903769001442,
+      "learning_rate": 5.759966550531764e-06,
+      "loss": 0.4194,
+      "step": 10457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5117019266635472,
+      "learning_rate": 5.759316542910557e-06,
+      "loss": 0.4498,
+      "step": 10458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9071571086636814,
+      "learning_rate": 5.758666529860718e-06,
+      "loss": 0.4716,
+      "step": 10459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9967938636210834,
+      "learning_rate": 5.758016511395855e-06,
+      "loss": 0.5869,
+      "step": 10460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1985565628658033,
+      "learning_rate": 5.757366487529579e-06,
+      "loss": 0.5383,
+      "step": 10461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5979185922774684,
+      "learning_rate": 5.756716458275493e-06,
+      "loss": 0.4134,
+      "step": 10462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9393959233613471,
+      "learning_rate": 5.756066423647208e-06,
+      "loss": 0.4976,
+      "step": 10463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7639238466346485,
+      "learning_rate": 5.7554163836583324e-06,
+      "loss": 0.4463,
+      "step": 10464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5129248713836423,
+      "learning_rate": 5.754766338322473e-06,
+      "loss": 0.4414,
+      "step": 10465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2832647924689062,
+      "learning_rate": 5.754116287653239e-06,
+      "loss": 0.3656,
+      "step": 10466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.109183037338092,
+      "learning_rate": 5.753466231664237e-06,
+      "loss": 0.5388,
+      "step": 10467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9079834460399658,
+      "learning_rate": 5.752816170369079e-06,
+      "loss": 0.499,
+      "step": 10468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7779539597473406,
+      "learning_rate": 5.75216610378137e-06,
+      "loss": 0.6962,
+      "step": 10469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4561262112474886,
+      "learning_rate": 5.75151603191472e-06,
+      "loss": 0.3553,
+      "step": 10470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7163175451752422,
+      "learning_rate": 5.750865954782739e-06,
+      "loss": 0.4473,
+      "step": 10471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.798130431947246,
+      "learning_rate": 5.750215872399034e-06,
+      "loss": 0.5533,
+      "step": 10472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7262404384229437,
+      "learning_rate": 5.749565784777216e-06,
+      "loss": 0.4091,
+      "step": 10473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.74696789664408,
+      "learning_rate": 5.748915691930894e-06,
+      "loss": 0.557,
+      "step": 10474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9306702409447838,
+      "learning_rate": 5.7482655938736735e-06,
+      "loss": 0.5892,
+      "step": 10475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1493683855934176,
+      "learning_rate": 5.747615490619171e-06,
+      "loss": 0.6762,
+      "step": 10476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2678942088397016,
+      "learning_rate": 5.74696538218099e-06,
+      "loss": 0.6157,
+      "step": 10477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.916421716899892,
+      "learning_rate": 5.746315268572742e-06,
+      "loss": 0.6581,
+      "step": 10478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7412197656112272,
+      "learning_rate": 5.745665149808038e-06,
+      "loss": 0.5065,
+      "step": 10479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5895334301402537,
+      "learning_rate": 5.745015025900485e-06,
+      "loss": 0.3936,
+      "step": 10480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.65922485148935,
+      "learning_rate": 5.744364896863696e-06,
+      "loss": 0.5043,
+      "step": 10481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5473282376298652,
+      "learning_rate": 5.743714762711278e-06,
+      "loss": 0.4462,
+      "step": 10482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7415422892515608,
+      "learning_rate": 5.743064623456843e-06,
+      "loss": 0.6252,
+      "step": 10483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7300958454996507,
+      "learning_rate": 5.742414479114002e-06,
+      "loss": 0.5493,
+      "step": 10484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9197220102275512,
+      "learning_rate": 5.7417643296963664e-06,
+      "loss": 0.4566,
+      "step": 10485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3549832913944277,
+      "learning_rate": 5.741114175217541e-06,
+      "loss": 0.3393,
+      "step": 10486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0404615878233603,
+      "learning_rate": 5.740464015691143e-06,
+      "loss": 0.4674,
+      "step": 10487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.208204894307987,
+      "learning_rate": 5.739813851130778e-06,
+      "loss": 0.4537,
+      "step": 10488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.719511539457364,
+      "learning_rate": 5.7391636815500606e-06,
+      "loss": 0.4284,
+      "step": 10489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9577123381334998,
+      "learning_rate": 5.7385135069625994e-06,
+      "loss": 0.5636,
+      "step": 10490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8527617756593449,
+      "learning_rate": 5.737863327382006e-06,
+      "loss": 0.5276,
+      "step": 10491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8993650490787208,
+      "learning_rate": 5.7372131428218925e-06,
+      "loss": 0.6249,
+      "step": 10492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5273326089731298,
+      "learning_rate": 5.736562953295869e-06,
+      "loss": 0.4452,
+      "step": 10493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6592851044592063,
+      "learning_rate": 5.7359127588175455e-06,
+      "loss": 0.5428,
+      "step": 10494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.660939630045155,
+      "learning_rate": 5.735262559400536e-06,
+      "loss": 0.4196,
+      "step": 10495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.661336636806073,
+      "learning_rate": 5.734612355058451e-06,
+      "loss": 0.6341,
+      "step": 10496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.274438781115821,
+      "learning_rate": 5.733962145804902e-06,
+      "loss": 0.6364,
+      "step": 10497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.216655764827946,
+      "learning_rate": 5.733311931653501e-06,
+      "loss": 0.6076,
+      "step": 10498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1101961622374086,
+      "learning_rate": 5.732661712617858e-06,
+      "loss": 0.61,
+      "step": 10499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.007849274021302,
+      "learning_rate": 5.732011488711588e-06,
+      "loss": 0.464,
+      "step": 10500
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.5274158120155334,
+      "eval_runtime": 67.9904,
+      "eval_samples_per_second": 50.154,
+      "eval_steps_per_second": 1.574,
+      "step": 10500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.789785970965716,
+      "learning_rate": 5.7313612599483026e-06,
+      "loss": 0.4725,
+      "step": 10501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1203303989231699,
+      "learning_rate": 5.730711026341611e-06,
+      "loss": 0.2767,
+      "step": 10502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4110205293327878,
+      "learning_rate": 5.730060787905128e-06,
+      "loss": 0.4209,
+      "step": 10503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5953984230567622,
+      "learning_rate": 5.729410544652465e-06,
+      "loss": 0.473,
+      "step": 10504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.658516978884679,
+      "learning_rate": 5.728760296597236e-06,
+      "loss": 0.4912,
+      "step": 10505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4611244093701092,
+      "learning_rate": 5.728110043753052e-06,
+      "loss": 0.5207,
+      "step": 10506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7839628411282473,
+      "learning_rate": 5.727459786133525e-06,
+      "loss": 0.5036,
+      "step": 10507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5967573428276218,
+      "learning_rate": 5.726809523752269e-06,
+      "loss": 0.5656,
+      "step": 10508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.638648646036394,
+      "learning_rate": 5.726159256622898e-06,
+      "loss": 0.4323,
+      "step": 10509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8636632562586573,
+      "learning_rate": 5.72550898475902e-06,
+      "loss": 0.5818,
+      "step": 10510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.599703102404121,
+      "learning_rate": 5.724858708174255e-06,
+      "loss": 0.4844,
+      "step": 10511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7100515957291786,
+      "learning_rate": 5.724208426882209e-06,
+      "loss": 0.4253,
+      "step": 10512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.807703006545834,
+      "learning_rate": 5.723558140896502e-06,
+      "loss": 0.6386,
+      "step": 10513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.801066519825143,
+      "learning_rate": 5.722907850230743e-06,
+      "loss": 0.5094,
+      "step": 10514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5659932135490506,
+      "learning_rate": 5.722257554898547e-06,
+      "loss": 0.5362,
+      "step": 10515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4880020671380654,
+      "learning_rate": 5.721607254913527e-06,
+      "loss": 0.3739,
+      "step": 10516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9248992257499824,
+      "learning_rate": 5.7209569502892955e-06,
+      "loss": 0.4588,
+      "step": 10517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6398175384731988,
+      "learning_rate": 5.720306641039469e-06,
+      "loss": 0.6024,
+      "step": 10518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.773295221090765,
+      "learning_rate": 5.71965632717766e-06,
+      "loss": 0.5488,
+      "step": 10519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4071517361893957,
+      "learning_rate": 5.719006008717481e-06,
+      "loss": 0.3951,
+      "step": 10520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9184981648733839,
+      "learning_rate": 5.718355685672549e-06,
+      "loss": 0.5846,
+      "step": 10521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3898586788197083,
+      "learning_rate": 5.7177053580564756e-06,
+      "loss": 0.4011,
+      "step": 10522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.023135144401915,
+      "learning_rate": 5.717055025882875e-06,
+      "loss": 0.5688,
+      "step": 10523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.88230639178296,
+      "learning_rate": 5.716404689165362e-06,
+      "loss": 0.5975,
+      "step": 10524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9411144187402574,
+      "learning_rate": 5.715754347917553e-06,
+      "loss": 0.5349,
+      "step": 10525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6852866549005505,
+      "learning_rate": 5.715104002153058e-06,
+      "loss": 0.4845,
+      "step": 10526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6711992698693836,
+      "learning_rate": 5.7144536518854965e-06,
+      "loss": 0.5744,
+      "step": 10527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1891843151132893,
+      "learning_rate": 5.7138032971284795e-06,
+      "loss": 0.5503,
+      "step": 10528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0253886635879104,
+      "learning_rate": 5.713152937895625e-06,
+      "loss": 0.6058,
+      "step": 10529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7794331809334385,
+      "learning_rate": 5.712502574200545e-06,
+      "loss": 0.5351,
+      "step": 10530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9154358004142849,
+      "learning_rate": 5.711852206056856e-06,
+      "loss": 0.597,
+      "step": 10531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3689147219692828,
+      "learning_rate": 5.711201833478173e-06,
+      "loss": 0.7079,
+      "step": 10532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6322186180181915,
+      "learning_rate": 5.71055145647811e-06,
+      "loss": 0.3936,
+      "step": 10533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7443366955014343,
+      "learning_rate": 5.709901075070283e-06,
+      "loss": 0.5335,
+      "step": 10534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5751383010985276,
+      "learning_rate": 5.7092506892683086e-06,
+      "loss": 0.3836,
+      "step": 10535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9875426591323873,
+      "learning_rate": 5.7086002990857996e-06,
+      "loss": 0.4658,
+      "step": 10536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7568474577983895,
+      "learning_rate": 5.707949904536374e-06,
+      "loss": 0.5739,
+      "step": 10537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6778679784951174,
+      "learning_rate": 5.707299505633646e-06,
+      "loss": 0.4171,
+      "step": 10538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.703454599854954,
+      "learning_rate": 5.706649102391231e-06,
+      "loss": 0.6472,
+      "step": 10539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9437015073247226,
+      "learning_rate": 5.705998694822747e-06,
+      "loss": 0.5117,
+      "step": 10540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6974163297710936,
+      "learning_rate": 5.705348282941805e-06,
+      "loss": 0.5105,
+      "step": 10541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5618456908508513,
+      "learning_rate": 5.7046978667620276e-06,
+      "loss": 0.5585,
+      "step": 10542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4368243426156078,
+      "learning_rate": 5.704047446297026e-06,
+      "loss": 0.3669,
+      "step": 10543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6582623681303237,
+      "learning_rate": 5.703397021560419e-06,
+      "loss": 0.364,
+      "step": 10544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5526018920985094,
+      "learning_rate": 5.70274659256582e-06,
+      "loss": 0.468,
+      "step": 10545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5611295471746616,
+      "learning_rate": 5.702096159326847e-06,
+      "loss": 0.4606,
+      "step": 10546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6334741327607232,
+      "learning_rate": 5.701445721857117e-06,
+      "loss": 0.5129,
+      "step": 10547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7171823107003226,
+      "learning_rate": 5.700795280170247e-06,
+      "loss": 0.5968,
+      "step": 10548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5440638872241097,
+      "learning_rate": 5.700144834279852e-06,
+      "loss": 0.3747,
+      "step": 10549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1253826895229193,
+      "learning_rate": 5.699494384199548e-06,
+      "loss": 0.5399,
+      "step": 10550
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.5266873836517334,
+      "eval_runtime": 67.9497,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 10550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9600012066739667,
+      "learning_rate": 5.698843929942954e-06,
+      "loss": 0.6088,
+      "step": 10551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4585390182250437,
+      "learning_rate": 5.6981934715236845e-06,
+      "loss": 0.6161,
+      "step": 10552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6741587158480722,
+      "learning_rate": 5.697543008955359e-06,
+      "loss": 0.6275,
+      "step": 10553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8377261502644429,
+      "learning_rate": 5.696892542251595e-06,
+      "loss": 0.418,
+      "step": 10554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8521463335103028,
+      "learning_rate": 5.696242071426005e-06,
+      "loss": 0.486,
+      "step": 10555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8054562172347066,
+      "learning_rate": 5.6955915964922114e-06,
+      "loss": 0.4766,
+      "step": 10556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.516354821970619,
+      "learning_rate": 5.694941117463827e-06,
+      "loss": 0.5199,
+      "step": 10557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9361813580353389,
+      "learning_rate": 5.6942906343544735e-06,
+      "loss": 0.6554,
+      "step": 10558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6438184302301009,
+      "learning_rate": 5.693640147177764e-06,
+      "loss": 0.6351,
+      "step": 10559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.691850990449328,
+      "learning_rate": 5.692989655947321e-06,
+      "loss": 0.4934,
+      "step": 10560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.964180153318229,
+      "learning_rate": 5.692339160676758e-06,
+      "loss": 0.453,
+      "step": 10561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9396766803054368,
+      "learning_rate": 5.691688661379694e-06,
+      "loss": 0.6511,
+      "step": 10562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7168983672042837,
+      "learning_rate": 5.691038158069747e-06,
+      "loss": 0.526,
+      "step": 10563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2311460499944678,
+      "learning_rate": 5.690387650760535e-06,
+      "loss": 0.3366,
+      "step": 10564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9517077731642598,
+      "learning_rate": 5.689737139465675e-06,
+      "loss": 0.5166,
+      "step": 10565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.626088999685853,
+      "learning_rate": 5.689086624198786e-06,
+      "loss": 0.4858,
+      "step": 10566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6940242492811526,
+      "learning_rate": 5.688436104973488e-06,
+      "loss": 0.4248,
+      "step": 10567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.163994252198955,
+      "learning_rate": 5.687785581803395e-06,
+      "loss": 0.6311,
+      "step": 10568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.67667762092717,
+      "learning_rate": 5.687135054702129e-06,
+      "loss": 0.5015,
+      "step": 10569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5984377661899474,
+      "learning_rate": 5.686484523683305e-06,
+      "loss": 0.4899,
+      "step": 10570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.962019678754312,
+      "learning_rate": 5.685833988760545e-06,
+      "loss": 0.6557,
+      "step": 10571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1154565852377436,
+      "learning_rate": 5.6851834499474654e-06,
+      "loss": 0.696,
+      "step": 10572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6543853049460413,
+      "learning_rate": 5.684532907257684e-06,
+      "loss": 0.5406,
+      "step": 10573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.521972410120544,
+      "learning_rate": 5.683882360704822e-06,
+      "loss": 0.4422,
+      "step": 10574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7525035976125574,
+      "learning_rate": 5.683231810302497e-06,
+      "loss": 0.5154,
+      "step": 10575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7222026066588951,
+      "learning_rate": 5.6825812560643264e-06,
+      "loss": 0.3889,
+      "step": 10576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4529162555258726,
+      "learning_rate": 5.6819306980039336e-06,
+      "loss": 0.4187,
+      "step": 10577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0595288190230674,
+      "learning_rate": 5.6812801361349324e-06,
+      "loss": 0.526,
+      "step": 10578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.878587689553998,
+      "learning_rate": 5.680629570470945e-06,
+      "loss": 0.6327,
+      "step": 10579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7917140538658936,
+      "learning_rate": 5.67997900102559e-06,
+      "loss": 0.5465,
+      "step": 10580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5525363485218608,
+      "learning_rate": 5.679328427812484e-06,
+      "loss": 0.4375,
+      "step": 10581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.700150509874295,
+      "learning_rate": 5.678677850845252e-06,
+      "loss": 0.3453,
+      "step": 10582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.777467996373546,
+      "learning_rate": 5.678027270137509e-06,
+      "loss": 0.4675,
+      "step": 10583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6495442672819394,
+      "learning_rate": 5.677376685702875e-06,
+      "loss": 0.4162,
+      "step": 10584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.951097258626604,
+      "learning_rate": 5.676726097554971e-06,
+      "loss": 0.4216,
+      "step": 10585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1231179522762997,
+      "learning_rate": 5.676075505707415e-06,
+      "loss": 0.5608,
+      "step": 10586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.733345164914469,
+      "learning_rate": 5.675424910173828e-06,
+      "loss": 0.4634,
+      "step": 10587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7595703615801197,
+      "learning_rate": 5.674774310967831e-06,
+      "loss": 0.5189,
+      "step": 10588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.626659463003894,
+      "learning_rate": 5.674123708103041e-06,
+      "loss": 0.4004,
+      "step": 10589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6682373270973574,
+      "learning_rate": 5.673473101593079e-06,
+      "loss": 0.4922,
+      "step": 10590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9007744802927795,
+      "learning_rate": 5.672822491451567e-06,
+      "loss": 0.4587,
+      "step": 10591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4808591994186637,
+      "learning_rate": 5.672171877692121e-06,
+      "loss": 0.4337,
+      "step": 10592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6407168628102953,
+      "learning_rate": 5.671521260328365e-06,
+      "loss": 0.5805,
+      "step": 10593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6877522117254256,
+      "learning_rate": 5.670870639373918e-06,
+      "loss": 0.3922,
+      "step": 10594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7340237999768107,
+      "learning_rate": 5.670220014842402e-06,
+      "loss": 0.5347,
+      "step": 10595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.791996919643356,
+      "learning_rate": 5.6695693867474325e-06,
+      "loss": 0.5043,
+      "step": 10596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0630100595991863,
+      "learning_rate": 5.6689187551026346e-06,
+      "loss": 0.4859,
+      "step": 10597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.630416848569348,
+      "learning_rate": 5.668268119921628e-06,
+      "loss": 0.4492,
+      "step": 10598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.474977415145372,
+      "learning_rate": 5.667617481218032e-06,
+      "loss": 0.3606,
+      "step": 10599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.73958203858857,
+      "learning_rate": 5.666966839005467e-06,
+      "loss": 0.676,
+      "step": 10600
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.5263962149620056,
+      "eval_runtime": 67.9239,
+      "eval_samples_per_second": 50.203,
+      "eval_steps_per_second": 1.575,
+      "step": 10600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5483079062798595,
+      "learning_rate": 5.666316193297558e-06,
+      "loss": 0.576,
+      "step": 10601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8525833088849688,
+      "learning_rate": 5.6656655441079215e-06,
+      "loss": 0.5167,
+      "step": 10602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8085303486747781,
+      "learning_rate": 5.6650148914501776e-06,
+      "loss": 0.4361,
+      "step": 10603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7932451131579261,
+      "learning_rate": 5.664364235337952e-06,
+      "loss": 0.5018,
+      "step": 10604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9421829658840937,
+      "learning_rate": 5.663713575784863e-06,
+      "loss": 0.5561,
+      "step": 10605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9898539175829693,
+      "learning_rate": 5.663062912804531e-06,
+      "loss": 0.3419,
+      "step": 10606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5951392553060748,
+      "learning_rate": 5.6624122464105805e-06,
+      "loss": 0.3654,
+      "step": 10607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8569911262027425,
+      "learning_rate": 5.661761576616628e-06,
+      "loss": 0.4966,
+      "step": 10608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9204451378536187,
+      "learning_rate": 5.6611109034362975e-06,
+      "loss": 0.5234,
+      "step": 10609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.68225933903588,
+      "learning_rate": 5.660460226883211e-06,
+      "loss": 0.4789,
+      "step": 10610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5247513739704697,
+      "learning_rate": 5.659809546970989e-06,
+      "loss": 0.4511,
+      "step": 10611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.727106209771502,
+      "learning_rate": 5.659158863713255e-06,
+      "loss": 0.5295,
+      "step": 10612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7782753984152635,
+      "learning_rate": 5.658508177123626e-06,
+      "loss": 0.6384,
+      "step": 10613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.596923162717744,
+      "learning_rate": 5.657857487215728e-06,
+      "loss": 0.3979,
+      "step": 10614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3276919496311885,
+      "learning_rate": 5.657206794003184e-06,
+      "loss": 0.388,
+      "step": 10615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8364081611718028,
+      "learning_rate": 5.65655609749961e-06,
+      "loss": 0.4657,
+      "step": 10616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8021912724245739,
+      "learning_rate": 5.655905397718633e-06,
+      "loss": 0.4954,
+      "step": 10617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.828254297290738,
+      "learning_rate": 5.655254694673873e-06,
+      "loss": 0.6957,
+      "step": 10618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6501046838659603,
+      "learning_rate": 5.654603988378953e-06,
+      "loss": 0.5576,
+      "step": 10619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.00063371082482,
+      "learning_rate": 5.6539532788474934e-06,
+      "loss": 0.5012,
+      "step": 10620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6184519077699822,
+      "learning_rate": 5.653302566093118e-06,
+      "loss": 0.4286,
+      "step": 10621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.558169557918933,
+      "learning_rate": 5.652651850129449e-06,
+      "loss": 0.5101,
+      "step": 10622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8384669593741063,
+      "learning_rate": 5.652001130970108e-06,
+      "loss": 0.5406,
+      "step": 10623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7759800630604903,
+      "learning_rate": 5.6513504086287165e-06,
+      "loss": 0.5506,
+      "step": 10624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9799625196034982,
+      "learning_rate": 5.6506996831188995e-06,
+      "loss": 0.4613,
+      "step": 10625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.905087810267142,
+      "learning_rate": 5.650048954454278e-06,
+      "loss": 0.5866,
+      "step": 10626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5507387394841663,
+      "learning_rate": 5.649398222648477e-06,
+      "loss": 0.4785,
+      "step": 10627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8815970546776786,
+      "learning_rate": 5.648747487715115e-06,
+      "loss": 0.4691,
+      "step": 10628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4371874169105971,
+      "learning_rate": 5.648096749667815e-06,
+      "loss": 0.3469,
+      "step": 10629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7752741600533917,
+      "learning_rate": 5.647446008520204e-06,
+      "loss": 0.5357,
+      "step": 10630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8375747579826922,
+      "learning_rate": 5.646795264285903e-06,
+      "loss": 0.4747,
+      "step": 10631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3430387744156376,
+      "learning_rate": 5.6461445169785324e-06,
+      "loss": 0.2953,
+      "step": 10632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8719351239171587,
+      "learning_rate": 5.645493766611718e-06,
+      "loss": 0.5884,
+      "step": 10633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3911853198123476,
+      "learning_rate": 5.6448430131990805e-06,
+      "loss": 0.5018,
+      "step": 10634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8939226134410734,
+      "learning_rate": 5.644192256754246e-06,
+      "loss": 0.5601,
+      "step": 10635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.72517558130324,
+      "learning_rate": 5.6435414972908376e-06,
+      "loss": 0.4012,
+      "step": 10636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.026549355397393,
+      "learning_rate": 5.6428907348224735e-06,
+      "loss": 0.6148,
+      "step": 10637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0150883745136055,
+      "learning_rate": 5.642239969362784e-06,
+      "loss": 0.5464,
+      "step": 10638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5525703339930128,
+      "learning_rate": 5.641589200925387e-06,
+      "loss": 0.4024,
+      "step": 10639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.686178258357059,
+      "learning_rate": 5.6409384295239085e-06,
+      "loss": 0.4857,
+      "step": 10640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.22127802331074,
+      "learning_rate": 5.640287655171972e-06,
+      "loss": 0.5812,
+      "step": 10641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7623566017327552,
+      "learning_rate": 5.639636877883199e-06,
+      "loss": 0.4939,
+      "step": 10642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5475241408451839,
+      "learning_rate": 5.6389860976712164e-06,
+      "loss": 0.4378,
+      "step": 10643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1353888188048553,
+      "learning_rate": 5.6383353145496455e-06,
+      "loss": 0.6904,
+      "step": 10644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8268801974909974,
+      "learning_rate": 5.637684528532111e-06,
+      "loss": 0.6229,
+      "step": 10645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.709527950781419,
+      "learning_rate": 5.637033739632237e-06,
+      "loss": 0.5415,
+      "step": 10646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7160965148612763,
+      "learning_rate": 5.636382947863644e-06,
+      "loss": 0.57,
+      "step": 10647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4702605498623589,
+      "learning_rate": 5.635732153239962e-06,
+      "loss": 0.4946,
+      "step": 10648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7972256344065407,
+      "learning_rate": 5.6350813557748105e-06,
+      "loss": 0.619,
+      "step": 10649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3176518684255083,
+      "learning_rate": 5.634430555481813e-06,
+      "loss": 0.3581,
+      "step": 10650
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.5268576145172119,
+      "eval_runtime": 67.9421,
+      "eval_samples_per_second": 50.19,
+      "eval_steps_per_second": 1.575,
+      "step": 10650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.408693464708101,
+      "learning_rate": 5.633779752374598e-06,
+      "loss": 0.3946,
+      "step": 10651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.708002108757489,
+      "learning_rate": 5.633128946466787e-06,
+      "loss": 0.457,
+      "step": 10652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9363904023649083,
+      "learning_rate": 5.632478137772002e-06,
+      "loss": 0.6177,
+      "step": 10653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.747082735924818,
+      "learning_rate": 5.631827326303872e-06,
+      "loss": 0.459,
+      "step": 10654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0649010057196735,
+      "learning_rate": 5.631176512076018e-06,
+      "loss": 0.5693,
+      "step": 10655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6958647037212304,
+      "learning_rate": 5.6305256951020635e-06,
+      "loss": 0.396,
+      "step": 10656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8712307482461552,
+      "learning_rate": 5.629874875395637e-06,
+      "loss": 0.703,
+      "step": 10657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6477273026962067,
+      "learning_rate": 5.629224052970358e-06,
+      "loss": 0.4756,
+      "step": 10658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.321129328256289,
+      "learning_rate": 5.628573227839855e-06,
+      "loss": 0.4981,
+      "step": 10659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2130830507781614,
+      "learning_rate": 5.627922400017752e-06,
+      "loss": 0.5374,
+      "step": 10660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.214155155461161,
+      "learning_rate": 5.627271569517672e-06,
+      "loss": 0.6167,
+      "step": 10661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7660608288005777,
+      "learning_rate": 5.626620736353242e-06,
+      "loss": 0.5082,
+      "step": 10662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8614808851658464,
+      "learning_rate": 5.625969900538084e-06,
+      "loss": 0.5308,
+      "step": 10663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8572488333817958,
+      "learning_rate": 5.625319062085826e-06,
+      "loss": 0.5853,
+      "step": 10664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.770816978349283,
+      "learning_rate": 5.624668221010091e-06,
+      "loss": 0.455,
+      "step": 10665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9589098470937825,
+      "learning_rate": 5.624017377324502e-06,
+      "loss": 0.6933,
+      "step": 10666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.786520853908827,
+      "learning_rate": 5.623366531042689e-06,
+      "loss": 0.4774,
+      "step": 10667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9602265340493095,
+      "learning_rate": 5.622715682178274e-06,
+      "loss": 0.4426,
+      "step": 10668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7588076843788951,
+      "learning_rate": 5.62206483074488e-06,
+      "loss": 0.4557,
+      "step": 10669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5898940024661121,
+      "learning_rate": 5.621413976756137e-06,
+      "loss": 0.5119,
+      "step": 10670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.917415708750859,
+      "learning_rate": 5.620763120225666e-06,
+      "loss": 0.5807,
+      "step": 10671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8758123390271768,
+      "learning_rate": 5.620112261167094e-06,
+      "loss": 0.5971,
+      "step": 10672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8974866531004333,
+      "learning_rate": 5.619461399594048e-06,
+      "loss": 0.5311,
+      "step": 10673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0247380304808993,
+      "learning_rate": 5.61881053552015e-06,
+      "loss": 0.6457,
+      "step": 10674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9111136922597538,
+      "learning_rate": 5.61815966895903e-06,
+      "loss": 0.6056,
+      "step": 10675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.611690820835847,
+      "learning_rate": 5.6175087999243106e-06,
+      "loss": 0.467,
+      "step": 10676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5331651535490627,
+      "learning_rate": 5.6168579284296145e-06,
+      "loss": 0.3751,
+      "step": 10677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6538433498715523,
+      "learning_rate": 5.616207054488573e-06,
+      "loss": 0.5949,
+      "step": 10678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5746144145185672,
+      "learning_rate": 5.615556178114807e-06,
+      "loss": 0.5364,
+      "step": 10679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6270189245528417,
+      "learning_rate": 5.6149052993219454e-06,
+      "loss": 0.424,
+      "step": 10680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9117923749129573,
+      "learning_rate": 5.614254418123613e-06,
+      "loss": 0.5341,
+      "step": 10681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8292278506837933,
+      "learning_rate": 5.613603534533434e-06,
+      "loss": 0.4546,
+      "step": 10682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1426103578047218,
+      "learning_rate": 5.612952648565037e-06,
+      "loss": 0.5262,
+      "step": 10683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9352745137357794,
+      "learning_rate": 5.612301760232045e-06,
+      "loss": 0.4638,
+      "step": 10684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4815114364853343,
+      "learning_rate": 5.611650869548087e-06,
+      "loss": 0.3309,
+      "step": 10685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9968427554607648,
+      "learning_rate": 5.610999976526787e-06,
+      "loss": 0.6142,
+      "step": 10686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1750918131881773,
+      "learning_rate": 5.61034908118177e-06,
+      "loss": 0.5602,
+      "step": 10687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7845680581022363,
+      "learning_rate": 5.609698183526664e-06,
+      "loss": 0.6198,
+      "step": 10688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.783806284017601,
+      "learning_rate": 5.609047283575095e-06,
+      "loss": 0.5233,
+      "step": 10689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8292579907048825,
+      "learning_rate": 5.608396381340689e-06,
+      "loss": 0.6011,
+      "step": 10690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6091124483739403,
+      "learning_rate": 5.6077454768370734e-06,
+      "loss": 0.4853,
+      "step": 10691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6672090366576755,
+      "learning_rate": 5.607094570077871e-06,
+      "loss": 0.5203,
+      "step": 10692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8607358020905653,
+      "learning_rate": 5.606443661076712e-06,
+      "loss": 0.5184,
+      "step": 10693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4937535521734355,
+      "learning_rate": 5.605792749847219e-06,
+      "loss": 0.4045,
+      "step": 10694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0638528331680996,
+      "learning_rate": 5.605141836403022e-06,
+      "loss": 0.5763,
+      "step": 10695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.676513209053796,
+      "learning_rate": 5.604490920757746e-06,
+      "loss": 0.4413,
+      "step": 10696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.355083693744341,
+      "learning_rate": 5.603840002925016e-06,
+      "loss": 0.363,
+      "step": 10697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.759976049121111,
+      "learning_rate": 5.603189082918461e-06,
+      "loss": 0.4227,
+      "step": 10698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4966501663837042,
+      "learning_rate": 5.602538160751707e-06,
+      "loss": 0.4156,
+      "step": 10699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5500174577529287,
+      "learning_rate": 5.6018872364383814e-06,
+      "loss": 0.5093,
+      "step": 10700
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.5268427729606628,
+      "eval_runtime": 67.9928,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 10700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4188842013463094,
+      "learning_rate": 5.601236309992108e-06,
+      "loss": 0.4466,
+      "step": 10701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6895477486973938,
+      "learning_rate": 5.600585381426515e-06,
+      "loss": 0.4411,
+      "step": 10702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5488602741331396,
+      "learning_rate": 5.59993445075523e-06,
+      "loss": 0.503,
+      "step": 10703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3374063387327333,
+      "learning_rate": 5.59928351799188e-06,
+      "loss": 0.3362,
+      "step": 10704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8428684355099507,
+      "learning_rate": 5.598632583150091e-06,
+      "loss": 0.4934,
+      "step": 10705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5529295327621795,
+      "learning_rate": 5.59798164624349e-06,
+      "loss": 0.4137,
+      "step": 10706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.111206648417762,
+      "learning_rate": 5.597330707285705e-06,
+      "loss": 0.416,
+      "step": 10707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5842019481778575,
+      "learning_rate": 5.596679766290359e-06,
+      "loss": 0.3786,
+      "step": 10708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9595546787455012,
+      "learning_rate": 5.596028823271086e-06,
+      "loss": 0.6363,
+      "step": 10709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6708567749371275,
+      "learning_rate": 5.595377878241508e-06,
+      "loss": 0.4526,
+      "step": 10710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7562535666781687,
+      "learning_rate": 5.594726931215252e-06,
+      "loss": 0.4162,
+      "step": 10711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5320408297290131,
+      "learning_rate": 5.5940759822059474e-06,
+      "loss": 0.4282,
+      "step": 10712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1785472739661293,
+      "learning_rate": 5.593425031227222e-06,
+      "loss": 0.7042,
+      "step": 10713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5853639376532038,
+      "learning_rate": 5.5927740782927e-06,
+      "loss": 0.4644,
+      "step": 10714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5603721805568183,
+      "learning_rate": 5.59212312341601e-06,
+      "loss": 0.4968,
+      "step": 10715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7845389322973186,
+      "learning_rate": 5.59147216661078e-06,
+      "loss": 0.4839,
+      "step": 10716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6561017617628657,
+      "learning_rate": 5.590821207890638e-06,
+      "loss": 0.4418,
+      "step": 10717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6916174552533905,
+      "learning_rate": 5.5901702472692095e-06,
+      "loss": 0.4536,
+      "step": 10718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.693232483299747,
+      "learning_rate": 5.589519284760122e-06,
+      "loss": 0.4279,
+      "step": 10719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6038612400018166,
+      "learning_rate": 5.588868320377005e-06,
+      "loss": 0.4464,
+      "step": 10720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.189712701489358,
+      "learning_rate": 5.588217354133485e-06,
+      "loss": 0.6233,
+      "step": 10721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6051921088822443,
+      "learning_rate": 5.58756638604319e-06,
+      "loss": 0.6099,
+      "step": 10722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9327276244275529,
+      "learning_rate": 5.586915416119745e-06,
+      "loss": 0.5208,
+      "step": 10723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8302808602429899,
+      "learning_rate": 5.586264444376782e-06,
+      "loss": 0.5152,
+      "step": 10724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6844336977130387,
+      "learning_rate": 5.585613470827925e-06,
+      "loss": 0.4861,
+      "step": 10725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8918518888588944,
+      "learning_rate": 5.584962495486804e-06,
+      "loss": 0.5191,
+      "step": 10726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4153291412148044,
+      "learning_rate": 5.5843115183670445e-06,
+      "loss": 0.3225,
+      "step": 10727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.696664699511182,
+      "learning_rate": 5.583660539482278e-06,
+      "loss": 0.4765,
+      "step": 10728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5147270838941995,
+      "learning_rate": 5.58300955884613e-06,
+      "loss": 0.4152,
+      "step": 10729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0180045539139235,
+      "learning_rate": 5.582358576472226e-06,
+      "loss": 0.6809,
+      "step": 10730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1736603244398247,
+      "learning_rate": 5.581707592374198e-06,
+      "loss": 0.7085,
+      "step": 10731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.000255980785831,
+      "learning_rate": 5.581056606565672e-06,
+      "loss": 0.5938,
+      "step": 10732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6742781553582804,
+      "learning_rate": 5.580405619060276e-06,
+      "loss": 0.4684,
+      "step": 10733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3714034302357119,
+      "learning_rate": 5.57975462987164e-06,
+      "loss": 0.3449,
+      "step": 10734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7432380238232295,
+      "learning_rate": 5.579103639013388e-06,
+      "loss": 0.5513,
+      "step": 10735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1301392942082966,
+      "learning_rate": 5.578452646499152e-06,
+      "loss": 0.6813,
+      "step": 10736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.661131353237045,
+      "learning_rate": 5.577801652342557e-06,
+      "loss": 0.5085,
+      "step": 10737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6212423582149806,
+      "learning_rate": 5.5771506565572345e-06,
+      "loss": 0.4268,
+      "step": 10738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6816238650548967,
+      "learning_rate": 5.576499659156811e-06,
+      "loss": 0.4684,
+      "step": 10739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6007724223215267,
+      "learning_rate": 5.575848660154912e-06,
+      "loss": 0.447,
+      "step": 10740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.598737646999654,
+      "learning_rate": 5.5751976595651705e-06,
+      "loss": 0.4682,
+      "step": 10741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6364199980644132,
+      "learning_rate": 5.574546657401214e-06,
+      "loss": 0.5761,
+      "step": 10742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9379850785594404,
+      "learning_rate": 5.573895653676666e-06,
+      "loss": 0.5025,
+      "step": 10743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.000725612472553,
+      "learning_rate": 5.573244648405162e-06,
+      "loss": 0.6228,
+      "step": 10744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2671196497885564,
+      "learning_rate": 5.5725936416003235e-06,
+      "loss": 0.5277,
+      "step": 10745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6477919719382745,
+      "learning_rate": 5.5719426332757845e-06,
+      "loss": 0.4906,
+      "step": 10746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.463929376427179,
+      "learning_rate": 5.571291623445171e-06,
+      "loss": 0.5085,
+      "step": 10747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7340598539011385,
+      "learning_rate": 5.57064061212211e-06,
+      "loss": 0.4849,
+      "step": 10748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.635188998175394,
+      "learning_rate": 5.569989599320234e-06,
+      "loss": 0.4951,
+      "step": 10749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7283041341190015,
+      "learning_rate": 5.569338585053168e-06,
+      "loss": 0.576,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5266112685203552,
+      "eval_runtime": 67.9384,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6977562610009884,
+      "learning_rate": 5.56868756933454e-06,
+      "loss": 0.4931,
+      "step": 10751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8323948696630956,
+      "learning_rate": 5.568036552177983e-06,
+      "loss": 0.4921,
+      "step": 10752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9027010730464062,
+      "learning_rate": 5.567385533597123e-06,
+      "loss": 0.5044,
+      "step": 10753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6104224524674648,
+      "learning_rate": 5.566734513605588e-06,
+      "loss": 0.5107,
+      "step": 10754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8391398592442152,
+      "learning_rate": 5.566083492217006e-06,
+      "loss": 0.5701,
+      "step": 10755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7545229414686163,
+      "learning_rate": 5.565432469445009e-06,
+      "loss": 0.5678,
+      "step": 10756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9030787005542618,
+      "learning_rate": 5.564781445303223e-06,
+      "loss": 0.4809,
+      "step": 10757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7118475621081748,
+      "learning_rate": 5.564130419805278e-06,
+      "loss": 0.4522,
+      "step": 10758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.081169244189924,
+      "learning_rate": 5.563479392964803e-06,
+      "loss": 0.4645,
+      "step": 10759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2291885022258393,
+      "learning_rate": 5.562828364795427e-06,
+      "loss": 0.4749,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8815895564478835,
+      "learning_rate": 5.562177335310776e-06,
+      "loss": 0.4869,
+      "step": 10761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.928573931616993,
+      "learning_rate": 5.5615263045244825e-06,
+      "loss": 0.5628,
+      "step": 10762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8760494363997835,
+      "learning_rate": 5.560875272450175e-06,
+      "loss": 0.3687,
+      "step": 10763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3249328996537955,
+      "learning_rate": 5.560224239101478e-06,
+      "loss": 0.5899,
+      "step": 10764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0415999375261804,
+      "learning_rate": 5.559573204492027e-06,
+      "loss": 0.5701,
+      "step": 10765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4485144559644414,
+      "learning_rate": 5.558922168635448e-06,
+      "loss": 0.4219,
+      "step": 10766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9242712596334772,
+      "learning_rate": 5.558271131545368e-06,
+      "loss": 0.5328,
+      "step": 10767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7422383156769634,
+      "learning_rate": 5.55762009323542e-06,
+      "loss": 0.4493,
+      "step": 10768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5263262258487436,
+      "learning_rate": 5.556969053719229e-06,
+      "loss": 0.4638,
+      "step": 10769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7792207231180721,
+      "learning_rate": 5.556318013010429e-06,
+      "loss": 0.612,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1109382991761745,
+      "learning_rate": 5.555666971122644e-06,
+      "loss": 0.5042,
+      "step": 10771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5651403968295272,
+      "learning_rate": 5.555015928069506e-06,
+      "loss": 0.7468,
+      "step": 10772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7008436082163036,
+      "learning_rate": 5.554364883864645e-06,
+      "loss": 0.3592,
+      "step": 10773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.556141043428205,
+      "learning_rate": 5.553713838521688e-06,
+      "loss": 0.2974,
+      "step": 10774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9288180418817928,
+      "learning_rate": 5.553062792054266e-06,
+      "loss": 0.5185,
+      "step": 10775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3875302023057496,
+      "learning_rate": 5.552411744476006e-06,
+      "loss": 0.3882,
+      "step": 10776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1026054310487567,
+      "learning_rate": 5.5517606958005395e-06,
+      "loss": 0.6159,
+      "step": 10777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7369135111035001,
+      "learning_rate": 5.551109646041493e-06,
+      "loss": 0.464,
+      "step": 10778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.782335363417897,
+      "learning_rate": 5.550458595212502e-06,
+      "loss": 0.4851,
+      "step": 10779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6585588746137632,
+      "learning_rate": 5.549807543327188e-06,
+      "loss": 0.4961,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5978810770951382,
+      "learning_rate": 5.549156490399186e-06,
+      "loss": 0.4513,
+      "step": 10781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7853454486752482,
+      "learning_rate": 5.548505436442123e-06,
+      "loss": 0.5055,
+      "step": 10782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0608240753520213,
+      "learning_rate": 5.547854381469628e-06,
+      "loss": 0.6109,
+      "step": 10783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9801149728138134,
+      "learning_rate": 5.5472033254953314e-06,
+      "loss": 0.4337,
+      "step": 10784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.859008381898776,
+      "learning_rate": 5.546552268532863e-06,
+      "loss": 0.5458,
+      "step": 10785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4577666806306757,
+      "learning_rate": 5.545901210595852e-06,
+      "loss": 0.4436,
+      "step": 10786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6491452743290371,
+      "learning_rate": 5.545250151697928e-06,
+      "loss": 0.5234,
+      "step": 10787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7914617582825005,
+      "learning_rate": 5.5445990918527195e-06,
+      "loss": 0.6277,
+      "step": 10788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8464123389401226,
+      "learning_rate": 5.543948031073857e-06,
+      "loss": 0.5003,
+      "step": 10789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8237279901389492,
+      "learning_rate": 5.543296969374969e-06,
+      "loss": 0.5264,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.675427570372963,
+      "learning_rate": 5.542645906769687e-06,
+      "loss": 0.5377,
+      "step": 10791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8872610699538241,
+      "learning_rate": 5.541994843271639e-06,
+      "loss": 0.6935,
+      "step": 10792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.836160836840236,
+      "learning_rate": 5.541343778894453e-06,
+      "loss": 0.5088,
+      "step": 10793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0713580710111525,
+      "learning_rate": 5.540692713651764e-06,
+      "loss": 0.5363,
+      "step": 10794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9516941809947543,
+      "learning_rate": 5.540041647557196e-06,
+      "loss": 0.8206,
+      "step": 10795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8661230015432084,
+      "learning_rate": 5.539390580624382e-06,
+      "loss": 0.547,
+      "step": 10796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.010418575099295,
+      "learning_rate": 5.5387395128669494e-06,
+      "loss": 0.5472,
+      "step": 10797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3110880961225906,
+      "learning_rate": 5.538088444298531e-06,
+      "loss": 0.7813,
+      "step": 10798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3796026371130463,
+      "learning_rate": 5.5374373749327535e-06,
+      "loss": 0.3274,
+      "step": 10799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9848129880762149,
+      "learning_rate": 5.536786304783248e-06,
+      "loss": 0.5565,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5257329940795898,
+      "eval_runtime": 67.9667,
+      "eval_samples_per_second": 50.172,
+      "eval_steps_per_second": 1.574,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7471287649107965,
+      "learning_rate": 5.5361352338636446e-06,
+      "loss": 0.4621,
+      "step": 10801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9695041778617595,
+      "learning_rate": 5.535484162187572e-06,
+      "loss": 0.4869,
+      "step": 10802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7713752715295004,
+      "learning_rate": 5.534833089768661e-06,
+      "loss": 0.4627,
+      "step": 10803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9980835235700887,
+      "learning_rate": 5.534182016620541e-06,
+      "loss": 0.5544,
+      "step": 10804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5610757485215803,
+      "learning_rate": 5.53353094275684e-06,
+      "loss": 0.4404,
+      "step": 10805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5299050709100739,
+      "learning_rate": 5.5328798681911925e-06,
+      "loss": 0.4494,
+      "step": 10806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8060547697204832,
+      "learning_rate": 5.532228792937223e-06,
+      "loss": 0.6164,
+      "step": 10807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2001308287936956,
+      "learning_rate": 5.531577717008565e-06,
+      "loss": 0.6187,
+      "step": 10808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6127805214108664,
+      "learning_rate": 5.530926640418847e-06,
+      "loss": 0.4978,
+      "step": 10809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0138874265425466,
+      "learning_rate": 5.530275563181698e-06,
+      "loss": 0.5654,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5980866568705852,
+      "learning_rate": 5.5296244853107494e-06,
+      "loss": 0.4403,
+      "step": 10811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.016016222270004,
+      "learning_rate": 5.52897340681963e-06,
+      "loss": 0.4053,
+      "step": 10812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.751838475551129,
+      "learning_rate": 5.528322327721971e-06,
+      "loss": 0.5644,
+      "step": 10813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5814222858117193,
+      "learning_rate": 5.527671248031401e-06,
+      "loss": 0.5962,
+      "step": 10814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8121563411585722,
+      "learning_rate": 5.527020167761552e-06,
+      "loss": 0.5602,
+      "step": 10815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.559888055613113,
+      "learning_rate": 5.526369086926052e-06,
+      "loss": 0.4247,
+      "step": 10816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9343052137953454,
+      "learning_rate": 5.525718005538532e-06,
+      "loss": 0.7343,
+      "step": 10817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9974968746761501,
+      "learning_rate": 5.525066923612619e-06,
+      "loss": 0.6001,
+      "step": 10818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4223059485139735,
+      "learning_rate": 5.524415841161949e-06,
+      "loss": 0.4436,
+      "step": 10819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0409439745022566,
+      "learning_rate": 5.523764758200145e-06,
+      "loss": 0.632,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5629168374531444,
+      "learning_rate": 5.523113674740845e-06,
+      "loss": 0.4418,
+      "step": 10821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.952759442965131,
+      "learning_rate": 5.522462590797671e-06,
+      "loss": 0.6061,
+      "step": 10822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7201440438572417,
+      "learning_rate": 5.521811506384257e-06,
+      "loss": 0.5224,
+      "step": 10823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7502960596176647,
+      "learning_rate": 5.521160421514234e-06,
+      "loss": 0.4587,
+      "step": 10824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.636135781784865,
+      "learning_rate": 5.5205093362012295e-06,
+      "loss": 0.4803,
+      "step": 10825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.730154832197086,
+      "learning_rate": 5.519858250458876e-06,
+      "loss": 0.4728,
+      "step": 10826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.040544763231843,
+      "learning_rate": 5.5192071643008015e-06,
+      "loss": 0.5381,
+      "step": 10827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4429978176614042,
+      "learning_rate": 5.518556077740637e-06,
+      "loss": 0.4256,
+      "step": 10828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.145566144138985,
+      "learning_rate": 5.517904990792012e-06,
+      "loss": 0.6567,
+      "step": 10829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5740240699244756,
+      "learning_rate": 5.517253903468559e-06,
+      "loss": 0.428,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4621799181814215,
+      "learning_rate": 5.516602815783905e-06,
+      "loss": 0.4396,
+      "step": 10831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6187477186634616,
+      "learning_rate": 5.515951727751681e-06,
+      "loss": 0.5626,
+      "step": 10832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2404249023041287,
+      "learning_rate": 5.5153006393855176e-06,
+      "loss": 0.6736,
+      "step": 10833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7074882933887146,
+      "learning_rate": 5.514649550699045e-06,
+      "loss": 0.4128,
+      "step": 10834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7912843502058828,
+      "learning_rate": 5.513998461705894e-06,
+      "loss": 0.5347,
+      "step": 10835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6146129971593477,
+      "learning_rate": 5.513347372419693e-06,
+      "loss": 0.3683,
+      "step": 10836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9828925630603733,
+      "learning_rate": 5.512696282854074e-06,
+      "loss": 0.5002,
+      "step": 10837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6587961805389224,
+      "learning_rate": 5.512045193022665e-06,
+      "loss": 0.4079,
+      "step": 10838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4415849099847158,
+      "learning_rate": 5.511394102939098e-06,
+      "loss": 0.4658,
+      "step": 10839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8178757195273996,
+      "learning_rate": 5.510743012617001e-06,
+      "loss": 0.5104,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.226452866388851,
+      "learning_rate": 5.510091922070007e-06,
+      "loss": 0.7072,
+      "step": 10841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7986116957047142,
+      "learning_rate": 5.509440831311747e-06,
+      "loss": 0.4971,
+      "step": 10842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.575172421906929,
+      "learning_rate": 5.5087897403558455e-06,
+      "loss": 0.4772,
+      "step": 10843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.518522644977581,
+      "learning_rate": 5.508138649215938e-06,
+      "loss": 0.3861,
+      "step": 10844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.455891902326253,
+      "learning_rate": 5.507487557905652e-06,
+      "loss": 0.3686,
+      "step": 10845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6438726653715254,
+      "learning_rate": 5.50683646643862e-06,
+      "loss": 0.3995,
+      "step": 10846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6347284757220788,
+      "learning_rate": 5.50618537482847e-06,
+      "loss": 0.5031,
+      "step": 10847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9836307666169524,
+      "learning_rate": 5.505534283088833e-06,
+      "loss": 0.4772,
+      "step": 10848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7361251347951485,
+      "learning_rate": 5.504883191233338e-06,
+      "loss": 0.5082,
+      "step": 10849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.66023162648659,
+      "learning_rate": 5.504232099275618e-06,
+      "loss": 0.5024,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5249655246734619,
+      "eval_runtime": 67.9087,
+      "eval_samples_per_second": 50.214,
+      "eval_steps_per_second": 1.576,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6456958987085313,
+      "learning_rate": 5.503581007229302e-06,
+      "loss": 0.3808,
+      "step": 10851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8055206140710067,
+      "learning_rate": 5.502929915108019e-06,
+      "loss": 0.5382,
+      "step": 10852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7379108374789072,
+      "learning_rate": 5.502278822925401e-06,
+      "loss": 0.6411,
+      "step": 10853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9286407862698376,
+      "learning_rate": 5.501627730695076e-06,
+      "loss": 0.4502,
+      "step": 10854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.014707635652678,
+      "learning_rate": 5.500976638430676e-06,
+      "loss": 0.4386,
+      "step": 10855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5904252404429886,
+      "learning_rate": 5.500325546145832e-06,
+      "loss": 0.5464,
+      "step": 10856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7976535887714955,
+      "learning_rate": 5.4996744538541705e-06,
+      "loss": 0.5734,
+      "step": 10857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3694728581791606,
+      "learning_rate": 5.4990233615693244e-06,
+      "loss": 0.384,
+      "step": 10858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2055157225723416,
+      "learning_rate": 5.498372269304925e-06,
+      "loss": 0.5732,
+      "step": 10859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.624625300080562,
+      "learning_rate": 5.4977211770746e-06,
+      "loss": 0.4867,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.848176576696968,
+      "learning_rate": 5.497070084891981e-06,
+      "loss": 0.5808,
+      "step": 10861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.624777514824512,
+      "learning_rate": 5.496418992770699e-06,
+      "loss": 0.5467,
+      "step": 10862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.919094274116563,
+      "learning_rate": 5.4957679007243826e-06,
+      "loss": 0.687,
+      "step": 10863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6277122956569867,
+      "learning_rate": 5.4951168087666615e-06,
+      "loss": 0.4812,
+      "step": 10864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8719819617618167,
+      "learning_rate": 5.494465716911168e-06,
+      "loss": 0.5858,
+      "step": 10865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7527356880289369,
+      "learning_rate": 5.493814625171531e-06,
+      "loss": 0.4756,
+      "step": 10866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.897413840424253,
+      "learning_rate": 5.493163533561382e-06,
+      "loss": 0.5236,
+      "step": 10867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6703868912726945,
+      "learning_rate": 5.492512442094349e-06,
+      "loss": 0.4793,
+      "step": 10868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7153588335708403,
+      "learning_rate": 5.491861350784063e-06,
+      "loss": 0.348,
+      "step": 10869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8344596682353547,
+      "learning_rate": 5.491210259644156e-06,
+      "loss": 0.5833,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.815289919563592,
+      "learning_rate": 5.4905591686882555e-06,
+      "loss": 0.4987,
+      "step": 10871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.742678089229063,
+      "learning_rate": 5.489908077929994e-06,
+      "loss": 0.446,
+      "step": 10872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3942183688877479,
+      "learning_rate": 5.4892569873829995e-06,
+      "loss": 0.3993,
+      "step": 10873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7215469394680831,
+      "learning_rate": 5.488605897060905e-06,
+      "loss": 0.5302,
+      "step": 10874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5257277798640358,
+      "learning_rate": 5.487954806977338e-06,
+      "loss": 0.3929,
+      "step": 10875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6475439005318935,
+      "learning_rate": 5.487303717145928e-06,
+      "loss": 0.4258,
+      "step": 10876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7276428345837922,
+      "learning_rate": 5.486652627580309e-06,
+      "loss": 0.5835,
+      "step": 10877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6749862079119784,
+      "learning_rate": 5.4860015382941076e-06,
+      "loss": 0.5091,
+      "step": 10878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.033660841920533,
+      "learning_rate": 5.485350449300957e-06,
+      "loss": 0.5365,
+      "step": 10879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6218985237452794,
+      "learning_rate": 5.484699360614484e-06,
+      "loss": 0.56,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6522840112247095,
+      "learning_rate": 5.484048272248321e-06,
+      "loss": 0.5091,
+      "step": 10881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3498581180299671,
+      "learning_rate": 5.483397184216097e-06,
+      "loss": 0.4157,
+      "step": 10882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8476730843725087,
+      "learning_rate": 5.482746096531442e-06,
+      "loss": 0.5201,
+      "step": 10883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6322340700932843,
+      "learning_rate": 5.482095009207989e-06,
+      "loss": 0.5683,
+      "step": 10884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6705979053881868,
+      "learning_rate": 5.481443922259365e-06,
+      "loss": 0.505,
+      "step": 10885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4206157265034407,
+      "learning_rate": 5.4807928356992e-06,
+      "loss": 0.3296,
+      "step": 10886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6467112335803424,
+      "learning_rate": 5.480141749541127e-06,
+      "loss": 0.4455,
+      "step": 10887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0716567890038853,
+      "learning_rate": 5.479490663798772e-06,
+      "loss": 0.4768,
+      "step": 10888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9712203705017801,
+      "learning_rate": 5.478839578485768e-06,
+      "loss": 0.6333,
+      "step": 10889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.687760534561524,
+      "learning_rate": 5.478188493615745e-06,
+      "loss": 0.5119,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6671135728411093,
+      "learning_rate": 5.477537409202332e-06,
+      "loss": 0.4659,
+      "step": 10891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0391749602878453,
+      "learning_rate": 5.4768863252591585e-06,
+      "loss": 0.5931,
+      "step": 10892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0747877465235565,
+      "learning_rate": 5.476235241799856e-06,
+      "loss": 0.5015,
+      "step": 10893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7569522402925783,
+      "learning_rate": 5.475584158838053e-06,
+      "loss": 0.4856,
+      "step": 10894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1737640001883882,
+      "learning_rate": 5.474933076387382e-06,
+      "loss": 0.599,
+      "step": 10895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8714488541542083,
+      "learning_rate": 5.47428199446147e-06,
+      "loss": 0.3752,
+      "step": 10896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.150919818787851,
+      "learning_rate": 5.4736309130739504e-06,
+      "loss": 0.603,
+      "step": 10897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8439748255165376,
+      "learning_rate": 5.472979832238451e-06,
+      "loss": 0.5447,
+      "step": 10898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7351792419921668,
+      "learning_rate": 5.472328751968601e-06,
+      "loss": 0.602,
+      "step": 10899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6652671795812617,
+      "learning_rate": 5.471677672278031e-06,
+      "loss": 0.477,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.524895429611206,
+      "eval_runtime": 67.9298,
+      "eval_samples_per_second": 50.199,
+      "eval_steps_per_second": 1.575,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5805752059256517,
+      "learning_rate": 5.471026593180373e-06,
+      "loss": 0.4158,
+      "step": 10901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.677257375687198,
+      "learning_rate": 5.470375514689254e-06,
+      "loss": 0.4594,
+      "step": 10902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6057166355323356,
+      "learning_rate": 5.4697244368183054e-06,
+      "loss": 0.5065,
+      "step": 10903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5614896151948674,
+      "learning_rate": 5.469073359581156e-06,
+      "loss": 0.4845,
+      "step": 10904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5697457463640223,
+      "learning_rate": 5.468422282991438e-06,
+      "loss": 0.4844,
+      "step": 10905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.615037882500592,
+      "learning_rate": 5.467771207062778e-06,
+      "loss": 0.5126,
+      "step": 10906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2332407662146205,
+      "learning_rate": 5.46712013180881e-06,
+      "loss": 0.6292,
+      "step": 10907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.056822196986843,
+      "learning_rate": 5.46646905724316e-06,
+      "loss": 0.7572,
+      "step": 10908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8840438939372144,
+      "learning_rate": 5.46581798337946e-06,
+      "loss": 0.4927,
+      "step": 10909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9091049724843434,
+      "learning_rate": 5.46516691023134e-06,
+      "loss": 0.4375,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6969234666555464,
+      "learning_rate": 5.464515837812429e-06,
+      "loss": 0.4074,
+      "step": 10911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.368429802147412,
+      "learning_rate": 5.463864766136357e-06,
+      "loss": 0.4204,
+      "step": 10912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7901730844886379,
+      "learning_rate": 5.463213695216752e-06,
+      "loss": 0.551,
+      "step": 10913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9158452380599165,
+      "learning_rate": 5.462562625067247e-06,
+      "loss": 0.6354,
+      "step": 10914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6590368191408005,
+      "learning_rate": 5.46191155570147e-06,
+      "loss": 0.5078,
+      "step": 10915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.661119918439415,
+      "learning_rate": 5.461260487133051e-06,
+      "loss": 0.4064,
+      "step": 10916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7540369098619795,
+      "learning_rate": 5.4606094193756196e-06,
+      "loss": 0.5572,
+      "step": 10917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7610950141502266,
+      "learning_rate": 5.459958352442804e-06,
+      "loss": 0.4762,
+      "step": 10918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.491142889492198,
+      "learning_rate": 5.459307286348238e-06,
+      "loss": 0.3662,
+      "step": 10919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6691222744003076,
+      "learning_rate": 5.458656221105547e-06,
+      "loss": 0.6311,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4955684737581063,
+      "learning_rate": 5.458005156728363e-06,
+      "loss": 0.4185,
+      "step": 10921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7654680377779053,
+      "learning_rate": 5.457354093230315e-06,
+      "loss": 0.4916,
+      "step": 10922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8070335688120818,
+      "learning_rate": 5.456703030625032e-06,
+      "loss": 0.5693,
+      "step": 10923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9424427808007254,
+      "learning_rate": 5.4560519689261445e-06,
+      "loss": 0.5198,
+      "step": 10924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4882463735378952,
+      "learning_rate": 5.455400908147283e-06,
+      "loss": 0.3718,
+      "step": 10925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5923211212703678,
+      "learning_rate": 5.454749848302073e-06,
+      "loss": 0.4067,
+      "step": 10926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.454501121832061,
+      "learning_rate": 5.454098789404149e-06,
+      "loss": 0.4156,
+      "step": 10927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5501585039199939,
+      "learning_rate": 5.4534477314671384e-06,
+      "loss": 0.4612,
+      "step": 10928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5926473715942193,
+      "learning_rate": 5.452796674504669e-06,
+      "loss": 0.5517,
+      "step": 10929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6046092946480937,
+      "learning_rate": 5.452145618530373e-06,
+      "loss": 0.4701,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0251240760470073,
+      "learning_rate": 5.451494563557879e-06,
+      "loss": 0.4731,
+      "step": 10931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4888993436820341,
+      "learning_rate": 5.450843509600816e-06,
+      "loss": 0.4104,
+      "step": 10932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.612024665732054,
+      "learning_rate": 5.450192456672814e-06,
+      "loss": 0.4695,
+      "step": 10933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8166152999391338,
+      "learning_rate": 5.449541404787501e-06,
+      "loss": 0.4329,
+      "step": 10934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.607348932610883,
+      "learning_rate": 5.448890353958507e-06,
+      "loss": 0.5464,
+      "step": 10935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5692062870766612,
+      "learning_rate": 5.448239304199462e-06,
+      "loss": 0.4584,
+      "step": 10936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2334460970849137,
+      "learning_rate": 5.447588255523995e-06,
+      "loss": 0.4869,
+      "step": 10937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6084591034151599,
+      "learning_rate": 5.446937207945737e-06,
+      "loss": 0.3399,
+      "step": 10938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.856842294821548,
+      "learning_rate": 5.4462861614783135e-06,
+      "loss": 0.5072,
+      "step": 10939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5868942687649095,
+      "learning_rate": 5.4456351161353575e-06,
+      "loss": 0.4367,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8913416063730095,
+      "learning_rate": 5.444984071930496e-06,
+      "loss": 0.5922,
+      "step": 10941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.834407532746862,
+      "learning_rate": 5.444333028877358e-06,
+      "loss": 0.6762,
+      "step": 10942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9825078208534415,
+      "learning_rate": 5.443681986989574e-06,
+      "loss": 0.6397,
+      "step": 10943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.470020513891553,
+      "learning_rate": 5.4430309462807716e-06,
+      "loss": 0.3965,
+      "step": 10944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5326120787572188,
+      "learning_rate": 5.442379906764583e-06,
+      "loss": 0.4383,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9461037799316865,
+      "learning_rate": 5.441728868454634e-06,
+      "loss": 0.5551,
+      "step": 10946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8710871020718634,
+      "learning_rate": 5.441077831364555e-06,
+      "loss": 0.5029,
+      "step": 10947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.776012473529251,
+      "learning_rate": 5.4404267955079745e-06,
+      "loss": 0.8443,
+      "step": 10948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.61614943359187,
+      "learning_rate": 5.4397757608985225e-06,
+      "loss": 0.634,
+      "step": 10949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7103593604603389,
+      "learning_rate": 5.439124727549828e-06,
+      "loss": 0.4027,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5255641341209412,
+      "eval_runtime": 67.949,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.906487983830258,
+      "learning_rate": 5.4384736954755206e-06,
+      "loss": 0.5395,
+      "step": 10951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1659931228558964,
+      "learning_rate": 5.437822664689226e-06,
+      "loss": 0.577,
+      "step": 10952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6676448410429259,
+      "learning_rate": 5.437171635204577e-06,
+      "loss": 0.4253,
+      "step": 10953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.617903452471723,
+      "learning_rate": 5.4365206070352e-06,
+      "loss": 0.4383,
+      "step": 10954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5753026836594644,
+      "learning_rate": 5.435869580194723e-06,
+      "loss": 0.5249,
+      "step": 10955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7324489664711404,
+      "learning_rate": 5.435218554696777e-06,
+      "loss": 0.5573,
+      "step": 10956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3498301692125143,
+      "learning_rate": 5.4345675305549914e-06,
+      "loss": 0.4247,
+      "step": 10957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.580699958074959,
+      "learning_rate": 5.4339165077829935e-06,
+      "loss": 0.5337,
+      "step": 10958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6949580924508159,
+      "learning_rate": 5.433265486394414e-06,
+      "loss": 0.4075,
+      "step": 10959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8382630301518987,
+      "learning_rate": 5.432614466402879e-06,
+      "loss": 0.6147,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7671423541311282,
+      "learning_rate": 5.4319634478220174e-06,
+      "loss": 0.7004,
+      "step": 10961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6963002955071103,
+      "learning_rate": 5.43131243066546e-06,
+      "loss": 0.4498,
+      "step": 10962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9203936046271177,
+      "learning_rate": 5.430661414946834e-06,
+      "loss": 0.5685,
+      "step": 10963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.553585374527286,
+      "learning_rate": 5.430010400679769e-06,
+      "loss": 0.4009,
+      "step": 10964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9526428872904549,
+      "learning_rate": 5.429359387877891e-06,
+      "loss": 0.587,
+      "step": 10965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7116854741546441,
+      "learning_rate": 5.428708376554831e-06,
+      "loss": 0.4508,
+      "step": 10966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.958021212119431,
+      "learning_rate": 5.428057366724217e-06,
+      "loss": 0.5976,
+      "step": 10967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5217562907488056,
+      "learning_rate": 5.427406358399679e-06,
+      "loss": 0.521,
+      "step": 10968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9025446731027826,
+      "learning_rate": 5.4267553515948405e-06,
+      "loss": 0.4352,
+      "step": 10969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.804348547954527,
+      "learning_rate": 5.426104346323335e-06,
+      "loss": 0.5761,
+      "step": 10970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5592674000289481,
+      "learning_rate": 5.425453342598789e-06,
+      "loss": 0.434,
+      "step": 10971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4623286161512021,
+      "learning_rate": 5.424802340434831e-06,
+      "loss": 0.3687,
+      "step": 10972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.689989958914628,
+      "learning_rate": 5.424151339845089e-06,
+      "loss": 0.661,
+      "step": 10973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8082871523268538,
+      "learning_rate": 5.423500340843192e-06,
+      "loss": 0.5463,
+      "step": 10974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5219658911137757,
+      "learning_rate": 5.422849343442768e-06,
+      "loss": 0.4306,
+      "step": 10975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.630602606349492,
+      "learning_rate": 5.422198347657444e-06,
+      "loss": 0.4667,
+      "step": 10976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8083283954761646,
+      "learning_rate": 5.42154735350085e-06,
+      "loss": 0.6227,
+      "step": 10977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7872074808408238,
+      "learning_rate": 5.420896360986614e-06,
+      "loss": 0.7356,
+      "step": 10978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1057437316203806,
+      "learning_rate": 5.420245370128363e-06,
+      "loss": 0.5549,
+      "step": 10979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8286995022598544,
+      "learning_rate": 5.4195943809397256e-06,
+      "loss": 0.5684,
+      "step": 10980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.560976818848312,
+      "learning_rate": 5.418943393434331e-06,
+      "loss": 0.5076,
+      "step": 10981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9004576388372716,
+      "learning_rate": 5.4182924076258036e-06,
+      "loss": 0.4875,
+      "step": 10982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8497796800759845,
+      "learning_rate": 5.417641423527775e-06,
+      "loss": 0.4581,
+      "step": 10983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4840031897713344,
+      "learning_rate": 5.416990441153873e-06,
+      "loss": 0.4932,
+      "step": 10984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.839585049851853,
+      "learning_rate": 5.416339460517724e-06,
+      "loss": 0.584,
+      "step": 10985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4079703527117347,
+      "learning_rate": 5.415688481632957e-06,
+      "loss": 0.5914,
+      "step": 10986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8392020430552227,
+      "learning_rate": 5.415037504513197e-06,
+      "loss": 0.4242,
+      "step": 10987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5882779964383733,
+      "learning_rate": 5.414386529172077e-06,
+      "loss": 0.4498,
+      "step": 10988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7933050605176808,
+      "learning_rate": 5.41373555562322e-06,
+      "loss": 0.4664,
+      "step": 10989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6936628684952197,
+      "learning_rate": 5.413084583880257e-06,
+      "loss": 0.4885,
+      "step": 10990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.209757863684857,
+      "learning_rate": 5.412433613956813e-06,
+      "loss": 0.6283,
+      "step": 10991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6344468186900392,
+      "learning_rate": 5.4117826458665164e-06,
+      "loss": 0.4228,
+      "step": 10992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.349682802021018,
+      "learning_rate": 5.411131679622997e-06,
+      "loss": 0.4284,
+      "step": 10993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8072461997294305,
+      "learning_rate": 5.410480715239881e-06,
+      "loss": 0.5393,
+      "step": 10994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1443961525336483,
+      "learning_rate": 5.409829752730793e-06,
+      "loss": 0.7124,
+      "step": 10995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.023748316111262,
+      "learning_rate": 5.4091787921093655e-06,
+      "loss": 0.483,
+      "step": 10996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8557433290123952,
+      "learning_rate": 5.408527833389221e-06,
+      "loss": 0.5518,
+      "step": 10997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1613411408956407,
+      "learning_rate": 5.407876876583992e-06,
+      "loss": 0.6953,
+      "step": 10998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5126293929579853,
+      "learning_rate": 5.407225921707303e-06,
+      "loss": 0.4628,
+      "step": 10999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.037581562304712,
+      "learning_rate": 5.406574968772781e-06,
+      "loss": 0.5497,
+      "step": 11000
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.5233793258666992,
+      "eval_runtime": 67.9727,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 11000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.001913302565474,
+      "learning_rate": 5.405924017794054e-06,
+      "loss": 0.468,
+      "step": 11001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6056631218905355,
+      "learning_rate": 5.405273068784749e-06,
+      "loss": 0.4226,
+      "step": 11002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7881175472772814,
+      "learning_rate": 5.404622121758495e-06,
+      "loss": 0.4742,
+      "step": 11003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.624942407429736,
+      "learning_rate": 5.4039711767289175e-06,
+      "loss": 0.3654,
+      "step": 11004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.167839225178481,
+      "learning_rate": 5.403320233709641e-06,
+      "loss": 0.5057,
+      "step": 11005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6367756132769509,
+      "learning_rate": 5.402669292714297e-06,
+      "loss": 0.558,
+      "step": 11006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8924598835073017,
+      "learning_rate": 5.402018353756511e-06,
+      "loss": 0.5775,
+      "step": 11007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9416311885051092,
+      "learning_rate": 5.4013674168499095e-06,
+      "loss": 0.6757,
+      "step": 11008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.529957359765102,
+      "learning_rate": 5.400716482008121e-06,
+      "loss": 0.4142,
+      "step": 11009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.514767600284363,
+      "learning_rate": 5.400065549244771e-06,
+      "loss": 0.5676,
+      "step": 11010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.490692745638796,
+      "learning_rate": 5.399414618573485e-06,
+      "loss": 0.5277,
+      "step": 11011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7195327742943258,
+      "learning_rate": 5.398763690007894e-06,
+      "loss": 0.5657,
+      "step": 11012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9738597190692373,
+      "learning_rate": 5.398112763561622e-06,
+      "loss": 0.5684,
+      "step": 11013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8211382852528804,
+      "learning_rate": 5.397461839248294e-06,
+      "loss": 0.5524,
+      "step": 11014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0767177587099805,
+      "learning_rate": 5.39681091708154e-06,
+      "loss": 0.4892,
+      "step": 11015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8939328240857503,
+      "learning_rate": 5.396159997074984e-06,
+      "loss": 0.5203,
+      "step": 11016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.501864871330139,
+      "learning_rate": 5.395509079242256e-06,
+      "loss": 0.3882,
+      "step": 11017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9078567026795252,
+      "learning_rate": 5.39485816359698e-06,
+      "loss": 0.4518,
+      "step": 11018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0381177581715813,
+      "learning_rate": 5.394207250152782e-06,
+      "loss": 0.6293,
+      "step": 11019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.662722164614394,
+      "learning_rate": 5.3935563389232915e-06,
+      "loss": 0.4818,
+      "step": 11020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.666629113630525,
+      "learning_rate": 5.3929054299221315e-06,
+      "loss": 0.434,
+      "step": 11021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7311964081295994,
+      "learning_rate": 5.39225452316293e-06,
+      "loss": 0.4109,
+      "step": 11022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8389082990938723,
+      "learning_rate": 5.391603618659312e-06,
+      "loss": 0.4913,
+      "step": 11023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9704554728912447,
+      "learning_rate": 5.390952716424906e-06,
+      "loss": 0.5719,
+      "step": 11024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7930099041697265,
+      "learning_rate": 5.390301816473338e-06,
+      "loss": 0.6332,
+      "step": 11025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7537128322270141,
+      "learning_rate": 5.3896509188182324e-06,
+      "loss": 0.5156,
+      "step": 11026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7779474727958298,
+      "learning_rate": 5.3890000234732155e-06,
+      "loss": 0.5601,
+      "step": 11027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7824915002192763,
+      "learning_rate": 5.388349130451916e-06,
+      "loss": 0.492,
+      "step": 11028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2030477475935744,
+      "learning_rate": 5.387698239767957e-06,
+      "loss": 0.6631,
+      "step": 11029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9127740296177993,
+      "learning_rate": 5.387047351434966e-06,
+      "loss": 0.5284,
+      "step": 11030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9983709506630438,
+      "learning_rate": 5.386396465466568e-06,
+      "loss": 0.5272,
+      "step": 11031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.053320502684797,
+      "learning_rate": 5.3857455818763894e-06,
+      "loss": 0.658,
+      "step": 11032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7130124479570987,
+      "learning_rate": 5.385094700678056e-06,
+      "loss": 0.7668,
+      "step": 11033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.043072757482817,
+      "learning_rate": 5.3844438218851944e-06,
+      "loss": 0.7856,
+      "step": 11034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9010764448843736,
+      "learning_rate": 5.383792945511429e-06,
+      "loss": 0.5593,
+      "step": 11035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6352622435511237,
+      "learning_rate": 5.383142071570388e-06,
+      "loss": 0.5247,
+      "step": 11036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6985455086162478,
+      "learning_rate": 5.3824912000756925e-06,
+      "loss": 0.6051,
+      "step": 11037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.453474186730988,
+      "learning_rate": 5.3818403310409716e-06,
+      "loss": 0.4102,
+      "step": 11038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.975811547234285,
+      "learning_rate": 5.38118946447985e-06,
+      "loss": 0.6455,
+      "step": 11039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4753477904422425,
+      "learning_rate": 5.380538600405953e-06,
+      "loss": 0.3644,
+      "step": 11040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.839779129630123,
+      "learning_rate": 5.379887738832907e-06,
+      "loss": 0.6718,
+      "step": 11041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5609528830366797,
+      "learning_rate": 5.379236879774335e-06,
+      "loss": 0.5918,
+      "step": 11042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.328799023013081,
+      "learning_rate": 5.378586023243866e-06,
+      "loss": 0.7398,
+      "step": 11043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8392745323260715,
+      "learning_rate": 5.377935169255123e-06,
+      "loss": 0.4564,
+      "step": 11044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6456337368647806,
+      "learning_rate": 5.3772843178217295e-06,
+      "loss": 0.4938,
+      "step": 11045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.610435842167065,
+      "learning_rate": 5.376633468957314e-06,
+      "loss": 0.5165,
+      "step": 11046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9716998106838814,
+      "learning_rate": 5.375982622675501e-06,
+      "loss": 0.6976,
+      "step": 11047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4957202306374726,
+      "learning_rate": 5.375331778989913e-06,
+      "loss": 0.4068,
+      "step": 11048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.321270343313629,
+      "learning_rate": 5.374680937914177e-06,
+      "loss": 0.5151,
+      "step": 11049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.655939927126412,
+      "learning_rate": 5.374030099461917e-06,
+      "loss": 0.5183,
+      "step": 11050
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.5215616226196289,
+      "eval_runtime": 67.9734,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 11050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.286590744118099,
+      "learning_rate": 5.37337926364676e-06,
+      "loss": 0.4998,
+      "step": 11051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8239408048526924,
+      "learning_rate": 5.372728430482331e-06,
+      "loss": 0.5039,
+      "step": 11052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5636608949769126,
+      "learning_rate": 5.37207759998225e-06,
+      "loss": 0.3676,
+      "step": 11053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.671826374531293,
+      "learning_rate": 5.371426772160145e-06,
+      "loss": 0.4235,
+      "step": 11054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.573647812141276,
+      "learning_rate": 5.370775947029643e-06,
+      "loss": 0.4567,
+      "step": 11055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7363814935764927,
+      "learning_rate": 5.370125124604365e-06,
+      "loss": 0.408,
+      "step": 11056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9163598947916385,
+      "learning_rate": 5.369474304897937e-06,
+      "loss": 0.5709,
+      "step": 11057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8722735846888752,
+      "learning_rate": 5.3688234879239845e-06,
+      "loss": 0.6159,
+      "step": 11058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7795296494661608,
+      "learning_rate": 5.368172673696129e-06,
+      "loss": 0.6353,
+      "step": 11059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.678049905444532,
+      "learning_rate": 5.367521862227998e-06,
+      "loss": 0.4432,
+      "step": 11060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0657176886102704,
+      "learning_rate": 5.366871053533215e-06,
+      "loss": 0.5834,
+      "step": 11061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4953630381949585,
+      "learning_rate": 5.366220247625402e-06,
+      "loss": 0.3906,
+      "step": 11062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.089682265205241,
+      "learning_rate": 5.365569444518187e-06,
+      "loss": 0.5041,
+      "step": 11063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.314708966620481,
+      "learning_rate": 5.364918644225191e-06,
+      "loss": 0.6945,
+      "step": 11064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4393804933176617,
+      "learning_rate": 5.364267846760039e-06,
+      "loss": 0.4155,
+      "step": 11065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8110069472663008,
+      "learning_rate": 5.3636170521363564e-06,
+      "loss": 0.5304,
+      "step": 11066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.803261956361254,
+      "learning_rate": 5.362966260367765e-06,
+      "loss": 0.61,
+      "step": 11067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8963533072012912,
+      "learning_rate": 5.36231547146789e-06,
+      "loss": 0.5552,
+      "step": 11068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7899404530777185,
+      "learning_rate": 5.361664685450355e-06,
+      "loss": 0.5036,
+      "step": 11069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1589214264738708,
+      "learning_rate": 5.361013902328784e-06,
+      "loss": 0.5041,
+      "step": 11070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8993148534917306,
+      "learning_rate": 5.360363122116802e-06,
+      "loss": 0.4523,
+      "step": 11071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.22271123156324,
+      "learning_rate": 5.359712344828029e-06,
+      "loss": 0.5707,
+      "step": 11072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.670022646717163,
+      "learning_rate": 5.359061570476094e-06,
+      "loss": 0.4418,
+      "step": 11073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.322764177972074,
+      "learning_rate": 5.358410799074616e-06,
+      "loss": 0.6831,
+      "step": 11074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.120486780422446,
+      "learning_rate": 5.357760030637219e-06,
+      "loss": 0.6637,
+      "step": 11075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.078868121987026,
+      "learning_rate": 5.357109265177528e-06,
+      "loss": 0.5962,
+      "step": 11076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.066957689399538,
+      "learning_rate": 5.3564585027091655e-06,
+      "loss": 0.4643,
+      "step": 11077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8423148171099484,
+      "learning_rate": 5.355807743245756e-06,
+      "loss": 0.5543,
+      "step": 11078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0871581316166554,
+      "learning_rate": 5.355156986800921e-06,
+      "loss": 0.5667,
+      "step": 11079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7202833092095289,
+      "learning_rate": 5.354506233388284e-06,
+      "loss": 0.5828,
+      "step": 11080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5997006768729052,
+      "learning_rate": 5.35385548302147e-06,
+      "loss": 0.5585,
+      "step": 11081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4651978109515709,
+      "learning_rate": 5.353204735714099e-06,
+      "loss": 0.3842,
+      "step": 11082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.945257986765747,
+      "learning_rate": 5.3525539914797974e-06,
+      "loss": 0.6182,
+      "step": 11083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.915167140215103,
+      "learning_rate": 5.351903250332185e-06,
+      "loss": 0.6585,
+      "step": 11084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5431904202484372,
+      "learning_rate": 5.351252512284888e-06,
+      "loss": 0.6353,
+      "step": 11085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6787708237055583,
+      "learning_rate": 5.350601777351526e-06,
+      "loss": 0.5698,
+      "step": 11086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8674467907639558,
+      "learning_rate": 5.349951045545724e-06,
+      "loss": 0.3984,
+      "step": 11087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1037733550765294,
+      "learning_rate": 5.349300316881101e-06,
+      "loss": 0.5254,
+      "step": 11088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7818514294824457,
+      "learning_rate": 5.348649591371284e-06,
+      "loss": 0.6216,
+      "step": 11089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7768564456836378,
+      "learning_rate": 5.347998869029894e-06,
+      "loss": 0.551,
+      "step": 11090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6350114047420319,
+      "learning_rate": 5.347348149870554e-06,
+      "loss": 0.4035,
+      "step": 11091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6799739885236673,
+      "learning_rate": 5.346697433906885e-06,
+      "loss": 0.6594,
+      "step": 11092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8274613155739565,
+      "learning_rate": 5.346046721152509e-06,
+      "loss": 0.4791,
+      "step": 11093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8539832968198962,
+      "learning_rate": 5.34539601162105e-06,
+      "loss": 0.5512,
+      "step": 11094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7167590336399448,
+      "learning_rate": 5.344745305326129e-06,
+      "loss": 0.449,
+      "step": 11095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4938215485181574,
+      "learning_rate": 5.344094602281369e-06,
+      "loss": 0.4277,
+      "step": 11096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.889325679156062,
+      "learning_rate": 5.343443902500393e-06,
+      "loss": 0.5266,
+      "step": 11097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7124399978213574,
+      "learning_rate": 5.3427932059968194e-06,
+      "loss": 0.569,
+      "step": 11098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.908702937173418,
+      "learning_rate": 5.342142512784274e-06,
+      "loss": 0.4829,
+      "step": 11099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7731797627099968,
+      "learning_rate": 5.341491822876376e-06,
+      "loss": 0.4741,
+      "step": 11100
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.5209984183311462,
+      "eval_runtime": 67.9914,
+      "eval_samples_per_second": 50.153,
+      "eval_steps_per_second": 1.574,
+      "step": 11100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.635653441090699,
+      "learning_rate": 5.3408411362867485e-06,
+      "loss": 0.4014,
+      "step": 11101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7263710108174473,
+      "learning_rate": 5.3401904530290135e-06,
+      "loss": 0.4982,
+      "step": 11102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3428537056494974,
+      "learning_rate": 5.339539773116791e-06,
+      "loss": 0.5793,
+      "step": 11103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.85327637492399,
+      "learning_rate": 5.338889096563703e-06,
+      "loss": 0.4122,
+      "step": 11104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5596996277136645,
+      "learning_rate": 5.338238423383374e-06,
+      "loss": 0.444,
+      "step": 11105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1051041797074235,
+      "learning_rate": 5.3375877535894226e-06,
+      "loss": 0.6192,
+      "step": 11106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9413152009558878,
+      "learning_rate": 5.33693708719547e-06,
+      "loss": 0.5311,
+      "step": 11107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2657474855776574,
+      "learning_rate": 5.336286424215138e-06,
+      "loss": 0.477,
+      "step": 11108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.049365532549588,
+      "learning_rate": 5.3356357646620485e-06,
+      "loss": 0.4355,
+      "step": 11109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9497604422929293,
+      "learning_rate": 5.334985108549823e-06,
+      "loss": 0.4754,
+      "step": 11110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6666314862852127,
+      "learning_rate": 5.334334455892081e-06,
+      "loss": 0.3419,
+      "step": 11111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5509285978072096,
+      "learning_rate": 5.333683806702443e-06,
+      "loss": 0.5129,
+      "step": 11112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.582602737682117,
+      "learning_rate": 5.333033160994533e-06,
+      "loss": 0.4915,
+      "step": 11113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8890984432940465,
+      "learning_rate": 5.33238251878197e-06,
+      "loss": 0.5517,
+      "step": 11114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3063288099657426,
+      "learning_rate": 5.331731880078373e-06,
+      "loss": 0.458,
+      "step": 11115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8816977193766768,
+      "learning_rate": 5.331081244897367e-06,
+      "loss": 0.4596,
+      "step": 11116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.801628564917014,
+      "learning_rate": 5.330430613252568e-06,
+      "loss": 0.5291,
+      "step": 11117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6478681963527204,
+      "learning_rate": 5.329779985157601e-06,
+      "loss": 0.5886,
+      "step": 11118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9160152461577946,
+      "learning_rate": 5.329129360626084e-06,
+      "loss": 0.571,
+      "step": 11119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9356657803605588,
+      "learning_rate": 5.328478739671636e-06,
+      "loss": 0.5259,
+      "step": 11120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0187720224817265,
+      "learning_rate": 5.327828122307881e-06,
+      "loss": 0.6447,
+      "step": 11121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.03086459172605,
+      "learning_rate": 5.327177508548436e-06,
+      "loss": 0.423,
+      "step": 11122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9908098741031546,
+      "learning_rate": 5.326526898406922e-06,
+      "loss": 0.4955,
+      "step": 11123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6484281140566532,
+      "learning_rate": 5.325876291896961e-06,
+      "loss": 0.6117,
+      "step": 11124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5181385321456522,
+      "learning_rate": 5.325225689032171e-06,
+      "loss": 0.4051,
+      "step": 11125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5716150014853696,
+      "learning_rate": 5.324575089826173e-06,
+      "loss": 0.3657,
+      "step": 11126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9496102335981482,
+      "learning_rate": 5.323924494292587e-06,
+      "loss": 0.5622,
+      "step": 11127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.272817884273995,
+      "learning_rate": 5.323273902445031e-06,
+      "loss": 0.6142,
+      "step": 11128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6062650266444078,
+      "learning_rate": 5.3226233142971275e-06,
+      "loss": 0.526,
+      "step": 11129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6936370211304523,
+      "learning_rate": 5.321972729862493e-06,
+      "loss": 0.4382,
+      "step": 11130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8486570567844398,
+      "learning_rate": 5.32132214915475e-06,
+      "loss": 0.469,
+      "step": 11131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0582004420087348,
+      "learning_rate": 5.3206715721875165e-06,
+      "loss": 0.5555,
+      "step": 11132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7108950267731047,
+      "learning_rate": 5.320020998974412e-06,
+      "loss": 0.4026,
+      "step": 11133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7025581015538704,
+      "learning_rate": 5.319370429529057e-06,
+      "loss": 0.4284,
+      "step": 11134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9909761228909082,
+      "learning_rate": 5.31871986386507e-06,
+      "loss": 0.5352,
+      "step": 11135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5448996119151033,
+      "learning_rate": 5.3180693019960695e-06,
+      "loss": 0.5042,
+      "step": 11136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5206165537476575,
+      "learning_rate": 5.317418743935675e-06,
+      "loss": 0.445,
+      "step": 11137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9035089494599116,
+      "learning_rate": 5.316768189697505e-06,
+      "loss": 0.653,
+      "step": 11138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.820171788318182,
+      "learning_rate": 5.316117639295181e-06,
+      "loss": 0.5728,
+      "step": 11139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4837974313672562,
+      "learning_rate": 5.315467092742318e-06,
+      "loss": 0.4263,
+      "step": 11140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5283477984222211,
+      "learning_rate": 5.314816550052538e-06,
+      "loss": 0.4198,
+      "step": 11141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1892807072755858,
+      "learning_rate": 5.314166011239459e-06,
+      "loss": 0.6627,
+      "step": 11142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9580813153811547,
+      "learning_rate": 5.313515476316697e-06,
+      "loss": 0.7396,
+      "step": 11143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9139462122008855,
+      "learning_rate": 5.312864945297874e-06,
+      "loss": 0.5677,
+      "step": 11144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.873350463546826,
+      "learning_rate": 5.312214418196608e-06,
+      "loss": 0.496,
+      "step": 11145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.777674885381212,
+      "learning_rate": 5.311563895026515e-06,
+      "loss": 0.5252,
+      "step": 11146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8341958607380628,
+      "learning_rate": 5.310913375801215e-06,
+      "loss": 0.4801,
+      "step": 11147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6696864996603165,
+      "learning_rate": 5.310262860534326e-06,
+      "loss": 0.4981,
+      "step": 11148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0061457773577738,
+      "learning_rate": 5.309612349239468e-06,
+      "loss": 0.6175,
+      "step": 11149
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.864782334228585,
+      "learning_rate": 5.308961841930257e-06,
+      "loss": 0.5277,
+      "step": 11150
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.5211642384529114,
+      "eval_runtime": 68.0463,
+      "eval_samples_per_second": 50.113,
+      "eval_steps_per_second": 1.572,
+      "step": 11150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7974715027831383,
+      "learning_rate": 5.308311338620309e-06,
+      "loss": 0.439,
+      "step": 11151
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6747657910422702,
+      "learning_rate": 5.307660839323243e-06,
+      "loss": 0.3479,
+      "step": 11152
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5681481628505667,
+      "learning_rate": 5.307010344052681e-06,
+      "loss": 0.4487,
+      "step": 11153
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.828552909230357,
+      "learning_rate": 5.306359852822236e-06,
+      "loss": 0.5265,
+      "step": 11154
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4851577679143988,
+      "learning_rate": 5.305709365645528e-06,
+      "loss": 0.4535,
+      "step": 11155
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9284409608978255,
+      "learning_rate": 5.3050588825361735e-06,
+      "loss": 0.4447,
+      "step": 11156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.019426647401912,
+      "learning_rate": 5.30440840350779e-06,
+      "loss": 0.7166,
+      "step": 11157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.536792686702918,
+      "learning_rate": 5.303757928573996e-06,
+      "loss": 0.5385,
+      "step": 11158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5391254945780948,
+      "learning_rate": 5.3031074577484065e-06,
+      "loss": 0.3644,
+      "step": 11159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8452888660134519,
+      "learning_rate": 5.302456991044641e-06,
+      "loss": 0.5396,
+      "step": 11160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8853943724929203,
+      "learning_rate": 5.301806528476315e-06,
+      "loss": 0.5381,
+      "step": 11161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5964672161690792,
+      "learning_rate": 5.301156070057046e-06,
+      "loss": 0.3889,
+      "step": 11162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6865455973183303,
+      "learning_rate": 5.300505615800453e-06,
+      "loss": 0.4124,
+      "step": 11163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.419936666493992,
+      "learning_rate": 5.2998551657201505e-06,
+      "loss": 0.4454,
+      "step": 11164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9506651621722624,
+      "learning_rate": 5.299204719829755e-06,
+      "loss": 0.542,
+      "step": 11165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8599414069388056,
+      "learning_rate": 5.298554278142883e-06,
+      "loss": 0.5602,
+      "step": 11166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.115784766547431,
+      "learning_rate": 5.297903840673154e-06,
+      "loss": 0.5579,
+      "step": 11167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8873599356981448,
+      "learning_rate": 5.297253407434182e-06,
+      "loss": 0.4109,
+      "step": 11168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4330443093265453,
+      "learning_rate": 5.2966029784395845e-06,
+      "loss": 0.4202,
+      "step": 11169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0628621619387006,
+      "learning_rate": 5.2959525537029754e-06,
+      "loss": 0.6535,
+      "step": 11170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8991608876843922,
+      "learning_rate": 5.295302133237974e-06,
+      "loss": 0.5808,
+      "step": 11171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.85072379043226,
+      "learning_rate": 5.294651717058197e-06,
+      "loss": 0.5401,
+      "step": 11172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9020388187804205,
+      "learning_rate": 5.294001305177256e-06,
+      "loss": 0.5212,
+      "step": 11173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.535463862314525,
+      "learning_rate": 5.293350897608771e-06,
+      "loss": 0.3893,
+      "step": 11174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0402313313644083,
+      "learning_rate": 5.292700494366357e-06,
+      "loss": 0.4251,
+      "step": 11175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8397302540793077,
+      "learning_rate": 5.292050095463628e-06,
+      "loss": 0.5868,
+      "step": 11176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4508666050730354,
+      "learning_rate": 5.291399700914203e-06,
+      "loss": 0.4494,
+      "step": 11177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5525465624743586,
+      "learning_rate": 5.290749310731693e-06,
+      "loss": 0.4259,
+      "step": 11178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6064077450198095,
+      "learning_rate": 5.290098924929719e-06,
+      "loss": 0.4798,
+      "step": 11179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.835412027130373,
+      "learning_rate": 5.289448543521891e-06,
+      "loss": 0.5786,
+      "step": 11180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.616920066922692,
+      "learning_rate": 5.2887981665218295e-06,
+      "loss": 0.5304,
+      "step": 11181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0520741033084393,
+      "learning_rate": 5.288147793943146e-06,
+      "loss": 0.5209,
+      "step": 11182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0517857523827607,
+      "learning_rate": 5.287497425799456e-06,
+      "loss": 0.5558,
+      "step": 11183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5758952557636594,
+      "learning_rate": 5.286847062104378e-06,
+      "loss": 0.5167,
+      "step": 11184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8821617979436822,
+      "learning_rate": 5.286196702871522e-06,
+      "loss": 0.6179,
+      "step": 11185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.59510933655286,
+      "learning_rate": 5.285546348114505e-06,
+      "loss": 0.4055,
+      "step": 11186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1431431561863907,
+      "learning_rate": 5.284895997846944e-06,
+      "loss": 0.5663,
+      "step": 11187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6529490127918969,
+      "learning_rate": 5.28424565208245e-06,
+      "loss": 0.334,
+      "step": 11188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7644089628525295,
+      "learning_rate": 5.283595310834641e-06,
+      "loss": 0.48,
+      "step": 11189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.948580501705175,
+      "learning_rate": 5.282944974117128e-06,
+      "loss": 0.616,
+      "step": 11190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9153874553698975,
+      "learning_rate": 5.2822946419435275e-06,
+      "loss": 0.6841,
+      "step": 11191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6915027265368368,
+      "learning_rate": 5.2816443143274545e-06,
+      "loss": 0.3908,
+      "step": 11192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6798013350297494,
+      "learning_rate": 5.28099399128252e-06,
+      "loss": 0.5451,
+      "step": 11193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9182410860740466,
+      "learning_rate": 5.280343672822342e-06,
+      "loss": 0.6287,
+      "step": 11194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6561826324096338,
+      "learning_rate": 5.279693358960534e-06,
+      "loss": 0.5536,
+      "step": 11195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9894797100291084,
+      "learning_rate": 5.279043049710706e-06,
+      "loss": 0.657,
+      "step": 11196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7978643141709418,
+      "learning_rate": 5.278392745086476e-06,
+      "loss": 0.5639,
+      "step": 11197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.712765479583182,
+      "learning_rate": 5.277742445101456e-06,
+      "loss": 0.6146,
+      "step": 11198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.633431799707683,
+      "learning_rate": 5.277092149769259e-06,
+      "loss": 0.4958,
+      "step": 11199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2232742616649386,
+      "learning_rate": 5.2764418591035e-06,
+      "loss": 0.302,
+      "step": 11200
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5204258561134338,
+      "eval_runtime": 68.0106,
+      "eval_samples_per_second": 50.139,
+      "eval_steps_per_second": 1.573,
+      "step": 11200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6311854593592547,
+      "learning_rate": 5.275791573117792e-06,
+      "loss": 0.46,
+      "step": 11201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6709140518827317,
+      "learning_rate": 5.275141291825746e-06,
+      "loss": 0.6865,
+      "step": 11202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6706018191350567,
+      "learning_rate": 5.27449101524098e-06,
+      "loss": 0.5675,
+      "step": 11203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.585688172525211,
+      "learning_rate": 5.2738407433771045e-06,
+      "loss": 0.3999,
+      "step": 11204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.398067893608156,
+      "learning_rate": 5.273190476247731e-06,
+      "loss": 0.3269,
+      "step": 11205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6944029629992265,
+      "learning_rate": 5.272540213866476e-06,
+      "loss": 0.485,
+      "step": 11206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8133152882394852,
+      "learning_rate": 5.27188995624695e-06,
+      "loss": 0.5169,
+      "step": 11207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.850650654709064,
+      "learning_rate": 5.271239703402765e-06,
+      "loss": 0.6171,
+      "step": 11208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.969665169853558,
+      "learning_rate": 5.2705894553475355e-06,
+      "loss": 0.437,
+      "step": 11209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6418214844231909,
+      "learning_rate": 5.2699392120948724e-06,
+      "loss": 0.4927,
+      "step": 11210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4290103635106046,
+      "learning_rate": 5.26928897365839e-06,
+      "loss": 0.4171,
+      "step": 11211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6792058443560554,
+      "learning_rate": 5.268638740051699e-06,
+      "loss": 0.6348,
+      "step": 11212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.039239112137097,
+      "learning_rate": 5.267988511288412e-06,
+      "loss": 0.6974,
+      "step": 11213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5609387889619357,
+      "learning_rate": 5.267338287382142e-06,
+      "loss": 0.4981,
+      "step": 11214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7044186441004268,
+      "learning_rate": 5.2666880683465e-06,
+      "loss": 0.6036,
+      "step": 11215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6340327670383148,
+      "learning_rate": 5.2660378541951e-06,
+      "loss": 0.5095,
+      "step": 11216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.604827842851227,
+      "learning_rate": 5.265387644941551e-06,
+      "loss": 0.5063,
+      "step": 11217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.436758640246844,
+      "learning_rate": 5.264737440599466e-06,
+      "loss": 0.4977,
+      "step": 11218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8044582390373607,
+      "learning_rate": 5.264087241182456e-06,
+      "loss": 0.4559,
+      "step": 11219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.640277983742384,
+      "learning_rate": 5.263437046704133e-06,
+      "loss": 0.5026,
+      "step": 11220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5224510965299354,
+      "learning_rate": 5.26278685717811e-06,
+      "loss": 0.5315,
+      "step": 11221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2427731725629347,
+      "learning_rate": 5.262136672617995e-06,
+      "loss": 0.5696,
+      "step": 11222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8012948824774169,
+      "learning_rate": 5.261486493037401e-06,
+      "loss": 0.4933,
+      "step": 11223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.675112961543166,
+      "learning_rate": 5.260836318449941e-06,
+      "loss": 0.5014,
+      "step": 11224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9359921593289195,
+      "learning_rate": 5.260186148869223e-06,
+      "loss": 0.4461,
+      "step": 11225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8703056415540615,
+      "learning_rate": 5.259535984308859e-06,
+      "loss": 0.463,
+      "step": 11226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.987721624965159,
+      "learning_rate": 5.25888582478246e-06,
+      "loss": 0.6052,
+      "step": 11227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6196461931087345,
+      "learning_rate": 5.258235670303637e-06,
+      "loss": 0.6727,
+      "step": 11228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7079359904044484,
+      "learning_rate": 5.2575855208859984e-06,
+      "loss": 0.4562,
+      "step": 11229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5925350186550646,
+      "learning_rate": 5.256935376543158e-06,
+      "loss": 0.3904,
+      "step": 11230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5483249208222079,
+      "learning_rate": 5.256285237288723e-06,
+      "loss": 0.4508,
+      "step": 11231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9744160074959292,
+      "learning_rate": 5.255635103136307e-06,
+      "loss": 0.5821,
+      "step": 11232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6725824985976872,
+      "learning_rate": 5.2549849740995175e-06,
+      "loss": 0.4381,
+      "step": 11233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0012744956178437,
+      "learning_rate": 5.254334850191965e-06,
+      "loss": 0.5645,
+      "step": 11234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.947951266623372,
+      "learning_rate": 5.253684731427261e-06,
+      "loss": 0.7423,
+      "step": 11235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8082182012212349,
+      "learning_rate": 5.2530346178190114e-06,
+      "loss": 0.4654,
+      "step": 11236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7492091008310817,
+      "learning_rate": 5.252384509380832e-06,
+      "loss": 0.4938,
+      "step": 11237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0567620116954797,
+      "learning_rate": 5.251734406126328e-06,
+      "loss": 0.5312,
+      "step": 11238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8542194268326715,
+      "learning_rate": 5.251084308069108e-06,
+      "loss": 0.5622,
+      "step": 11239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4414820093650775,
+      "learning_rate": 5.2504342152227865e-06,
+      "loss": 0.451,
+      "step": 11240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.526339369549197,
+      "learning_rate": 5.249784127600967e-06,
+      "loss": 0.5249,
+      "step": 11241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2068570579014457,
+      "learning_rate": 5.249134045217263e-06,
+      "loss": 0.5897,
+      "step": 11242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6379329303543906,
+      "learning_rate": 5.248483968085283e-06,
+      "loss": 0.4115,
+      "step": 11243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6059488100596067,
+      "learning_rate": 5.247833896218633e-06,
+      "loss": 0.5055,
+      "step": 11244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6918037549914748,
+      "learning_rate": 5.247183829630924e-06,
+      "loss": 0.5121,
+      "step": 11245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9711350460227721,
+      "learning_rate": 5.246533768335764e-06,
+      "loss": 0.478,
+      "step": 11246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7083838907920557,
+      "learning_rate": 5.245883712346763e-06,
+      "loss": 0.529,
+      "step": 11247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6307852338556106,
+      "learning_rate": 5.245233661677529e-06,
+      "loss": 0.4032,
+      "step": 11248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9982976633276213,
+      "learning_rate": 5.24458361634167e-06,
+      "loss": 0.5719,
+      "step": 11249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5958260129344684,
+      "learning_rate": 5.243933576352793e-06,
+      "loss": 0.5347,
+      "step": 11250
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5205143690109253,
+      "eval_runtime": 67.9524,
+      "eval_samples_per_second": 50.182,
+      "eval_steps_per_second": 1.575,
+      "step": 11250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0003556967297422,
+      "learning_rate": 5.243283541724508e-06,
+      "loss": 0.6551,
+      "step": 11251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8607556326795736,
+      "learning_rate": 5.2426335124704235e-06,
+      "loss": 0.4178,
+      "step": 11252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.896108064564834,
+      "learning_rate": 5.241983488604146e-06,
+      "loss": 0.5036,
+      "step": 11253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9203557369926552,
+      "learning_rate": 5.241333470139283e-06,
+      "loss": 0.5404,
+      "step": 11254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8837645912316772,
+      "learning_rate": 5.240683457089444e-06,
+      "loss": 0.6764,
+      "step": 11255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5859079427412879,
+      "learning_rate": 5.240033449468236e-06,
+      "loss": 0.6237,
+      "step": 11256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4064530850873174,
+      "learning_rate": 5.239383447289267e-06,
+      "loss": 0.4359,
+      "step": 11257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1002423670965906,
+      "learning_rate": 5.238733450566141e-06,
+      "loss": 0.5755,
+      "step": 11258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9516584298862314,
+      "learning_rate": 5.2380834593124706e-06,
+      "loss": 0.571,
+      "step": 11259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9975446420074863,
+      "learning_rate": 5.2374334735418595e-06,
+      "loss": 0.6263,
+      "step": 11260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8112325810382555,
+      "learning_rate": 5.236783493267916e-06,
+      "loss": 0.5095,
+      "step": 11261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8158089541377915,
+      "learning_rate": 5.236133518504246e-06,
+      "loss": 0.5421,
+      "step": 11262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9248453447503306,
+      "learning_rate": 5.235483549264456e-06,
+      "loss": 0.624,
+      "step": 11263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9184315748958571,
+      "learning_rate": 5.234833585562156e-06,
+      "loss": 0.6241,
+      "step": 11264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9581646541191238,
+      "learning_rate": 5.234183627410949e-06,
+      "loss": 0.5467,
+      "step": 11265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6948356525974082,
+      "learning_rate": 5.233533674824442e-06,
+      "loss": 0.5476,
+      "step": 11266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9207828183819748,
+      "learning_rate": 5.232883727816243e-06,
+      "loss": 0.4364,
+      "step": 11267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8118999878844058,
+      "learning_rate": 5.232233786399955e-06,
+      "loss": 0.4174,
+      "step": 11268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6310594829159324,
+      "learning_rate": 5.231583850589189e-06,
+      "loss": 0.4677,
+      "step": 11269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9640322250625313,
+      "learning_rate": 5.2309339203975485e-06,
+      "loss": 0.5381,
+      "step": 11270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3295253551391206,
+      "learning_rate": 5.230283995838637e-06,
+      "loss": 0.5725,
+      "step": 11271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.548036913501766,
+      "learning_rate": 5.229634076926065e-06,
+      "loss": 0.5705,
+      "step": 11272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0788383371548167,
+      "learning_rate": 5.228984163673433e-06,
+      "loss": 0.5817,
+      "step": 11273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6999976703155695,
+      "learning_rate": 5.22833425609435e-06,
+      "loss": 0.4672,
+      "step": 11274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8378106287965268,
+      "learning_rate": 5.227684354202423e-06,
+      "loss": 0.5035,
+      "step": 11275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6291730441102439,
+      "learning_rate": 5.227034458011252e-06,
+      "loss": 0.4864,
+      "step": 11276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1290711194074152,
+      "learning_rate": 5.226384567534446e-06,
+      "loss": 0.6309,
+      "step": 11277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7486595256642912,
+      "learning_rate": 5.22573468278561e-06,
+      "loss": 0.4324,
+      "step": 11278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6282104841911789,
+      "learning_rate": 5.225084803778346e-06,
+      "loss": 0.4339,
+      "step": 11279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0295366233623975,
+      "learning_rate": 5.224434930526263e-06,
+      "loss": 0.4992,
+      "step": 11280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.789153763973368,
+      "learning_rate": 5.223785063042962e-06,
+      "loss": 0.5627,
+      "step": 11281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.877105640353868,
+      "learning_rate": 5.2231352013420515e-06,
+      "loss": 0.4448,
+      "step": 11282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.797271304993683,
+      "learning_rate": 5.222485345437132e-06,
+      "loss": 0.5875,
+      "step": 11283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8774755226518136,
+      "learning_rate": 5.221835495341809e-06,
+      "loss": 0.7009,
+      "step": 11284
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6652002598132902,
+      "learning_rate": 5.221185651069688e-06,
+      "loss": 0.3445,
+      "step": 11285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6540917838823863,
+      "learning_rate": 5.220535812634371e-06,
+      "loss": 0.4173,
+      "step": 11286
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8937312759153044,
+      "learning_rate": 5.2198859800494635e-06,
+      "loss": 0.602,
+      "step": 11287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4887920078474477,
+      "learning_rate": 5.219236153328571e-06,
+      "loss": 0.4383,
+      "step": 11288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.518578373695226,
+      "learning_rate": 5.2185863324852926e-06,
+      "loss": 0.2935,
+      "step": 11289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.590555742402881,
+      "learning_rate": 5.217936517533237e-06,
+      "loss": 0.4265,
+      "step": 11290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4732519470039431,
+      "learning_rate": 5.2172867084860034e-06,
+      "loss": 0.4592,
+      "step": 11291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9861471102954318,
+      "learning_rate": 5.216636905357196e-06,
+      "loss": 0.5438,
+      "step": 11292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.041846981300252,
+      "learning_rate": 5.215987108160422e-06,
+      "loss": 0.5307,
+      "step": 11293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.89544022283746,
+      "learning_rate": 5.215337316909279e-06,
+      "loss": 0.4118,
+      "step": 11294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8639084532894463,
+      "learning_rate": 5.214687531617374e-06,
+      "loss": 0.6233,
+      "step": 11295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.731561877320037,
+      "learning_rate": 5.214037752298307e-06,
+      "loss": 0.4146,
+      "step": 11296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7909090064531523,
+      "learning_rate": 5.213387978965683e-06,
+      "loss": 0.5508,
+      "step": 11297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8046371572488746,
+      "learning_rate": 5.212738211633104e-06,
+      "loss": 0.5142,
+      "step": 11298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7573768146987223,
+      "learning_rate": 5.212088450314171e-06,
+      "loss": 0.5089,
+      "step": 11299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.800311478820836,
+      "learning_rate": 5.211438695022486e-06,
+      "loss": 0.6059,
+      "step": 11300
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5203261971473694,
+      "eval_runtime": 67.9586,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.574,
+      "step": 11300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.511526810487472,
+      "learning_rate": 5.210788945771654e-06,
+      "loss": 0.4122,
+      "step": 11301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.037687057865894,
+      "learning_rate": 5.2101392025752785e-06,
+      "loss": 0.5569,
+      "step": 11302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.735149007125267,
+      "learning_rate": 5.209489465446956e-06,
+      "loss": 0.4374,
+      "step": 11303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3085673243310367,
+      "learning_rate": 5.208839734400292e-06,
+      "loss": 0.4315,
+      "step": 11304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.972868538596752,
+      "learning_rate": 5.208190009448887e-06,
+      "loss": 0.5522,
+      "step": 11305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0837810297248316,
+      "learning_rate": 5.2075402906063425e-06,
+      "loss": 0.5039,
+      "step": 11306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8070592828911023,
+      "learning_rate": 5.206890577886261e-06,
+      "loss": 0.4376,
+      "step": 11307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8697579269593398,
+      "learning_rate": 5.2062408713022425e-06,
+      "loss": 0.5735,
+      "step": 11308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0854562794522136,
+      "learning_rate": 5.20559117086789e-06,
+      "loss": 0.5779,
+      "step": 11309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7234593546947516,
+      "learning_rate": 5.204941476596803e-06,
+      "loss": 0.4323,
+      "step": 11310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7112294022388688,
+      "learning_rate": 5.2042917885025825e-06,
+      "loss": 0.504,
+      "step": 11311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4815137101990123,
+      "learning_rate": 5.203642106598831e-06,
+      "loss": 0.4214,
+      "step": 11312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7061423431617055,
+      "learning_rate": 5.202992430899146e-06,
+      "loss": 0.6632,
+      "step": 11313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3503309649792894,
+      "learning_rate": 5.20234276141713e-06,
+      "loss": 0.4343,
+      "step": 11314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.753249262470768,
+      "learning_rate": 5.201693098166386e-06,
+      "loss": 0.5456,
+      "step": 11315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7904197884614026,
+      "learning_rate": 5.2010434411605095e-06,
+      "loss": 0.5309,
+      "step": 11316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9124122195490516,
+      "learning_rate": 5.200393790413104e-06,
+      "loss": 0.6791,
+      "step": 11317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8503571508844425,
+      "learning_rate": 5.199744145937769e-06,
+      "loss": 0.5444,
+      "step": 11318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.876157424395559,
+      "learning_rate": 5.199094507748102e-06,
+      "loss": 0.4825,
+      "step": 11319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.637414855084788,
+      "learning_rate": 5.198444875857704e-06,
+      "loss": 0.6055,
+      "step": 11320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.828856719093492,
+      "learning_rate": 5.197795250280176e-06,
+      "loss": 0.4836,
+      "step": 11321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7192620434391355,
+      "learning_rate": 5.197145631029118e-06,
+      "loss": 0.4381,
+      "step": 11322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2273224804346987,
+      "learning_rate": 5.196496018118128e-06,
+      "loss": 0.7682,
+      "step": 11323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9732134893413966,
+      "learning_rate": 5.195846411560803e-06,
+      "loss": 0.5298,
+      "step": 11324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8136053315755216,
+      "learning_rate": 5.195196811370747e-06,
+      "loss": 0.613,
+      "step": 11325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7673691687822086,
+      "learning_rate": 5.194547217561555e-06,
+      "loss": 0.5899,
+      "step": 11326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.83564041772088,
+      "learning_rate": 5.193897630146828e-06,
+      "loss": 0.6026,
+      "step": 11327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7204776080248443,
+      "learning_rate": 5.193248049140165e-06,
+      "loss": 0.5008,
+      "step": 11328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6493202505707656,
+      "learning_rate": 5.192598474555162e-06,
+      "loss": 0.5503,
+      "step": 11329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9666084669351933,
+      "learning_rate": 5.191948906405418e-06,
+      "loss": 0.5342,
+      "step": 11330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.545177612761598,
+      "learning_rate": 5.191299344704536e-06,
+      "loss": 0.5225,
+      "step": 11331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9028332528211855,
+      "learning_rate": 5.190649789466107e-06,
+      "loss": 0.6043,
+      "step": 11332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.558102420650366,
+      "learning_rate": 5.190000240703734e-06,
+      "loss": 0.4064,
+      "step": 11333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5165851886336912,
+      "learning_rate": 5.189350698431013e-06,
+      "loss": 0.4312,
+      "step": 11334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.827366044783293,
+      "learning_rate": 5.188701162661543e-06,
+      "loss": 0.556,
+      "step": 11335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4087030813950994,
+      "learning_rate": 5.188051633408919e-06,
+      "loss": 0.3628,
+      "step": 11336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.730998335276022,
+      "learning_rate": 5.187402110686741e-06,
+      "loss": 0.5001,
+      "step": 11337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9937360921699632,
+      "learning_rate": 5.1867525945086075e-06,
+      "loss": 0.6908,
+      "step": 11338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6360247041996803,
+      "learning_rate": 5.1861030848881115e-06,
+      "loss": 0.4114,
+      "step": 11339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.349459473020302,
+      "learning_rate": 5.185453581838853e-06,
+      "loss": 0.5645,
+      "step": 11340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4866044056204837,
+      "learning_rate": 5.1848040853744294e-06,
+      "loss": 0.7099,
+      "step": 11341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4399791114250013,
+      "learning_rate": 5.184154595508435e-06,
+      "loss": 0.4186,
+      "step": 11342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1371655428485377,
+      "learning_rate": 5.1835051122544685e-06,
+      "loss": 0.632,
+      "step": 11343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8200804407940876,
+      "learning_rate": 5.182855635626128e-06,
+      "loss": 0.4912,
+      "step": 11344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.59122481091175,
+      "learning_rate": 5.182206165637005e-06,
+      "loss": 0.4711,
+      "step": 11345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.503751893880375,
+      "learning_rate": 5.1815567023007e-06,
+      "loss": 0.5617,
+      "step": 11346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.045605676856927,
+      "learning_rate": 5.180907245630806e-06,
+      "loss": 0.7698,
+      "step": 11347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4675699487855534,
+      "learning_rate": 5.1802577956409215e-06,
+      "loss": 0.3859,
+      "step": 11348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.094235276537224,
+      "learning_rate": 5.179608352344641e-06,
+      "loss": 0.6545,
+      "step": 11349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7732560878952324,
+      "learning_rate": 5.178958915755562e-06,
+      "loss": 0.5404,
+      "step": 11350
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5199688673019409,
+      "eval_runtime": 67.997,
+      "eval_samples_per_second": 50.149,
+      "eval_steps_per_second": 1.574,
+      "step": 11350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9578807948687103,
+      "learning_rate": 5.178309485887278e-06,
+      "loss": 0.6648,
+      "step": 11351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4736168901453706,
+      "learning_rate": 5.177660062753385e-06,
+      "loss": 0.4021,
+      "step": 11352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6939300822477563,
+      "learning_rate": 5.177010646367477e-06,
+      "loss": 0.4055,
+      "step": 11353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4482826939066247,
+      "learning_rate": 5.1763612367431525e-06,
+      "loss": 0.44,
+      "step": 11354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.963922779123796,
+      "learning_rate": 5.175711833894005e-06,
+      "loss": 0.4523,
+      "step": 11355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5310655555396757,
+      "learning_rate": 5.175062437833627e-06,
+      "loss": 0.4389,
+      "step": 11356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7927273328383881,
+      "learning_rate": 5.174413048575616e-06,
+      "loss": 0.4152,
+      "step": 11357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8113969005039963,
+      "learning_rate": 5.173763666133566e-06,
+      "loss": 0.5352,
+      "step": 11358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4816487316472005,
+      "learning_rate": 5.17311429052107e-06,
+      "loss": 0.5107,
+      "step": 11359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.829017533506114,
+      "learning_rate": 5.172464921751724e-06,
+      "loss": 0.4488,
+      "step": 11360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.031587637541732,
+      "learning_rate": 5.17181555983912e-06,
+      "loss": 0.5744,
+      "step": 11361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7361945955269014,
+      "learning_rate": 5.171166204796856e-06,
+      "loss": 0.5616,
+      "step": 11362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7081144112356874,
+      "learning_rate": 5.170516856638522e-06,
+      "loss": 0.4509,
+      "step": 11363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3880337075936984,
+      "learning_rate": 5.169867515377713e-06,
+      "loss": 0.3386,
+      "step": 11364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.353058972198095,
+      "learning_rate": 5.169218181028022e-06,
+      "loss": 0.3287,
+      "step": 11365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.639199235363635,
+      "learning_rate": 5.168568853603042e-06,
+      "loss": 0.4866,
+      "step": 11366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1815313064707755,
+      "learning_rate": 5.167919533116369e-06,
+      "loss": 0.5036,
+      "step": 11367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9092845480211962,
+      "learning_rate": 5.167270219581595e-06,
+      "loss": 0.5471,
+      "step": 11368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.061523268955708,
+      "learning_rate": 5.166620913012309e-06,
+      "loss": 0.6963,
+      "step": 11369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6234279057359327,
+      "learning_rate": 5.165971613422111e-06,
+      "loss": 0.5082,
+      "step": 11370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5792810292090216,
+      "learning_rate": 5.165322320824587e-06,
+      "loss": 0.3689,
+      "step": 11371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5993096418304977,
+      "learning_rate": 5.164673035233334e-06,
+      "loss": 0.5643,
+      "step": 11372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.066339415618535,
+      "learning_rate": 5.164023756661942e-06,
+      "loss": 0.5927,
+      "step": 11373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7894382180932669,
+      "learning_rate": 5.163374485124003e-06,
+      "loss": 0.6703,
+      "step": 11374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6955755798777885,
+      "learning_rate": 5.1627252206331115e-06,
+      "loss": 0.5381,
+      "step": 11375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5889571351216996,
+      "learning_rate": 5.162075963202859e-06,
+      "loss": 0.4644,
+      "step": 11376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6330161547697364,
+      "learning_rate": 5.161426712846833e-06,
+      "loss": 0.5154,
+      "step": 11377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.674669890693562,
+      "learning_rate": 5.16077746957863e-06,
+      "loss": 0.6554,
+      "step": 11378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4995823055835695,
+      "learning_rate": 5.160128233411841e-06,
+      "loss": 0.428,
+      "step": 11379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7822984601678131,
+      "learning_rate": 5.159479004360056e-06,
+      "loss": 0.5421,
+      "step": 11380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.678916610246368,
+      "learning_rate": 5.158829782436867e-06,
+      "loss": 0.4643,
+      "step": 11381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.728585210322077,
+      "learning_rate": 5.158180567655863e-06,
+      "loss": 0.5398,
+      "step": 11382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.98350663381574,
+      "learning_rate": 5.157531360030638e-06,
+      "loss": 0.5736,
+      "step": 11383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6168731119152704,
+      "learning_rate": 5.156882159574781e-06,
+      "loss": 0.4482,
+      "step": 11384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8950779008680247,
+      "learning_rate": 5.156232966301882e-06,
+      "loss": 0.4101,
+      "step": 11385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7978086894809777,
+      "learning_rate": 5.155583780225533e-06,
+      "loss": 0.5666,
+      "step": 11386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7537032270127628,
+      "learning_rate": 5.154934601359322e-06,
+      "loss": 0.4809,
+      "step": 11387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8781555063953717,
+      "learning_rate": 5.1542854297168435e-06,
+      "loss": 0.4573,
+      "step": 11388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5980664606950605,
+      "learning_rate": 5.153636265311683e-06,
+      "loss": 0.4042,
+      "step": 11389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2405795228330727,
+      "learning_rate": 5.152987108157432e-06,
+      "loss": 0.578,
+      "step": 11390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.469211739762237,
+      "learning_rate": 5.152337958267682e-06,
+      "loss": 0.4217,
+      "step": 11391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6994817651128353,
+      "learning_rate": 5.1516888156560185e-06,
+      "loss": 0.5119,
+      "step": 11392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7450905812290747,
+      "learning_rate": 5.151039680336036e-06,
+      "loss": 0.5582,
+      "step": 11393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5335360748244853,
+      "learning_rate": 5.1503905523213206e-06,
+      "loss": 0.4696,
+      "step": 11394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5681177739772223,
+      "learning_rate": 5.14974143162546e-06,
+      "loss": 0.4391,
+      "step": 11395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9446753122572242,
+      "learning_rate": 5.149092318262046e-06,
+      "loss": 0.5184,
+      "step": 11396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.599994725416508,
+      "learning_rate": 5.148443212244666e-06,
+      "loss": 0.4488,
+      "step": 11397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8866692601022463,
+      "learning_rate": 5.147794113586909e-06,
+      "loss": 0.6461,
+      "step": 11398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4650448520457373,
+      "learning_rate": 5.147145022302363e-06,
+      "loss": 0.4045,
+      "step": 11399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5067828867147193,
+      "learning_rate": 5.1464959384046175e-06,
+      "loss": 0.3905,
+      "step": 11400
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.518957257270813,
+      "eval_runtime": 67.9951,
+      "eval_samples_per_second": 50.151,
+      "eval_steps_per_second": 1.574,
+      "step": 11400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.974488008926492,
+      "learning_rate": 5.14584686190726e-06,
+      "loss": 0.6813,
+      "step": 11401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.767506102817398,
+      "learning_rate": 5.145197792823878e-06,
+      "loss": 0.5443,
+      "step": 11402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7643847532734993,
+      "learning_rate": 5.144548731168062e-06,
+      "loss": 0.5313,
+      "step": 11403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6637753797639263,
+      "learning_rate": 5.143899676953394e-06,
+      "loss": 0.6184,
+      "step": 11404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7796794522516668,
+      "learning_rate": 5.1432506301934685e-06,
+      "loss": 0.5193,
+      "step": 11405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9120062755945995,
+      "learning_rate": 5.142601590901867e-06,
+      "loss": 0.4635,
+      "step": 11406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5164911947990931,
+      "learning_rate": 5.1419525590921805e-06,
+      "loss": 0.5122,
+      "step": 11407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.853948562288852,
+      "learning_rate": 5.141303534777995e-06,
+      "loss": 0.6346,
+      "step": 11408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4296546773010768,
+      "learning_rate": 5.140654517972896e-06,
+      "loss": 0.4039,
+      "step": 11409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8631404807643963,
+      "learning_rate": 5.1400055086904725e-06,
+      "loss": 0.4974,
+      "step": 11410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8764007195230652,
+      "learning_rate": 5.1393565069443095e-06,
+      "loss": 0.4658,
+      "step": 11411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.750442852140126,
+      "learning_rate": 5.138707512747993e-06,
+      "loss": 0.5216,
+      "step": 11412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6452643790935815,
+      "learning_rate": 5.138058526115113e-06,
+      "loss": 0.5518,
+      "step": 11413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6090435898965156,
+      "learning_rate": 5.137409547059251e-06,
+      "loss": 0.5897,
+      "step": 11414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6128705795145049,
+      "learning_rate": 5.136760575593996e-06,
+      "loss": 0.5102,
+      "step": 11415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8455258376992174,
+      "learning_rate": 5.136111611732932e-06,
+      "loss": 0.4676,
+      "step": 11416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7652996400523062,
+      "learning_rate": 5.135462655489645e-06,
+      "loss": 0.5164,
+      "step": 11417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6689550136290854,
+      "learning_rate": 5.134813706877722e-06,
+      "loss": 0.539,
+      "step": 11418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8472203249911192,
+      "learning_rate": 5.134164765910745e-06,
+      "loss": 0.4869,
+      "step": 11419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2308116371034306,
+      "learning_rate": 5.133515832602304e-06,
+      "loss": 0.5762,
+      "step": 11420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8416845753866349,
+      "learning_rate": 5.1328669069659805e-06,
+      "loss": 0.5554,
+      "step": 11421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7262892585100904,
+      "learning_rate": 5.13221798901536e-06,
+      "loss": 0.4788,
+      "step": 11422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0205479318130357,
+      "learning_rate": 5.131569078764028e-06,
+      "loss": 0.5826,
+      "step": 11423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5133522669882462,
+      "learning_rate": 5.130920176225569e-06,
+      "loss": 0.4763,
+      "step": 11424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.013312665702664,
+      "learning_rate": 5.130271281413565e-06,
+      "loss": 0.5041,
+      "step": 11425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4474241143314601,
+      "learning_rate": 5.129622394341604e-06,
+      "loss": 0.3855,
+      "step": 11426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8957997749067728,
+      "learning_rate": 5.1289735150232656e-06,
+      "loss": 0.5905,
+      "step": 11427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8500705590172635,
+      "learning_rate": 5.128324643472139e-06,
+      "loss": 0.4693,
+      "step": 11428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1039174425451694,
+      "learning_rate": 5.127675779701804e-06,
+      "loss": 0.5062,
+      "step": 11429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7626015675592486,
+      "learning_rate": 5.127026923725845e-06,
+      "loss": 0.4609,
+      "step": 11430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0089025598161263,
+      "learning_rate": 5.1263780755578465e-06,
+      "loss": 0.5523,
+      "step": 11431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9379676797549839,
+      "learning_rate": 5.12572923521139e-06,
+      "loss": 0.7075,
+      "step": 11432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5643920789666237,
+      "learning_rate": 5.12508040270006e-06,
+      "loss": 0.421,
+      "step": 11433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1093580856367065,
+      "learning_rate": 5.1244315780374384e-06,
+      "loss": 0.5424,
+      "step": 11434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4562625217713032,
+      "learning_rate": 5.123782761237109e-06,
+      "loss": 0.405,
+      "step": 11435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6742092750872308,
+      "learning_rate": 5.123133952312653e-06,
+      "loss": 0.4054,
+      "step": 11436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9289293209320577,
+      "learning_rate": 5.122485151277654e-06,
+      "loss": 0.6497,
+      "step": 11437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6868033295151637,
+      "learning_rate": 5.1218363581456945e-06,
+      "loss": 0.4174,
+      "step": 11438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8666129329692158,
+      "learning_rate": 5.121187572930356e-06,
+      "loss": 0.5824,
+      "step": 11439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6868999067419341,
+      "learning_rate": 5.1205387956452195e-06,
+      "loss": 0.4258,
+      "step": 11440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7773574722003802,
+      "learning_rate": 5.119890026303869e-06,
+      "loss": 0.4823,
+      "step": 11441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4695694871882,
+      "learning_rate": 5.119241264919883e-06,
+      "loss": 0.3761,
+      "step": 11442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.742030962448058,
+      "learning_rate": 5.118592511506846e-06,
+      "loss": 0.536,
+      "step": 11443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9032119761642157,
+      "learning_rate": 5.117943766078337e-06,
+      "loss": 0.5041,
+      "step": 11444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.909791492204333,
+      "learning_rate": 5.117295028647937e-06,
+      "loss": 0.5194,
+      "step": 11445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7831853553473795,
+      "learning_rate": 5.11664629922923e-06,
+      "loss": 0.5809,
+      "step": 11446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9171576243234403,
+      "learning_rate": 5.115997577835794e-06,
+      "loss": 0.4058,
+      "step": 11447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7612085260064214,
+      "learning_rate": 5.115348864481208e-06,
+      "loss": 0.5816,
+      "step": 11448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6302265328090777,
+      "learning_rate": 5.114700159179057e-06,
+      "loss": 0.469,
+      "step": 11449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4652579070040332,
+      "learning_rate": 5.114051461942919e-06,
+      "loss": 0.4026,
+      "step": 11450
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.5180988907814026,
+      "eval_runtime": 68.0098,
+      "eval_samples_per_second": 50.14,
+      "eval_steps_per_second": 1.573,
+      "step": 11450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8847041663030188,
+      "learning_rate": 5.113402772786372e-06,
+      "loss": 0.4322,
+      "step": 11451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6891727512777657,
+      "learning_rate": 5.112754091722999e-06,
+      "loss": 0.4743,
+      "step": 11452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7140736800182979,
+      "learning_rate": 5.112105418766379e-06,
+      "loss": 0.5237,
+      "step": 11453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8999382598754524,
+      "learning_rate": 5.111456753930091e-06,
+      "loss": 0.4928,
+      "step": 11454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3235492873361405,
+      "learning_rate": 5.110808097227714e-06,
+      "loss": 0.419,
+      "step": 11455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6036272002740628,
+      "learning_rate": 5.110159448672829e-06,
+      "loss": 0.4611,
+      "step": 11456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7426138748022708,
+      "learning_rate": 5.109510808279011e-06,
+      "loss": 0.6715,
+      "step": 11457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5942902085155344,
+      "learning_rate": 5.108862176059844e-06,
+      "loss": 0.458,
+      "step": 11458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5730991635727616,
+      "learning_rate": 5.108213552028903e-06,
+      "loss": 0.5004,
+      "step": 11459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9446109810334664,
+      "learning_rate": 5.10756493619977e-06,
+      "loss": 0.5831,
+      "step": 11460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.791284677388978,
+      "learning_rate": 5.1069163285860216e-06,
+      "loss": 0.4652,
+      "step": 11461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.939216786886762,
+      "learning_rate": 5.106267729201233e-06,
+      "loss": 0.5038,
+      "step": 11462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9657160070203878,
+      "learning_rate": 5.105619138058987e-06,
+      "loss": 0.7703,
+      "step": 11463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.231809009862881,
+      "learning_rate": 5.104970555172859e-06,
+      "loss": 0.4975,
+      "step": 11464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6402688049868357,
+      "learning_rate": 5.104321980556426e-06,
+      "loss": 0.4558,
+      "step": 11465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.498994847335367,
+      "learning_rate": 5.103673414223267e-06,
+      "loss": 0.4526,
+      "step": 11466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7807280635376195,
+      "learning_rate": 5.103024856186959e-06,
+      "loss": 0.5026,
+      "step": 11467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7461900605841034,
+      "learning_rate": 5.10237630646108e-06,
+      "loss": 0.5004,
+      "step": 11468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.368966087423844,
+      "learning_rate": 5.101727765059205e-06,
+      "loss": 0.3888,
+      "step": 11469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7458582282036799,
+      "learning_rate": 5.101079231994912e-06,
+      "loss": 0.4493,
+      "step": 11470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.975865563133833,
+      "learning_rate": 5.100430707281778e-06,
+      "loss": 0.66,
+      "step": 11471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0345111409990504,
+      "learning_rate": 5.099782190933378e-06,
+      "loss": 0.6021,
+      "step": 11472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6956162778377388,
+      "learning_rate": 5.09913368296329e-06,
+      "loss": 0.552,
+      "step": 11473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5263220150509067,
+      "learning_rate": 5.0984851833850905e-06,
+      "loss": 0.4176,
+      "step": 11474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.93912300481932,
+      "learning_rate": 5.097836692212351e-06,
+      "loss": 0.5675,
+      "step": 11475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.028028221608822,
+      "learning_rate": 5.097188209458653e-06,
+      "loss": 0.6059,
+      "step": 11476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8636235643097268,
+      "learning_rate": 5.096539735137569e-06,
+      "loss": 0.5111,
+      "step": 11477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6683581301545856,
+      "learning_rate": 5.095891269262675e-06,
+      "loss": 0.4837,
+      "step": 11478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8344741629193924,
+      "learning_rate": 5.0952428118475474e-06,
+      "loss": 0.461,
+      "step": 11479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3830221199097164,
+      "learning_rate": 5.0945943629057586e-06,
+      "loss": 0.4443,
+      "step": 11480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0292665592543546,
+      "learning_rate": 5.093945922450887e-06,
+      "loss": 0.8571,
+      "step": 11481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.513914824312418,
+      "learning_rate": 5.093297490496504e-06,
+      "loss": 0.4657,
+      "step": 11482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3363729063150296,
+      "learning_rate": 5.092649067056185e-06,
+      "loss": 0.3839,
+      "step": 11483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6214501901199863,
+      "learning_rate": 5.092000652143505e-06,
+      "loss": 0.4971,
+      "step": 11484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8896072641211705,
+      "learning_rate": 5.091352245772038e-06,
+      "loss": 0.5337,
+      "step": 11485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.733819784775642,
+      "learning_rate": 5.090703847955359e-06,
+      "loss": 0.5535,
+      "step": 11486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6549818577863664,
+      "learning_rate": 5.090055458707039e-06,
+      "loss": 0.5173,
+      "step": 11487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8442439725202149,
+      "learning_rate": 5.089407078040654e-06,
+      "loss": 0.4342,
+      "step": 11488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8713328232330841,
+      "learning_rate": 5.088758705969776e-06,
+      "loss": 0.5106,
+      "step": 11489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.206111257166057,
+      "learning_rate": 5.08811034250798e-06,
+      "loss": 0.4907,
+      "step": 11490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.040881356808639,
+      "learning_rate": 5.087461987668838e-06,
+      "loss": 0.6404,
+      "step": 11491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6199496793808539,
+      "learning_rate": 5.086813641465924e-06,
+      "loss": 0.5487,
+      "step": 11492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8235854550125121,
+      "learning_rate": 5.086165303912808e-06,
+      "loss": 0.551,
+      "step": 11493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6663843075202194,
+      "learning_rate": 5.085516975023066e-06,
+      "loss": 0.4902,
+      "step": 11494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.503704732780273,
+      "learning_rate": 5.084868654810268e-06,
+      "loss": 0.4055,
+      "step": 11495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.676735789074945,
+      "learning_rate": 5.084220343287986e-06,
+      "loss": 0.4908,
+      "step": 11496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453249093868845,
+      "learning_rate": 5.083572040469792e-06,
+      "loss": 0.651,
+      "step": 11497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6357560243975393,
+      "learning_rate": 5.08292374636926e-06,
+      "loss": 0.5222,
+      "step": 11498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7803033053310449,
+      "learning_rate": 5.082275460999961e-06,
+      "loss": 0.3655,
+      "step": 11499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2807421906115795,
+      "learning_rate": 5.081627184375466e-06,
+      "loss": 0.4933,
+      "step": 11500
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.5186848640441895,
+      "eval_runtime": 67.9162,
+      "eval_samples_per_second": 50.209,
+      "eval_steps_per_second": 1.575,
+      "step": 11500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8729468237615914,
+      "learning_rate": 5.080978916509346e-06,
+      "loss": 0.526,
+      "step": 11501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8541013973116547,
+      "learning_rate": 5.08033065741517e-06,
+      "loss": 0.5197,
+      "step": 11502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6370124130624428,
+      "learning_rate": 5.079682407106513e-06,
+      "loss": 0.4902,
+      "step": 11503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8109076720271626,
+      "learning_rate": 5.079034165596944e-06,
+      "loss": 0.4092,
+      "step": 11504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.640421827215906,
+      "learning_rate": 5.0783859329000315e-06,
+      "loss": 0.4646,
+      "step": 11505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.750200246774802,
+      "learning_rate": 5.07773770902935e-06,
+      "loss": 0.5914,
+      "step": 11506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6551881000608926,
+      "learning_rate": 5.077089493998465e-06,
+      "loss": 0.5188,
+      "step": 11507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7376067444663534,
+      "learning_rate": 5.07644128782095e-06,
+      "loss": 0.6377,
+      "step": 11508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.29603575253447,
+      "learning_rate": 5.075793090510374e-06,
+      "loss": 0.6364,
+      "step": 11509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8508935116583227,
+      "learning_rate": 5.075144902080304e-06,
+      "loss": 0.4901,
+      "step": 11510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9366668222669794,
+      "learning_rate": 5.0744967225443144e-06,
+      "loss": 0.5377,
+      "step": 11511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7231507619466118,
+      "learning_rate": 5.07384855191597e-06,
+      "loss": 0.4442,
+      "step": 11512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6686330753928265,
+      "learning_rate": 5.073200390208841e-06,
+      "loss": 0.4289,
+      "step": 11513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6968723845198264,
+      "learning_rate": 5.072552237436499e-06,
+      "loss": 0.5399,
+      "step": 11514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.003533688572227,
+      "learning_rate": 5.071904093612507e-06,
+      "loss": 0.4141,
+      "step": 11515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9208337641514956,
+      "learning_rate": 5.07125595875044e-06,
+      "loss": 0.4463,
+      "step": 11516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.518300705308072,
+      "learning_rate": 5.070607832863861e-06,
+      "loss": 0.4444,
+      "step": 11517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7651550572678725,
+      "learning_rate": 5.069959715966342e-06,
+      "loss": 0.5093,
+      "step": 11518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6706814249463837,
+      "learning_rate": 5.069311608071449e-06,
+      "loss": 0.5288,
+      "step": 11519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3933696807442493,
+      "learning_rate": 5.068663509192749e-06,
+      "loss": 0.5703,
+      "step": 11520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3574042879429964,
+      "learning_rate": 5.068015419343811e-06,
+      "loss": 0.5333,
+      "step": 11521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9324692147422744,
+      "learning_rate": 5.067367338538203e-06,
+      "loss": 0.7002,
+      "step": 11522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5777105146408823,
+      "learning_rate": 5.066719266789489e-06,
+      "loss": 0.4475,
+      "step": 11523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1354722170931875,
+      "learning_rate": 5.066071204111239e-06,
+      "loss": 0.5138,
+      "step": 11524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8681437342614589,
+      "learning_rate": 5.065423150517018e-06,
+      "loss": 0.4243,
+      "step": 11525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2765515035029478,
+      "learning_rate": 5.064775106020395e-06,
+      "loss": 0.4046,
+      "step": 11526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6837556271476404,
+      "learning_rate": 5.064127070634935e-06,
+      "loss": 0.4759,
+      "step": 11527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.633692362738061,
+      "learning_rate": 5.063479044374202e-06,
+      "loss": 0.5091,
+      "step": 11528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9965625995042031,
+      "learning_rate": 5.062831027251765e-06,
+      "loss": 0.471,
+      "step": 11529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.562437886124392,
+      "learning_rate": 5.0621830192811894e-06,
+      "loss": 0.4897,
+      "step": 11530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5284918401053187,
+      "learning_rate": 5.06153502047604e-06,
+      "loss": 0.4686,
+      "step": 11531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6901707303401616,
+      "learning_rate": 5.060887030849884e-06,
+      "loss": 0.3479,
+      "step": 11532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.521488951436481,
+      "learning_rate": 5.060239050416282e-06,
+      "loss": 0.4022,
+      "step": 11533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9805718646946016,
+      "learning_rate": 5.0595910791888055e-06,
+      "loss": 0.5557,
+      "step": 11534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7297826613294165,
+      "learning_rate": 5.058943117181016e-06,
+      "loss": 0.4252,
+      "step": 11535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.812946316648035,
+      "learning_rate": 5.058295164406476e-06,
+      "loss": 0.378,
+      "step": 11536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5834197686161635,
+      "learning_rate": 5.057647220878754e-06,
+      "loss": 0.4198,
+      "step": 11537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6780176679814163,
+      "learning_rate": 5.0569992866114114e-06,
+      "loss": 0.4861,
+      "step": 11538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6335224392466854,
+      "learning_rate": 5.056351361618015e-06,
+      "loss": 0.5325,
+      "step": 11539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8804017001361863,
+      "learning_rate": 5.055703445912128e-06,
+      "loss": 0.7206,
+      "step": 11540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.788822044811231,
+      "learning_rate": 5.055055539507311e-06,
+      "loss": 0.5293,
+      "step": 11541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5026702516630253,
+      "learning_rate": 5.054407642417131e-06,
+      "loss": 0.4046,
+      "step": 11542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9405959337908107,
+      "learning_rate": 5.053759754655151e-06,
+      "loss": 0.5411,
+      "step": 11543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.703000304913576,
+      "learning_rate": 5.053111876234933e-06,
+      "loss": 0.6116,
+      "step": 11544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5295806863664807,
+      "learning_rate": 5.05246400717004e-06,
+      "loss": 0.4745,
+      "step": 11545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9565747731587313,
+      "learning_rate": 5.051816147474033e-06,
+      "loss": 0.5327,
+      "step": 11546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015370125289531,
+      "learning_rate": 5.051168297160479e-06,
+      "loss": 0.5624,
+      "step": 11547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7545850690147993,
+      "learning_rate": 5.050520456242937e-06,
+      "loss": 0.5445,
+      "step": 11548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9171027286861848,
+      "learning_rate": 5.049872624734972e-06,
+      "loss": 0.4339,
+      "step": 11549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7234499290007577,
+      "learning_rate": 5.0492248026501425e-06,
+      "loss": 0.4176,
+      "step": 11550
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.5180816650390625,
+      "eval_runtime": 68.0062,
+      "eval_samples_per_second": 50.142,
+      "eval_steps_per_second": 1.573,
+      "step": 11550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8934275458546705,
+      "learning_rate": 5.048576990002012e-06,
+      "loss": 0.5007,
+      "step": 11551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359455407619262,
+      "learning_rate": 5.047929186804141e-06,
+      "loss": 0.6441,
+      "step": 11552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8453741212439159,
+      "learning_rate": 5.0472813930700925e-06,
+      "loss": 0.4871,
+      "step": 11553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6281678498239553,
+      "learning_rate": 5.046633608813427e-06,
+      "loss": 0.5386,
+      "step": 11554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2855439508897115,
+      "learning_rate": 5.0459858340477054e-06,
+      "loss": 0.6895,
+      "step": 11555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.351665532396386,
+      "learning_rate": 5.045338068786488e-06,
+      "loss": 0.4033,
+      "step": 11556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4169791325538101,
+      "learning_rate": 5.044690313043336e-06,
+      "loss": 0.3681,
+      "step": 11557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2357133585944737,
+      "learning_rate": 5.044042566831809e-06,
+      "loss": 0.6402,
+      "step": 11558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.606615660299667,
+      "learning_rate": 5.043394830165469e-06,
+      "loss": 0.3584,
+      "step": 11559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7165429145803504,
+      "learning_rate": 5.042747103057872e-06,
+      "loss": 0.4895,
+      "step": 11560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8393924532662087,
+      "learning_rate": 5.042099385522583e-06,
+      "loss": 0.7115,
+      "step": 11561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7178823037078133,
+      "learning_rate": 5.041451677573158e-06,
+      "loss": 0.5172,
+      "step": 11562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8905474744609336,
+      "learning_rate": 5.040803979223157e-06,
+      "loss": 0.4432,
+      "step": 11563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7552422095359819,
+      "learning_rate": 5.04015629048614e-06,
+      "loss": 0.5168,
+      "step": 11564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.156329782980664,
+      "learning_rate": 5.039508611375664e-06,
+      "loss": 0.5064,
+      "step": 11565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4962858587420895,
+      "learning_rate": 5.038860941905292e-06,
+      "loss": 0.4835,
+      "step": 11566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.718960129947219,
+      "learning_rate": 5.038213282088578e-06,
+      "loss": 0.4491,
+      "step": 11567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8356981434425137,
+      "learning_rate": 5.037565631939081e-06,
+      "loss": 0.5217,
+      "step": 11568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0840157608123495,
+      "learning_rate": 5.036917991470362e-06,
+      "loss": 0.4458,
+      "step": 11569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5132119733659493,
+      "learning_rate": 5.036270360695976e-06,
+      "loss": 0.6506,
+      "step": 11570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.01095861063208,
+      "learning_rate": 5.035622739629483e-06,
+      "loss": 0.5926,
+      "step": 11571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.877599896160818,
+      "learning_rate": 5.034975128284439e-06,
+      "loss": 0.5697,
+      "step": 11572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8702791294456478,
+      "learning_rate": 5.034327526674402e-06,
+      "loss": 0.5138,
+      "step": 11573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.551168945323145,
+      "learning_rate": 5.03367993481293e-06,
+      "loss": 0.4306,
+      "step": 11574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7512764396052412,
+      "learning_rate": 5.033032352713578e-06,
+      "loss": 0.5651,
+      "step": 11575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8975034425968862,
+      "learning_rate": 5.0323847803899025e-06,
+      "loss": 0.4436,
+      "step": 11576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8570299530553755,
+      "learning_rate": 5.031737217855464e-06,
+      "loss": 0.4574,
+      "step": 11577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.440867063168405,
+      "learning_rate": 5.031089665123813e-06,
+      "loss": 0.6186,
+      "step": 11578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9012075039776342,
+      "learning_rate": 5.03044212220851e-06,
+      "loss": 0.4648,
+      "step": 11579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7811423679002962,
+      "learning_rate": 5.029794589123111e-06,
+      "loss": 0.5458,
+      "step": 11580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9191622600650007,
+      "learning_rate": 5.029147065881167e-06,
+      "loss": 0.405,
+      "step": 11581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4886511812784433,
+      "learning_rate": 5.028499552496239e-06,
+      "loss": 0.4555,
+      "step": 11582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6611955401018843,
+      "learning_rate": 5.027852048981878e-06,
+      "loss": 0.4373,
+      "step": 11583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.110620086700642,
+      "learning_rate": 5.027204555351644e-06,
+      "loss": 0.8048,
+      "step": 11584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7972878721547518,
+      "learning_rate": 5.026557071619087e-06,
+      "loss": 0.6635,
+      "step": 11585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.888843463000319,
+      "learning_rate": 5.025909597797763e-06,
+      "loss": 0.5954,
+      "step": 11586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7304655755850327,
+      "learning_rate": 5.025262133901228e-06,
+      "loss": 0.4715,
+      "step": 11587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7374398165312879,
+      "learning_rate": 5.024614679943037e-06,
+      "loss": 0.4854,
+      "step": 11588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7431246083756091,
+      "learning_rate": 5.02396723593674e-06,
+      "loss": 0.6721,
+      "step": 11589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.567149970284704,
+      "learning_rate": 5.023319801895895e-06,
+      "loss": 0.4625,
+      "step": 11590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7225400742152024,
+      "learning_rate": 5.0226723778340524e-06,
+      "loss": 0.4898,
+      "step": 11591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7086554986474103,
+      "learning_rate": 5.022024963764768e-06,
+      "loss": 0.4769,
+      "step": 11592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.849997429211358,
+      "learning_rate": 5.021377559701594e-06,
+      "loss": 0.4706,
+      "step": 11593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.996345693788508,
+      "learning_rate": 5.020730165658083e-06,
+      "loss": 0.5601,
+      "step": 11594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9009819769404364,
+      "learning_rate": 5.020082781647789e-06,
+      "loss": 0.5391,
+      "step": 11595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6686553480262316,
+      "learning_rate": 5.019435407684261e-06,
+      "loss": 0.6167,
+      "step": 11596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4626352898750596,
+      "learning_rate": 5.018788043781057e-06,
+      "loss": 0.4187,
+      "step": 11597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8964927184125935,
+      "learning_rate": 5.018140689951726e-06,
+      "loss": 0.5027,
+      "step": 11598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5251153303849383,
+      "learning_rate": 5.017493346209821e-06,
+      "loss": 0.3899,
+      "step": 11599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7840283340265943,
+      "learning_rate": 5.016846012568892e-06,
+      "loss": 0.4425,
+      "step": 11600
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.5186576843261719,
+      "eval_runtime": 68.0155,
+      "eval_samples_per_second": 50.136,
+      "eval_steps_per_second": 1.573,
+      "step": 11600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0377618260502848,
+      "learning_rate": 5.016198689042493e-06,
+      "loss": 0.4053,
+      "step": 11601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6889178284617903,
+      "learning_rate": 5.015551375644174e-06,
+      "loss": 0.4682,
+      "step": 11602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9182636458175715,
+      "learning_rate": 5.014904072387484e-06,
+      "loss": 0.6717,
+      "step": 11603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8277628583816445,
+      "learning_rate": 5.014256779285977e-06,
+      "loss": 0.4022,
+      "step": 11604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9075460189980948,
+      "learning_rate": 5.013609496353202e-06,
+      "loss": 0.4934,
+      "step": 11605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7473815315746728,
+      "learning_rate": 5.012962223602712e-06,
+      "loss": 0.4517,
+      "step": 11606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9052737259859693,
+      "learning_rate": 5.0123149610480545e-06,
+      "loss": 0.5139,
+      "step": 11607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8615136556403107,
+      "learning_rate": 5.01166770870278e-06,
+      "loss": 0.5287,
+      "step": 11608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.434671953815195,
+      "learning_rate": 5.011020466580438e-06,
+      "loss": 0.3521,
+      "step": 11609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8918110629830531,
+      "learning_rate": 5.01037323469458e-06,
+      "loss": 0.599,
+      "step": 11610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7884153875729298,
+      "learning_rate": 5.009726013058754e-06,
+      "loss": 0.5029,
+      "step": 11611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1231867316908564,
+      "learning_rate": 5.00907880168651e-06,
+      "loss": 0.6147,
+      "step": 11612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6263684621075882,
+      "learning_rate": 5.008431600591395e-06,
+      "loss": 0.4643,
+      "step": 11613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.439640809127431,
+      "learning_rate": 5.00778440978696e-06,
+      "loss": 0.3428,
+      "step": 11614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7729974695020672,
+      "learning_rate": 5.007137229286753e-06,
+      "loss": 0.5288,
+      "step": 11615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5758937384837783,
+      "learning_rate": 5.0064900591043205e-06,
+      "loss": 0.4058,
+      "step": 11616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2515709921458327,
+      "learning_rate": 5.005842899253212e-06,
+      "loss": 0.6501,
+      "step": 11617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8582616509394954,
+      "learning_rate": 5.005195749746977e-06,
+      "loss": 0.5479,
+      "step": 11618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0226904592798927,
+      "learning_rate": 5.0045486105991615e-06,
+      "loss": 0.7187,
+      "step": 11619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9128105550232068,
+      "learning_rate": 5.003901481823314e-06,
+      "loss": 0.4232,
+      "step": 11620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.003505548804445,
+      "learning_rate": 5.003254363432978e-06,
+      "loss": 0.5417,
+      "step": 11621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5552695435066368,
+      "learning_rate": 5.002607255441705e-06,
+      "loss": 0.4479,
+      "step": 11622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8865998047744252,
+      "learning_rate": 5.001960157863039e-06,
+      "loss": 0.6055,
+      "step": 11623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4753445708592805,
+      "learning_rate": 5.001313070710529e-06,
+      "loss": 0.5743,
+      "step": 11624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2846947736636536,
+      "learning_rate": 5.000665993997721e-06,
+      "loss": 0.3872,
+      "step": 11625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8750120819228917,
+      "learning_rate": 5.000018927738158e-06,
+      "loss": 0.4317,
+      "step": 11626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4735275488310742,
+      "learning_rate": 4.99937187194539e-06,
+      "loss": 0.3968,
+      "step": 11627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7026600783171806,
+      "learning_rate": 4.998724826632962e-06,
+      "loss": 0.4458,
+      "step": 11628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5554399813586757,
+      "learning_rate": 4.9980777918144155e-06,
+      "loss": 0.3923,
+      "step": 11629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2660745966184805,
+      "learning_rate": 4.9974307675032995e-06,
+      "loss": 0.3056,
+      "step": 11630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.785316197855852,
+      "learning_rate": 4.996783753713158e-06,
+      "loss": 0.5928,
+      "step": 11631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9763704939818287,
+      "learning_rate": 4.9961367504575376e-06,
+      "loss": 0.59,
+      "step": 11632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.616992744786155,
+      "learning_rate": 4.995489757749981e-06,
+      "loss": 0.4681,
+      "step": 11633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6259656015651553,
+      "learning_rate": 4.994842775604031e-06,
+      "loss": 0.4242,
+      "step": 11634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6599994565019263,
+      "learning_rate": 4.994195804033236e-06,
+      "loss": 0.525,
+      "step": 11635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7522359812586688,
+      "learning_rate": 4.993548843051136e-06,
+      "loss": 0.6037,
+      "step": 11636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9013599653743363,
+      "learning_rate": 4.992901892671278e-06,
+      "loss": 0.6633,
+      "step": 11637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8367753829958555,
+      "learning_rate": 4.9922549529072025e-06,
+      "loss": 0.4677,
+      "step": 11638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9409073958063434,
+      "learning_rate": 4.991608023772454e-06,
+      "loss": 0.6198,
+      "step": 11639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.643413565411127,
+      "learning_rate": 4.990961105280577e-06,
+      "loss": 0.5031,
+      "step": 11640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7979471244278222,
+      "learning_rate": 4.990314197445113e-06,
+      "loss": 0.4066,
+      "step": 11641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.510156928967038,
+      "learning_rate": 4.989667300279603e-06,
+      "loss": 0.3771,
+      "step": 11642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.843405362421622,
+      "learning_rate": 4.989020413797594e-06,
+      "loss": 0.7349,
+      "step": 11643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.781272127057521,
+      "learning_rate": 4.9883735380126215e-06,
+      "loss": 0.6596,
+      "step": 11644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5946264700938202,
+      "learning_rate": 4.987726672938232e-06,
+      "loss": 0.4789,
+      "step": 11645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5951069742819375,
+      "learning_rate": 4.987079818587969e-06,
+      "loss": 0.3928,
+      "step": 11646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9205822021795056,
+      "learning_rate": 4.98643297497537e-06,
+      "loss": 0.436,
+      "step": 11647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4395552579723423,
+      "learning_rate": 4.985786142113977e-06,
+      "loss": 0.3622,
+      "step": 11648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1168400208942755,
+      "learning_rate": 4.985139320017334e-06,
+      "loss": 0.5836,
+      "step": 11649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6584691731792554,
+      "learning_rate": 4.984492508698977e-06,
+      "loss": 0.3563,
+      "step": 11650
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.5183089971542358,
+      "eval_runtime": 67.9651,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 11650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6266535653418197,
+      "learning_rate": 4.98384570817245e-06,
+      "loss": 0.4794,
+      "step": 11651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9587225035048708,
+      "learning_rate": 4.983198918451294e-06,
+      "loss": 0.4856,
+      "step": 11652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7405050843623642,
+      "learning_rate": 4.982552139549046e-06,
+      "loss": 0.5016,
+      "step": 11653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8112580939432648,
+      "learning_rate": 4.981905371479248e-06,
+      "loss": 0.5479,
+      "step": 11654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7612704693886758,
+      "learning_rate": 4.98125861425544e-06,
+      "loss": 0.5808,
+      "step": 11655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5494360855672213,
+      "learning_rate": 4.9806118678911585e-06,
+      "loss": 0.5543,
+      "step": 11656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3494031566755387,
+      "learning_rate": 4.979965132399947e-06,
+      "loss": 0.4372,
+      "step": 11657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.106159295739507,
+      "learning_rate": 4.979318407795341e-06,
+      "loss": 0.4958,
+      "step": 11658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7255247402233882,
+      "learning_rate": 4.978671694090882e-06,
+      "loss": 0.5138,
+      "step": 11659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6827944108057615,
+      "learning_rate": 4.978024991300108e-06,
+      "loss": 0.4865,
+      "step": 11660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7741273008830998,
+      "learning_rate": 4.977378299436554e-06,
+      "loss": 0.4159,
+      "step": 11661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6720794419919407,
+      "learning_rate": 4.976731618513763e-06,
+      "loss": 0.5475,
+      "step": 11662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1365308883744167,
+      "learning_rate": 4.976084948545269e-06,
+      "loss": 0.4278,
+      "step": 11663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.450145215983818,
+      "learning_rate": 4.975438289544612e-06,
+      "loss": 0.4972,
+      "step": 11664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6011149449091848,
+      "learning_rate": 4.974791641525328e-06,
+      "loss": 0.3503,
+      "step": 11665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3551938152226302,
+      "learning_rate": 4.974145004500955e-06,
+      "loss": 0.3154,
+      "step": 11666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.990417773724122,
+      "learning_rate": 4.97349837848503e-06,
+      "loss": 0.5305,
+      "step": 11667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8019745329415164,
+      "learning_rate": 4.97285176349109e-06,
+      "loss": 0.5698,
+      "step": 11668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7886741774958734,
+      "learning_rate": 4.9722051595326696e-06,
+      "loss": 0.478,
+      "step": 11669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4386983870809207,
+      "learning_rate": 4.971558566623308e-06,
+      "loss": 0.4051,
+      "step": 11670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3435467591401575,
+      "learning_rate": 4.970911984776537e-06,
+      "loss": 0.4135,
+      "step": 11671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0143627806109157,
+      "learning_rate": 4.970265414005897e-06,
+      "loss": 0.7272,
+      "step": 11672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1899445834504014,
+      "learning_rate": 4.969618854324922e-06,
+      "loss": 0.4661,
+      "step": 11673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2996639335890072,
+      "learning_rate": 4.968972305747145e-06,
+      "loss": 0.637,
+      "step": 11674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6310663156068501,
+      "learning_rate": 4.968325768286104e-06,
+      "loss": 0.4043,
+      "step": 11675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7917222350240072,
+      "learning_rate": 4.967679241955332e-06,
+      "loss": 0.5748,
+      "step": 11676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5986919853898622,
+      "learning_rate": 4.967032726768366e-06,
+      "loss": 0.49,
+      "step": 11677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7961909003915408,
+      "learning_rate": 4.966386222738738e-06,
+      "loss": 0.5378,
+      "step": 11678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5792758665119804,
+      "learning_rate": 4.965739729879983e-06,
+      "loss": 0.4191,
+      "step": 11679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.004280399001436,
+      "learning_rate": 4.965093248205634e-06,
+      "loss": 0.6199,
+      "step": 11680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9072281755911589,
+      "learning_rate": 4.964446777729227e-06,
+      "loss": 0.6879,
+      "step": 11681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7952117322331043,
+      "learning_rate": 4.963800318464294e-06,
+      "loss": 0.7137,
+      "step": 11682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6312314210719057,
+      "learning_rate": 4.963153870424368e-06,
+      "loss": 0.4659,
+      "step": 11683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.857084827284497,
+      "learning_rate": 4.9625074336229805e-06,
+      "loss": 0.4948,
+      "step": 11684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8970923573698093,
+      "learning_rate": 4.961861008073668e-06,
+      "loss": 0.6289,
+      "step": 11685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.068418034814677,
+      "learning_rate": 4.9612145937899615e-06,
+      "loss": 0.731,
+      "step": 11686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3735613282866437,
+      "learning_rate": 4.9605681907853906e-06,
+      "loss": 0.4927,
+      "step": 11687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6907967404884705,
+      "learning_rate": 4.959921799073492e-06,
+      "loss": 0.4194,
+      "step": 11688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.60134189339294,
+      "learning_rate": 4.959275418667793e-06,
+      "loss": 0.7078,
+      "step": 11689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8678192423021505,
+      "learning_rate": 4.958629049581829e-06,
+      "loss": 0.4788,
+      "step": 11690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7590775985149667,
+      "learning_rate": 4.95798269182913e-06,
+      "loss": 0.4362,
+      "step": 11691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7933122109347956,
+      "learning_rate": 4.957336345423223e-06,
+      "loss": 0.3756,
+      "step": 11692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.500171173636713,
+      "learning_rate": 4.956690010377646e-06,
+      "loss": 0.3456,
+      "step": 11693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.88018371820163,
+      "learning_rate": 4.956043686705924e-06,
+      "loss": 0.5678,
+      "step": 11694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8303475621536833,
+      "learning_rate": 4.955397374421591e-06,
+      "loss": 0.3951,
+      "step": 11695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.694174418352139,
+      "learning_rate": 4.954751073538175e-06,
+      "loss": 0.5152,
+      "step": 11696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.581915725965412,
+      "learning_rate": 4.954104784069208e-06,
+      "loss": 0.4449,
+      "step": 11697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8228814746500903,
+      "learning_rate": 4.953458506028217e-06,
+      "loss": 0.5463,
+      "step": 11698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9332460872079869,
+      "learning_rate": 4.952812239428734e-06,
+      "loss": 0.4363,
+      "step": 11699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1148731363029856,
+      "learning_rate": 4.9521659842842855e-06,
+      "loss": 0.4643,
+      "step": 11700
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.5179631114006042,
+      "eval_runtime": 68.0446,
+      "eval_samples_per_second": 50.114,
+      "eval_steps_per_second": 1.572,
+      "step": 11700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4370553286504077,
+      "learning_rate": 4.951519740608402e-06,
+      "loss": 0.4211,
+      "step": 11701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6100679812924312,
+      "learning_rate": 4.950873508414612e-06,
+      "loss": 0.5104,
+      "step": 11702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.153417858223534,
+      "learning_rate": 4.950227287716445e-06,
+      "loss": 0.6035,
+      "step": 11703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8756732710352706,
+      "learning_rate": 4.9495810785274264e-06,
+      "loss": 0.5519,
+      "step": 11704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1744344021867508,
+      "learning_rate": 4.948934880861089e-06,
+      "loss": 0.5271,
+      "step": 11705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9647965627853812,
+      "learning_rate": 4.948288694730955e-06,
+      "loss": 0.6422,
+      "step": 11706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5296633069276238,
+      "learning_rate": 4.947642520150557e-06,
+      "loss": 0.4393,
+      "step": 11707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5175617018766194,
+      "learning_rate": 4.9469963571334175e-06,
+      "loss": 0.3152,
+      "step": 11708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6724953307192862,
+      "learning_rate": 4.9463502056930655e-06,
+      "loss": 0.5409,
+      "step": 11709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8213831267724967,
+      "learning_rate": 4.94570406584303e-06,
+      "loss": 0.7404,
+      "step": 11710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6721552769462473,
+      "learning_rate": 4.945057937596834e-06,
+      "loss": 0.518,
+      "step": 11711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7944579145323556,
+      "learning_rate": 4.944411820968006e-06,
+      "loss": 0.5435,
+      "step": 11712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.158661477344818,
+      "learning_rate": 4.9437657159700715e-06,
+      "loss": 0.5215,
+      "step": 11713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0513521341958127,
+      "learning_rate": 4.943119622616556e-06,
+      "loss": 0.5744,
+      "step": 11714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4362523411832966,
+      "learning_rate": 4.9424735409209855e-06,
+      "loss": 0.4026,
+      "step": 11715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6448379209872086,
+      "learning_rate": 4.941827470896885e-06,
+      "loss": 0.4043,
+      "step": 11716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4825608217723651,
+      "learning_rate": 4.94118141255778e-06,
+      "loss": 0.3764,
+      "step": 11717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8362578765586404,
+      "learning_rate": 4.940535365917195e-06,
+      "loss": 0.785,
+      "step": 11718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7722983605616576,
+      "learning_rate": 4.939889330988654e-06,
+      "loss": 0.5323,
+      "step": 11719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8908351969046238,
+      "learning_rate": 4.939243307785683e-06,
+      "loss": 0.7219,
+      "step": 11720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8856641386352528,
+      "learning_rate": 4.9385972963218056e-06,
+      "loss": 0.577,
+      "step": 11721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5813163518833686,
+      "learning_rate": 4.9379512966105425e-06,
+      "loss": 0.4123,
+      "step": 11722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9664825863676991,
+      "learning_rate": 4.937305308665421e-06,
+      "loss": 0.4829,
+      "step": 11723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6558968524923836,
+      "learning_rate": 4.936659332499964e-06,
+      "loss": 0.5338,
+      "step": 11724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.858969202995168,
+      "learning_rate": 4.9360133681276945e-06,
+      "loss": 0.5053,
+      "step": 11725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.446115648359795,
+      "learning_rate": 4.935367415562134e-06,
+      "loss": 0.4862,
+      "step": 11726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7276244032390664,
+      "learning_rate": 4.934721474816806e-06,
+      "loss": 0.7296,
+      "step": 11727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4108527345751147,
+      "learning_rate": 4.934075545905232e-06,
+      "loss": 0.4564,
+      "step": 11728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7064142781808644,
+      "learning_rate": 4.9334296288409355e-06,
+      "loss": 0.4801,
+      "step": 11729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.718203629169395,
+      "learning_rate": 4.932783723637439e-06,
+      "loss": 0.531,
+      "step": 11730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9622028461407754,
+      "learning_rate": 4.932137830308262e-06,
+      "loss": 0.6672,
+      "step": 11731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7035329332810922,
+      "learning_rate": 4.931491948866926e-06,
+      "loss": 0.5072,
+      "step": 11732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5814923649628319,
+      "learning_rate": 4.930846079326954e-06,
+      "loss": 0.4488,
+      "step": 11733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6925579209527513,
+      "learning_rate": 4.9302002217018656e-06,
+      "loss": 0.4465,
+      "step": 11734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.743675241567515,
+      "learning_rate": 4.929554376005181e-06,
+      "loss": 0.5691,
+      "step": 11735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7262598682441115,
+      "learning_rate": 4.928908542250424e-06,
+      "loss": 0.497,
+      "step": 11736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5617254590049192,
+      "learning_rate": 4.928262720451108e-06,
+      "loss": 0.3409,
+      "step": 11737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.77753597305334,
+      "learning_rate": 4.92761691062076e-06,
+      "loss": 0.6705,
+      "step": 11738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9253318558712769,
+      "learning_rate": 4.9269711127728945e-06,
+      "loss": 0.4776,
+      "step": 11739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7097786221539857,
+      "learning_rate": 4.926325326921032e-06,
+      "loss": 0.4393,
+      "step": 11740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9877021595211823,
+      "learning_rate": 4.925679553078695e-06,
+      "loss": 0.6206,
+      "step": 11741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3404093926074794,
+      "learning_rate": 4.9250337912593985e-06,
+      "loss": 0.4386,
+      "step": 11742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.69399043330631,
+      "learning_rate": 4.924388041476661e-06,
+      "loss": 0.3758,
+      "step": 11743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7478839083671949,
+      "learning_rate": 4.923742303744004e-06,
+      "loss": 0.5353,
+      "step": 11744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4244936215280868,
+      "learning_rate": 4.923096578074944e-06,
+      "loss": 0.34,
+      "step": 11745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8593297560135986,
+      "learning_rate": 4.922450864482998e-06,
+      "loss": 0.739,
+      "step": 11746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.743734205590497,
+      "learning_rate": 4.921805162981686e-06,
+      "loss": 0.5212,
+      "step": 11747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.071312798968305,
+      "learning_rate": 4.921159473584523e-06,
+      "loss": 0.5676,
+      "step": 11748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8697427744521689,
+      "learning_rate": 4.920513796305026e-06,
+      "loss": 0.5632,
+      "step": 11749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9661553201970003,
+      "learning_rate": 4.9198681311567145e-06,
+      "loss": 0.4733,
+      "step": 11750
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.5159168839454651,
+      "eval_runtime": 68.008,
+      "eval_samples_per_second": 50.141,
+      "eval_steps_per_second": 1.573,
+      "step": 11750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9915902337161218,
+      "learning_rate": 4.919222478153102e-06,
+      "loss": 0.5274,
+      "step": 11751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6558373445366679,
+      "learning_rate": 4.918576837307707e-06,
+      "loss": 0.4986,
+      "step": 11752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.825614917544944,
+      "learning_rate": 4.917931208634044e-06,
+      "loss": 0.4998,
+      "step": 11753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6489094354744318,
+      "learning_rate": 4.91728559214563e-06,
+      "loss": 0.4938,
+      "step": 11754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6172328184243177,
+      "learning_rate": 4.91663998785598e-06,
+      "loss": 0.446,
+      "step": 11755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6468728447355152,
+      "learning_rate": 4.91599439577861e-06,
+      "loss": 0.4987,
+      "step": 11756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9672410870261505,
+      "learning_rate": 4.915348815927034e-06,
+      "loss": 0.6167,
+      "step": 11757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7516828985725181,
+      "learning_rate": 4.914703248314768e-06,
+      "loss": 0.3875,
+      "step": 11758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4774292468191381,
+      "learning_rate": 4.914057692955325e-06,
+      "loss": 0.3737,
+      "step": 11759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.678589347318125,
+      "learning_rate": 4.913412149862221e-06,
+      "loss": 0.5352,
+      "step": 11760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6797701207375002,
+      "learning_rate": 4.9127666190489694e-06,
+      "loss": 0.4824,
+      "step": 11761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7955963820815586,
+      "learning_rate": 4.912121100529084e-06,
+      "loss": 0.5796,
+      "step": 11762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4409090101839188,
+      "learning_rate": 4.911475594316078e-06,
+      "loss": 0.3842,
+      "step": 11763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5967225645291512,
+      "learning_rate": 4.910830100423464e-06,
+      "loss": 0.4194,
+      "step": 11764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6546857849975551,
+      "learning_rate": 4.910184618864757e-06,
+      "loss": 0.5711,
+      "step": 11765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.414189717694349,
+      "learning_rate": 4.909539149653469e-06,
+      "loss": 0.4267,
+      "step": 11766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5022725222572828,
+      "learning_rate": 4.90889369280311e-06,
+      "loss": 0.4505,
+      "step": 11767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7954070943921174,
+      "learning_rate": 4.9082482483271965e-06,
+      "loss": 0.5032,
+      "step": 11768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7508794111465598,
+      "learning_rate": 4.907602816239237e-06,
+      "loss": 0.5282,
+      "step": 11769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5783911999971543,
+      "learning_rate": 4.906957396552746e-06,
+      "loss": 0.5618,
+      "step": 11770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8669453865184331,
+      "learning_rate": 4.906311989281233e-06,
+      "loss": 0.5433,
+      "step": 11771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2262584764843987,
+      "learning_rate": 4.90566659443821e-06,
+      "loss": 0.8637,
+      "step": 11772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5300110860650107,
+      "learning_rate": 4.905021212037187e-06,
+      "loss": 0.4774,
+      "step": 11773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9111619942913518,
+      "learning_rate": 4.9043758420916755e-06,
+      "loss": 0.6176,
+      "step": 11774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7230681652871054,
+      "learning_rate": 4.903730484615186e-06,
+      "loss": 0.5191,
+      "step": 11775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6993147368861796,
+      "learning_rate": 4.9030851396212285e-06,
+      "loss": 0.5284,
+      "step": 11776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6001697863076603,
+      "learning_rate": 4.902439807123313e-06,
+      "loss": 0.4843,
+      "step": 11777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.101124596410044,
+      "learning_rate": 4.901794487134949e-06,
+      "loss": 0.6317,
+      "step": 11778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.777919550719853,
+      "learning_rate": 4.901149179669645e-06,
+      "loss": 0.561,
+      "step": 11779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6889383509292024,
+      "learning_rate": 4.900503884740912e-06,
+      "loss": 0.4221,
+      "step": 11780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7972410593874506,
+      "learning_rate": 4.899858602362257e-06,
+      "loss": 0.5193,
+      "step": 11781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8721143316293372,
+      "learning_rate": 4.89921333254719e-06,
+      "loss": 0.4732,
+      "step": 11782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5236732278724383,
+      "learning_rate": 4.898568075309219e-06,
+      "loss": 0.3852,
+      "step": 11783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7957586313285518,
+      "learning_rate": 4.8979228306618505e-06,
+      "loss": 0.5696,
+      "step": 11784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6003580369131551,
+      "learning_rate": 4.897277598618594e-06,
+      "loss": 0.4892,
+      "step": 11785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7739745381735068,
+      "learning_rate": 4.896632379192958e-06,
+      "loss": 0.5278,
+      "step": 11786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8906500215956272,
+      "learning_rate": 4.895987172398447e-06,
+      "loss": 0.5646,
+      "step": 11787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2670289727349613,
+      "learning_rate": 4.89534197824857e-06,
+      "loss": 0.6576,
+      "step": 11788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1393300516392535,
+      "learning_rate": 4.894696796756833e-06,
+      "loss": 0.7235,
+      "step": 11789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4899914245844597,
+      "learning_rate": 4.894051627936741e-06,
+      "loss": 0.6072,
+      "step": 11790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8046342307846903,
+      "learning_rate": 4.893406471801804e-06,
+      "loss": 0.5443,
+      "step": 11791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.35472025130394,
+      "learning_rate": 4.892761328365523e-06,
+      "loss": 0.5982,
+      "step": 11792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6216796614910707,
+      "learning_rate": 4.892116197641409e-06,
+      "loss": 0.532,
+      "step": 11793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.055351893714213,
+      "learning_rate": 4.891471079642964e-06,
+      "loss": 0.6374,
+      "step": 11794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7162841963643178,
+      "learning_rate": 4.890825974383694e-06,
+      "loss": 0.4367,
+      "step": 11795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4691550001071518,
+      "learning_rate": 4.890180881877103e-06,
+      "loss": 0.4574,
+      "step": 11796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.089678463789025,
+      "learning_rate": 4.889535802136699e-06,
+      "loss": 0.5597,
+      "step": 11797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.051748333428037,
+      "learning_rate": 4.888890735175983e-06,
+      "loss": 0.5381,
+      "step": 11798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9727705374995885,
+      "learning_rate": 4.888245681008458e-06,
+      "loss": 0.486,
+      "step": 11799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3990554151922578,
+      "learning_rate": 4.887600639647632e-06,
+      "loss": 0.3073,
+      "step": 11800
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.5168783068656921,
+      "eval_runtime": 67.9581,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.574,
+      "step": 11800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7393618536586375,
+      "learning_rate": 4.886955611107005e-06,
+      "loss": 0.4798,
+      "step": 11801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1266702598038822,
+      "learning_rate": 4.886310595400081e-06,
+      "loss": 0.669,
+      "step": 11802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6173125368958392,
+      "learning_rate": 4.885665592540364e-06,
+      "loss": 0.5277,
+      "step": 11803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4824301655549506,
+      "learning_rate": 4.885020602541356e-06,
+      "loss": 0.4629,
+      "step": 11804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.581039167945152,
+      "learning_rate": 4.884375625416561e-06,
+      "loss": 0.4802,
+      "step": 11805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8961707362174907,
+      "learning_rate": 4.883730661179479e-06,
+      "loss": 0.5485,
+      "step": 11806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5721303534716549,
+      "learning_rate": 4.883085709843612e-06,
+      "loss": 0.4103,
+      "step": 11807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8788501838624645,
+      "learning_rate": 4.882440771422464e-06,
+      "loss": 0.5185,
+      "step": 11808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.832374794496835,
+      "learning_rate": 4.881795845929534e-06,
+      "loss": 0.5076,
+      "step": 11809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.113991893116632,
+      "learning_rate": 4.8811509333783245e-06,
+      "loss": 0.658,
+      "step": 11810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6513851918899634,
+      "learning_rate": 4.880506033782335e-06,
+      "loss": 0.5164,
+      "step": 11811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9568671311971197,
+      "learning_rate": 4.8798611471550674e-06,
+      "loss": 0.4918,
+      "step": 11812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9644972627430002,
+      "learning_rate": 4.879216273510022e-06,
+      "loss": 0.5418,
+      "step": 11813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4572430418112852,
+      "learning_rate": 4.8785714128606975e-06,
+      "loss": 0.3573,
+      "step": 11814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.695909088939337,
+      "learning_rate": 4.877926565220595e-06,
+      "loss": 0.5198,
+      "step": 11815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9832371677700318,
+      "learning_rate": 4.877281730603214e-06,
+      "loss": 0.4808,
+      "step": 11816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6160379636824742,
+      "learning_rate": 4.876636909022051e-06,
+      "loss": 0.5512,
+      "step": 11817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9188108211565158,
+      "learning_rate": 4.87599210049061e-06,
+      "loss": 0.5306,
+      "step": 11818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.321543683218787,
+      "learning_rate": 4.875347305022385e-06,
+      "loss": 0.6883,
+      "step": 11819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.738032138382169,
+      "learning_rate": 4.874702522630877e-06,
+      "loss": 0.5211,
+      "step": 11820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9187517487148527,
+      "learning_rate": 4.874057753329582e-06,
+      "loss": 0.4338,
+      "step": 11821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3543684852751465,
+      "learning_rate": 4.873412997132e-06,
+      "loss": 0.3269,
+      "step": 11822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.957443747707886,
+      "learning_rate": 4.872768254051629e-06,
+      "loss": 0.5605,
+      "step": 11823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.189715472028773,
+      "learning_rate": 4.872123524101963e-06,
+      "loss": 0.6255,
+      "step": 11824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.133726780929894,
+      "learning_rate": 4.871478807296501e-06,
+      "loss": 0.7055,
+      "step": 11825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.969483698605488,
+      "learning_rate": 4.8708341036487414e-06,
+      "loss": 0.5775,
+      "step": 11826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.862751404552439,
+      "learning_rate": 4.870189413172177e-06,
+      "loss": 0.5441,
+      "step": 11827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6153415921241276,
+      "learning_rate": 4.8695447358803085e-06,
+      "loss": 0.4661,
+      "step": 11828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7801121991394544,
+      "learning_rate": 4.8689000717866285e-06,
+      "loss": 0.5464,
+      "step": 11829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7425586536344115,
+      "learning_rate": 4.868255420904632e-06,
+      "loss": 0.5149,
+      "step": 11830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.997259297093634,
+      "learning_rate": 4.867610783247818e-06,
+      "loss": 0.5865,
+      "step": 11831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6551024212545271,
+      "learning_rate": 4.866966158829678e-06,
+      "loss": 0.4005,
+      "step": 11832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6764710865807935,
+      "learning_rate": 4.8663215476637086e-06,
+      "loss": 0.4432,
+      "step": 11833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9013064654701295,
+      "learning_rate": 4.8656769497634045e-06,
+      "loss": 0.4212,
+      "step": 11834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7438398011574934,
+      "learning_rate": 4.865032365142258e-06,
+      "loss": 0.5775,
+      "step": 11835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8932181256090321,
+      "learning_rate": 4.864387793813766e-06,
+      "loss": 0.491,
+      "step": 11836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8732991815384785,
+      "learning_rate": 4.863743235791422e-06,
+      "loss": 0.5103,
+      "step": 11837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7286887836731317,
+      "learning_rate": 4.863098691088715e-06,
+      "loss": 0.5746,
+      "step": 11838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6463981441456095,
+      "learning_rate": 4.8624541597191435e-06,
+      "loss": 0.427,
+      "step": 11839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.634373652659293,
+      "learning_rate": 4.861809641696197e-06,
+      "loss": 0.5037,
+      "step": 11840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3611852046139394,
+      "learning_rate": 4.86116513703337e-06,
+      "loss": 0.3524,
+      "step": 11841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8628794801877548,
+      "learning_rate": 4.860520645744152e-06,
+      "loss": 0.4905,
+      "step": 11842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2019966795626265,
+      "learning_rate": 4.859876167842039e-06,
+      "loss": 0.5049,
+      "step": 11843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3561724064810639,
+      "learning_rate": 4.859231703340521e-06,
+      "loss": 0.3668,
+      "step": 11844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6688899844402432,
+      "learning_rate": 4.8585872522530895e-06,
+      "loss": 0.5177,
+      "step": 11845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7958797161399236,
+      "learning_rate": 4.857942814593235e-06,
+      "loss": 0.5967,
+      "step": 11846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5125606937289362,
+      "learning_rate": 4.857298390374448e-06,
+      "loss": 0.3802,
+      "step": 11847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.101048391957198,
+      "learning_rate": 4.856653979610222e-06,
+      "loss": 0.6537,
+      "step": 11848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.743612300895372,
+      "learning_rate": 4.856009582314044e-06,
+      "loss": 0.4903,
+      "step": 11849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.787174673227342,
+      "learning_rate": 4.855365198499406e-06,
+      "loss": 0.7484,
+      "step": 11850
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5162047147750854,
+      "eval_runtime": 67.9927,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 11850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7267783676620105,
+      "learning_rate": 4.854720828179797e-06,
+      "loss": 0.3837,
+      "step": 11851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9499316876779074,
+      "learning_rate": 4.854076471368706e-06,
+      "loss": 0.6298,
+      "step": 11852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8383162621072129,
+      "learning_rate": 4.853432128079624e-06,
+      "loss": 0.4225,
+      "step": 11853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.806907357074638,
+      "learning_rate": 4.852787798326038e-06,
+      "loss": 0.4672,
+      "step": 11854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5755594736354956,
+      "learning_rate": 4.852143482121437e-06,
+      "loss": 0.4337,
+      "step": 11855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5436974267165477,
+      "learning_rate": 4.851499179479311e-06,
+      "loss": 0.4301,
+      "step": 11856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25899683566582,
+      "learning_rate": 4.8508548904131455e-06,
+      "loss": 0.7147,
+      "step": 11857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.716369738529152,
+      "learning_rate": 4.850210614936432e-06,
+      "loss": 0.4852,
+      "step": 11858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7849773278832397,
+      "learning_rate": 4.849566353062654e-06,
+      "loss": 0.48,
+      "step": 11859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7711345098796232,
+      "learning_rate": 4.8489221048053005e-06,
+      "loss": 0.4074,
+      "step": 11860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6988325879070436,
+      "learning_rate": 4.848277870177858e-06,
+      "loss": 0.5396,
+      "step": 11861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7350395192678778,
+      "learning_rate": 4.847633649193813e-06,
+      "loss": 0.4565,
+      "step": 11862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.182586329087827,
+      "learning_rate": 4.846989441866654e-06,
+      "loss": 0.6477,
+      "step": 11863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5792929999722873,
+      "learning_rate": 4.846345248209865e-06,
+      "loss": 0.6474,
+      "step": 11864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8932310310897527,
+      "learning_rate": 4.845701068236932e-06,
+      "loss": 0.596,
+      "step": 11865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2913397764615633,
+      "learning_rate": 4.84505690196134e-06,
+      "loss": 0.681,
+      "step": 11866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9615758975436541,
+      "learning_rate": 4.844412749396575e-06,
+      "loss": 0.5397,
+      "step": 11867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4928393940320632,
+      "learning_rate": 4.843768610556123e-06,
+      "loss": 0.3598,
+      "step": 11868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.685376972938598,
+      "learning_rate": 4.843124485453468e-06,
+      "loss": 0.5342,
+      "step": 11869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7804263929067936,
+      "learning_rate": 4.842480374102092e-06,
+      "loss": 0.5051,
+      "step": 11870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.885713600833652,
+      "learning_rate": 4.841836276515483e-06,
+      "loss": 0.5776,
+      "step": 11871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.686740879781686,
+      "learning_rate": 4.841192192707122e-06,
+      "loss": 0.4496,
+      "step": 11872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.444266378326097,
+      "learning_rate": 4.8405481226904915e-06,
+      "loss": 0.3666,
+      "step": 11873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5148903066372357,
+      "learning_rate": 4.839904066479079e-06,
+      "loss": 0.4215,
+      "step": 11874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6703732249561714,
+      "learning_rate": 4.839260024086362e-06,
+      "loss": 0.4177,
+      "step": 11875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5063896750160968,
+      "learning_rate": 4.838615995525828e-06,
+      "loss": 0.4643,
+      "step": 11876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.011021633906461,
+      "learning_rate": 4.837971980810958e-06,
+      "loss": 0.5879,
+      "step": 11877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5029848391564284,
+      "learning_rate": 4.837327979955232e-06,
+      "loss": 0.4142,
+      "step": 11878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6803771886875403,
+      "learning_rate": 4.836683992972133e-06,
+      "loss": 0.4303,
+      "step": 11879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.069349675396812,
+      "learning_rate": 4.836040019875143e-06,
+      "loss": 0.5135,
+      "step": 11880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5761199646525577,
+      "learning_rate": 4.835396060677744e-06,
+      "loss": 0.4225,
+      "step": 11881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8546097811084687,
+      "learning_rate": 4.834752115393414e-06,
+      "loss": 0.5236,
+      "step": 11882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2947390721353154,
+      "learning_rate": 4.8341081840356364e-06,
+      "loss": 0.5444,
+      "step": 11883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5368995583870122,
+      "learning_rate": 4.8334642666178895e-06,
+      "loss": 0.5256,
+      "step": 11884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1080467843679647,
+      "learning_rate": 4.832820363153655e-06,
+      "loss": 0.5779,
+      "step": 11885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9118714651628304,
+      "learning_rate": 4.832176473656411e-06,
+      "loss": 0.5426,
+      "step": 11886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7181697549628234,
+      "learning_rate": 4.831532598139638e-06,
+      "loss": 0.4294,
+      "step": 11887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.544009419053724,
+      "learning_rate": 4.830888736616814e-06,
+      "loss": 0.3421,
+      "step": 11888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6116584452625757,
+      "learning_rate": 4.830244889101421e-06,
+      "loss": 0.5112,
+      "step": 11889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7522333674783543,
+      "learning_rate": 4.829601055606934e-06,
+      "loss": 0.5827,
+      "step": 11890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6573160290790205,
+      "learning_rate": 4.828957236146831e-06,
+      "loss": 0.4826,
+      "step": 11891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7925651011133652,
+      "learning_rate": 4.828313430734592e-06,
+      "loss": 0.6915,
+      "step": 11892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0655988680121666,
+      "learning_rate": 4.827669639383694e-06,
+      "loss": 0.5501,
+      "step": 11893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6809695112274883,
+      "learning_rate": 4.827025862107616e-06,
+      "loss": 0.4571,
+      "step": 11894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0276064106784317,
+      "learning_rate": 4.826382098919833e-06,
+      "loss": 0.6094,
+      "step": 11895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6543486252309798,
+      "learning_rate": 4.825738349833823e-06,
+      "loss": 0.5109,
+      "step": 11896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6680425928988958,
+      "learning_rate": 4.82509461486306e-06,
+      "loss": 0.5075,
+      "step": 11897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4020068367531577,
+      "learning_rate": 4.824450894021024e-06,
+      "loss": 0.3198,
+      "step": 11898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5032474308749033,
+      "learning_rate": 4.823807187321189e-06,
+      "loss": 0.3841,
+      "step": 11899
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6615508060565423,
+      "learning_rate": 4.823163494777029e-06,
+      "loss": 0.4291,
+      "step": 11900
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5160362720489502,
+      "eval_runtime": 68.0533,
+      "eval_samples_per_second": 50.108,
+      "eval_steps_per_second": 1.572,
+      "step": 11900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5151710728459453,
+      "learning_rate": 4.822519816402021e-06,
+      "loss": 0.4151,
+      "step": 11901
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8152487203233365,
+      "learning_rate": 4.82187615220964e-06,
+      "loss": 0.5012,
+      "step": 11902
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.666024439650599,
+      "learning_rate": 4.821232502213362e-06,
+      "loss": 0.5236,
+      "step": 11903
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6600215629697912,
+      "learning_rate": 4.820588866426659e-06,
+      "loss": 0.4983,
+      "step": 11904
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.538517624292994,
+      "learning_rate": 4.819945244863006e-06,
+      "loss": 0.475,
+      "step": 11905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8489232979719867,
+      "learning_rate": 4.819301637535877e-06,
+      "loss": 0.5665,
+      "step": 11906
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5989682935994343,
+      "learning_rate": 4.8186580444587446e-06,
+      "loss": 0.345,
+      "step": 11907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1868401825254655,
+      "learning_rate": 4.818014465645083e-06,
+      "loss": 0.6052,
+      "step": 11908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.510189435586291,
+      "learning_rate": 4.817370901108367e-06,
+      "loss": 0.6529,
+      "step": 11909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6794284744220174,
+      "learning_rate": 4.816727350862065e-06,
+      "loss": 0.4373,
+      "step": 11910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7171594644772796,
+      "learning_rate": 4.816083814919653e-06,
+      "loss": 0.4385,
+      "step": 11911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.482287499594495,
+      "learning_rate": 4.815440293294601e-06,
+      "loss": 0.4345,
+      "step": 11912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9685673000167028,
+      "learning_rate": 4.8147967860003795e-06,
+      "loss": 0.4588,
+      "step": 11913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7441298610070737,
+      "learning_rate": 4.814153293050464e-06,
+      "loss": 0.5774,
+      "step": 11914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7060980065625566,
+      "learning_rate": 4.813509814458323e-06,
+      "loss": 0.4472,
+      "step": 11915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4080267854828727,
+      "learning_rate": 4.812866350237426e-06,
+      "loss": 0.3409,
+      "step": 11916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9577005861507895,
+      "learning_rate": 4.812222900401248e-06,
+      "loss": 0.566,
+      "step": 11917
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8099380487027332,
+      "learning_rate": 4.811579464963253e-06,
+      "loss": 0.6984,
+      "step": 11918
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.992997256314306,
+      "learning_rate": 4.810936043936917e-06,
+      "loss": 0.6331,
+      "step": 11919
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9815246293113138,
+      "learning_rate": 4.810292637335705e-06,
+      "loss": 0.493,
+      "step": 11920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.379116159973461,
+      "learning_rate": 4.809649245173091e-06,
+      "loss": 0.7862,
+      "step": 11921
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9056652778358822,
+      "learning_rate": 4.809005867462539e-06,
+      "loss": 0.456,
+      "step": 11922
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9233849728998826,
+      "learning_rate": 4.80836250421752e-06,
+      "loss": 0.5147,
+      "step": 11923
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6705553713466803,
+      "learning_rate": 4.807719155451503e-06,
+      "loss": 0.4402,
+      "step": 11924
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8626137866096983,
+      "learning_rate": 4.807075821177955e-06,
+      "loss": 0.6653,
+      "step": 11925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5566261021162195,
+      "learning_rate": 4.806432501410344e-06,
+      "loss": 0.5334,
+      "step": 11926
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.120752594798559,
+      "learning_rate": 4.805789196162139e-06,
+      "loss": 0.5606,
+      "step": 11927
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.30547410803905,
+      "learning_rate": 4.8051459054468056e-06,
+      "loss": 0.5325,
+      "step": 11928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9963377240603573,
+      "learning_rate": 4.804502629277811e-06,
+      "loss": 0.5784,
+      "step": 11929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8457472086355102,
+      "learning_rate": 4.803859367668622e-06,
+      "loss": 0.5314,
+      "step": 11930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7165707629746842,
+      "learning_rate": 4.803216120632703e-06,
+      "loss": 0.5361,
+      "step": 11931
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.75496913689946,
+      "learning_rate": 4.802572888183524e-06,
+      "loss": 0.5043,
+      "step": 11932
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8599834499094987,
+      "learning_rate": 4.8019296703345475e-06,
+      "loss": 0.4011,
+      "step": 11933
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.517260927767698,
+      "learning_rate": 4.801286467099241e-06,
+      "loss": 0.4732,
+      "step": 11934
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6912375168581801,
+      "learning_rate": 4.800643278491067e-06,
+      "loss": 0.4327,
+      "step": 11935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8246748862307502,
+      "learning_rate": 4.800000104523491e-06,
+      "loss": 0.5647,
+      "step": 11936
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9536027809543197,
+      "learning_rate": 4.79935694520998e-06,
+      "loss": 0.5565,
+      "step": 11937
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4092578810376264,
+      "learning_rate": 4.798713800563995e-06,
+      "loss": 0.4128,
+      "step": 11938
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6179941003586469,
+      "learning_rate": 4.798070670599e-06,
+      "loss": 0.454,
+      "step": 11939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9963547890595492,
+      "learning_rate": 4.797427555328459e-06,
+      "loss": 0.4945,
+      "step": 11940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7208668633577036,
+      "learning_rate": 4.796784454765837e-06,
+      "loss": 0.4058,
+      "step": 11941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3462228537624028,
+      "learning_rate": 4.7961413689245964e-06,
+      "loss": 0.3946,
+      "step": 11942
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.100411909062216,
+      "learning_rate": 4.795498297818198e-06,
+      "loss": 0.5752,
+      "step": 11943
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1214485670355687,
+      "learning_rate": 4.794855241460106e-06,
+      "loss": 0.6191,
+      "step": 11944
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0333420717646784,
+      "learning_rate": 4.794212199863781e-06,
+      "loss": 0.6757,
+      "step": 11945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7488226565103875,
+      "learning_rate": 4.793569173042686e-06,
+      "loss": 0.586,
+      "step": 11946
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9368971508748736,
+      "learning_rate": 4.79292616101028e-06,
+      "loss": 0.6385,
+      "step": 11947
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9028416729723372,
+      "learning_rate": 4.792283163780027e-06,
+      "loss": 0.5318,
+      "step": 11948
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5192963092416085,
+      "learning_rate": 4.791640181365387e-06,
+      "loss": 0.5816,
+      "step": 11949
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1024181813814478,
+      "learning_rate": 4.790997213779818e-06,
+      "loss": 0.5558,
+      "step": 11950
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5161501169204712,
+      "eval_runtime": 67.9797,
+      "eval_samples_per_second": 50.162,
+      "eval_steps_per_second": 1.574,
+      "step": 11950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.637330879811165,
+      "learning_rate": 4.790354261036784e-06,
+      "loss": 0.4876,
+      "step": 11951
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8845151937829494,
+      "learning_rate": 4.789711323149742e-06,
+      "loss": 0.4982,
+      "step": 11952
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9427119496403096,
+      "learning_rate": 4.789068400132151e-06,
+      "loss": 0.5989,
+      "step": 11953
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5673208116198711,
+      "learning_rate": 4.788425491997473e-06,
+      "loss": 0.3997,
+      "step": 11954
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6692143109122666,
+      "learning_rate": 4.787782598759164e-06,
+      "loss": 0.5013,
+      "step": 11955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6824784544487958,
+      "learning_rate": 4.7871397204306845e-06,
+      "loss": 0.4494,
+      "step": 11956
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.133136739691327,
+      "learning_rate": 4.786496857025493e-06,
+      "loss": 0.5945,
+      "step": 11957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.14226709722103,
+      "learning_rate": 4.785854008557045e-06,
+      "loss": 0.5891,
+      "step": 11958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.816740836965751,
+      "learning_rate": 4.785211175038801e-06,
+      "loss": 0.6061,
+      "step": 11959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.585518798934923,
+      "learning_rate": 4.7845683564842164e-06,
+      "loss": 0.4819,
+      "step": 11960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7944724363731022,
+      "learning_rate": 4.78392555290675e-06,
+      "loss": 0.5291,
+      "step": 11961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.236858860530569,
+      "learning_rate": 4.783282764319857e-06,
+      "loss": 0.6388,
+      "step": 11962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6048591608623246,
+      "learning_rate": 4.7826399907369925e-06,
+      "loss": 0.4504,
+      "step": 11963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6722734403237853,
+      "learning_rate": 4.781997232171616e-06,
+      "loss": 0.4034,
+      "step": 11964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1111676342992975,
+      "learning_rate": 4.781354488637182e-06,
+      "loss": 0.5953,
+      "step": 11965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5088866246569264,
+      "learning_rate": 4.780711760147143e-06,
+      "loss": 0.3548,
+      "step": 11966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.450316188026567,
+      "learning_rate": 4.7800690467149576e-06,
+      "loss": 0.3744,
+      "step": 11967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0816379472642517,
+      "learning_rate": 4.779426348354081e-06,
+      "loss": 0.6346,
+      "step": 11968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6535166124263274,
+      "learning_rate": 4.778783665077965e-06,
+      "loss": 0.461,
+      "step": 11969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3032417732788826,
+      "learning_rate": 4.778140996900065e-06,
+      "loss": 0.4108,
+      "step": 11970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8599230895114396,
+      "learning_rate": 4.777498343833835e-06,
+      "loss": 0.5023,
+      "step": 11971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0462901553411172,
+      "learning_rate": 4.776855705892729e-06,
+      "loss": 0.4748,
+      "step": 11972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.059868907739072,
+      "learning_rate": 4.776213083090197e-06,
+      "loss": 0.5574,
+      "step": 11973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4528239943270487,
+      "learning_rate": 4.775570475439698e-06,
+      "loss": 0.318,
+      "step": 11974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8834564511355991,
+      "learning_rate": 4.77492788295468e-06,
+      "loss": 0.4465,
+      "step": 11975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6090505650234315,
+      "learning_rate": 4.774285305648595e-06,
+      "loss": 0.4191,
+      "step": 11976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9412706299245452,
+      "learning_rate": 4.773642743534899e-06,
+      "loss": 0.4425,
+      "step": 11977
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.931756646784295,
+      "learning_rate": 4.773000196627039e-06,
+      "loss": 0.5525,
+      "step": 11978
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.275681248500431,
+      "learning_rate": 4.7723576649384684e-06,
+      "loss": 0.758,
+      "step": 11979
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7309852475524012,
+      "learning_rate": 4.771715148482638e-06,
+      "loss": 0.6043,
+      "step": 11980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5298699483948555,
+      "learning_rate": 4.771072647273e-06,
+      "loss": 0.4727,
+      "step": 11981
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0182396928745376,
+      "learning_rate": 4.770430161323003e-06,
+      "loss": 0.4324,
+      "step": 11982
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9364531045036069,
+      "learning_rate": 4.769787690646097e-06,
+      "loss": 0.6042,
+      "step": 11983
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.556069622297289,
+      "learning_rate": 4.769145235255731e-06,
+      "loss": 0.4505,
+      "step": 11984
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4412980742624497,
+      "learning_rate": 4.768502795165357e-06,
+      "loss": 0.3078,
+      "step": 11985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0637957439807324,
+      "learning_rate": 4.767860370388422e-06,
+      "loss": 0.5092,
+      "step": 11986
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7573429444384652,
+      "learning_rate": 4.767217960938376e-06,
+      "loss": 0.4992,
+      "step": 11987
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.690989297334625,
+      "learning_rate": 4.7665755668286675e-06,
+      "loss": 0.4713,
+      "step": 11988
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8280349439114911,
+      "learning_rate": 4.765933188072741e-06,
+      "loss": 0.6143,
+      "step": 11989
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7887297004709544,
+      "learning_rate": 4.765290824684049e-06,
+      "loss": 0.6473,
+      "step": 11990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6031296704878901,
+      "learning_rate": 4.764648476676037e-06,
+      "loss": 0.397,
+      "step": 11991
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8960724399899869,
+      "learning_rate": 4.764006144062154e-06,
+      "loss": 0.439,
+      "step": 11992
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7370432308147432,
+      "learning_rate": 4.763363826855844e-06,
+      "loss": 0.4352,
+      "step": 11993
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.884229257846463,
+      "learning_rate": 4.762721525070555e-06,
+      "loss": 0.6769,
+      "step": 11994
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8676362959174106,
+      "learning_rate": 4.7620792387197315e-06,
+      "loss": 0.4836,
+      "step": 11995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5119976203193144,
+      "learning_rate": 4.7614369678168225e-06,
+      "loss": 0.4873,
+      "step": 11996
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.948133717838364,
+      "learning_rate": 4.760794712375272e-06,
+      "loss": 0.6713,
+      "step": 11997
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0106204384885356,
+      "learning_rate": 4.760152472408523e-06,
+      "loss": 0.647,
+      "step": 11998
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7563787588767361,
+      "learning_rate": 4.759510247930025e-06,
+      "loss": 0.4121,
+      "step": 11999
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6220242846906932,
+      "learning_rate": 4.7588680389532185e-06,
+      "loss": 0.5396,
+      "step": 12000
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5164961814880371,
+      "eval_runtime": 67.9386,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 12000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7221235528918617,
+      "learning_rate": 4.758225845491551e-06,
+      "loss": 0.6171,
+      "step": 12001
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5351034623849087,
+      "learning_rate": 4.757583667558463e-06,
+      "loss": 0.6054,
+      "step": 12002
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3234122169369646,
+      "learning_rate": 4.756941505167399e-06,
+      "loss": 0.6896,
+      "step": 12003
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.59022488620948,
+      "learning_rate": 4.756299358331805e-06,
+      "loss": 0.4643,
+      "step": 12004
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.73573544405432,
+      "learning_rate": 4.7556572270651224e-06,
+      "loss": 0.4152,
+      "step": 12005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9477742582274478,
+      "learning_rate": 4.755015111380791e-06,
+      "loss": 0.4846,
+      "step": 12006
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.647846098851426,
+      "learning_rate": 4.754373011292257e-06,
+      "loss": 0.3832,
+      "step": 12007
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.230049916121682,
+      "learning_rate": 4.753730926812961e-06,
+      "loss": 0.6142,
+      "step": 12008
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9317872088065084,
+      "learning_rate": 4.753088857956344e-06,
+      "loss": 0.5191,
+      "step": 12009
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6738309897561496,
+      "learning_rate": 4.752446804735848e-06,
+      "loss": 0.5445,
+      "step": 12010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4927381600438911,
+      "learning_rate": 4.751804767164914e-06,
+      "loss": 0.394,
+      "step": 12011
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9787664273922454,
+      "learning_rate": 4.751162745256983e-06,
+      "loss": 0.6032,
+      "step": 12012
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.935765753903036,
+      "learning_rate": 4.750520739025494e-06,
+      "loss": 0.5117,
+      "step": 12013
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.778465085475297,
+      "learning_rate": 4.749878748483888e-06,
+      "loss": 0.4559,
+      "step": 12014
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5675155998952743,
+      "learning_rate": 4.749236773645605e-06,
+      "loss": 0.4491,
+      "step": 12015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.759590193029184,
+      "learning_rate": 4.748594814524082e-06,
+      "loss": 0.6743,
+      "step": 12016
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.952921279310216,
+      "learning_rate": 4.747952871132761e-06,
+      "loss": 0.6221,
+      "step": 12017
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8868249010749065,
+      "learning_rate": 4.74731094348508e-06,
+      "loss": 0.426,
+      "step": 12018
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7543194964440558,
+      "learning_rate": 4.746669031594476e-06,
+      "loss": 0.523,
+      "step": 12019
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2911579265951305,
+      "learning_rate": 4.746027135474389e-06,
+      "loss": 0.5895,
+      "step": 12020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890473894859042,
+      "learning_rate": 4.745385255138255e-06,
+      "loss": 0.5327,
+      "step": 12021
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9278929043664743,
+      "learning_rate": 4.744743390599512e-06,
+      "loss": 0.479,
+      "step": 12022
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.79883002096031,
+      "learning_rate": 4.744101541871597e-06,
+      "loss": 0.4195,
+      "step": 12023
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9279159017455623,
+      "learning_rate": 4.743459708967946e-06,
+      "loss": 0.4911,
+      "step": 12024
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6017152351943906,
+      "learning_rate": 4.742817891901998e-06,
+      "loss": 0.4099,
+      "step": 12025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.630127611595498,
+      "learning_rate": 4.742176090687185e-06,
+      "loss": 0.4617,
+      "step": 12026
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1202015542076293,
+      "learning_rate": 4.741534305336947e-06,
+      "loss": 0.7844,
+      "step": 12027
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9095811975401669,
+      "learning_rate": 4.740892535864716e-06,
+      "loss": 0.5825,
+      "step": 12028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5623915281133938,
+      "learning_rate": 4.7402507822839286e-06,
+      "loss": 0.3127,
+      "step": 12029
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9396721116887758,
+      "learning_rate": 4.7396090446080186e-06,
+      "loss": 0.6589,
+      "step": 12030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6133043869457224,
+      "learning_rate": 4.738967322850421e-06,
+      "loss": 0.5062,
+      "step": 12031
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.682427690138348,
+      "learning_rate": 4.738325617024571e-06,
+      "loss": 0.4744,
+      "step": 12032
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7881433895798968,
+      "learning_rate": 4.737683927143899e-06,
+      "loss": 0.571,
+      "step": 12033
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6690592776381086,
+      "learning_rate": 4.737042253221842e-06,
+      "loss": 0.3806,
+      "step": 12034
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.946604303449652,
+      "learning_rate": 4.736400595271831e-06,
+      "loss": 0.5625,
+      "step": 12035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4792704034362207,
+      "learning_rate": 4.7357589533073e-06,
+      "loss": 0.4753,
+      "step": 12036
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6598578004789264,
+      "learning_rate": 4.735117327341679e-06,
+      "loss": 0.3652,
+      "step": 12037
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4616815265962757,
+      "learning_rate": 4.734475717388404e-06,
+      "loss": 0.538,
+      "step": 12038
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6702115518820846,
+      "learning_rate": 4.733834123460901e-06,
+      "loss": 0.3694,
+      "step": 12039
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7664508970167543,
+      "learning_rate": 4.733192545572607e-06,
+      "loss": 0.4791,
+      "step": 12040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9389339735442543,
+      "learning_rate": 4.732550983736951e-06,
+      "loss": 0.5384,
+      "step": 12041
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1157158306977224,
+      "learning_rate": 4.731909437967364e-06,
+      "loss": 0.6053,
+      "step": 12042
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.922998279690859,
+      "learning_rate": 4.731267908277274e-06,
+      "loss": 0.4579,
+      "step": 12043
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.516966187596325,
+      "learning_rate": 4.730626394680113e-06,
+      "loss": 0.3672,
+      "step": 12044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8203409866010984,
+      "learning_rate": 4.729984897189313e-06,
+      "loss": 0.4549,
+      "step": 12045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4220431285840143,
+      "learning_rate": 4.7293434158182984e-06,
+      "loss": 0.3949,
+      "step": 12046
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3550875999161909,
+      "learning_rate": 4.728701950580501e-06,
+      "loss": 0.3138,
+      "step": 12047
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5373276910699933,
+      "learning_rate": 4.72806050148935e-06,
+      "loss": 0.3335,
+      "step": 12048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5685824461495996,
+      "learning_rate": 4.727419068558272e-06,
+      "loss": 0.4226,
+      "step": 12049
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5482465266216332,
+      "learning_rate": 4.7267776518006966e-06,
+      "loss": 0.4221,
+      "step": 12050
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5154851675033569,
+      "eval_runtime": 67.9851,
+      "eval_samples_per_second": 50.158,
+      "eval_steps_per_second": 1.574,
+      "step": 12050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.751639107585009,
+      "learning_rate": 4.7261362512300514e-06,
+      "loss": 0.5499,
+      "step": 12051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7832602892110045,
+      "learning_rate": 4.725494866859762e-06,
+      "loss": 0.5159,
+      "step": 12052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.567414804020602,
+      "learning_rate": 4.724853498703257e-06,
+      "loss": 0.4814,
+      "step": 12053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0554257536964387,
+      "learning_rate": 4.724212146773963e-06,
+      "loss": 0.6303,
+      "step": 12054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0375982813564826,
+      "learning_rate": 4.7235708110853065e-06,
+      "loss": 0.552,
+      "step": 12055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9339253499633962,
+      "learning_rate": 4.722929491650711e-06,
+      "loss": 0.618,
+      "step": 12056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7455682240001977,
+      "learning_rate": 4.722288188483605e-06,
+      "loss": 0.4363,
+      "step": 12057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7376481381490354,
+      "learning_rate": 4.721646901597412e-06,
+      "loss": 0.4751,
+      "step": 12058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6148386308879081,
+      "learning_rate": 4.721005631005557e-06,
+      "loss": 0.432,
+      "step": 12059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.778561392158119,
+      "learning_rate": 4.720364376721467e-06,
+      "loss": 0.5609,
+      "step": 12060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.669763043207116,
+      "learning_rate": 4.719723138758564e-06,
+      "loss": 0.5175,
+      "step": 12061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.47617703710124,
+      "learning_rate": 4.719081917130272e-06,
+      "loss": 0.3838,
+      "step": 12062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15443732254752,
+      "learning_rate": 4.718440711850017e-06,
+      "loss": 0.5198,
+      "step": 12063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.151169313144446,
+      "learning_rate": 4.717799522931217e-06,
+      "loss": 0.5655,
+      "step": 12064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.949052319786599,
+      "learning_rate": 4.7171583503873e-06,
+      "loss": 0.5188,
+      "step": 12065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8526815821644298,
+      "learning_rate": 4.7165171942316855e-06,
+      "loss": 0.4592,
+      "step": 12066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5678414519438348,
+      "learning_rate": 4.715876054477798e-06,
+      "loss": 0.4023,
+      "step": 12067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.697335366684724,
+      "learning_rate": 4.715234931139059e-06,
+      "loss": 0.5089,
+      "step": 12068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8094410032634243,
+      "learning_rate": 4.714593824228888e-06,
+      "loss": 0.471,
+      "step": 12069
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4747469631262244,
+      "learning_rate": 4.713952733760707e-06,
+      "loss": 0.3723,
+      "step": 12070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.121509479498362,
+      "learning_rate": 4.713311659747938e-06,
+      "loss": 0.4884,
+      "step": 12071
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.600669525249108,
+      "learning_rate": 4.712670602204002e-06,
+      "loss": 0.4732,
+      "step": 12072
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6522383955582387,
+      "learning_rate": 4.712029561142316e-06,
+      "loss": 0.5186,
+      "step": 12073
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9068509725423104,
+      "learning_rate": 4.711388536576302e-06,
+      "loss": 0.6658,
+      "step": 12074
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.518661973953965,
+      "learning_rate": 4.71074752851938e-06,
+      "loss": 0.4087,
+      "step": 12075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7174962757264474,
+      "learning_rate": 4.710106536984969e-06,
+      "loss": 0.5132,
+      "step": 12076
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6222994594072575,
+      "learning_rate": 4.7094655619864845e-06,
+      "loss": 0.3861,
+      "step": 12077
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7366061970662163,
+      "learning_rate": 4.708824603537349e-06,
+      "loss": 0.5255,
+      "step": 12078
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7495189131756133,
+      "learning_rate": 4.708183661650979e-06,
+      "loss": 0.4803,
+      "step": 12079
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9758169195479751,
+      "learning_rate": 4.707542736340793e-06,
+      "loss": 0.4318,
+      "step": 12080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5582718720019106,
+      "learning_rate": 4.706901827620207e-06,
+      "loss": 0.4241,
+      "step": 12081
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1002474040949624,
+      "learning_rate": 4.706260935502638e-06,
+      "loss": 0.9313,
+      "step": 12082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6092509035331508,
+      "learning_rate": 4.705620060001503e-06,
+      "loss": 0.5175,
+      "step": 12083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7678803568852686,
+      "learning_rate": 4.704979201130219e-06,
+      "loss": 0.6602,
+      "step": 12084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.842846509687218,
+      "learning_rate": 4.7043383589022015e-06,
+      "loss": 0.5356,
+      "step": 12085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.82394673014517,
+      "learning_rate": 4.703697533330868e-06,
+      "loss": 0.4667,
+      "step": 12086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0947462567185484,
+      "learning_rate": 4.7030567244296296e-06,
+      "loss": 0.5573,
+      "step": 12087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4795186883560234,
+      "learning_rate": 4.702415932211904e-06,
+      "loss": 0.3712,
+      "step": 12088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.100461073985514,
+      "learning_rate": 4.701775156691107e-06,
+      "loss": 0.6222,
+      "step": 12089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7797402500089443,
+      "learning_rate": 4.70113439788065e-06,
+      "loss": 0.4508,
+      "step": 12090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.451377656226112,
+      "learning_rate": 4.700493655793948e-06,
+      "loss": 0.3693,
+      "step": 12091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8207867094481132,
+      "learning_rate": 4.699852930444416e-06,
+      "loss": 0.5274,
+      "step": 12092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6866574146324875,
+      "learning_rate": 4.699212221845465e-06,
+      "loss": 0.5071,
+      "step": 12093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.751789723197815,
+      "learning_rate": 4.698571530010509e-06,
+      "loss": 0.5814,
+      "step": 12094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5699426071517206,
+      "learning_rate": 4.697930854952961e-06,
+      "loss": 0.4312,
+      "step": 12095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5735436979599795,
+      "learning_rate": 4.697290196686231e-06,
+      "loss": 0.3642,
+      "step": 12096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.247106018607124,
+      "learning_rate": 4.696649555223733e-06,
+      "loss": 0.3911,
+      "step": 12097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.551211354330986,
+      "learning_rate": 4.696008930578878e-06,
+      "loss": 0.3492,
+      "step": 12098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0898599674536897,
+      "learning_rate": 4.695368322765076e-06,
+      "loss": 0.5381,
+      "step": 12099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.94809056413324,
+      "learning_rate": 4.694727731795738e-06,
+      "loss": 0.5259,
+      "step": 12100
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.5157272815704346,
+      "eval_runtime": 67.9743,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 12100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7834060651787809,
+      "learning_rate": 4.694087157684275e-06,
+      "loss": 0.4723,
+      "step": 12101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6036313790998926,
+      "learning_rate": 4.693446600444098e-06,
+      "loss": 0.481,
+      "step": 12102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5938882189209327,
+      "learning_rate": 4.692806060088614e-06,
+      "loss": 0.4051,
+      "step": 12103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6546041826697706,
+      "learning_rate": 4.692165536631234e-06,
+      "loss": 0.464,
+      "step": 12104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.463498413690016,
+      "learning_rate": 4.691525030085368e-06,
+      "loss": 0.3703,
+      "step": 12105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.847395259967534,
+      "learning_rate": 4.690884540464421e-06,
+      "loss": 0.4324,
+      "step": 12106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5698347600357367,
+      "learning_rate": 4.690244067781805e-06,
+      "loss": 0.4317,
+      "step": 12107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5646378814829058,
+      "learning_rate": 4.689603612050927e-06,
+      "loss": 0.4576,
+      "step": 12108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8564571814799704,
+      "learning_rate": 4.688963173285192e-06,
+      "loss": 0.5534,
+      "step": 12109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.611766157960277,
+      "learning_rate": 4.688322751498011e-06,
+      "loss": 0.5028,
+      "step": 12110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9938558729669664,
+      "learning_rate": 4.6876823467027885e-06,
+      "loss": 0.5405,
+      "step": 12111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7315354294577423,
+      "learning_rate": 4.687041958912932e-06,
+      "loss": 0.4068,
+      "step": 12112
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.772454928134875,
+      "learning_rate": 4.6864015881418475e-06,
+      "loss": 0.514,
+      "step": 12113
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1075374284457893,
+      "learning_rate": 4.685761234402939e-06,
+      "loss": 0.5894,
+      "step": 12114
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0286550972285897,
+      "learning_rate": 4.685120897709614e-06,
+      "loss": 0.483,
+      "step": 12115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5019038269056386,
+      "learning_rate": 4.684480578075278e-06,
+      "loss": 0.4213,
+      "step": 12116
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5947984365898162,
+      "learning_rate": 4.683840275513333e-06,
+      "loss": 0.4404,
+      "step": 12117
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25652194048055,
+      "learning_rate": 4.683199990037186e-06,
+      "loss": 0.5117,
+      "step": 12118
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7225604889805026,
+      "learning_rate": 4.682559721660238e-06,
+      "loss": 0.4146,
+      "step": 12119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9328982866836721,
+      "learning_rate": 4.681919470395896e-06,
+      "loss": 0.4364,
+      "step": 12120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5620269006878753,
+      "learning_rate": 4.6812792362575624e-06,
+      "loss": 0.4989,
+      "step": 12121
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9343024362223473,
+      "learning_rate": 4.680639019258638e-06,
+      "loss": 0.5553,
+      "step": 12122
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6266211619177997,
+      "learning_rate": 4.679998819412528e-06,
+      "loss": 0.4555,
+      "step": 12123
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.44919178864391,
+      "learning_rate": 4.679358636732632e-06,
+      "loss": 0.5668,
+      "step": 12124
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4214203668739371,
+      "learning_rate": 4.678718471232354e-06,
+      "loss": 0.4254,
+      "step": 12125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.854193926033104,
+      "learning_rate": 4.678078322925095e-06,
+      "loss": 0.5711,
+      "step": 12126
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9780718185130837,
+      "learning_rate": 4.677438191824255e-06,
+      "loss": 0.5953,
+      "step": 12127
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7784541517905719,
+      "learning_rate": 4.676798077943237e-06,
+      "loss": 0.5574,
+      "step": 12128
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3983833674877357,
+      "learning_rate": 4.6761579812954385e-06,
+      "loss": 0.3782,
+      "step": 12129
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.040903127477676,
+      "learning_rate": 4.675517901894261e-06,
+      "loss": 0.6181,
+      "step": 12130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6296835996061556,
+      "learning_rate": 4.674877839753105e-06,
+      "loss": 0.3905,
+      "step": 12131
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8563891225599827,
+      "learning_rate": 4.674237794885368e-06,
+      "loss": 0.4231,
+      "step": 12132
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.609845772238922,
+      "learning_rate": 4.67359776730445e-06,
+      "loss": 0.4104,
+      "step": 12133
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.911338812298708,
+      "learning_rate": 4.672957757023751e-06,
+      "loss": 0.4759,
+      "step": 12134
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5702252188810704,
+      "learning_rate": 4.6723177640566654e-06,
+      "loss": 0.488,
+      "step": 12135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4407020385446545,
+      "learning_rate": 4.6716777884165945e-06,
+      "loss": 0.3571,
+      "step": 12136
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5223306119821371,
+      "learning_rate": 4.671037830116932e-06,
+      "loss": 0.4861,
+      "step": 12137
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.441274725517821,
+      "learning_rate": 4.670397889171081e-06,
+      "loss": 0.3908,
+      "step": 12138
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5683091765384876,
+      "learning_rate": 4.669757965592434e-06,
+      "loss": 0.3847,
+      "step": 12139
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4066951703509811,
+      "learning_rate": 4.669118059394388e-06,
+      "loss": 0.3344,
+      "step": 12140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7204853960548632,
+      "learning_rate": 4.668478170590339e-06,
+      "loss": 0.4607,
+      "step": 12141
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7442787393667274,
+      "learning_rate": 4.667838299193684e-06,
+      "loss": 0.5055,
+      "step": 12142
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0397290021381203,
+      "learning_rate": 4.667198445217817e-06,
+      "loss": 0.6344,
+      "step": 12143
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7595926750260718,
+      "learning_rate": 4.666558608676133e-06,
+      "loss": 0.537,
+      "step": 12144
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6394463163945696,
+      "learning_rate": 4.665918789582028e-06,
+      "loss": 0.5313,
+      "step": 12145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.063452257672542,
+      "learning_rate": 4.665278987948894e-06,
+      "loss": 0.4593,
+      "step": 12146
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6426131272244138,
+      "learning_rate": 4.664639203790125e-06,
+      "loss": 0.403,
+      "step": 12147
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.312007038404432,
+      "learning_rate": 4.6639994371191175e-06,
+      "loss": 0.4828,
+      "step": 12148
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8129768765924108,
+      "learning_rate": 4.663359687949261e-06,
+      "loss": 0.4848,
+      "step": 12149
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9021821971114763,
+      "learning_rate": 4.66271995629395e-06,
+      "loss": 0.5195,
+      "step": 12150
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.5158646702766418,
+      "eval_runtime": 67.9575,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.575,
+      "step": 12150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4750311354566081,
+      "learning_rate": 4.6620802421665766e-06,
+      "loss": 0.4527,
+      "step": 12151
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0084798243912454,
+      "learning_rate": 4.661440545580534e-06,
+      "loss": 0.5179,
+      "step": 12152
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6901694327125913,
+      "learning_rate": 4.6608008665492135e-06,
+      "loss": 0.5422,
+      "step": 12153
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3611738069348651,
+      "learning_rate": 4.660161205086002e-06,
+      "loss": 0.3792,
+      "step": 12154
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9772972765597645,
+      "learning_rate": 4.6595215612042965e-06,
+      "loss": 0.4466,
+      "step": 12155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0683921292348155,
+      "learning_rate": 4.658881934917485e-06,
+      "loss": 0.5607,
+      "step": 12156
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5538491901246898,
+      "learning_rate": 4.6582423262389575e-06,
+      "loss": 0.4372,
+      "step": 12157
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8000522476113703,
+      "learning_rate": 4.657602735182104e-06,
+      "loss": 0.5155,
+      "step": 12158
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8034069985073988,
+      "learning_rate": 4.656963161760314e-06,
+      "loss": 0.4226,
+      "step": 12159
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6343909124600102,
+      "learning_rate": 4.656323605986977e-06,
+      "loss": 0.4954,
+      "step": 12160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.691055780601486,
+      "learning_rate": 4.6556840678754815e-06,
+      "loss": 0.4027,
+      "step": 12161
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8224659922804127,
+      "learning_rate": 4.655044547439214e-06,
+      "loss": 0.5229,
+      "step": 12162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7120301163567202,
+      "learning_rate": 4.654405044691566e-06,
+      "loss": 0.5846,
+      "step": 12163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8019299301072722,
+      "learning_rate": 4.6537655596459206e-06,
+      "loss": 0.4579,
+      "step": 12164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.81430331987034,
+      "learning_rate": 4.6531260923156694e-06,
+      "loss": 0.5676,
+      "step": 12165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.097294448818273,
+      "learning_rate": 4.652486642714199e-06,
+      "loss": 0.5525,
+      "step": 12166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6490484413672812,
+      "learning_rate": 4.651847210854892e-06,
+      "loss": 0.4009,
+      "step": 12167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6038940577986947,
+      "learning_rate": 4.651207796751137e-06,
+      "loss": 0.438,
+      "step": 12168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7372374074238055,
+      "learning_rate": 4.650568400416321e-06,
+      "loss": 0.4558,
+      "step": 12169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.538332944305103,
+      "learning_rate": 4.649929021863826e-06,
+      "loss": 0.4598,
+      "step": 12170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7101546779760577,
+      "learning_rate": 4.649289661107041e-06,
+      "loss": 0.6465,
+      "step": 12171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.868190965322715,
+      "learning_rate": 4.648650318159346e-06,
+      "loss": 0.6598,
+      "step": 12172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.727229219293307,
+      "learning_rate": 4.64801099303413e-06,
+      "loss": 0.4695,
+      "step": 12173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.711034361394893,
+      "learning_rate": 4.6473716857447735e-06,
+      "loss": 0.5194,
+      "step": 12174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.838694761928603,
+      "learning_rate": 4.646732396304662e-06,
+      "loss": 0.5017,
+      "step": 12175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.63926851041874,
+      "learning_rate": 4.6460931247271764e-06,
+      "loss": 0.4816,
+      "step": 12176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6019403372891734,
+      "learning_rate": 4.645453871025701e-06,
+      "loss": 0.5291,
+      "step": 12177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5904493585647943,
+      "learning_rate": 4.644814635213621e-06,
+      "loss": 0.5064,
+      "step": 12178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5616608178589555,
+      "learning_rate": 4.644175417304313e-06,
+      "loss": 0.592,
+      "step": 12179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4408832612298337,
+      "learning_rate": 4.643536217311161e-06,
+      "loss": 0.3889,
+      "step": 12180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6910863502688,
+      "learning_rate": 4.642897035247547e-06,
+      "loss": 0.5146,
+      "step": 12181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7999150324214963,
+      "learning_rate": 4.6422578711268506e-06,
+      "loss": 0.5628,
+      "step": 12182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.931775706758941,
+      "learning_rate": 4.641618724962453e-06,
+      "loss": 0.458,
+      "step": 12183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0700573467194965,
+      "learning_rate": 4.640979596767735e-06,
+      "loss": 0.5723,
+      "step": 12184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1010220431084754,
+      "learning_rate": 4.6403404865560746e-06,
+      "loss": 0.7406,
+      "step": 12185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9001972986603355,
+      "learning_rate": 4.639701394340851e-06,
+      "loss": 0.4329,
+      "step": 12186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.895473134113773,
+      "learning_rate": 4.6390623201354455e-06,
+      "loss": 0.4859,
+      "step": 12187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8891206272615229,
+      "learning_rate": 4.638423263953237e-06,
+      "loss": 0.6154,
+      "step": 12188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8958944934926645,
+      "learning_rate": 4.6377842258076e-06,
+      "loss": 0.5762,
+      "step": 12189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6126109672914641,
+      "learning_rate": 4.637145205711917e-06,
+      "loss": 0.4937,
+      "step": 12190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7560348317215198,
+      "learning_rate": 4.636506203679562e-06,
+      "loss": 0.606,
+      "step": 12191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.214861464839572,
+      "learning_rate": 4.635867219723914e-06,
+      "loss": 0.5827,
+      "step": 12192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3206610097352054,
+      "learning_rate": 4.6352282538583485e-06,
+      "loss": 0.6627,
+      "step": 12193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.045948880683335,
+      "learning_rate": 4.634589306096241e-06,
+      "loss": 0.6238,
+      "step": 12194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6458958906741625,
+      "learning_rate": 4.633950376450971e-06,
+      "loss": 0.4728,
+      "step": 12195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6786963949368074,
+      "learning_rate": 4.6333114649359115e-06,
+      "loss": 0.6524,
+      "step": 12196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7031592435117306,
+      "learning_rate": 4.632672571564437e-06,
+      "loss": 0.442,
+      "step": 12197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.579041244900152,
+      "learning_rate": 4.632033696349925e-06,
+      "loss": 0.385,
+      "step": 12198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.639104438678272,
+      "learning_rate": 4.631394839305747e-06,
+      "loss": 0.5383,
+      "step": 12199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8917821383909974,
+      "learning_rate": 4.63075600044528e-06,
+      "loss": 0.5227,
+      "step": 12200
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.5157405734062195,
+      "eval_runtime": 67.9838,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 12200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6741235657037437,
+      "learning_rate": 4.630117179781897e-06,
+      "loss": 0.5956,
+      "step": 12201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0082934176199165,
+      "learning_rate": 4.629478377328967e-06,
+      "loss": 0.5426,
+      "step": 12202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8091274605252352,
+      "learning_rate": 4.62883959309987e-06,
+      "loss": 0.5236,
+      "step": 12203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5762624670022833,
+      "learning_rate": 4.628200827107973e-06,
+      "loss": 0.3783,
+      "step": 12204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6901733810645292,
+      "learning_rate": 4.62756207936665e-06,
+      "loss": 0.5078,
+      "step": 12205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8194683186615148,
+      "learning_rate": 4.626923349889274e-06,
+      "loss": 0.4382,
+      "step": 12206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7383416776224232,
+      "learning_rate": 4.626284638689214e-06,
+      "loss": 0.4485,
+      "step": 12207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7734311369119162,
+      "learning_rate": 4.625645945779845e-06,
+      "loss": 0.4646,
+      "step": 12208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9304359901456125,
+      "learning_rate": 4.625007271174534e-06,
+      "loss": 0.5165,
+      "step": 12209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.465667922039754,
+      "learning_rate": 4.62436861488665e-06,
+      "loss": 0.4551,
+      "step": 12210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6346743349013608,
+      "learning_rate": 4.6237299769295685e-06,
+      "loss": 0.4802,
+      "step": 12211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.50266651477881,
+      "learning_rate": 4.623091357316654e-06,
+      "loss": 0.3685,
+      "step": 12212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4417688052408069,
+      "learning_rate": 4.622452756061277e-06,
+      "loss": 0.414,
+      "step": 12213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8400062306720073,
+      "learning_rate": 4.621814173176807e-06,
+      "loss": 0.5832,
+      "step": 12214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8824095222343338,
+      "learning_rate": 4.621175608676613e-06,
+      "loss": 0.5655,
+      "step": 12215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5575382419832093,
+      "learning_rate": 4.62053706257406e-06,
+      "loss": 0.3701,
+      "step": 12216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9237653206711507,
+      "learning_rate": 4.619898534882517e-06,
+      "loss": 0.4902,
+      "step": 12217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8283332410585587,
+      "learning_rate": 4.619260025615353e-06,
+      "loss": 0.438,
+      "step": 12218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.703832747990675,
+      "learning_rate": 4.618621534785933e-06,
+      "loss": 0.47,
+      "step": 12219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1458223906502303,
+      "learning_rate": 4.617983062407623e-06,
+      "loss": 0.5581,
+      "step": 12220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.788584934524354,
+      "learning_rate": 4.61734460849379e-06,
+      "loss": 0.5498,
+      "step": 12221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.414121104100082,
+      "learning_rate": 4.616706173057801e-06,
+      "loss": 0.6899,
+      "step": 12222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.783893879796849,
+      "learning_rate": 4.616067756113017e-06,
+      "loss": 0.4779,
+      "step": 12223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6504161511908684,
+      "learning_rate": 4.615429357672807e-06,
+      "loss": 0.4931,
+      "step": 12224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8124590267952132,
+      "learning_rate": 4.614790977750533e-06,
+      "loss": 0.4685,
+      "step": 12225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5361326296974445,
+      "learning_rate": 4.614152616359562e-06,
+      "loss": 0.3362,
+      "step": 12226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7036539002142834,
+      "learning_rate": 4.613514273513254e-06,
+      "loss": 0.6171,
+      "step": 12227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3388179159709304,
+      "learning_rate": 4.612875949224974e-06,
+      "loss": 0.3664,
+      "step": 12228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7533365783359451,
+      "learning_rate": 4.612237643508086e-06,
+      "loss": 0.4279,
+      "step": 12229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.588919433430441,
+      "learning_rate": 4.611599356375951e-06,
+      "loss": 0.4834,
+      "step": 12230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1759714680683557,
+      "learning_rate": 4.610961087841932e-06,
+      "loss": 0.5252,
+      "step": 12231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4768899172786234,
+      "learning_rate": 4.61032283791939e-06,
+      "loss": 0.4176,
+      "step": 12232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2830054807363878,
+      "learning_rate": 4.609684606621687e-06,
+      "loss": 0.4739,
+      "step": 12233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7698205740720694,
+      "learning_rate": 4.609046393962183e-06,
+      "loss": 0.5089,
+      "step": 12234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.829241976187153,
+      "learning_rate": 4.6084081999542415e-06,
+      "loss": 0.514,
+      "step": 12235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0212822940126833,
+      "learning_rate": 4.607770024611216e-06,
+      "loss": 0.5504,
+      "step": 12236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4896236092980475,
+      "learning_rate": 4.607131867946475e-06,
+      "loss": 0.5807,
+      "step": 12237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7827399688929217,
+      "learning_rate": 4.6064937299733724e-06,
+      "loss": 0.5035,
+      "step": 12238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9224291470955766,
+      "learning_rate": 4.605855610705268e-06,
+      "loss": 0.5287,
+      "step": 12239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.076845306961399,
+      "learning_rate": 4.605217510155523e-06,
+      "loss": 0.5272,
+      "step": 12240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.899905606846224,
+      "learning_rate": 4.604579428337494e-06,
+      "loss": 0.608,
+      "step": 12241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6294303640748664,
+      "learning_rate": 4.603941365264536e-06,
+      "loss": 0.4337,
+      "step": 12242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5718181330423528,
+      "learning_rate": 4.60330332095001e-06,
+      "loss": 0.5169,
+      "step": 12243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1379790498118822,
+      "learning_rate": 4.602665295407271e-06,
+      "loss": 0.4292,
+      "step": 12244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2225330621137442,
+      "learning_rate": 4.6020272886496785e-06,
+      "loss": 0.6337,
+      "step": 12245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7229101607545259,
+      "learning_rate": 4.601389300690586e-06,
+      "loss": 0.5164,
+      "step": 12246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7475793927382473,
+      "learning_rate": 4.60075133154335e-06,
+      "loss": 0.5728,
+      "step": 12247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6110264689238358,
+      "learning_rate": 4.600113381221328e-06,
+      "loss": 0.5004,
+      "step": 12248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9286901966490175,
+      "learning_rate": 4.599475449737872e-06,
+      "loss": 0.4085,
+      "step": 12249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6033382787555523,
+      "learning_rate": 4.598837537106337e-06,
+      "loss": 0.339,
+      "step": 12250
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.5160401463508606,
+      "eval_runtime": 67.9125,
+      "eval_samples_per_second": 50.212,
+      "eval_steps_per_second": 1.576,
+      "step": 12250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.957785292377674,
+      "learning_rate": 4.598199643340081e-06,
+      "loss": 0.6074,
+      "step": 12251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.340350273882454,
+      "learning_rate": 4.5975617684524546e-06,
+      "loss": 0.6401,
+      "step": 12252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9018722591935142,
+      "learning_rate": 4.596923912456811e-06,
+      "loss": 0.5012,
+      "step": 12253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4867334229712075,
+      "learning_rate": 4.596286075366506e-06,
+      "loss": 0.4924,
+      "step": 12254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5598101439048746,
+      "learning_rate": 4.595648257194889e-06,
+      "loss": 0.4386,
+      "step": 12255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4629994899458623,
+      "learning_rate": 4.595010457955315e-06,
+      "loss": 0.4313,
+      "step": 12256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7451915366499013,
+      "learning_rate": 4.594372677661134e-06,
+      "loss": 0.6012,
+      "step": 12257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7117816972369129,
+      "learning_rate": 4.593734916325699e-06,
+      "loss": 0.5981,
+      "step": 12258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8554409027756975,
+      "learning_rate": 4.593097173962362e-06,
+      "loss": 0.4564,
+      "step": 12259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.366306344407331,
+      "learning_rate": 4.592459450584471e-06,
+      "loss": 0.3639,
+      "step": 12260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4997370910474044,
+      "learning_rate": 4.591821746205378e-06,
+      "loss": 0.4348,
+      "step": 12261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.833463869602267,
+      "learning_rate": 4.5911840608384326e-06,
+      "loss": 0.5732,
+      "step": 12262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7000059430740804,
+      "learning_rate": 4.590546394496984e-06,
+      "loss": 0.5297,
+      "step": 12263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6009714853565993,
+      "learning_rate": 4.589908747194383e-06,
+      "loss": 0.4581,
+      "step": 12264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.014818950832485,
+      "learning_rate": 4.589271118943976e-06,
+      "loss": 0.6479,
+      "step": 12265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5745576487752715,
+      "learning_rate": 4.588633509759112e-06,
+      "loss": 0.47,
+      "step": 12266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.646710190711981,
+      "learning_rate": 4.5879959196531396e-06,
+      "loss": 0.4009,
+      "step": 12267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4606422132825443,
+      "learning_rate": 4.587358348639407e-06,
+      "loss": 0.3995,
+      "step": 12268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5760773851596093,
+      "learning_rate": 4.58672079673126e-06,
+      "loss": 0.424,
+      "step": 12269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8254283120164545,
+      "learning_rate": 4.586083263942046e-06,
+      "loss": 0.5406,
+      "step": 12270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7984143244246398,
+      "learning_rate": 4.585445750285112e-06,
+      "loss": 0.4431,
+      "step": 12271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7228698722590265,
+      "learning_rate": 4.5848082557738016e-06,
+      "loss": 0.4345,
+      "step": 12272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6206174014175496,
+      "learning_rate": 4.584170780421462e-06,
+      "loss": 0.5008,
+      "step": 12273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9671985278298199,
+      "learning_rate": 4.5835333242414395e-06,
+      "loss": 0.6428,
+      "step": 12274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.713199511402438,
+      "learning_rate": 4.582895887247078e-06,
+      "loss": 0.3868,
+      "step": 12275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8839422290804193,
+      "learning_rate": 4.5822584694517195e-06,
+      "loss": 0.4139,
+      "step": 12276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8668142004932395,
+      "learning_rate": 4.581621070868712e-06,
+      "loss": 0.4554,
+      "step": 12277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9426677156159933,
+      "learning_rate": 4.5809836915113945e-06,
+      "loss": 0.4997,
+      "step": 12278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4463632494878575,
+      "learning_rate": 4.580346331393114e-06,
+      "loss": 0.391,
+      "step": 12279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9256339520290764,
+      "learning_rate": 4.579708990527212e-06,
+      "loss": 0.4329,
+      "step": 12280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9858405030431563,
+      "learning_rate": 4.57907166892703e-06,
+      "loss": 0.485,
+      "step": 12281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8569814605579322,
+      "learning_rate": 4.578434366605912e-06,
+      "loss": 0.4355,
+      "step": 12282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7545380287328876,
+      "learning_rate": 4.577797083577196e-06,
+      "loss": 0.4751,
+      "step": 12283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5610482691712495,
+      "learning_rate": 4.577159819854227e-06,
+      "loss": 0.4266,
+      "step": 12284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7910388172132836,
+      "learning_rate": 4.5765225754503416e-06,
+      "loss": 0.4851,
+      "step": 12285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.664184033508571,
+      "learning_rate": 4.575885350378885e-06,
+      "loss": 0.552,
+      "step": 12286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6964434526892473,
+      "learning_rate": 4.575248144653194e-06,
+      "loss": 0.4438,
+      "step": 12287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7437510586016822,
+      "learning_rate": 4.574610958286608e-06,
+      "loss": 0.5256,
+      "step": 12288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.584981997304609,
+      "learning_rate": 4.573973791292468e-06,
+      "loss": 0.5159,
+      "step": 12289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1039143929289,
+      "learning_rate": 4.573336643684111e-06,
+      "loss": 0.5771,
+      "step": 12290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8443452893723118,
+      "learning_rate": 4.572699515474876e-06,
+      "loss": 0.4237,
+      "step": 12291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.35487194367733,
+      "learning_rate": 4.572062406678099e-06,
+      "loss": 0.3587,
+      "step": 12292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9473431614881642,
+      "learning_rate": 4.571425317307121e-06,
+      "loss": 0.5234,
+      "step": 12293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5984387910649192,
+      "learning_rate": 4.570788247375278e-06,
+      "loss": 0.4941,
+      "step": 12294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.568379248739168,
+      "learning_rate": 4.570151196895904e-06,
+      "loss": 0.457,
+      "step": 12295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7304022555631018,
+      "learning_rate": 4.569514165882339e-06,
+      "loss": 0.4407,
+      "step": 12296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2688240467049487,
+      "learning_rate": 4.568877154347914e-06,
+      "loss": 0.564,
+      "step": 12297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.867233158747734,
+      "learning_rate": 4.568240162305969e-06,
+      "loss": 0.5172,
+      "step": 12298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8101568181401493,
+      "learning_rate": 4.567603189769838e-06,
+      "loss": 0.5085,
+      "step": 12299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.048388586302161,
+      "learning_rate": 4.566966236752855e-06,
+      "loss": 0.625,
+      "step": 12300
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5141030550003052,
+      "eval_runtime": 67.9492,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 12300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.626791554324591,
+      "learning_rate": 4.566329303268354e-06,
+      "loss": 0.5127,
+      "step": 12301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4517160075357882,
+      "learning_rate": 4.56569238932967e-06,
+      "loss": 0.7767,
+      "step": 12302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8090247853807666,
+      "learning_rate": 4.565055494950133e-06,
+      "loss": 0.4376,
+      "step": 12303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8774697410918242,
+      "learning_rate": 4.564418620143082e-06,
+      "loss": 0.5675,
+      "step": 12304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.879515010841408,
+      "learning_rate": 4.563781764921843e-06,
+      "loss": 0.5661,
+      "step": 12305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8004136065315657,
+      "learning_rate": 4.563144929299753e-06,
+      "loss": 0.6281,
+      "step": 12306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9643745455298358,
+      "learning_rate": 4.562508113290142e-06,
+      "loss": 0.6519,
+      "step": 12307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1234557085245616,
+      "learning_rate": 4.561871316906341e-06,
+      "loss": 0.676,
+      "step": 12308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.315428700955107,
+      "learning_rate": 4.561234540161681e-06,
+      "loss": 0.3468,
+      "step": 12309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8689870459950169,
+      "learning_rate": 4.560597783069492e-06,
+      "loss": 0.5229,
+      "step": 12310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3723911328366694,
+      "learning_rate": 4.559961045643107e-06,
+      "loss": 0.4162,
+      "step": 12311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9882760649279252,
+      "learning_rate": 4.559324327895853e-06,
+      "loss": 0.5151,
+      "step": 12312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7383470827105956,
+      "learning_rate": 4.558687629841059e-06,
+      "loss": 0.5964,
+      "step": 12313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5214608072552103,
+      "learning_rate": 4.558050951492056e-06,
+      "loss": 0.3977,
+      "step": 12314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8650674557905011,
+      "learning_rate": 4.55741429286217e-06,
+      "loss": 0.5762,
+      "step": 12315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7386626133200944,
+      "learning_rate": 4.55677765396473e-06,
+      "loss": 0.5833,
+      "step": 12316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7025347968659852,
+      "learning_rate": 4.556141034813065e-06,
+      "loss": 0.5446,
+      "step": 12317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.137228057445583,
+      "learning_rate": 4.555504435420501e-06,
+      "loss": 0.4427,
+      "step": 12318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5886647340553772,
+      "learning_rate": 4.554867855800364e-06,
+      "loss": 0.3495,
+      "step": 12319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7622528934860109,
+      "learning_rate": 4.554231295965984e-06,
+      "loss": 0.5358,
+      "step": 12320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7341049506453494,
+      "learning_rate": 4.553594755930682e-06,
+      "loss": 0.4974,
+      "step": 12321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3860504512272982,
+      "learning_rate": 4.552958235707787e-06,
+      "loss": 0.3919,
+      "step": 12322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.138899166821137,
+      "learning_rate": 4.552321735310622e-06,
+      "loss": 0.5571,
+      "step": 12323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7801155363275294,
+      "learning_rate": 4.551685254752515e-06,
+      "loss": 0.5137,
+      "step": 12324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.020636675738991,
+      "learning_rate": 4.551048794046787e-06,
+      "loss": 0.5099,
+      "step": 12325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8043520696041975,
+      "learning_rate": 4.550412353206762e-06,
+      "loss": 0.4577,
+      "step": 12326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7513002281344572,
+      "learning_rate": 4.549775932245766e-06,
+      "loss": 0.3968,
+      "step": 12327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4510083863410237,
+      "learning_rate": 4.54913953117712e-06,
+      "loss": 0.3473,
+      "step": 12328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1117325737299035,
+      "learning_rate": 4.5485031500141475e-06,
+      "loss": 0.6042,
+      "step": 12329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4411715121106823,
+      "learning_rate": 4.547866788770171e-06,
+      "loss": 0.3282,
+      "step": 12330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6472883588521825,
+      "learning_rate": 4.54723044745851e-06,
+      "loss": 0.443,
+      "step": 12331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8427757871547672,
+      "learning_rate": 4.546594126092491e-06,
+      "loss": 0.5671,
+      "step": 12332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.442010463711657,
+      "learning_rate": 4.54595782468543e-06,
+      "loss": 0.3881,
+      "step": 12333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9744515110411125,
+      "learning_rate": 4.545321543250647e-06,
+      "loss": 0.5352,
+      "step": 12334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7666512954475377,
+      "learning_rate": 4.544685281801466e-06,
+      "loss": 0.5169,
+      "step": 12335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7167167895358544,
+      "learning_rate": 4.544049040351207e-06,
+      "loss": 0.4835,
+      "step": 12336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0199400216315166,
+      "learning_rate": 4.543412818913186e-06,
+      "loss": 0.4585,
+      "step": 12337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8244710306901182,
+      "learning_rate": 4.5427766175007245e-06,
+      "loss": 0.4815,
+      "step": 12338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6999976218634403,
+      "learning_rate": 4.54214043612714e-06,
+      "loss": 0.4995,
+      "step": 12339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.07888886149415,
+      "learning_rate": 4.541504274805749e-06,
+      "loss": 0.6653,
+      "step": 12340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7988531942130859,
+      "learning_rate": 4.540868133549873e-06,
+      "loss": 0.387,
+      "step": 12341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4428936836085522,
+      "learning_rate": 4.5402320123728265e-06,
+      "loss": 0.5167,
+      "step": 12342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6534414903697807,
+      "learning_rate": 4.539595911287925e-06,
+      "loss": 0.4889,
+      "step": 12343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8941493188156004,
+      "learning_rate": 4.538959830308488e-06,
+      "loss": 0.5472,
+      "step": 12344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.432522540217135,
+      "learning_rate": 4.5383237694478275e-06,
+      "loss": 0.315,
+      "step": 12345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4501063850963443,
+      "learning_rate": 4.537687728719264e-06,
+      "loss": 0.4011,
+      "step": 12346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4714819224633777,
+      "learning_rate": 4.537051708136111e-06,
+      "loss": 0.5047,
+      "step": 12347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3936279552458615,
+      "learning_rate": 4.536415707711681e-06,
+      "loss": 0.3846,
+      "step": 12348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7552649696388274,
+      "learning_rate": 4.53577972745929e-06,
+      "loss": 0.4551,
+      "step": 12349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8445919544845473,
+      "learning_rate": 4.535143767392251e-06,
+      "loss": 0.573,
+      "step": 12350
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5139021277427673,
+      "eval_runtime": 67.9215,
+      "eval_samples_per_second": 50.205,
+      "eval_steps_per_second": 1.575,
+      "step": 12350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4152805489460059,
+      "learning_rate": 4.534507827523877e-06,
+      "loss": 0.5184,
+      "step": 12351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.55994336384964,
+      "learning_rate": 4.533871907867484e-06,
+      "loss": 0.4212,
+      "step": 12352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5867955816725112,
+      "learning_rate": 4.5332360084363815e-06,
+      "loss": 0.4228,
+      "step": 12353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.018518824575166,
+      "learning_rate": 4.532600129243884e-06,
+      "loss": 0.6397,
+      "step": 12354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7085888242129093,
+      "learning_rate": 4.531964270303301e-06,
+      "loss": 0.534,
+      "step": 12355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1221585185605525,
+      "learning_rate": 4.531328431627944e-06,
+      "loss": 0.4892,
+      "step": 12356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.911674522199076,
+      "learning_rate": 4.530692613231125e-06,
+      "loss": 0.4749,
+      "step": 12357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.121883764656834,
+      "learning_rate": 4.5300568151261535e-06,
+      "loss": 0.5401,
+      "step": 12358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.104173496318747,
+      "learning_rate": 4.529421037326342e-06,
+      "loss": 0.5063,
+      "step": 12359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.623709988239414,
+      "learning_rate": 4.528785279844997e-06,
+      "loss": 0.4072,
+      "step": 12360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6361879911036292,
+      "learning_rate": 4.528149542695428e-06,
+      "loss": 0.4999,
+      "step": 12361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7087203735480736,
+      "learning_rate": 4.5275138258909464e-06,
+      "loss": 0.489,
+      "step": 12362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5676848396215757,
+      "learning_rate": 4.526878129444857e-06,
+      "loss": 0.4195,
+      "step": 12363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9819549123633944,
+      "learning_rate": 4.526242453370472e-06,
+      "loss": 0.4714,
+      "step": 12364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4070108444407583,
+      "learning_rate": 4.5256067976810944e-06,
+      "loss": 0.5427,
+      "step": 12365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8548981104864928,
+      "learning_rate": 4.524971162390034e-06,
+      "loss": 0.4506,
+      "step": 12366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8115337810031842,
+      "learning_rate": 4.524335547510597e-06,
+      "loss": 0.4318,
+      "step": 12367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.648196523321305,
+      "learning_rate": 4.523699953056088e-06,
+      "loss": 0.4948,
+      "step": 12368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6150073690486064,
+      "learning_rate": 4.523064379039815e-06,
+      "loss": 0.3625,
+      "step": 12369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6838258346507464,
+      "learning_rate": 4.522428825475084e-06,
+      "loss": 0.5402,
+      "step": 12370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8476448325320816,
+      "learning_rate": 4.521793292375195e-06,
+      "loss": 0.4387,
+      "step": 12371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7133578973655124,
+      "learning_rate": 4.521157779753458e-06,
+      "loss": 0.4984,
+      "step": 12372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9228790709071508,
+      "learning_rate": 4.520522287623176e-06,
+      "loss": 0.5926,
+      "step": 12373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0122982135950367,
+      "learning_rate": 4.519886815997648e-06,
+      "loss": 0.4967,
+      "step": 12374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.900430935837052,
+      "learning_rate": 4.519251364890184e-06,
+      "loss": 0.6049,
+      "step": 12375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4227708321393953,
+      "learning_rate": 4.518615934314081e-06,
+      "loss": 0.3834,
+      "step": 12376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7244523437129151,
+      "learning_rate": 4.5179805242826454e-06,
+      "loss": 0.3931,
+      "step": 12377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9096462862632533,
+      "learning_rate": 4.5173451348091794e-06,
+      "loss": 0.3298,
+      "step": 12378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8239239196709343,
+      "learning_rate": 4.516709765906979e-06,
+      "loss": 0.5043,
+      "step": 12379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7614441794204319,
+      "learning_rate": 4.516074417589351e-06,
+      "loss": 0.6018,
+      "step": 12380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6180711869813904,
+      "learning_rate": 4.5154390898695935e-06,
+      "loss": 0.4586,
+      "step": 12381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.523171086238769,
+      "learning_rate": 4.514803782761007e-06,
+      "loss": 0.5835,
+      "step": 12382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5723281533697298,
+      "learning_rate": 4.514168496276889e-06,
+      "loss": 0.5027,
+      "step": 12383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5700853714158518,
+      "learning_rate": 4.513533230430545e-06,
+      "loss": 0.5359,
+      "step": 12384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5323055448260794,
+      "learning_rate": 4.512897985235267e-06,
+      "loss": 0.4708,
+      "step": 12385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4790541096115462,
+      "learning_rate": 4.51226276070436e-06,
+      "loss": 0.4031,
+      "step": 12386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7107202342626366,
+      "learning_rate": 4.511627556851117e-06,
+      "loss": 0.4483,
+      "step": 12387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9311563677494672,
+      "learning_rate": 4.510992373688836e-06,
+      "loss": 0.4453,
+      "step": 12388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8464435326465813,
+      "learning_rate": 4.510357211230817e-06,
+      "loss": 0.5485,
+      "step": 12389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7951190328470978,
+      "learning_rate": 4.509722069490355e-06,
+      "loss": 0.5588,
+      "step": 12390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4985891439097208,
+      "learning_rate": 4.509086948480746e-06,
+      "loss": 0.442,
+      "step": 12391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7106164656682503,
+      "learning_rate": 4.508451848215286e-06,
+      "loss": 0.3936,
+      "step": 12392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7095508742142167,
+      "learning_rate": 4.50781676870727e-06,
+      "loss": 0.448,
+      "step": 12393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8106896736630282,
+      "learning_rate": 4.5071817099699944e-06,
+      "loss": 0.6392,
+      "step": 12394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5950384048859894,
+      "learning_rate": 4.506546672016754e-06,
+      "loss": 0.4835,
+      "step": 12395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.559331114289707,
+      "learning_rate": 4.505911654860841e-06,
+      "loss": 0.4544,
+      "step": 12396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6244038847541968,
+      "learning_rate": 4.5052766585155505e-06,
+      "loss": 0.487,
+      "step": 12397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6332659606092734,
+      "learning_rate": 4.504641682994174e-06,
+      "loss": 0.437,
+      "step": 12398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3748463395374175,
+      "learning_rate": 4.5040067283100075e-06,
+      "loss": 0.351,
+      "step": 12399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7939050257763793,
+      "learning_rate": 4.503371794476341e-06,
+      "loss": 0.4197,
+      "step": 12400
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5129477381706238,
+      "eval_runtime": 67.9015,
+      "eval_samples_per_second": 50.22,
+      "eval_steps_per_second": 1.576,
+      "step": 12400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6621515935721225,
+      "learning_rate": 4.502736881506465e-06,
+      "loss": 0.4511,
+      "step": 12401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5565461735826427,
+      "learning_rate": 4.502101989413676e-06,
+      "loss": 0.4005,
+      "step": 12402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4975125617053193,
+      "learning_rate": 4.5014671182112606e-06,
+      "loss": 0.4255,
+      "step": 12403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6852295138540363,
+      "learning_rate": 4.500832267912511e-06,
+      "loss": 0.5911,
+      "step": 12404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7563774169820912,
+      "learning_rate": 4.500197438530717e-06,
+      "loss": 0.4396,
+      "step": 12405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4967149948964245,
+      "learning_rate": 4.499562630079168e-06,
+      "loss": 0.3866,
+      "step": 12406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.892551989090611,
+      "learning_rate": 4.498927842571155e-06,
+      "loss": 0.4686,
+      "step": 12407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.488005687067376,
+      "learning_rate": 4.498293076019965e-06,
+      "loss": 0.3871,
+      "step": 12408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7388900252287813,
+      "learning_rate": 4.497658330438889e-06,
+      "loss": 0.4051,
+      "step": 12409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9203465844386025,
+      "learning_rate": 4.497023605841212e-06,
+      "loss": 0.4716,
+      "step": 12410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8473323306266756,
+      "learning_rate": 4.4963889022402215e-06,
+      "loss": 0.5374,
+      "step": 12411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.71848524392148,
+      "learning_rate": 4.495754219649209e-06,
+      "loss": 0.6239,
+      "step": 12412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7008102087146306,
+      "learning_rate": 4.495119558081457e-06,
+      "loss": 0.6194,
+      "step": 12413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.509982463360597,
+      "learning_rate": 4.494484917550251e-06,
+      "loss": 0.3991,
+      "step": 12414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3778796400468885,
+      "learning_rate": 4.49385029806888e-06,
+      "loss": 0.6116,
+      "step": 12415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.159072286861902,
+      "learning_rate": 4.493215699650628e-06,
+      "loss": 0.697,
+      "step": 12416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7192911848279504,
+      "learning_rate": 4.4925811223087815e-06,
+      "loss": 0.5658,
+      "step": 12417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6661400627058427,
+      "learning_rate": 4.491946566056622e-06,
+      "loss": 0.3687,
+      "step": 12418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8085697379319814,
+      "learning_rate": 4.491312030907434e-06,
+      "loss": 0.5686,
+      "step": 12419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7385878370654233,
+      "learning_rate": 4.490677516874504e-06,
+      "loss": 0.5772,
+      "step": 12420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.809935829985637,
+      "learning_rate": 4.490043023971112e-06,
+      "loss": 0.5831,
+      "step": 12421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.137307673505339,
+      "learning_rate": 4.489408552210544e-06,
+      "loss": 0.4348,
+      "step": 12422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0810520255968905,
+      "learning_rate": 4.488774101606078e-06,
+      "loss": 0.4339,
+      "step": 12423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5250041101838743,
+      "learning_rate": 4.488139672170998e-06,
+      "loss": 0.3935,
+      "step": 12424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.787947774382038,
+      "learning_rate": 4.487505263918587e-06,
+      "loss": 0.405,
+      "step": 12425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9708894611204157,
+      "learning_rate": 4.486870876862124e-06,
+      "loss": 0.5916,
+      "step": 12426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8366476044138242,
+      "learning_rate": 4.4862365110148895e-06,
+      "loss": 0.6207,
+      "step": 12427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9882168366351525,
+      "learning_rate": 4.485602166390165e-06,
+      "loss": 0.5875,
+      "step": 12428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.995935084203544,
+      "learning_rate": 4.484967843001228e-06,
+      "loss": 0.5989,
+      "step": 12429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0782399724432166,
+      "learning_rate": 4.4843335408613585e-06,
+      "loss": 0.6079,
+      "step": 12430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.081738027608156,
+      "learning_rate": 4.483699259983837e-06,
+      "loss": 0.595,
+      "step": 12431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5334541933543593,
+      "learning_rate": 4.483065000381939e-06,
+      "loss": 0.3811,
+      "step": 12432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5605142858882464,
+      "learning_rate": 4.482430762068941e-06,
+      "loss": 0.3767,
+      "step": 12433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.691261919595403,
+      "learning_rate": 4.481796545058127e-06,
+      "loss": 0.6106,
+      "step": 12434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.988437591177018,
+      "learning_rate": 4.481162349362769e-06,
+      "loss": 0.6562,
+      "step": 12435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.671253512258811,
+      "learning_rate": 4.480528174996143e-06,
+      "loss": 0.4825,
+      "step": 12436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8486633840279827,
+      "learning_rate": 4.4798940219715285e-06,
+      "loss": 0.6248,
+      "step": 12437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.747132630531847,
+      "learning_rate": 4.479259890302196e-06,
+      "loss": 0.6051,
+      "step": 12438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3283487752745375,
+      "learning_rate": 4.478625780001427e-06,
+      "loss": 0.5261,
+      "step": 12439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5449333741593125,
+      "learning_rate": 4.4779916910824915e-06,
+      "loss": 0.4709,
+      "step": 12440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6094122746227577,
+      "learning_rate": 4.477357623558665e-06,
+      "loss": 0.4463,
+      "step": 12441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.821768125237558,
+      "learning_rate": 4.476723577443222e-06,
+      "loss": 0.4556,
+      "step": 12442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.707546923766246,
+      "learning_rate": 4.476089552749435e-06,
+      "loss": 0.5175,
+      "step": 12443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.618171786575948,
+      "learning_rate": 4.475455549490577e-06,
+      "loss": 0.4447,
+      "step": 12444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.079657018118014,
+      "learning_rate": 4.4748215676799226e-06,
+      "loss": 0.6266,
+      "step": 12445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1044913808992707,
+      "learning_rate": 4.474187607330739e-06,
+      "loss": 0.619,
+      "step": 12446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8752207373367997,
+      "learning_rate": 4.473553668456302e-06,
+      "loss": 0.6587,
+      "step": 12447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6558732141793344,
+      "learning_rate": 4.47291975106988e-06,
+      "loss": 0.443,
+      "step": 12448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.032654031090134,
+      "learning_rate": 4.4722858551847485e-06,
+      "loss": 0.5346,
+      "step": 12449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5637446414390077,
+      "learning_rate": 4.471651980814172e-06,
+      "loss": 0.387,
+      "step": 12450
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5136176347732544,
+      "eval_runtime": 67.9378,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 12450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.659174290589412,
+      "learning_rate": 4.471018127971422e-06,
+      "loss": 0.5485,
+      "step": 12451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4470635958899343,
+      "learning_rate": 4.4703842966697695e-06,
+      "loss": 0.4079,
+      "step": 12452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.592214932567677,
+      "learning_rate": 4.4697504869224825e-06,
+      "loss": 0.4105,
+      "step": 12453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2703581520363822,
+      "learning_rate": 4.469116698742828e-06,
+      "loss": 0.3568,
+      "step": 12454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6015473418231498,
+      "learning_rate": 4.468482932144076e-06,
+      "loss": 0.3874,
+      "step": 12455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7588068633224252,
+      "learning_rate": 4.467849187139492e-06,
+      "loss": 0.4394,
+      "step": 12456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5657372308902433,
+      "learning_rate": 4.467215463742346e-06,
+      "loss": 0.366,
+      "step": 12457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.544456772410071,
+      "learning_rate": 4.466581761965901e-06,
+      "loss": 0.4372,
+      "step": 12458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.591820990144071,
+      "learning_rate": 4.465948081823425e-06,
+      "loss": 0.4817,
+      "step": 12459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8614655899901476,
+      "learning_rate": 4.465314423328185e-06,
+      "loss": 0.5328,
+      "step": 12460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4141868300792018,
+      "learning_rate": 4.4646807864934435e-06,
+      "loss": 0.3756,
+      "step": 12461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6639014095850562,
+      "learning_rate": 4.464047171332468e-06,
+      "loss": 0.5178,
+      "step": 12462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8128728482945011,
+      "learning_rate": 4.463413577858522e-06,
+      "loss": 0.4414,
+      "step": 12463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2735831452680317,
+      "learning_rate": 4.462780006084867e-06,
+      "loss": 0.7287,
+      "step": 12464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5258059267889235,
+      "learning_rate": 4.46214645602477e-06,
+      "loss": 0.4298,
+      "step": 12465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.153096101794534,
+      "learning_rate": 4.461512927691492e-06,
+      "loss": 0.6107,
+      "step": 12466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.640882285710841,
+      "learning_rate": 4.460879421098293e-06,
+      "loss": 0.313,
+      "step": 12467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9555803662061173,
+      "learning_rate": 4.4602459362584415e-06,
+      "loss": 0.6166,
+      "step": 12468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7516195642915133,
+      "learning_rate": 4.459612473185193e-06,
+      "loss": 0.4449,
+      "step": 12469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6709996756174583,
+      "learning_rate": 4.458979031891813e-06,
+      "loss": 0.4232,
+      "step": 12470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6458627509077284,
+      "learning_rate": 4.4583456123915605e-06,
+      "loss": 0.4757,
+      "step": 12471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7106526149948311,
+      "learning_rate": 4.457712214697694e-06,
+      "loss": 0.5409,
+      "step": 12472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.095425345980934,
+      "learning_rate": 4.457078838823475e-06,
+      "loss": 0.512,
+      "step": 12473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.16501242207257,
+      "learning_rate": 4.456445484782163e-06,
+      "loss": 0.642,
+      "step": 12474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8949793036596383,
+      "learning_rate": 4.455812152587018e-06,
+      "loss": 0.4531,
+      "step": 12475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.727472069469402,
+      "learning_rate": 4.455178842251296e-06,
+      "loss": 0.5478,
+      "step": 12476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.886124417958946,
+      "learning_rate": 4.4545455537882555e-06,
+      "loss": 0.6167,
+      "step": 12477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7165195159675335,
+      "learning_rate": 4.453912287211155e-06,
+      "loss": 0.4893,
+      "step": 12478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7678062263188272,
+      "learning_rate": 4.4532790425332515e-06,
+      "loss": 0.4163,
+      "step": 12479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7493958353010128,
+      "learning_rate": 4.4526458197678005e-06,
+      "loss": 0.4491,
+      "step": 12480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.14473357395536,
+      "learning_rate": 4.452012618928059e-06,
+      "loss": 0.4918,
+      "step": 12481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7622729369306007,
+      "learning_rate": 4.451379440027281e-06,
+      "loss": 0.4545,
+      "step": 12482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7323910791787387,
+      "learning_rate": 4.450746283078723e-06,
+      "loss": 0.484,
+      "step": 12483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1278808426278015,
+      "learning_rate": 4.450113148095642e-06,
+      "loss": 0.5892,
+      "step": 12484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0029321282541197,
+      "learning_rate": 4.449480035091289e-06,
+      "loss": 0.5899,
+      "step": 12485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.730159100581805,
+      "learning_rate": 4.4488469440789194e-06,
+      "loss": 0.434,
+      "step": 12486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4362733444369618,
+      "learning_rate": 4.4482138750717865e-06,
+      "loss": 0.3566,
+      "step": 12487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0066223534748446,
+      "learning_rate": 4.447580828083141e-06,
+      "loss": 0.5989,
+      "step": 12488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9839044817624873,
+      "learning_rate": 4.446947803126239e-06,
+      "loss": 0.5609,
+      "step": 12489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0258713388952794,
+      "learning_rate": 4.446314800214331e-06,
+      "loss": 0.4984,
+      "step": 12490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6691851795772836,
+      "learning_rate": 4.4456818193606665e-06,
+      "loss": 0.3611,
+      "step": 12491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6322095660334444,
+      "learning_rate": 4.445048860578499e-06,
+      "loss": 0.5164,
+      "step": 12492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7489488254768966,
+      "learning_rate": 4.444415923881079e-06,
+      "loss": 0.4575,
+      "step": 12493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5466327760774323,
+      "learning_rate": 4.443783009281656e-06,
+      "loss": 0.5021,
+      "step": 12494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7540905048258124,
+      "learning_rate": 4.443150116793478e-06,
+      "loss": 0.5951,
+      "step": 12495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0498944365114458,
+      "learning_rate": 4.4425172464297976e-06,
+      "loss": 0.5365,
+      "step": 12496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0079994869670696,
+      "learning_rate": 4.441884398203861e-06,
+      "loss": 0.5872,
+      "step": 12497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8861951571380875,
+      "learning_rate": 4.441251572128919e-06,
+      "loss": 0.5756,
+      "step": 12498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7220070029085266,
+      "learning_rate": 4.440618768218214e-06,
+      "loss": 0.4664,
+      "step": 12499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6677740914263846,
+      "learning_rate": 4.4399859864850005e-06,
+      "loss": 0.4455,
+      "step": 12500
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5133107304573059,
+      "eval_runtime": 67.9938,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 12500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.367395946017059,
+      "learning_rate": 4.4393532269425196e-06,
+      "loss": 0.3225,
+      "step": 12501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5637040008073662,
+      "learning_rate": 4.438720489604022e-06,
+      "loss": 0.5091,
+      "step": 12502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9531161579894183,
+      "learning_rate": 4.438087774482752e-06,
+      "loss": 0.4846,
+      "step": 12503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4343485038362156,
+      "learning_rate": 4.437455081591953e-06,
+      "loss": 0.4629,
+      "step": 12504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.722080138171036,
+      "learning_rate": 4.436822410944873e-06,
+      "loss": 0.5424,
+      "step": 12505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7266395305328246,
+      "learning_rate": 4.436189762554756e-06,
+      "loss": 0.642,
+      "step": 12506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8530112245314259,
+      "learning_rate": 4.435557136434844e-06,
+      "loss": 0.6699,
+      "step": 12507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2063530520673376,
+      "learning_rate": 4.434924532598382e-06,
+      "loss": 0.6027,
+      "step": 12508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9309500848486174,
+      "learning_rate": 4.434291951058615e-06,
+      "loss": 0.6678,
+      "step": 12509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8178079836748169,
+      "learning_rate": 4.433659391828784e-06,
+      "loss": 0.443,
+      "step": 12510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6745930849216346,
+      "learning_rate": 4.433026854922131e-06,
+      "loss": 0.5404,
+      "step": 12511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6101876949467555,
+      "learning_rate": 4.432394340351898e-06,
+      "loss": 0.4651,
+      "step": 12512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5053534773156736,
+      "learning_rate": 4.431761848131325e-06,
+      "loss": 0.4251,
+      "step": 12513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8450116416085869,
+      "learning_rate": 4.4311293782736555e-06,
+      "loss": 0.5528,
+      "step": 12514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.021567111900749,
+      "learning_rate": 4.430496930792129e-06,
+      "loss": 0.5125,
+      "step": 12515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6382405593659466,
+      "learning_rate": 4.429864505699985e-06,
+      "loss": 0.4897,
+      "step": 12516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6774999660073975,
+      "learning_rate": 4.4292321030104625e-06,
+      "loss": 0.4572,
+      "step": 12517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4241369533611976,
+      "learning_rate": 4.428599722736801e-06,
+      "loss": 0.3995,
+      "step": 12518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7298276902155654,
+      "learning_rate": 4.427967364892239e-06,
+      "loss": 0.4845,
+      "step": 12519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3842210357300537,
+      "learning_rate": 4.427335029490012e-06,
+      "loss": 0.3781,
+      "step": 12520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.082367714369043,
+      "learning_rate": 4.426702716543363e-06,
+      "loss": 0.6328,
+      "step": 12521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8044917274482122,
+      "learning_rate": 4.426070426065524e-06,
+      "loss": 0.5445,
+      "step": 12522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8642809479325042,
+      "learning_rate": 4.425438158069735e-06,
+      "loss": 0.5849,
+      "step": 12523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0314705718859427,
+      "learning_rate": 4.42480591256923e-06,
+      "loss": 0.4843,
+      "step": 12524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.749067554604906,
+      "learning_rate": 4.424173689577245e-06,
+      "loss": 0.6087,
+      "step": 12525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9178899293912213,
+      "learning_rate": 4.423541489107016e-06,
+      "loss": 0.6935,
+      "step": 12526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4721841206836497,
+      "learning_rate": 4.422909311171777e-06,
+      "loss": 0.4321,
+      "step": 12527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7704336921543136,
+      "learning_rate": 4.4222771557847636e-06,
+      "loss": 0.484,
+      "step": 12528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3711628390948691,
+      "learning_rate": 4.421645022959207e-06,
+      "loss": 0.4195,
+      "step": 12529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5579357419666378,
+      "learning_rate": 4.421012912708343e-06,
+      "loss": 0.39,
+      "step": 12530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8159478448825408,
+      "learning_rate": 4.420380825045401e-06,
+      "loss": 0.4205,
+      "step": 12531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8697363100270605,
+      "learning_rate": 4.41974875998362e-06,
+      "loss": 0.4978,
+      "step": 12532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6998452447656622,
+      "learning_rate": 4.419116717536226e-06,
+      "loss": 0.4981,
+      "step": 12533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.767617846367489,
+      "learning_rate": 4.418484697716452e-06,
+      "loss": 0.6159,
+      "step": 12534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.722200127576703,
+      "learning_rate": 4.417852700537529e-06,
+      "loss": 0.4268,
+      "step": 12535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8863603247007275,
+      "learning_rate": 4.4172207260126875e-06,
+      "loss": 0.6133,
+      "step": 12536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6252258907812887,
+      "learning_rate": 4.416588774155158e-06,
+      "loss": 0.4062,
+      "step": 12537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8367237671046148,
+      "learning_rate": 4.41595684497817e-06,
+      "loss": 0.5231,
+      "step": 12538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1180879323657567,
+      "learning_rate": 4.4153249384949505e-06,
+      "loss": 0.6095,
+      "step": 12539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8733626452793173,
+      "learning_rate": 4.414693054718732e-06,
+      "loss": 0.5458,
+      "step": 12540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7367422843332132,
+      "learning_rate": 4.414061193662739e-06,
+      "loss": 0.4439,
+      "step": 12541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0280932904589655,
+      "learning_rate": 4.4134293553402e-06,
+      "loss": 0.5692,
+      "step": 12542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4202205499257623,
+      "learning_rate": 4.412797539764343e-06,
+      "loss": 0.4065,
+      "step": 12543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4894864389644509,
+      "learning_rate": 4.412165746948394e-06,
+      "loss": 0.4814,
+      "step": 12544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6701178503977743,
+      "learning_rate": 4.4115339769055795e-06,
+      "loss": 0.4789,
+      "step": 12545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6447607242654994,
+      "learning_rate": 4.410902229649127e-06,
+      "loss": 0.4699,
+      "step": 12546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8494805058795871,
+      "learning_rate": 4.410270505192257e-06,
+      "loss": 0.5267,
+      "step": 12547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7517392231843842,
+      "learning_rate": 4.409638803548198e-06,
+      "loss": 0.4867,
+      "step": 12548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6348959566148293,
+      "learning_rate": 4.409007124730173e-06,
+      "loss": 0.5266,
+      "step": 12549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1809970949489785,
+      "learning_rate": 4.408375468751408e-06,
+      "loss": 0.5516,
+      "step": 12550
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5121986269950867,
+      "eval_runtime": 68.0133,
+      "eval_samples_per_second": 50.137,
+      "eval_steps_per_second": 1.573,
+      "step": 12550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.531288315534489,
+      "learning_rate": 4.407743835625123e-06,
+      "loss": 0.5284,
+      "step": 12551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7821411541607832,
+      "learning_rate": 4.407112225364543e-06,
+      "loss": 0.5028,
+      "step": 12552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6290945396847176,
+      "learning_rate": 4.4064806379828894e-06,
+      "loss": 0.5185,
+      "step": 12553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.159102139826804,
+      "learning_rate": 4.405849073493385e-06,
+      "loss": 0.7994,
+      "step": 12554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7334093201452243,
+      "learning_rate": 4.405217531909249e-06,
+      "loss": 0.4341,
+      "step": 12555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9883246346675099,
+      "learning_rate": 4.404586013243707e-06,
+      "loss": 0.4444,
+      "step": 12556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8428317405072001,
+      "learning_rate": 4.403954517509973e-06,
+      "loss": 0.3989,
+      "step": 12557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6388434289669467,
+      "learning_rate": 4.403323044721273e-06,
+      "loss": 0.4032,
+      "step": 12558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9604578717694228,
+      "learning_rate": 4.402691594890822e-06,
+      "loss": 0.4927,
+      "step": 12559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.43118124805676,
+      "learning_rate": 4.402060168031842e-06,
+      "loss": 0.3717,
+      "step": 12560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.532446751903519,
+      "learning_rate": 4.40142876415755e-06,
+      "loss": 0.4743,
+      "step": 12561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7864425855159953,
+      "learning_rate": 4.400797383281165e-06,
+      "loss": 0.5755,
+      "step": 12562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.444969713473236,
+      "learning_rate": 4.400166025415904e-06,
+      "loss": 0.5019,
+      "step": 12563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.867286402642676,
+      "learning_rate": 4.399534690574984e-06,
+      "loss": 0.4432,
+      "step": 12564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7952457963885926,
+      "learning_rate": 4.3989033787716204e-06,
+      "loss": 0.637,
+      "step": 12565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6313829377180424,
+      "learning_rate": 4.398272090019034e-06,
+      "loss": 0.523,
+      "step": 12566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7628773818937489,
+      "learning_rate": 4.3976408243304325e-06,
+      "loss": 0.5569,
+      "step": 12567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7405607870929714,
+      "learning_rate": 4.39700958171904e-06,
+      "loss": 0.4412,
+      "step": 12568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.684848768352383,
+      "learning_rate": 4.3963783621980634e-06,
+      "loss": 0.5622,
+      "step": 12569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.655585855111108,
+      "learning_rate": 4.395747165780721e-06,
+      "loss": 0.3798,
+      "step": 12570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7599947741364947,
+      "learning_rate": 4.395115992480228e-06,
+      "loss": 0.5146,
+      "step": 12571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.092817345434225,
+      "learning_rate": 4.394484842309794e-06,
+      "loss": 0.4264,
+      "step": 12572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7299221402760265,
+      "learning_rate": 4.393853715282632e-06,
+      "loss": 0.5592,
+      "step": 12573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7816080840030597,
+      "learning_rate": 4.393222611411958e-06,
+      "loss": 0.507,
+      "step": 12574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9561298003474707,
+      "learning_rate": 4.392591530710979e-06,
+      "loss": 0.5002,
+      "step": 12575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5765170144192748,
+      "learning_rate": 4.39196047319291e-06,
+      "loss": 0.4343,
+      "step": 12576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0232079634517897,
+      "learning_rate": 4.391329438870959e-06,
+      "loss": 0.5903,
+      "step": 12577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.88177204970354,
+      "learning_rate": 4.390698427758337e-06,
+      "loss": 0.5026,
+      "step": 12578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7576608654427461,
+      "learning_rate": 4.390067439868256e-06,
+      "loss": 0.4604,
+      "step": 12579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.704055573231311,
+      "learning_rate": 4.389436475213922e-06,
+      "loss": 0.5108,
+      "step": 12580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.468587318299274,
+      "learning_rate": 4.388805533808547e-06,
+      "loss": 0.3649,
+      "step": 12581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9578222629495536,
+      "learning_rate": 4.388174615665338e-06,
+      "loss": 0.4957,
+      "step": 12582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6950658190226588,
+      "learning_rate": 4.387543720797503e-06,
+      "loss": 0.4707,
+      "step": 12583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5761568965354864,
+      "learning_rate": 4.386912849218248e-06,
+      "loss": 0.4754,
+      "step": 12584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.481867639195436,
+      "learning_rate": 4.386282000940782e-06,
+      "loss": 0.4588,
+      "step": 12585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3913952411111283,
+      "learning_rate": 4.385651175978312e-06,
+      "loss": 0.5033,
+      "step": 12586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.27981460540736,
+      "learning_rate": 4.38502037434404e-06,
+      "loss": 0.5396,
+      "step": 12587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1102574046848477,
+      "learning_rate": 4.384389596051175e-06,
+      "loss": 0.4523,
+      "step": 12588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.689498411956782,
+      "learning_rate": 4.383758841112921e-06,
+      "loss": 0.4787,
+      "step": 12589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6111805694812913,
+      "learning_rate": 4.383128109542481e-06,
+      "loss": 0.4166,
+      "step": 12590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9965336950997474,
+      "learning_rate": 4.382497401353063e-06,
+      "loss": 0.5928,
+      "step": 12591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8084621039167847,
+      "learning_rate": 4.3818667165578655e-06,
+      "loss": 0.5448,
+      "step": 12592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.748976214885678,
+      "learning_rate": 4.381236055170095e-06,
+      "loss": 0.4461,
+      "step": 12593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9601526949685648,
+      "learning_rate": 4.380605417202951e-06,
+      "loss": 0.5547,
+      "step": 12594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6506263789108064,
+      "learning_rate": 4.37997480266964e-06,
+      "loss": 0.508,
+      "step": 12595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9737915515592326,
+      "learning_rate": 4.379344211583359e-06,
+      "loss": 0.5525,
+      "step": 12596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9481173818936728,
+      "learning_rate": 4.3787136439573104e-06,
+      "loss": 0.6181,
+      "step": 12597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.359050686857484,
+      "learning_rate": 4.378083099804698e-06,
+      "loss": 0.3405,
+      "step": 12598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3196662766346727,
+      "learning_rate": 4.377452579138717e-06,
+      "loss": 0.3882,
+      "step": 12599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8330331022496096,
+      "learning_rate": 4.376822081972569e-06,
+      "loss": 0.5453,
+      "step": 12600
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5141906142234802,
+      "eval_runtime": 67.9785,
+      "eval_samples_per_second": 50.163,
+      "eval_steps_per_second": 1.574,
+      "step": 12600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.971526430504114,
+      "learning_rate": 4.376191608319453e-06,
+      "loss": 0.5778,
+      "step": 12601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7840416636715426,
+      "learning_rate": 4.375561158192569e-06,
+      "loss": 0.511,
+      "step": 12602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5364481533908987,
+      "learning_rate": 4.374930731605112e-06,
+      "loss": 0.4118,
+      "step": 12603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1676462283296494,
+      "learning_rate": 4.374300328570282e-06,
+      "loss": 0.5884,
+      "step": 12604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6795547337557346,
+      "learning_rate": 4.373669949101276e-06,
+      "loss": 0.513,
+      "step": 12605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5709541486205318,
+      "learning_rate": 4.373039593211289e-06,
+      "loss": 0.4137,
+      "step": 12606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8168803756815,
+      "learning_rate": 4.372409260913518e-06,
+      "loss": 0.4766,
+      "step": 12607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0808077565088645,
+      "learning_rate": 4.3717789522211596e-06,
+      "loss": 0.5614,
+      "step": 12608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4939286223988146,
+      "learning_rate": 4.371148667147408e-06,
+      "loss": 0.5668,
+      "step": 12609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4835538952873555,
+      "learning_rate": 4.370518405705457e-06,
+      "loss": 0.4421,
+      "step": 12610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6314000834144176,
+      "learning_rate": 4.369888167908504e-06,
+      "loss": 0.5175,
+      "step": 12611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.618493791938106,
+      "learning_rate": 4.369257953769739e-06,
+      "loss": 0.3904,
+      "step": 12612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7618252115779705,
+      "learning_rate": 4.368627763302354e-06,
+      "loss": 0.42,
+      "step": 12613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8633075719510415,
+      "learning_rate": 4.367997596519548e-06,
+      "loss": 0.5124,
+      "step": 12614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9568222200841505,
+      "learning_rate": 4.367367453434505e-06,
+      "loss": 0.5743,
+      "step": 12615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9756930278864988,
+      "learning_rate": 4.366737334060425e-06,
+      "loss": 0.61,
+      "step": 12616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.746514305310846,
+      "learning_rate": 4.366107238410493e-06,
+      "loss": 0.4898,
+      "step": 12617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3900986579734902,
+      "learning_rate": 4.3654771664979016e-06,
+      "loss": 0.4141,
+      "step": 12618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.29884364865013,
+      "learning_rate": 4.364847118335842e-06,
+      "loss": 0.5662,
+      "step": 12619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.349080651126033,
+      "learning_rate": 4.364217093937502e-06,
+      "loss": 0.3936,
+      "step": 12620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1085606267689703,
+      "learning_rate": 4.363587093316071e-06,
+      "loss": 0.5443,
+      "step": 12621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.149348910808773,
+      "learning_rate": 4.3629571164847394e-06,
+      "loss": 0.5443,
+      "step": 12622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0027165340352235,
+      "learning_rate": 4.362327163456694e-06,
+      "loss": 0.5784,
+      "step": 12623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4730321183618542,
+      "learning_rate": 4.361697234245124e-06,
+      "loss": 0.4741,
+      "step": 12624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8863724724517652,
+      "learning_rate": 4.3610673288632135e-06,
+      "loss": 0.3852,
+      "step": 12625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0974623024656736,
+      "learning_rate": 4.3604374473241525e-06,
+      "loss": 0.4728,
+      "step": 12626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9466859869529944,
+      "learning_rate": 4.359807589641125e-06,
+      "loss": 0.6509,
+      "step": 12627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7209458182737405,
+      "learning_rate": 4.359177755827317e-06,
+      "loss": 0.4887,
+      "step": 12628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4347418845693365,
+      "learning_rate": 4.358547945895912e-06,
+      "loss": 0.5211,
+      "step": 12629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8071524134270003,
+      "learning_rate": 4.357918159860101e-06,
+      "loss": 0.5023,
+      "step": 12630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.789787619103664,
+      "learning_rate": 4.357288397733063e-06,
+      "loss": 0.5562,
+      "step": 12631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.681661902243337,
+      "learning_rate": 4.356658659527981e-06,
+      "loss": 0.4464,
+      "step": 12632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7676282209274654,
+      "learning_rate": 4.3560289452580416e-06,
+      "loss": 0.4878,
+      "step": 12633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9859107193339076,
+      "learning_rate": 4.3553992549364245e-06,
+      "loss": 0.5432,
+      "step": 12634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.190304938717755,
+      "learning_rate": 4.354769588576313e-06,
+      "loss": 0.6536,
+      "step": 12635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5215887115249227,
+      "learning_rate": 4.354139946190889e-06,
+      "loss": 0.4152,
+      "step": 12636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6550708556624325,
+      "learning_rate": 4.353510327793334e-06,
+      "loss": 0.3843,
+      "step": 12637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3878914103065996,
+      "learning_rate": 4.352880733396828e-06,
+      "loss": 0.4204,
+      "step": 12638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5504898193913441,
+      "learning_rate": 4.352251163014553e-06,
+      "loss": 0.4718,
+      "step": 12639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8887961485671463,
+      "learning_rate": 4.351621616659683e-06,
+      "loss": 0.5647,
+      "step": 12640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6455058065219088,
+      "learning_rate": 4.350992094345405e-06,
+      "loss": 0.4555,
+      "step": 12641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5734078738426873,
+      "learning_rate": 4.350362596084891e-06,
+      "loss": 0.5218,
+      "step": 12642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8133279811927954,
+      "learning_rate": 4.3497331218913245e-06,
+      "loss": 0.5792,
+      "step": 12643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3559821305655282,
+      "learning_rate": 4.349103671777879e-06,
+      "loss": 0.3776,
+      "step": 12644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.017853920264862,
+      "learning_rate": 4.348474245757733e-06,
+      "loss": 0.6644,
+      "step": 12645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6390791937867364,
+      "learning_rate": 4.347844843844064e-06,
+      "loss": 0.4449,
+      "step": 12646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6445248339349863,
+      "learning_rate": 4.347215466050047e-06,
+      "loss": 0.4434,
+      "step": 12647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7673940330662927,
+      "learning_rate": 4.34658611238886e-06,
+      "loss": 0.5098,
+      "step": 12648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9144732315596085,
+      "learning_rate": 4.345956782873676e-06,
+      "loss": 0.7377,
+      "step": 12649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0120576398807892,
+      "learning_rate": 4.345327477517668e-06,
+      "loss": 0.432,
+      "step": 12650
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5126391053199768,
+      "eval_runtime": 67.9909,
+      "eval_samples_per_second": 50.154,
+      "eval_steps_per_second": 1.574,
+      "step": 12650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.471490773806323,
+      "learning_rate": 4.344698196334014e-06,
+      "loss": 0.4458,
+      "step": 12651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8255477729714773,
+      "learning_rate": 4.344068939335885e-06,
+      "loss": 0.6192,
+      "step": 12652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7501019366244057,
+      "learning_rate": 4.343439706536454e-06,
+      "loss": 0.4814,
+      "step": 12653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5451813033162578,
+      "learning_rate": 4.342810497948896e-06,
+      "loss": 0.4411,
+      "step": 12654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.832824740208683,
+      "learning_rate": 4.34218131358638e-06,
+      "loss": 0.4339,
+      "step": 12655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7214158104403767,
+      "learning_rate": 4.34155215346208e-06,
+      "loss": 0.4303,
+      "step": 12656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9085049130525211,
+      "learning_rate": 4.340923017589166e-06,
+      "loss": 0.5096,
+      "step": 12657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8028675872072513,
+      "learning_rate": 4.340293905980808e-06,
+      "loss": 0.4501,
+      "step": 12658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8608217158540972,
+      "learning_rate": 4.339664818650177e-06,
+      "loss": 0.5716,
+      "step": 12659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6382139140102607,
+      "learning_rate": 4.339035755610442e-06,
+      "loss": 0.4838,
+      "step": 12660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.305985665360685,
+      "learning_rate": 4.338406716874773e-06,
+      "loss": 0.6023,
+      "step": 12661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7568304761718943,
+      "learning_rate": 4.337777702456338e-06,
+      "loss": 0.5765,
+      "step": 12662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8784871601784376,
+      "learning_rate": 4.337148712368304e-06,
+      "loss": 0.4981,
+      "step": 12663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5877534144020322,
+      "learning_rate": 4.336519746623841e-06,
+      "loss": 0.4824,
+      "step": 12664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109041148320402,
+      "learning_rate": 4.335890805236113e-06,
+      "loss": 0.6487,
+      "step": 12665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2408995415118107,
+      "learning_rate": 4.3352618882182885e-06,
+      "loss": 0.5133,
+      "step": 12666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4695475805675342,
+      "learning_rate": 4.334632995583533e-06,
+      "loss": 0.3527,
+      "step": 12667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8433662080103612,
+      "learning_rate": 4.3340041273450115e-06,
+      "loss": 0.4303,
+      "step": 12668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0222654782012497,
+      "learning_rate": 4.333375283515889e-06,
+      "loss": 0.6171,
+      "step": 12669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6588896851563992,
+      "learning_rate": 4.332746464109332e-06,
+      "loss": 0.348,
+      "step": 12670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6543038468423794,
+      "learning_rate": 4.332117669138502e-06,
+      "loss": 0.4858,
+      "step": 12671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.129358491844393,
+      "learning_rate": 4.331488898616563e-06,
+      "loss": 0.4917,
+      "step": 12672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1179267798296366,
+      "learning_rate": 4.3308601525566775e-06,
+      "loss": 0.5127,
+      "step": 12673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6128919305177953,
+      "learning_rate": 4.330231430972009e-06,
+      "loss": 0.485,
+      "step": 12674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6459604905616125,
+      "learning_rate": 4.32960273387572e-06,
+      "loss": 0.4979,
+      "step": 12675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.899165865464035,
+      "learning_rate": 4.328974061280969e-06,
+      "loss": 0.5941,
+      "step": 12676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8000159759023775,
+      "learning_rate": 4.32834541320092e-06,
+      "loss": 0.5758,
+      "step": 12677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.878125822964451,
+      "learning_rate": 4.3277167896487306e-06,
+      "loss": 0.529,
+      "step": 12678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8364886531430116,
+      "learning_rate": 4.327088190637563e-06,
+      "loss": 0.5761,
+      "step": 12679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6942218324916898,
+      "learning_rate": 4.326459616180574e-06,
+      "loss": 0.4755,
+      "step": 12680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.538491421887533,
+      "learning_rate": 4.325831066290926e-06,
+      "loss": 0.5074,
+      "step": 12681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6904473540542915,
+      "learning_rate": 4.325202540981775e-06,
+      "loss": 0.478,
+      "step": 12682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6276538280689286,
+      "learning_rate": 4.324574040266278e-06,
+      "loss": 0.4659,
+      "step": 12683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9625230950745223,
+      "learning_rate": 4.3239455641575955e-06,
+      "loss": 0.4426,
+      "step": 12684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2750489763993997,
+      "learning_rate": 4.32331711266888e-06,
+      "loss": 0.3171,
+      "step": 12685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0448561869902453,
+      "learning_rate": 4.322688685813292e-06,
+      "loss": 0.5177,
+      "step": 12686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5673743235928617,
+      "learning_rate": 4.322060283603984e-06,
+      "loss": 0.461,
+      "step": 12687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.396262209063762,
+      "learning_rate": 4.321431906054112e-06,
+      "loss": 0.3929,
+      "step": 12688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6983883607897778,
+      "learning_rate": 4.320803553176832e-06,
+      "loss": 0.5527,
+      "step": 12689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4377400285463606,
+      "learning_rate": 4.3201752249852955e-06,
+      "loss": 0.5675,
+      "step": 12690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.691975508388796,
+      "learning_rate": 4.31954692149266e-06,
+      "loss": 0.4643,
+      "step": 12691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.675947783619254,
+      "learning_rate": 4.318918642712075e-06,
+      "loss": 0.5012,
+      "step": 12692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.778029852282607,
+      "learning_rate": 4.318290388656695e-06,
+      "loss": 0.5854,
+      "step": 12693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4690016090623694,
+      "learning_rate": 4.317662159339672e-06,
+      "loss": 0.3981,
+      "step": 12694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8406610209921603,
+      "learning_rate": 4.317033954774157e-06,
+      "loss": 0.4215,
+      "step": 12695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0756844782298085,
+      "learning_rate": 4.3164057749733025e-06,
+      "loss": 0.5794,
+      "step": 12696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6717557454124452,
+      "learning_rate": 4.315777619950256e-06,
+      "loss": 0.532,
+      "step": 12697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.680962999873337,
+      "learning_rate": 4.31514948971817e-06,
+      "loss": 0.4542,
+      "step": 12698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5210856919615323,
+      "learning_rate": 4.314521384290195e-06,
+      "loss": 0.466,
+      "step": 12699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.69419314732775,
+      "learning_rate": 4.313893303679476e-06,
+      "loss": 0.4265,
+      "step": 12700
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5127339363098145,
+      "eval_runtime": 67.9631,
+      "eval_samples_per_second": 50.174,
+      "eval_steps_per_second": 1.574,
+      "step": 12700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0739698963069775,
+      "learning_rate": 4.313265247899166e-06,
+      "loss": 0.5826,
+      "step": 12701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5645233195968746,
+      "learning_rate": 4.31263721696241e-06,
+      "loss": 0.4029,
+      "step": 12702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.628966674972122,
+      "learning_rate": 4.3120092108823565e-06,
+      "loss": 0.4119,
+      "step": 12703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6021167018938358,
+      "learning_rate": 4.311381229672152e-06,
+      "loss": 0.7037,
+      "step": 12704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9285270885063757,
+      "learning_rate": 4.3107532733449445e-06,
+      "loss": 0.5021,
+      "step": 12705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4251418097543223,
+      "learning_rate": 4.310125341913877e-06,
+      "loss": 0.4151,
+      "step": 12706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.327013725774137,
+      "learning_rate": 4.3094974353920984e-06,
+      "loss": 0.344,
+      "step": 12707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7024688258603318,
+      "learning_rate": 4.308869553792749e-06,
+      "loss": 0.5639,
+      "step": 12708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.146040949494078,
+      "learning_rate": 4.308241697128977e-06,
+      "loss": 0.5959,
+      "step": 12709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9367968396921875,
+      "learning_rate": 4.307613865413925e-06,
+      "loss": 0.5978,
+      "step": 12710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4672412018044985,
+      "learning_rate": 4.306986058660736e-06,
+      "loss": 0.3776,
+      "step": 12711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6439297178983105,
+      "learning_rate": 4.306358276882553e-06,
+      "loss": 0.6101,
+      "step": 12712
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4689870294016965,
+      "learning_rate": 4.305730520092516e-06,
+      "loss": 0.3976,
+      "step": 12713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6342699541726742,
+      "learning_rate": 4.3051027883037715e-06,
+      "loss": 0.522,
+      "step": 12714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.039629334806542,
+      "learning_rate": 4.304475081529458e-06,
+      "loss": 0.4907,
+      "step": 12715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9015296941614215,
+      "learning_rate": 4.303847399782713e-06,
+      "loss": 0.4714,
+      "step": 12716
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7371657365990414,
+      "learning_rate": 4.3032197430766835e-06,
+      "loss": 0.439,
+      "step": 12717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6785386970831144,
+      "learning_rate": 4.302592111424503e-06,
+      "loss": 0.4733,
+      "step": 12718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0671585052976686,
+      "learning_rate": 4.3019645048393136e-06,
+      "loss": 0.6127,
+      "step": 12719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7733911037254488,
+      "learning_rate": 4.301336923334254e-06,
+      "loss": 0.4284,
+      "step": 12720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5648012739586008,
+      "learning_rate": 4.300709366922458e-06,
+      "loss": 0.4442,
+      "step": 12721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1707624206405653,
+      "learning_rate": 4.30008183561707e-06,
+      "loss": 0.4479,
+      "step": 12722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.680386087219296,
+      "learning_rate": 4.299454329431224e-06,
+      "loss": 0.4752,
+      "step": 12723
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6331722735295164,
+      "learning_rate": 4.298826848378053e-06,
+      "loss": 0.4006,
+      "step": 12724
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7559337484222746,
+      "learning_rate": 4.298199392470698e-06,
+      "loss": 0.5241,
+      "step": 12725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.033520074742832,
+      "learning_rate": 4.29757196172229e-06,
+      "loss": 0.7456,
+      "step": 12726
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.166521531969662,
+      "learning_rate": 4.296944556145969e-06,
+      "loss": 0.6312,
+      "step": 12727
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4923213721516866,
+      "learning_rate": 4.296317175754863e-06,
+      "loss": 0.4577,
+      "step": 12728
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3965879108244594,
+      "learning_rate": 4.295689820562112e-06,
+      "loss": 0.3811,
+      "step": 12729
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8976030378627073,
+      "learning_rate": 4.295062490580846e-06,
+      "loss": 0.429,
+      "step": 12730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7932156746953702,
+      "learning_rate": 4.294435185824199e-06,
+      "loss": 0.4135,
+      "step": 12731
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6415861522458035,
+      "learning_rate": 4.293807906305302e-06,
+      "loss": 0.427,
+      "step": 12732
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4121376588103376,
+      "learning_rate": 4.293180652037287e-06,
+      "loss": 0.4125,
+      "step": 12733
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.76132291842618,
+      "learning_rate": 4.2925534230332845e-06,
+      "loss": 0.5906,
+      "step": 12734
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.612033333709704,
+      "learning_rate": 4.291926219306427e-06,
+      "loss": 0.4616,
+      "step": 12735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7513082180071662,
+      "learning_rate": 4.291299040869844e-06,
+      "loss": 0.5209,
+      "step": 12736
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.635292103290784,
+      "learning_rate": 4.290671887736666e-06,
+      "loss": 0.5936,
+      "step": 12737
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8493144156762413,
+      "learning_rate": 4.290044759920019e-06,
+      "loss": 0.5056,
+      "step": 12738
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7746557918044785,
+      "learning_rate": 4.289417657433034e-06,
+      "loss": 0.5012,
+      "step": 12739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.228603602727646,
+      "learning_rate": 4.288790580288839e-06,
+      "loss": 0.5635,
+      "step": 12740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5420816753986004,
+      "learning_rate": 4.28816352850056e-06,
+      "loss": 0.3801,
+      "step": 12741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7257294814955153,
+      "learning_rate": 4.287536502081325e-06,
+      "loss": 0.5604,
+      "step": 12742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9766435463130083,
+      "learning_rate": 4.286909501044261e-06,
+      "loss": 0.4332,
+      "step": 12743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8631958924765644,
+      "learning_rate": 4.286282525402492e-06,
+      "loss": 0.4599,
+      "step": 12744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.559444726667864,
+      "learning_rate": 4.2856555751691455e-06,
+      "loss": 0.4554,
+      "step": 12745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5938990241215094,
+      "learning_rate": 4.2850286503573446e-06,
+      "loss": 0.4722,
+      "step": 12746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6626396061015642,
+      "learning_rate": 4.284401750980215e-06,
+      "loss": 0.5431,
+      "step": 12747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6423234172307082,
+      "learning_rate": 4.283774877050878e-06,
+      "loss": 0.4727,
+      "step": 12748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6844327595868216,
+      "learning_rate": 4.283148028582461e-06,
+      "loss": 0.5936,
+      "step": 12749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6703969585975103,
+      "learning_rate": 4.282521205588084e-06,
+      "loss": 0.4842,
+      "step": 12750
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.5111120343208313,
+      "eval_runtime": 67.9882,
+      "eval_samples_per_second": 50.156,
+      "eval_steps_per_second": 1.574,
+      "step": 12750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8047653431444548,
+      "learning_rate": 4.281894408080867e-06,
+      "loss": 0.5366,
+      "step": 12751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6943829188158335,
+      "learning_rate": 4.281267636073936e-06,
+      "loss": 0.5933,
+      "step": 12752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8939914344525273,
+      "learning_rate": 4.280640889580409e-06,
+      "loss": 0.609,
+      "step": 12753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7007950184855554,
+      "learning_rate": 4.280014168613408e-06,
+      "loss": 0.4756,
+      "step": 12754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8388075932508559,
+      "learning_rate": 4.279387473186053e-06,
+      "loss": 0.4896,
+      "step": 12755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6929074672316091,
+      "learning_rate": 4.2787608033114614e-06,
+      "loss": 0.4553,
+      "step": 12756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.631467746755874,
+      "learning_rate": 4.278134159002756e-06,
+      "loss": 0.4185,
+      "step": 12757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8070406239724275,
+      "learning_rate": 4.277507540273051e-06,
+      "loss": 0.4709,
+      "step": 12758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.680415748635079,
+      "learning_rate": 4.276880947135468e-06,
+      "loss": 0.547,
+      "step": 12759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5628213143440632,
+      "learning_rate": 4.276254379603123e-06,
+      "loss": 0.6887,
+      "step": 12760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4884018638101881,
+      "learning_rate": 4.27562783768913e-06,
+      "loss": 0.2983,
+      "step": 12761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0161346910921236,
+      "learning_rate": 4.27500132140661e-06,
+      "loss": 0.5383,
+      "step": 12762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8063558405327071,
+      "learning_rate": 4.274374830768677e-06,
+      "loss": 0.6372,
+      "step": 12763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6999097198710027,
+      "learning_rate": 4.273748365788444e-06,
+      "loss": 0.5013,
+      "step": 12764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.658456781709679,
+      "learning_rate": 4.273121926479028e-06,
+      "loss": 0.5042,
+      "step": 12765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7731990608456367,
+      "learning_rate": 4.272495512853541e-06,
+      "loss": 0.4857,
+      "step": 12766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.573678701989896,
+      "learning_rate": 4.2718691249251e-06,
+      "loss": 0.4739,
+      "step": 12767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.70538728629295,
+      "learning_rate": 4.271242762706816e-06,
+      "loss": 0.4196,
+      "step": 12768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.720439715514655,
+      "learning_rate": 4.2706164262118e-06,
+      "loss": 0.5731,
+      "step": 12769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8796928932781558,
+      "learning_rate": 4.2699901154531665e-06,
+      "loss": 0.5215,
+      "step": 12770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7568386439192472,
+      "learning_rate": 4.2693638304440244e-06,
+      "loss": 0.641,
+      "step": 12771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9755016250160522,
+      "learning_rate": 4.268737571197489e-06,
+      "loss": 0.671,
+      "step": 12772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3668194899867687,
+      "learning_rate": 4.268111337726666e-06,
+      "loss": 0.3268,
+      "step": 12773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9799072642708433,
+      "learning_rate": 4.267485130044668e-06,
+      "loss": 0.516,
+      "step": 12774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7692676404718894,
+      "learning_rate": 4.266858948164602e-06,
+      "loss": 0.4272,
+      "step": 12775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8878490065053966,
+      "learning_rate": 4.266232792099579e-06,
+      "loss": 0.4356,
+      "step": 12776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.067070731354681,
+      "learning_rate": 4.265606661862704e-06,
+      "loss": 0.5745,
+      "step": 12777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6976274337552717,
+      "learning_rate": 4.264980557467087e-06,
+      "loss": 0.4329,
+      "step": 12778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4846340590661067,
+      "learning_rate": 4.264354478925837e-06,
+      "loss": 0.4437,
+      "step": 12779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5975158504941394,
+      "learning_rate": 4.263728426252058e-06,
+      "loss": 0.3839,
+      "step": 12780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.767509926127083,
+      "learning_rate": 4.263102399458857e-06,
+      "loss": 0.4601,
+      "step": 12781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6884599910155507,
+      "learning_rate": 4.262476398559339e-06,
+      "loss": 0.5274,
+      "step": 12782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5149925271213351,
+      "learning_rate": 4.261850423566607e-06,
+      "loss": 0.425,
+      "step": 12783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7162811549119377,
+      "learning_rate": 4.261224474493771e-06,
+      "loss": 0.5981,
+      "step": 12784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9516119516637402,
+      "learning_rate": 4.2605985513539275e-06,
+      "loss": 0.6313,
+      "step": 12785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8978745307454408,
+      "learning_rate": 4.259972654160185e-06,
+      "loss": 0.4701,
+      "step": 12786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.044214540895543,
+      "learning_rate": 4.259346782925645e-06,
+      "loss": 0.6502,
+      "step": 12787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9464667079702187,
+      "learning_rate": 4.2587209376634085e-06,
+      "loss": 0.5353,
+      "step": 12788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6451323279073533,
+      "learning_rate": 4.258095118386579e-06,
+      "loss": 0.4141,
+      "step": 12789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4604968389330792,
+      "learning_rate": 4.2574693251082575e-06,
+      "loss": 0.4067,
+      "step": 12790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.223391715364172,
+      "learning_rate": 4.2568435578415425e-06,
+      "loss": 0.6048,
+      "step": 12791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7199611045514291,
+      "learning_rate": 4.256217816599537e-06,
+      "loss": 0.5225,
+      "step": 12792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1554429362233796,
+      "learning_rate": 4.255592101395337e-06,
+      "loss": 0.6604,
+      "step": 12793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8137907818448145,
+      "learning_rate": 4.254966412242045e-06,
+      "loss": 0.4677,
+      "step": 12794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8337385250511502,
+      "learning_rate": 4.254340749152758e-06,
+      "loss": 0.4781,
+      "step": 12795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7299041781310456,
+      "learning_rate": 4.253715112140572e-06,
+      "loss": 0.5308,
+      "step": 12796
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5477273007666756,
+      "learning_rate": 4.253089501218588e-06,
+      "loss": 0.5761,
+      "step": 12797
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5081365342900765,
+      "learning_rate": 4.252463916399899e-06,
+      "loss": 0.5186,
+      "step": 12798
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.213096282760434,
+      "learning_rate": 4.251838357697604e-06,
+      "loss": 0.6605,
+      "step": 12799
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9961823220581625,
+      "learning_rate": 4.251212825124799e-06,
+      "loss": 0.4857,
+      "step": 12800
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.5116164684295654,
+      "eval_runtime": 67.9493,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 12800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7640869780752242,
+      "learning_rate": 4.250587318694576e-06,
+      "loss": 0.4625,
+      "step": 12801
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3610958284448416,
+      "learning_rate": 4.249961838420032e-06,
+      "loss": 0.2827,
+      "step": 12802
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8523899747922163,
+      "learning_rate": 4.249336384314261e-06,
+      "loss": 0.6242,
+      "step": 12803
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.50439421497141,
+      "learning_rate": 4.248710956390354e-06,
+      "loss": 0.4737,
+      "step": 12804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.550745445949329,
+      "learning_rate": 4.248085554661409e-06,
+      "loss": 0.4506,
+      "step": 12805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.569036033596089,
+      "learning_rate": 4.247460179140513e-06,
+      "loss": 0.4168,
+      "step": 12806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8309301017027226,
+      "learning_rate": 4.24683482984076e-06,
+      "loss": 0.5657,
+      "step": 12807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9312613381016797,
+      "learning_rate": 4.246209506775243e-06,
+      "loss": 0.5785,
+      "step": 12808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.499008708926385,
+      "learning_rate": 4.24558420995705e-06,
+      "loss": 0.3767,
+      "step": 12809
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1431783184939976,
+      "learning_rate": 4.244958939399274e-06,
+      "loss": 0.4866,
+      "step": 12810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6044656000174513,
+      "learning_rate": 4.244333695115e-06,
+      "loss": 0.4439,
+      "step": 12811
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2092455205503123,
+      "learning_rate": 4.243708477117323e-06,
+      "loss": 0.5211,
+      "step": 12812
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9089869885175919,
+      "learning_rate": 4.243083285419328e-06,
+      "loss": 0.4901,
+      "step": 12813
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6630865351212212,
+      "learning_rate": 4.242458120034103e-06,
+      "loss": 0.5716,
+      "step": 12814
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8934776487142466,
+      "learning_rate": 4.241832980974737e-06,
+      "loss": 0.5038,
+      "step": 12815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.656662605299253,
+      "learning_rate": 4.241207868254318e-06,
+      "loss": 0.4623,
+      "step": 12816
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7244081340395088,
+      "learning_rate": 4.240582781885927e-06,
+      "loss": 0.6345,
+      "step": 12817
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4019279618864988,
+      "learning_rate": 4.239957721882655e-06,
+      "loss": 0.4415,
+      "step": 12818
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.601460182858265,
+      "learning_rate": 4.239332688257584e-06,
+      "loss": 0.5342,
+      "step": 12819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9761180191784347,
+      "learning_rate": 4.238707681023802e-06,
+      "loss": 0.5715,
+      "step": 12820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7839880378324993,
+      "learning_rate": 4.23808270019439e-06,
+      "loss": 0.5996,
+      "step": 12821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7850999080529926,
+      "learning_rate": 4.237457745782433e-06,
+      "loss": 0.4475,
+      "step": 12822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6201822505798007,
+      "learning_rate": 4.236832817801014e-06,
+      "loss": 0.4159,
+      "step": 12823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0541103346201237,
+      "learning_rate": 4.236207916263215e-06,
+      "loss": 0.6718,
+      "step": 12824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9105102815903514,
+      "learning_rate": 4.23558304118212e-06,
+      "loss": 0.4631,
+      "step": 12825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3173222619081133,
+      "learning_rate": 4.2349581925708046e-06,
+      "loss": 0.5333,
+      "step": 12826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7380283262082692,
+      "learning_rate": 4.234333370442357e-06,
+      "loss": 0.4844,
+      "step": 12827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0992190148260828,
+      "learning_rate": 4.233708574809853e-06,
+      "loss": 0.5939,
+      "step": 12828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8293314410458967,
+      "learning_rate": 4.233083805686375e-06,
+      "loss": 0.6374,
+      "step": 12829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9911163031266375,
+      "learning_rate": 4.2324590630850006e-06,
+      "loss": 0.4996,
+      "step": 12830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.830240011543804,
+      "learning_rate": 4.2318343470188076e-06,
+      "loss": 0.5481,
+      "step": 12831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3491208753668655,
+      "learning_rate": 4.231209657500874e-06,
+      "loss": 0.3278,
+      "step": 12832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7391735697684556,
+      "learning_rate": 4.230584994544279e-06,
+      "loss": 0.5309,
+      "step": 12833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1307443883024133,
+      "learning_rate": 4.229960358162098e-06,
+      "loss": 0.5753,
+      "step": 12834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3739315126858547,
+      "learning_rate": 4.22933574836741e-06,
+      "loss": 0.339,
+      "step": 12835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6596774321251964,
+      "learning_rate": 4.228711165173287e-06,
+      "loss": 0.5043,
+      "step": 12836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8462004014558726,
+      "learning_rate": 4.228086608592808e-06,
+      "loss": 0.4163,
+      "step": 12837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6590725090056901,
+      "learning_rate": 4.227462078639043e-06,
+      "loss": 0.4858,
+      "step": 12838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2691891015592598,
+      "learning_rate": 4.226837575325071e-06,
+      "loss": 0.2958,
+      "step": 12839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9343124466709523,
+      "learning_rate": 4.226213098663964e-06,
+      "loss": 0.4697,
+      "step": 12840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8098974530321175,
+      "learning_rate": 4.225588648668793e-06,
+      "loss": 0.5599,
+      "step": 12841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6504856083590378,
+      "learning_rate": 4.224964225352632e-06,
+      "loss": 0.499,
+      "step": 12842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.95295309141486,
+      "learning_rate": 4.224339828728554e-06,
+      "loss": 0.5815,
+      "step": 12843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6371305729919998,
+      "learning_rate": 4.2237154588096295e-06,
+      "loss": 0.5054,
+      "step": 12844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.798275205148373,
+      "learning_rate": 4.22309111560893e-06,
+      "loss": 0.6391,
+      "step": 12845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.529387502443051,
+      "learning_rate": 4.222466799139522e-06,
+      "loss": 0.3847,
+      "step": 12846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8380973492836459,
+      "learning_rate": 4.221842509414481e-06,
+      "loss": 0.5411,
+      "step": 12847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9591381348241086,
+      "learning_rate": 4.221218246446872e-06,
+      "loss": 0.6172,
+      "step": 12848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.884027928316401,
+      "learning_rate": 4.220594010249764e-06,
+      "loss": 0.6898,
+      "step": 12849
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5093083648304062,
+      "learning_rate": 4.2199698008362265e-06,
+      "loss": 0.3533,
+      "step": 12850
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.5106036067008972,
+      "eval_runtime": 67.9204,
+      "eval_samples_per_second": 50.206,
+      "eval_steps_per_second": 1.575,
+      "step": 12850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.842291800262284,
+      "learning_rate": 4.219345618219325e-06,
+      "loss": 0.5126,
+      "step": 12851
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.562246900071894,
+      "learning_rate": 4.21872146241213e-06,
+      "loss": 0.4044,
+      "step": 12852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6919147224805733,
+      "learning_rate": 4.218097333427703e-06,
+      "loss": 0.4356,
+      "step": 12853
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5921496232573507,
+      "learning_rate": 4.217473231279113e-06,
+      "loss": 0.4713,
+      "step": 12854
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6571060446628223,
+      "learning_rate": 4.216849155979425e-06,
+      "loss": 0.4453,
+      "step": 12855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4400733159001193,
+      "learning_rate": 4.216225107541701e-06,
+      "loss": 0.3463,
+      "step": 12856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.413267434731753,
+      "learning_rate": 4.215601085979009e-06,
+      "loss": 0.354,
+      "step": 12857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7624390014050482,
+      "learning_rate": 4.2149770913044096e-06,
+      "loss": 0.4664,
+      "step": 12858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6135889351765298,
+      "learning_rate": 4.214353123530966e-06,
+      "loss": 0.3852,
+      "step": 12859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9446291335542627,
+      "learning_rate": 4.213729182671743e-06,
+      "loss": 0.4338,
+      "step": 12860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.259475445458742,
+      "learning_rate": 4.2131052687398e-06,
+      "loss": 0.5665,
+      "step": 12861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8885788563617376,
+      "learning_rate": 4.212481381748198e-06,
+      "loss": 0.4811,
+      "step": 12862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7447891088362992,
+      "learning_rate": 4.211857521709999e-06,
+      "loss": 0.5158,
+      "step": 12863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7985152941746587,
+      "learning_rate": 4.211233688638262e-06,
+      "loss": 0.567,
+      "step": 12864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6698403506113173,
+      "learning_rate": 4.210609882546048e-06,
+      "loss": 0.5238,
+      "step": 12865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.606513296078358,
+      "learning_rate": 4.2099861034464165e-06,
+      "loss": 0.4077,
+      "step": 12866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8478839740367068,
+      "learning_rate": 4.209362351352422e-06,
+      "loss": 0.4921,
+      "step": 12867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.726407892523928,
+      "learning_rate": 4.2087386262771265e-06,
+      "loss": 0.6095,
+      "step": 12868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7381462635095248,
+      "learning_rate": 4.208114928233587e-06,
+      "loss": 0.3819,
+      "step": 12869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.614201046200978,
+      "learning_rate": 4.207491257234857e-06,
+      "loss": 0.429,
+      "step": 12870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9901699121575995,
+      "learning_rate": 4.2068676132939954e-06,
+      "loss": 0.5439,
+      "step": 12871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7021764911625763,
+      "learning_rate": 4.2062439964240565e-06,
+      "loss": 0.3833,
+      "step": 12872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9541668582679226,
+      "learning_rate": 4.205620406638097e-06,
+      "loss": 0.5668,
+      "step": 12873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8262955922589272,
+      "learning_rate": 4.204996843949171e-06,
+      "loss": 0.5677,
+      "step": 12874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6682130440190954,
+      "learning_rate": 4.20437330837033e-06,
+      "loss": 0.4848,
+      "step": 12875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03744554030902,
+      "learning_rate": 4.203749799914628e-06,
+      "loss": 0.544,
+      "step": 12876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8389503762524533,
+      "learning_rate": 4.203126318595121e-06,
+      "loss": 0.4618,
+      "step": 12877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7083975337952315,
+      "learning_rate": 4.2025028644248575e-06,
+      "loss": 0.5941,
+      "step": 12878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6162312541750614,
+      "learning_rate": 4.201879437416893e-06,
+      "loss": 0.3668,
+      "step": 12879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6061159925117992,
+      "learning_rate": 4.2012560375842744e-06,
+      "loss": 0.4436,
+      "step": 12880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0251606098173105,
+      "learning_rate": 4.200632664940054e-06,
+      "loss": 0.5271,
+      "step": 12881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8083692805600828,
+      "learning_rate": 4.200009319497284e-06,
+      "loss": 0.5226,
+      "step": 12882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5736767978504935,
+      "learning_rate": 4.199386001269009e-06,
+      "loss": 0.531,
+      "step": 12883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2337545828180416,
+      "learning_rate": 4.198762710268279e-06,
+      "loss": 0.5717,
+      "step": 12884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5628602226484636,
+      "learning_rate": 4.198139446508147e-06,
+      "loss": 0.3679,
+      "step": 12885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9552073258528155,
+      "learning_rate": 4.1975162100016544e-06,
+      "loss": 0.5169,
+      "step": 12886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.658794581158048,
+      "learning_rate": 4.196893000761852e-06,
+      "loss": 0.4516,
+      "step": 12887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2569597010631988,
+      "learning_rate": 4.196269818801786e-06,
+      "loss": 0.6961,
+      "step": 12888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0561115483974604,
+      "learning_rate": 4.1956466641344995e-06,
+      "loss": 0.5593,
+      "step": 12889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.944805373184377,
+      "learning_rate": 4.1950235367730416e-06,
+      "loss": 0.6224,
+      "step": 12890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6168095651871355,
+      "learning_rate": 4.1944004367304544e-06,
+      "loss": 0.5045,
+      "step": 12891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9619702561571877,
+      "learning_rate": 4.193777364019784e-06,
+      "loss": 0.4495,
+      "step": 12892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2871732623292775,
+      "learning_rate": 4.193154318654073e-06,
+      "loss": 0.4911,
+      "step": 12893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6511077380763015,
+      "learning_rate": 4.192531300646364e-06,
+      "loss": 0.4064,
+      "step": 12894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.936675771479009,
+      "learning_rate": 4.191908310009701e-06,
+      "loss": 0.5432,
+      "step": 12895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8163243906845188,
+      "learning_rate": 4.191285346757125e-06,
+      "loss": 0.5069,
+      "step": 12896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.540282379584447,
+      "learning_rate": 4.190662410901676e-06,
+      "loss": 0.3522,
+      "step": 12897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9695634146444967,
+      "learning_rate": 4.1900395024563975e-06,
+      "loss": 0.6731,
+      "step": 12898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.057422446884762,
+      "learning_rate": 4.189416621434328e-06,
+      "loss": 0.5374,
+      "step": 12899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.92523010484389,
+      "learning_rate": 4.188793767848507e-06,
+      "loss": 0.6109,
+      "step": 12900
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.511380672454834,
+      "eval_runtime": 67.9379,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 12900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.381455844998315,
+      "learning_rate": 4.188170941711976e-06,
+      "loss": 0.5614,
+      "step": 12901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7532677709059783,
+      "learning_rate": 4.187548143037768e-06,
+      "loss": 0.6171,
+      "step": 12902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7718122973988866,
+      "learning_rate": 4.186925371838928e-06,
+      "loss": 0.521,
+      "step": 12903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1303716324809554,
+      "learning_rate": 4.186302628128487e-06,
+      "loss": 0.5988,
+      "step": 12904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7748927550353661,
+      "learning_rate": 4.1856799119194865e-06,
+      "loss": 0.5005,
+      "step": 12905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7329399166528991,
+      "learning_rate": 4.1850572232249596e-06,
+      "loss": 0.4432,
+      "step": 12906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.840457050501693,
+      "learning_rate": 4.184434562057943e-06,
+      "loss": 0.6343,
+      "step": 12907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8001724972013649,
+      "learning_rate": 4.183811928431472e-06,
+      "loss": 0.5201,
+      "step": 12908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7508442212605426,
+      "learning_rate": 4.183189322358582e-06,
+      "loss": 0.5271,
+      "step": 12909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.053196331790732,
+      "learning_rate": 4.182566743852305e-06,
+      "loss": 0.588,
+      "step": 12910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7227581679104331,
+      "learning_rate": 4.181944192925674e-06,
+      "loss": 0.4765,
+      "step": 12911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5105030851444905,
+      "learning_rate": 4.181321669591723e-06,
+      "loss": 0.4416,
+      "step": 12912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5317065678585284,
+      "learning_rate": 4.180699173863484e-06,
+      "loss": 0.3306,
+      "step": 12913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9871192545781045,
+      "learning_rate": 4.180076705753988e-06,
+      "loss": 0.5967,
+      "step": 12914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5841328800932506,
+      "learning_rate": 4.179454265276266e-06,
+      "loss": 0.4211,
+      "step": 12915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.56081911421206,
+      "learning_rate": 4.178831852443349e-06,
+      "loss": 0.4077,
+      "step": 12916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7846376260144758,
+      "learning_rate": 4.178209467268266e-06,
+      "loss": 0.6126,
+      "step": 12917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5414193107492193,
+      "learning_rate": 4.1775871097640485e-06,
+      "loss": 0.5375,
+      "step": 12918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5987159780834117,
+      "learning_rate": 4.176964779943723e-06,
+      "loss": 0.4316,
+      "step": 12919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4318732883271503,
+      "learning_rate": 4.176342477820317e-06,
+      "loss": 0.365,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7174225595543817,
+      "learning_rate": 4.17572020340686e-06,
+      "loss": 0.4007,
+      "step": 12921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4167456767904874,
+      "learning_rate": 4.175097956716377e-06,
+      "loss": 0.3614,
+      "step": 12922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8992575079344576,
+      "learning_rate": 4.174475737761897e-06,
+      "loss": 0.5033,
+      "step": 12923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1333205340731665,
+      "learning_rate": 4.173853546556444e-06,
+      "loss": 0.5996,
+      "step": 12924
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.736343518426057,
+      "learning_rate": 4.173231383113041e-06,
+      "loss": 0.5315,
+      "step": 12925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8954417816851394,
+      "learning_rate": 4.172609247444716e-06,
+      "loss": 0.5339,
+      "step": 12926
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.764333801658509,
+      "learning_rate": 4.171987139564492e-06,
+      "loss": 0.5075,
+      "step": 12927
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9313999066617418,
+      "learning_rate": 4.171365059485393e-06,
+      "loss": 0.6865,
+      "step": 12928
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.743496152775985,
+      "learning_rate": 4.17074300722044e-06,
+      "loss": 0.5777,
+      "step": 12929
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3673511025142,
+      "learning_rate": 4.170120982782659e-06,
+      "loss": 0.4008,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1648813012707624,
+      "learning_rate": 4.1694989861850665e-06,
+      "loss": 0.5596,
+      "step": 12931
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6219225374221105,
+      "learning_rate": 4.1688770174406875e-06,
+      "loss": 0.4593,
+      "step": 12932
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6748281623859689,
+      "learning_rate": 4.168255076562542e-06,
+      "loss": 0.5205,
+      "step": 12933
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.034576087505843,
+      "learning_rate": 4.167633163563648e-06,
+      "loss": 0.5428,
+      "step": 12934
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5181301365516595,
+      "learning_rate": 4.167011278457027e-06,
+      "loss": 0.4393,
+      "step": 12935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.021036919191038,
+      "learning_rate": 4.166389421255697e-06,
+      "loss": 0.4893,
+      "step": 12936
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4582630098742713,
+      "learning_rate": 4.165767591972675e-06,
+      "loss": 0.4112,
+      "step": 12937
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.65990309984696,
+      "learning_rate": 4.1651457906209815e-06,
+      "loss": 0.5234,
+      "step": 12938
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.843642124305904,
+      "learning_rate": 4.16452401721363e-06,
+      "loss": 0.4673,
+      "step": 12939
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.213953523095298,
+      "learning_rate": 4.16390227176364e-06,
+      "loss": 0.5011,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5960372724126308,
+      "learning_rate": 4.163280554284026e-06,
+      "loss": 0.3583,
+      "step": 12941
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4129422071621678,
+      "learning_rate": 4.162658864787802e-06,
+      "loss": 0.4799,
+      "step": 12942
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4231279836092061,
+      "learning_rate": 4.162037203287985e-06,
+      "loss": 0.4084,
+      "step": 12943
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5013394520179968,
+      "learning_rate": 4.161415569797588e-06,
+      "loss": 0.4543,
+      "step": 12944
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.729457226024976,
+      "learning_rate": 4.160793964329626e-06,
+      "loss": 0.4588,
+      "step": 12945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8295269737897761,
+      "learning_rate": 4.160172386897109e-06,
+      "loss": 0.5636,
+      "step": 12946
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2380711004869758,
+      "learning_rate": 4.15955083751305e-06,
+      "loss": 0.6744,
+      "step": 12947
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.12066908077826,
+      "learning_rate": 4.158929316190464e-06,
+      "loss": 0.5129,
+      "step": 12948
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.403253863121965,
+      "learning_rate": 4.158307822942359e-06,
+      "loss": 0.4143,
+      "step": 12949
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7354481215649153,
+      "learning_rate": 4.157686357781745e-06,
+      "loss": 0.4904,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5113911628723145,
+      "eval_runtime": 67.9842,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.726574276450158,
+      "learning_rate": 4.157064920721635e-06,
+      "loss": 0.5543,
+      "step": 12951
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9080972259066902,
+      "learning_rate": 4.156443511775035e-06,
+      "loss": 0.7638,
+      "step": 12952
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.567051008031067,
+      "learning_rate": 4.155822130954957e-06,
+      "loss": 0.5918,
+      "step": 12953
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.692440373873129,
+      "learning_rate": 4.155200778274408e-06,
+      "loss": 0.493,
+      "step": 12954
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.560877802281553,
+      "learning_rate": 4.1545794537463935e-06,
+      "loss": 0.5081,
+      "step": 12955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.477527405411249,
+      "learning_rate": 4.1539581573839246e-06,
+      "loss": 0.5617,
+      "step": 12956
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4626928397732635,
+      "learning_rate": 4.153336889200004e-06,
+      "loss": 0.3645,
+      "step": 12957
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5418309561190529,
+      "learning_rate": 4.15271564920764e-06,
+      "loss": 0.418,
+      "step": 12958
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7945787678632648,
+      "learning_rate": 4.152094437419838e-06,
+      "loss": 0.6013,
+      "step": 12959
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6679695129882182,
+      "learning_rate": 4.1514732538496e-06,
+      "loss": 0.4783,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7784081691392273,
+      "learning_rate": 4.150852098509933e-06,
+      "loss": 0.4449,
+      "step": 12961
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.488423718384487,
+      "learning_rate": 4.150230971413838e-06,
+      "loss": 0.5584,
+      "step": 12962
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9968449862640074,
+      "learning_rate": 4.149609872574321e-06,
+      "loss": 0.4015,
+      "step": 12963
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7569622112455872,
+      "learning_rate": 4.148988802004382e-06,
+      "loss": 0.5008,
+      "step": 12964
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671051156717613,
+      "learning_rate": 4.148367759717022e-06,
+      "loss": 0.6942,
+      "step": 12965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671538766957347,
+      "learning_rate": 4.1477467457252455e-06,
+      "loss": 0.484,
+      "step": 12966
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9351875894018273,
+      "learning_rate": 4.14712576004205e-06,
+      "loss": 0.7528,
+      "step": 12967
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9844880035779922,
+      "learning_rate": 4.146504802680435e-06,
+      "loss": 0.5388,
+      "step": 12968
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7129957397249038,
+      "learning_rate": 4.1458838736534036e-06,
+      "loss": 0.5472,
+      "step": 12969
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9304877461551715,
+      "learning_rate": 4.1452629729739505e-06,
+      "loss": 0.6381,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.937866730120218,
+      "learning_rate": 4.144642100655077e-06,
+      "loss": 0.6239,
+      "step": 12971
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5946196904281158,
+      "learning_rate": 4.1440212567097785e-06,
+      "loss": 0.4514,
+      "step": 12972
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1349491900349156,
+      "learning_rate": 4.143400441151053e-06,
+      "loss": 0.524,
+      "step": 12973
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7979399834101844,
+      "learning_rate": 4.142779653991894e-06,
+      "loss": 0.5163,
+      "step": 12974
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8507879251082382,
+      "learning_rate": 4.142158895245303e-06,
+      "loss": 0.5691,
+      "step": 12975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1709785041403853,
+      "learning_rate": 4.141538164924272e-06,
+      "loss": 0.5644,
+      "step": 12976
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5110387512255161,
+      "learning_rate": 4.140917463041795e-06,
+      "loss": 0.4086,
+      "step": 12977
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4979993554033129,
+      "learning_rate": 4.140296789610867e-06,
+      "loss": 0.4036,
+      "step": 12978
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5488812962713274,
+      "learning_rate": 4.139676144644481e-06,
+      "loss": 0.4165,
+      "step": 12979
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3774080830313893,
+      "learning_rate": 4.139055528155631e-06,
+      "loss": 0.6764,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0350964376364473,
+      "learning_rate": 4.138434940157308e-06,
+      "loss": 0.6544,
+      "step": 12981
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7925244105918363,
+      "learning_rate": 4.137814380662502e-06,
+      "loss": 0.5025,
+      "step": 12982
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8175793219446421,
+      "learning_rate": 4.137193849684208e-06,
+      "loss": 0.5526,
+      "step": 12983
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5184077132480762,
+      "learning_rate": 4.136573347235412e-06,
+      "loss": 0.4406,
+      "step": 12984
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8517136481922376,
+      "learning_rate": 4.135952873329109e-06,
+      "loss": 0.4524,
+      "step": 12985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4715990879621221,
+      "learning_rate": 4.135332427978283e-06,
+      "loss": 0.387,
+      "step": 12986
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9017246463091106,
+      "learning_rate": 4.134712011195926e-06,
+      "loss": 0.4952,
+      "step": 12987
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7541047023772993,
+      "learning_rate": 4.134091622995026e-06,
+      "loss": 0.4896,
+      "step": 12988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8404326628348864,
+      "learning_rate": 4.133471263388568e-06,
+      "loss": 0.577,
+      "step": 12989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1468750659556144,
+      "learning_rate": 4.13285093238954e-06,
+      "loss": 0.4922,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0651055768884152,
+      "learning_rate": 4.13223063001093e-06,
+      "loss": 0.4799,
+      "step": 12991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7986434013491601,
+      "learning_rate": 4.1316103562657204e-06,
+      "loss": 0.5671,
+      "step": 12992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5654514455151127,
+      "learning_rate": 4.1309901111668995e-06,
+      "loss": 0.3298,
+      "step": 12993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5497010567704903,
+      "learning_rate": 4.13036989472745e-06,
+      "loss": 0.3442,
+      "step": 12994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.816052369191501,
+      "learning_rate": 4.129749706960357e-06,
+      "loss": 0.4544,
+      "step": 12995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6046179733786379,
+      "learning_rate": 4.129129547878601e-06,
+      "loss": 0.4994,
+      "step": 12996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.744117773796909,
+      "learning_rate": 4.128509417495167e-06,
+      "loss": 0.4906,
+      "step": 12997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8248710546959765,
+      "learning_rate": 4.127889315823036e-06,
+      "loss": 0.5028,
+      "step": 12998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7214480934196372,
+      "learning_rate": 4.127269242875191e-06,
+      "loss": 0.5054,
+      "step": 12999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.130529382237908,
+      "learning_rate": 4.12664919866461e-06,
+      "loss": 0.4955,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5108364224433899,
+      "eval_runtime": 67.9844,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.757265549828176,
+      "learning_rate": 4.126029183204278e-06,
+      "loss": 0.4393,
+      "step": 13001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6907767528283997,
+      "learning_rate": 4.12540919650717e-06,
+      "loss": 0.4283,
+      "step": 13002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4885061189702884,
+      "learning_rate": 4.124789238586266e-06,
+      "loss": 0.356,
+      "step": 13003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7975975756822138,
+      "learning_rate": 4.1241693094545465e-06,
+      "loss": 0.4733,
+      "step": 13004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5755328216989968,
+      "learning_rate": 4.123549409124987e-06,
+      "loss": 0.4309,
+      "step": 13005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.218337981076605,
+      "learning_rate": 4.122929537610566e-06,
+      "loss": 0.6943,
+      "step": 13006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8069159919343734,
+      "learning_rate": 4.1223096949242605e-06,
+      "loss": 0.5059,
+      "step": 13007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.075851933731946,
+      "learning_rate": 4.121689881079045e-06,
+      "loss": 0.4239,
+      "step": 13008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.768435488108565,
+      "learning_rate": 4.121070096087896e-06,
+      "loss": 0.6872,
+      "step": 13009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5724807968804322,
+      "learning_rate": 4.120450339963786e-06,
+      "loss": 0.4808,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7414880844710146,
+      "learning_rate": 4.119830612719694e-06,
+      "loss": 0.5516,
+      "step": 13011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2516179088543296,
+      "learning_rate": 4.11921091436859e-06,
+      "loss": 0.6234,
+      "step": 13012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5482612788378807,
+      "learning_rate": 4.118591244923448e-06,
+      "loss": 0.4347,
+      "step": 13013
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6049339540094327,
+      "learning_rate": 4.117971604397241e-06,
+      "loss": 0.4204,
+      "step": 13014
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6879563031366491,
+      "learning_rate": 4.117351992802938e-06,
+      "loss": 0.509,
+      "step": 13015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.813680919070016,
+      "learning_rate": 4.116732410153514e-06,
+      "loss": 0.6087,
+      "step": 13016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.536739353619457,
+      "learning_rate": 4.1161128564619365e-06,
+      "loss": 0.406,
+      "step": 13017
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0968554964762545,
+      "learning_rate": 4.115493331741179e-06,
+      "loss": 0.5912,
+      "step": 13018
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6409971647057504,
+      "learning_rate": 4.114873836004206e-06,
+      "loss": 0.4841,
+      "step": 13019
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5264450405215948,
+      "learning_rate": 4.11425436926399e-06,
+      "loss": 0.528,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9028735409992592,
+      "learning_rate": 4.113634931533499e-06,
+      "loss": 0.5094,
+      "step": 13021
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9216369070562622,
+      "learning_rate": 4.113015522825698e-06,
+      "loss": 0.4961,
+      "step": 13022
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2333483018650164,
+      "learning_rate": 4.112396143153555e-06,
+      "loss": 0.4946,
+      "step": 13023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6382031687647092,
+      "learning_rate": 4.1117767925300375e-06,
+      "loss": 0.4547,
+      "step": 13024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6841777530529576,
+      "learning_rate": 4.111157470968111e-06,
+      "loss": 0.5994,
+      "step": 13025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6083892712028467,
+      "learning_rate": 4.11053817848074e-06,
+      "loss": 0.4259,
+      "step": 13026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8034761043451386,
+      "learning_rate": 4.10991891508089e-06,
+      "loss": 0.4832,
+      "step": 13027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6818048303355349,
+      "learning_rate": 4.1092996807815246e-06,
+      "loss": 0.3837,
+      "step": 13028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.660117711829497,
+      "learning_rate": 4.108680475595607e-06,
+      "loss": 0.3796,
+      "step": 13029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5798215957393005,
+      "learning_rate": 4.1080612995360965e-06,
+      "loss": 0.455,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.940844403379353,
+      "learning_rate": 4.10744215261596e-06,
+      "loss": 0.5187,
+      "step": 13031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7517100430940211,
+      "learning_rate": 4.1068230348481575e-06,
+      "loss": 0.6138,
+      "step": 13032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9155599653229927,
+      "learning_rate": 4.106203946245648e-06,
+      "loss": 0.6544,
+      "step": 13033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9105093771051491,
+      "learning_rate": 4.105584886821395e-06,
+      "loss": 0.6252,
+      "step": 13034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8398442776053265,
+      "learning_rate": 4.1049658565883545e-06,
+      "loss": 0.4651,
+      "step": 13035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6220464084882311,
+      "learning_rate": 4.104346855559489e-06,
+      "loss": 0.4551,
+      "step": 13036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.831422033473444,
+      "learning_rate": 4.103727883747754e-06,
+      "loss": 0.5196,
+      "step": 13037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.201392306205965,
+      "learning_rate": 4.103108941166108e-06,
+      "loss": 0.5405,
+      "step": 13038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.823879650735527,
+      "learning_rate": 4.10249002782751e-06,
+      "loss": 0.6244,
+      "step": 13039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3073769355540175,
+      "learning_rate": 4.101871143744914e-06,
+      "loss": 0.5831,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0380498823156623,
+      "learning_rate": 4.101252288931279e-06,
+      "loss": 0.5547,
+      "step": 13041
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8165225272211143,
+      "learning_rate": 4.100633463399557e-06,
+      "loss": 0.4493,
+      "step": 13042
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9237636394368627,
+      "learning_rate": 4.100014667162705e-06,
+      "loss": 0.4643,
+      "step": 13043
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7885405108576584,
+      "learning_rate": 4.099395900233676e-06,
+      "loss": 0.6275,
+      "step": 13044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7578979903211622,
+      "learning_rate": 4.098777162625423e-06,
+      "loss": 0.5498,
+      "step": 13045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.71003829547397,
+      "learning_rate": 4.0981584543509015e-06,
+      "loss": 0.3822,
+      "step": 13046
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.708271054517565,
+      "learning_rate": 4.097539775423062e-06,
+      "loss": 0.4465,
+      "step": 13047
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0086357587314243,
+      "learning_rate": 4.0969211258548544e-06,
+      "loss": 0.4562,
+      "step": 13048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.885817667515129,
+      "learning_rate": 4.096302505659233e-06,
+      "loss": 0.5433,
+      "step": 13049
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9138012972957874,
+      "learning_rate": 4.095683914849146e-06,
+      "loss": 0.4875,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5102566480636597,
+      "eval_runtime": 67.9383,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7938309442323812,
+      "learning_rate": 4.095065353437545e-06,
+      "loss": 0.4725,
+      "step": 13051
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9440038650090028,
+      "learning_rate": 4.094446821437378e-06,
+      "loss": 0.5078,
+      "step": 13052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.839004623852961,
+      "learning_rate": 4.093828318861593e-06,
+      "loss": 0.5752,
+      "step": 13053
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.070852453060591,
+      "learning_rate": 4.093209845723139e-06,
+      "loss": 0.5708,
+      "step": 13054
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.578657588242431,
+      "learning_rate": 4.092591402034963e-06,
+      "loss": 0.501,
+      "step": 13055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8117799668964178,
+      "learning_rate": 4.091972987810012e-06,
+      "loss": 0.4745,
+      "step": 13056
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2204390959686453,
+      "learning_rate": 4.091354603061233e-06,
+      "loss": 0.7619,
+      "step": 13057
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5682119017042162,
+      "learning_rate": 4.090736247801568e-06,
+      "loss": 0.4767,
+      "step": 13058
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.798204899210479,
+      "learning_rate": 4.0901179220439665e-06,
+      "loss": 0.5945,
+      "step": 13059
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4863243265970083,
+      "learning_rate": 4.08949962580137e-06,
+      "loss": 0.3634,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5144962572094187,
+      "learning_rate": 4.0888813590867215e-06,
+      "loss": 0.3953,
+      "step": 13061
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8666086462570948,
+      "learning_rate": 4.0882631219129675e-06,
+      "loss": 0.6379,
+      "step": 13062
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6691250225783034,
+      "learning_rate": 4.087644914293046e-06,
+      "loss": 0.4701,
+      "step": 13063
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5828111534382758,
+      "learning_rate": 4.087026736239902e-06,
+      "loss": 0.4589,
+      "step": 13064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7884588831071557,
+      "learning_rate": 4.086408587766476e-06,
+      "loss": 0.4578,
+      "step": 13065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.001783925003731,
+      "learning_rate": 4.085790468885707e-06,
+      "loss": 0.5146,
+      "step": 13066
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.856391966649036,
+      "learning_rate": 4.085172379610538e-06,
+      "loss": 0.4867,
+      "step": 13067
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3748713645985327,
+      "learning_rate": 4.0845543199539055e-06,
+      "loss": 0.3472,
+      "step": 13068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.896950262394898,
+      "learning_rate": 4.083936289928749e-06,
+      "loss": 0.4728,
+      "step": 13069
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.53650805840977,
+      "learning_rate": 4.0833182895480076e-06,
+      "loss": 0.5167,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9245892614732962,
+      "learning_rate": 4.082700318824616e-06,
+      "loss": 0.5214,
+      "step": 13071
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6253959063989776,
+      "learning_rate": 4.082082377771514e-06,
+      "loss": 0.4512,
+      "step": 13072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9474036541910815,
+      "learning_rate": 4.081464466401635e-06,
+      "loss": 0.5152,
+      "step": 13073
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8498760286478615,
+      "learning_rate": 4.0808465847279185e-06,
+      "loss": 0.4591,
+      "step": 13074
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6539085870965495,
+      "learning_rate": 4.080228732763296e-06,
+      "loss": 0.5507,
+      "step": 13075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.905903019825326,
+      "learning_rate": 4.079610910520705e-06,
+      "loss": 0.6448,
+      "step": 13076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9550910185489474,
+      "learning_rate": 4.078993118013076e-06,
+      "loss": 0.5366,
+      "step": 13077
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.002732826674489,
+      "learning_rate": 4.078375355253344e-06,
+      "loss": 0.4444,
+      "step": 13078
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5997105057764442,
+      "learning_rate": 4.07775762225444e-06,
+      "loss": 0.3486,
+      "step": 13079
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3709579348475516,
+      "learning_rate": 4.077139919029297e-06,
+      "loss": 0.271,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8083840141947862,
+      "learning_rate": 4.076522245590847e-06,
+      "loss": 0.5135,
+      "step": 13081
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.877246639872419,
+      "learning_rate": 4.075904601952019e-06,
+      "loss": 0.5936,
+      "step": 13082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7790913477268036,
+      "learning_rate": 4.075286988125744e-06,
+      "loss": 0.6165,
+      "step": 13083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.201399690817197,
+      "learning_rate": 4.0746694041249514e-06,
+      "loss": 0.5827,
+      "step": 13084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7517929682533802,
+      "learning_rate": 4.074051849962567e-06,
+      "loss": 0.4821,
+      "step": 13085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7736777084879194,
+      "learning_rate": 4.073434325651524e-06,
+      "loss": 0.413,
+      "step": 13086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4649474977409553,
+      "learning_rate": 4.0728168312047465e-06,
+      "loss": 0.4226,
+      "step": 13087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.424893718660538,
+      "learning_rate": 4.072199366635162e-06,
+      "loss": 0.4167,
+      "step": 13088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5081709250445097,
+      "learning_rate": 4.071581931955697e-06,
+      "loss": 0.4546,
+      "step": 13089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7912661125458678,
+      "learning_rate": 4.070964527179277e-06,
+      "loss": 0.4832,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.124177915822543,
+      "learning_rate": 4.070347152318829e-06,
+      "loss": 0.4788,
+      "step": 13091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7871421546799133,
+      "learning_rate": 4.069729807387274e-06,
+      "loss": 0.6204,
+      "step": 13092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.863457234039915,
+      "learning_rate": 4.069112492397536e-06,
+      "loss": 0.4054,
+      "step": 13093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7817811139028585,
+      "learning_rate": 4.06849520736254e-06,
+      "loss": 0.6354,
+      "step": 13094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9767735004800517,
+      "learning_rate": 4.067877952295206e-06,
+      "loss": 0.5512,
+      "step": 13095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2356086029224642,
+      "learning_rate": 4.067260727208461e-06,
+      "loss": 0.3299,
+      "step": 13096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0196597711766864,
+      "learning_rate": 4.0666435321152195e-06,
+      "loss": 0.5442,
+      "step": 13097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8793436328483468,
+      "learning_rate": 4.066026367028407e-06,
+      "loss": 0.5369,
+      "step": 13098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0568464686708547,
+      "learning_rate": 4.065409231960942e-06,
+      "loss": 0.7229,
+      "step": 13099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0485222682801294,
+      "learning_rate": 4.064792126925744e-06,
+      "loss": 0.483,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5104198455810547,
+      "eval_runtime": 67.9865,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.130520633063255,
+      "learning_rate": 4.064175051935729e-06,
+      "loss": 0.46,
+      "step": 13101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4790437952993607,
+      "learning_rate": 4.06355800700382e-06,
+      "loss": 0.4469,
+      "step": 13102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6584051538157618,
+      "learning_rate": 4.062940992142931e-06,
+      "loss": 0.5306,
+      "step": 13103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5301122395753688,
+      "learning_rate": 4.06232400736598e-06,
+      "loss": 0.4543,
+      "step": 13104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7428112161978557,
+      "learning_rate": 4.0617070526858826e-06,
+      "loss": 0.4675,
+      "step": 13105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8092298451801332,
+      "learning_rate": 4.0610901281155545e-06,
+      "loss": 0.3728,
+      "step": 13106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9825732617882745,
+      "learning_rate": 4.060473233667912e-06,
+      "loss": 0.5458,
+      "step": 13107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5666799452120448,
+      "learning_rate": 4.059856369355867e-06,
+      "loss": 0.4784,
+      "step": 13108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4380669491938995,
+      "learning_rate": 4.059239535192335e-06,
+      "loss": 0.3553,
+      "step": 13109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2060746444538037,
+      "learning_rate": 4.058622731190228e-06,
+      "loss": 0.5879,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6005970303141603,
+      "learning_rate": 4.058005957362458e-06,
+      "loss": 0.5334,
+      "step": 13111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6155588843764928,
+      "learning_rate": 4.05738921372194e-06,
+      "loss": 0.3653,
+      "step": 13112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5607554707345264,
+      "learning_rate": 4.056772500281581e-06,
+      "loss": 0.4339,
+      "step": 13113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.783117027938276,
+      "learning_rate": 4.0561558170542935e-06,
+      "loss": 0.4347,
+      "step": 13114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3625898556353944,
+      "learning_rate": 4.0555391640529876e-06,
+      "loss": 0.3657,
+      "step": 13115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0621306638679813,
+      "learning_rate": 4.054922541290572e-06,
+      "loss": 0.6729,
+      "step": 13116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6738518241523936,
+      "learning_rate": 4.054305948779955e-06,
+      "loss": 0.364,
+      "step": 13117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6358618763419852,
+      "learning_rate": 4.053689386534045e-06,
+      "loss": 0.4266,
+      "step": 13118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7476920708934427,
+      "learning_rate": 4.0530728545657505e-06,
+      "loss": 0.505,
+      "step": 13119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.051074787495192,
+      "learning_rate": 4.052456352887977e-06,
+      "loss": 0.5255,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8462208370908766,
+      "learning_rate": 4.05183988151363e-06,
+      "loss": 0.5388,
+      "step": 13121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6719822874156813,
+      "learning_rate": 4.051223440455613e-06,
+      "loss": 0.4022,
+      "step": 13122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4798115724187624,
+      "learning_rate": 4.050607029726836e-06,
+      "loss": 0.4638,
+      "step": 13123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7637642732603191,
+      "learning_rate": 4.049990649340201e-06,
+      "loss": 0.5686,
+      "step": 13124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.076907101023456,
+      "learning_rate": 4.04937429930861e-06,
+      "loss": 0.6912,
+      "step": 13125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1426350304456507,
+      "learning_rate": 4.048757979644968e-06,
+      "loss": 0.5823,
+      "step": 13126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.849054145349166,
+      "learning_rate": 4.0481416903621765e-06,
+      "loss": 0.5617,
+      "step": 13127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4377465793078392,
+      "learning_rate": 4.047525431473136e-06,
+      "loss": 0.455,
+      "step": 13128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8789142310854818,
+      "learning_rate": 4.0469092029907485e-06,
+      "loss": 0.5496,
+      "step": 13129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5754181709880226,
+      "learning_rate": 4.046293004927915e-06,
+      "loss": 0.4794,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7957035497892717,
+      "learning_rate": 4.0456768372975345e-06,
+      "loss": 0.5945,
+      "step": 13131
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.625913126958117,
+      "learning_rate": 4.045060700112505e-06,
+      "loss": 0.5746,
+      "step": 13132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7348031587418997,
+      "learning_rate": 4.044444593385727e-06,
+      "loss": 0.4917,
+      "step": 13133
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5011892064646988,
+      "learning_rate": 4.043828517130098e-06,
+      "loss": 0.5257,
+      "step": 13134
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0461634261065345,
+      "learning_rate": 4.043212471358512e-06,
+      "loss": 0.5731,
+      "step": 13135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9284052668119918,
+      "learning_rate": 4.0425964560838706e-06,
+      "loss": 0.5486,
+      "step": 13136
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8725027295486534,
+      "learning_rate": 4.041980471319067e-06,
+      "loss": 0.6523,
+      "step": 13137
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.949571035771303,
+      "learning_rate": 4.041364517076994e-06,
+      "loss": 0.3617,
+      "step": 13138
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7146847568781145,
+      "learning_rate": 4.040748593370553e-06,
+      "loss": 0.4635,
+      "step": 13139
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7531715432998944,
+      "learning_rate": 4.040132700212631e-06,
+      "loss": 0.4908,
+      "step": 13140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.901389376210766,
+      "learning_rate": 4.039516837616124e-06,
+      "loss": 0.5107,
+      "step": 13141
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6425839011739483,
+      "learning_rate": 4.038901005593926e-06,
+      "loss": 0.5576,
+      "step": 13142
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8108561984820377,
+      "learning_rate": 4.038285204158927e-06,
+      "loss": 0.4601,
+      "step": 13143
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9439763638362058,
+      "learning_rate": 4.03766943332402e-06,
+      "loss": 0.5748,
+      "step": 13144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.23499155159485,
+      "learning_rate": 4.037053693102096e-06,
+      "loss": 0.59,
+      "step": 13145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9159414646958024,
+      "learning_rate": 4.0364379835060415e-06,
+      "loss": 0.4927,
+      "step": 13146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3258521368894134,
+      "learning_rate": 4.035822304548751e-06,
+      "loss": 0.6582,
+      "step": 13147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.802321007046129,
+      "learning_rate": 4.035206656243111e-06,
+      "loss": 0.4612,
+      "step": 13148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7700902943732455,
+      "learning_rate": 4.03459103860201e-06,
+      "loss": 0.5683,
+      "step": 13149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.65084152123263,
+      "learning_rate": 4.033975451638336e-06,
+      "loss": 0.3868,
+      "step": 13150
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5099422335624695,
+      "eval_runtime": 67.9384,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 13150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6514814269756821,
+      "learning_rate": 4.033359895364973e-06,
+      "loss": 0.4764,
+      "step": 13151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6103307608732584,
+      "learning_rate": 4.032744369794811e-06,
+      "loss": 0.4089,
+      "step": 13152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9475892638360814,
+      "learning_rate": 4.032128874940734e-06,
+      "loss": 0.4439,
+      "step": 13153
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9440234242466838,
+      "learning_rate": 4.031513410815627e-06,
+      "loss": 0.5477,
+      "step": 13154
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5854416206084014,
+      "learning_rate": 4.030897977432375e-06,
+      "loss": 0.5425,
+      "step": 13155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7239742929453472,
+      "learning_rate": 4.030282574803861e-06,
+      "loss": 0.5019,
+      "step": 13156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5027370810187535,
+      "learning_rate": 4.0296672029429696e-06,
+      "loss": 0.436,
+      "step": 13157
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4867530580290924,
+      "learning_rate": 4.029051861862582e-06,
+      "loss": 0.4404,
+      "step": 13158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6520647302096254,
+      "learning_rate": 4.028436551575579e-06,
+      "loss": 0.4198,
+      "step": 13159
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7502504796807412,
+      "learning_rate": 4.027821272094843e-06,
+      "loss": 0.4744,
+      "step": 13160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9200187141316296,
+      "learning_rate": 4.027206023433254e-06,
+      "loss": 0.5393,
+      "step": 13161
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.158419008320145,
+      "learning_rate": 4.026590805603693e-06,
+      "loss": 0.5487,
+      "step": 13162
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7355289219102656,
+      "learning_rate": 4.025975618619037e-06,
+      "loss": 0.3903,
+      "step": 13163
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5391036982681272,
+      "learning_rate": 4.025360462492167e-06,
+      "loss": 0.3638,
+      "step": 13164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4328378053555226,
+      "learning_rate": 4.024745337235959e-06,
+      "loss": 0.3649,
+      "step": 13165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6943736374874254,
+      "learning_rate": 4.024130242863292e-06,
+      "loss": 0.4979,
+      "step": 13166
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8062933151921874,
+      "learning_rate": 4.0235151793870405e-06,
+      "loss": 0.6441,
+      "step": 13167
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8471786491571087,
+      "learning_rate": 4.022900146820083e-06,
+      "loss": 0.544,
+      "step": 13168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6676216937374737,
+      "learning_rate": 4.022285145175292e-06,
+      "loss": 0.5129,
+      "step": 13169
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4399633283295779,
+      "learning_rate": 4.021670174465545e-06,
+      "loss": 0.4021,
+      "step": 13170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6786799910429764,
+      "learning_rate": 4.021055234703713e-06,
+      "loss": 0.4182,
+      "step": 13171
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7850275361829715,
+      "learning_rate": 4.020440325902671e-06,
+      "loss": 0.5156,
+      "step": 13172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.470410537478779,
+      "learning_rate": 4.019825448075293e-06,
+      "loss": 0.53,
+      "step": 13173
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.821557333688833,
+      "learning_rate": 4.0192106012344505e-06,
+      "loss": 0.5264,
+      "step": 13174
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8741737000424776,
+      "learning_rate": 4.0185957853930125e-06,
+      "loss": 0.6299,
+      "step": 13175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7291466702468832,
+      "learning_rate": 4.017981000563853e-06,
+      "loss": 0.451,
+      "step": 13176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.980590879759026,
+      "learning_rate": 4.017366246759841e-06,
+      "loss": 0.754,
+      "step": 13177
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.873591862062339,
+      "learning_rate": 4.016751523993845e-06,
+      "loss": 0.3932,
+      "step": 13178
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7744901601256198,
+      "learning_rate": 4.016136832278735e-06,
+      "loss": 0.4917,
+      "step": 13179
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.506518575058064,
+      "learning_rate": 4.015522171627379e-06,
+      "loss": 0.4603,
+      "step": 13180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8026833781961362,
+      "learning_rate": 4.014907542052643e-06,
+      "loss": 0.4541,
+      "step": 13181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1876304674607296,
+      "learning_rate": 4.014292943567396e-06,
+      "loss": 0.5243,
+      "step": 13182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7705376427895003,
+      "learning_rate": 4.013678376184503e-06,
+      "loss": 0.3906,
+      "step": 13183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4137063185050742,
+      "learning_rate": 4.0130638399168305e-06,
+      "loss": 0.5387,
+      "step": 13184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1276380948421045,
+      "learning_rate": 4.012449334777243e-06,
+      "loss": 0.5869,
+      "step": 13185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9157393363110549,
+      "learning_rate": 4.011834860778604e-06,
+      "loss": 0.4907,
+      "step": 13186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6662223855933693,
+      "learning_rate": 4.011220417933779e-06,
+      "loss": 0.4204,
+      "step": 13187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9204253738407095,
+      "learning_rate": 4.010606006255629e-06,
+      "loss": 0.6387,
+      "step": 13188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.714179605911492,
+      "learning_rate": 4.009991625757017e-06,
+      "loss": 0.3995,
+      "step": 13189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.752466946997335,
+      "learning_rate": 4.0093772764508064e-06,
+      "loss": 0.4723,
+      "step": 13190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1087880336739646,
+      "learning_rate": 4.008762958349854e-06,
+      "loss": 0.6429,
+      "step": 13191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6069788012337087,
+      "learning_rate": 4.008148671467026e-06,
+      "loss": 0.5165,
+      "step": 13192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8438756168909025,
+      "learning_rate": 4.007534415815176e-06,
+      "loss": 0.505,
+      "step": 13193
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.013279844514564,
+      "learning_rate": 4.006920191407168e-06,
+      "loss": 0.5049,
+      "step": 13194
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.442138471241453,
+      "learning_rate": 4.006305998255857e-06,
+      "loss": 0.3398,
+      "step": 13195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3352243650854407,
+      "learning_rate": 4.005691836374102e-06,
+      "loss": 0.6678,
+      "step": 13196
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.807809941116442,
+      "learning_rate": 4.005077705774762e-06,
+      "loss": 0.4862,
+      "step": 13197
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9179174300181054,
+      "learning_rate": 4.004463606470691e-06,
+      "loss": 0.3736,
+      "step": 13198
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6802243984724414,
+      "learning_rate": 4.0038495384747455e-06,
+      "loss": 0.5001,
+      "step": 13199
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9374488943910841,
+      "learning_rate": 4.00323550179978e-06,
+      "loss": 0.5234,
+      "step": 13200
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5091068148612976,
+      "eval_runtime": 67.8894,
+      "eval_samples_per_second": 50.229,
+      "eval_steps_per_second": 1.576,
+      "step": 13200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6252663234027969,
+      "learning_rate": 4.002621496458649e-06,
+      "loss": 0.5403,
+      "step": 13201
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9578076032960974,
+      "learning_rate": 4.002007522464209e-06,
+      "loss": 0.4132,
+      "step": 13202
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6970158767071448,
+      "learning_rate": 4.0013935798293105e-06,
+      "loss": 0.5701,
+      "step": 13203
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.151729914073116,
+      "learning_rate": 4.000779668566805e-06,
+      "loss": 0.4953,
+      "step": 13204
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8142756782001088,
+      "learning_rate": 4.000165788689547e-06,
+      "loss": 0.4417,
+      "step": 13205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6648563437487693,
+      "learning_rate": 3.999551940210388e-06,
+      "loss": 0.4822,
+      "step": 13206
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.488968034759153,
+      "learning_rate": 3.998938123142174e-06,
+      "loss": 0.3864,
+      "step": 13207
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2238815350872407,
+      "learning_rate": 3.998324337497761e-06,
+      "loss": 0.6234,
+      "step": 13208
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6814710861285045,
+      "learning_rate": 3.997710583289993e-06,
+      "loss": 0.4597,
+      "step": 13209
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8199493420968489,
+      "learning_rate": 3.997096860531723e-06,
+      "loss": 0.4748,
+      "step": 13210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.853718008266346,
+      "learning_rate": 3.996483169235796e-06,
+      "loss": 0.5325,
+      "step": 13211
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.805606559693227,
+      "learning_rate": 3.995869509415059e-06,
+      "loss": 0.5231,
+      "step": 13212
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7970282955535182,
+      "learning_rate": 3.9952558810823604e-06,
+      "loss": 0.4959,
+      "step": 13213
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.395263254919815,
+      "learning_rate": 3.994642284250545e-06,
+      "loss": 0.4925,
+      "step": 13214
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5198849207887313,
+      "learning_rate": 3.994028718932459e-06,
+      "loss": 0.5008,
+      "step": 13215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9052736231529575,
+      "learning_rate": 3.993415185140946e-06,
+      "loss": 0.5411,
+      "step": 13216
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1079309049895354,
+      "learning_rate": 3.992801682888849e-06,
+      "loss": 0.5508,
+      "step": 13217
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0223090768053402,
+      "learning_rate": 3.992188212189014e-06,
+      "loss": 0.5834,
+      "step": 13218
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.068336157965434,
+      "learning_rate": 3.991574773054282e-06,
+      "loss": 0.6978,
+      "step": 13219
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7688574081289798,
+      "learning_rate": 3.9909613654974946e-06,
+      "loss": 0.4975,
+      "step": 13220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8167411130323334,
+      "learning_rate": 3.990347989531494e-06,
+      "loss": 0.4921,
+      "step": 13221
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3497556341818167,
+      "learning_rate": 3.989734645169122e-06,
+      "loss": 0.3418,
+      "step": 13222
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.033530262141373,
+      "learning_rate": 3.9891213324232144e-06,
+      "loss": 0.6111,
+      "step": 13223
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.948383484352404,
+      "learning_rate": 3.9885080513066154e-06,
+      "loss": 0.5953,
+      "step": 13224
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9194620265704967,
+      "learning_rate": 3.987894801832162e-06,
+      "loss": 0.4606,
+      "step": 13225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5368093223697832,
+      "learning_rate": 3.98728158401269e-06,
+      "loss": 0.4148,
+      "step": 13226
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8100873785353648,
+      "learning_rate": 3.98666839786104e-06,
+      "loss": 0.4894,
+      "step": 13227
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6908150954046484,
+      "learning_rate": 3.986055243390045e-06,
+      "loss": 0.4333,
+      "step": 13228
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6792685517697572,
+      "learning_rate": 3.985442120612545e-06,
+      "loss": 0.4543,
+      "step": 13229
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8357229381141107,
+      "learning_rate": 3.984829029541374e-06,
+      "loss": 0.5182,
+      "step": 13230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.558730727645225,
+      "learning_rate": 3.984215970189364e-06,
+      "loss": 0.4487,
+      "step": 13231
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4341671217052778,
+      "learning_rate": 3.983602942569354e-06,
+      "loss": 0.3573,
+      "step": 13232
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.750295687724945,
+      "learning_rate": 3.982989946694173e-06,
+      "loss": 0.5929,
+      "step": 13233
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8345867857754579,
+      "learning_rate": 3.982376982576653e-06,
+      "loss": 0.4668,
+      "step": 13234
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.835090802571397,
+      "learning_rate": 3.981764050229632e-06,
+      "loss": 0.4042,
+      "step": 13235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8018060602168051,
+      "learning_rate": 3.981151149665935e-06,
+      "loss": 0.5977,
+      "step": 13236
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.769511335269914,
+      "learning_rate": 3.980538280898396e-06,
+      "loss": 0.5569,
+      "step": 13237
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9632540420059903,
+      "learning_rate": 3.979925443939845e-06,
+      "loss": 0.5608,
+      "step": 13238
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0350315504018646,
+      "learning_rate": 3.97931263880311e-06,
+      "loss": 0.6165,
+      "step": 13239
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8448405025558225,
+      "learning_rate": 3.978699865501021e-06,
+      "loss": 0.47,
+      "step": 13240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.756885762326226,
+      "learning_rate": 3.978087124046403e-06,
+      "loss": 0.4487,
+      "step": 13241
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5319987211516404,
+      "learning_rate": 3.97747441445209e-06,
+      "loss": 0.4888,
+      "step": 13242
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1366279445772083,
+      "learning_rate": 3.9768617367309014e-06,
+      "loss": 0.6199,
+      "step": 13243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7321547028664506,
+      "learning_rate": 3.976249090895666e-06,
+      "loss": 0.4844,
+      "step": 13244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1890495311116434,
+      "learning_rate": 3.975636476959212e-06,
+      "loss": 0.6625,
+      "step": 13245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7139601544417162,
+      "learning_rate": 3.97502389493436e-06,
+      "loss": 0.4731,
+      "step": 13246
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4419472404518905,
+      "learning_rate": 3.974411344833935e-06,
+      "loss": 0.4987,
+      "step": 13247
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.805240147585372,
+      "learning_rate": 3.973798826670763e-06,
+      "loss": 0.5234,
+      "step": 13248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8387241026611008,
+      "learning_rate": 3.973186340457662e-06,
+      "loss": 0.4974,
+      "step": 13249
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.928880735365129,
+      "learning_rate": 3.972573886207458e-06,
+      "loss": 0.4906,
+      "step": 13250
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5087615847587585,
+      "eval_runtime": 67.9055,
+      "eval_samples_per_second": 50.217,
+      "eval_steps_per_second": 1.576,
+      "step": 13250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0555219334266424,
+      "learning_rate": 3.97196146393297e-06,
+      "loss": 0.6125,
+      "step": 13251
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6150196821357787,
+      "learning_rate": 3.9713490736470205e-06,
+      "loss": 0.6398,
+      "step": 13252
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0540931791056707,
+      "learning_rate": 3.970736715362429e-06,
+      "loss": 0.5273,
+      "step": 13253
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.978986283731586,
+      "learning_rate": 3.970124389092014e-06,
+      "loss": 0.5248,
+      "step": 13254
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.123462594276873,
+      "learning_rate": 3.969512094848594e-06,
+      "loss": 0.5376,
+      "step": 13255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5723215113610682,
+      "learning_rate": 3.968899832644989e-06,
+      "loss": 0.4325,
+      "step": 13256
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2734334234647773,
+      "learning_rate": 3.9682876024940145e-06,
+      "loss": 0.3302,
+      "step": 13257
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4776753789259287,
+      "learning_rate": 3.967675404408488e-06,
+      "loss": 0.4568,
+      "step": 13258
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7443910551524369,
+      "learning_rate": 3.967063238401223e-06,
+      "loss": 0.531,
+      "step": 13259
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5726238052624884,
+      "learning_rate": 3.96645110448504e-06,
+      "loss": 0.5169,
+      "step": 13260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3157707322651275,
+      "learning_rate": 3.965839002672749e-06,
+      "loss": 0.3365,
+      "step": 13261
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9730120305200218,
+      "learning_rate": 3.965226932977166e-06,
+      "loss": 0.5565,
+      "step": 13262
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8024903308143394,
+      "learning_rate": 3.964614895411103e-06,
+      "loss": 0.5289,
+      "step": 13263
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4294603202831775,
+      "learning_rate": 3.964002889987375e-06,
+      "loss": 0.3665,
+      "step": 13264
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5167305859093037,
+      "learning_rate": 3.963390916718791e-06,
+      "loss": 0.4258,
+      "step": 13265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6811962200463133,
+      "learning_rate": 3.9627789756181656e-06,
+      "loss": 0.5601,
+      "step": 13266
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7020288857550427,
+      "learning_rate": 3.962167066698305e-06,
+      "loss": 0.42,
+      "step": 13267
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6248055847092262,
+      "learning_rate": 3.961555189972024e-06,
+      "loss": 0.556,
+      "step": 13268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6198803275901192,
+      "learning_rate": 3.960943345452128e-06,
+      "loss": 0.5392,
+      "step": 13269
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4185418931751055,
+      "learning_rate": 3.9603315331514266e-06,
+      "loss": 0.3865,
+      "step": 13270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5351809895384252,
+      "learning_rate": 3.959719753082729e-06,
+      "loss": 0.4526,
+      "step": 13271
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8877759021754086,
+      "learning_rate": 3.959108005258842e-06,
+      "loss": 0.553,
+      "step": 13272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9565044102816682,
+      "learning_rate": 3.958496289692573e-06,
+      "loss": 0.5619,
+      "step": 13273
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.643855509297049,
+      "learning_rate": 3.957884606396725e-06,
+      "loss": 0.3964,
+      "step": 13274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.80162102515623,
+      "learning_rate": 3.957272955384105e-06,
+      "loss": 0.6827,
+      "step": 13275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4759036017269365,
+      "learning_rate": 3.956661336667518e-06,
+      "loss": 0.4532,
+      "step": 13276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6923396766855296,
+      "learning_rate": 3.956049750259767e-06,
+      "loss": 0.5345,
+      "step": 13277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.839127747447994,
+      "learning_rate": 3.9554381961736565e-06,
+      "loss": 0.5105,
+      "step": 13278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5018517833819605,
+      "learning_rate": 3.9548266744219865e-06,
+      "loss": 0.3814,
+      "step": 13279
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.863589581948683,
+      "learning_rate": 3.954215185017561e-06,
+      "loss": 0.4794,
+      "step": 13280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5049893582452052,
+      "learning_rate": 3.95360372797318e-06,
+      "loss": 0.4485,
+      "step": 13281
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5947695313582406,
+      "learning_rate": 3.952992303301646e-06,
+      "loss": 0.4756,
+      "step": 13282
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828266485661123,
+      "learning_rate": 3.952380911015756e-06,
+      "loss": 0.5222,
+      "step": 13283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4253191451791754,
+      "learning_rate": 3.95176955112831e-06,
+      "loss": 0.3449,
+      "step": 13284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1328944065280973,
+      "learning_rate": 3.951158223652107e-06,
+      "loss": 0.6189,
+      "step": 13285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8593153640155278,
+      "learning_rate": 3.950546928599946e-06,
+      "loss": 0.6177,
+      "step": 13286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.582040571459916,
+      "learning_rate": 3.949935665984621e-06,
+      "loss": 0.4766,
+      "step": 13287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8503946002471998,
+      "learning_rate": 3.949324435818931e-06,
+      "loss": 0.4323,
+      "step": 13288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.546063828872779,
+      "learning_rate": 3.94871323811567e-06,
+      "loss": 0.3359,
+      "step": 13289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7916970471271239,
+      "learning_rate": 3.948102072887634e-06,
+      "loss": 0.5702,
+      "step": 13290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0155728529686217,
+      "learning_rate": 3.947490940147618e-06,
+      "loss": 0.6261,
+      "step": 13291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.881766346893894,
+      "learning_rate": 3.946879839908414e-06,
+      "loss": 0.4901,
+      "step": 13292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.285574422959749,
+      "learning_rate": 3.946268772182816e-06,
+      "loss": 0.3222,
+      "step": 13293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5625733524122674,
+      "learning_rate": 3.9456577369836155e-06,
+      "loss": 0.4328,
+      "step": 13294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8638647686527818,
+      "learning_rate": 3.945046734323608e-06,
+      "loss": 0.5324,
+      "step": 13295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528449607732159,
+      "learning_rate": 3.944435764215579e-06,
+      "loss": 0.4508,
+      "step": 13296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6909542733444214,
+      "learning_rate": 3.94382482667232e-06,
+      "loss": 0.4889,
+      "step": 13297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7653136012417823,
+      "learning_rate": 3.9432139217066235e-06,
+      "loss": 0.475,
+      "step": 13298
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.809373536477976,
+      "learning_rate": 3.942603049331275e-06,
+      "loss": 0.4192,
+      "step": 13299
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0426472338922057,
+      "learning_rate": 3.941992209559066e-06,
+      "loss": 0.4892,
+      "step": 13300
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5101577043533325,
+      "eval_runtime": 67.95,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 13300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6690440941046274,
+      "learning_rate": 3.941381402402782e-06,
+      "loss": 0.6125,
+      "step": 13301
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6298241268146307,
+      "learning_rate": 3.940770627875208e-06,
+      "loss": 0.5918,
+      "step": 13302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6373281296904556,
+      "learning_rate": 3.9401598859891355e-06,
+      "loss": 0.4627,
+      "step": 13303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8783057152130491,
+      "learning_rate": 3.9395491767573455e-06,
+      "loss": 0.4448,
+      "step": 13304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4508621225437086,
+      "learning_rate": 3.938938500192624e-06,
+      "loss": 0.3913,
+      "step": 13305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9505589919792758,
+      "learning_rate": 3.938327856307756e-06,
+      "loss": 0.546,
+      "step": 13306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.891833473173275,
+      "learning_rate": 3.9377172451155245e-06,
+      "loss": 0.6373,
+      "step": 13307
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.008756782780875,
+      "learning_rate": 3.937106666628711e-06,
+      "loss": 0.516,
+      "step": 13308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.980598562213332,
+      "learning_rate": 3.936496120860101e-06,
+      "loss": 0.5912,
+      "step": 13309
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8263534018550733,
+      "learning_rate": 3.9358856078224705e-06,
+      "loss": 0.4325,
+      "step": 13310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7715804492875442,
+      "learning_rate": 3.935275127528605e-06,
+      "loss": 0.5638,
+      "step": 13311
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.533992440009013,
+      "learning_rate": 3.934664679991283e-06,
+      "loss": 0.5256,
+      "step": 13312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8665646588178808,
+      "learning_rate": 3.934054265223283e-06,
+      "loss": 0.4908,
+      "step": 13313
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.393638414022685,
+      "learning_rate": 3.933443883237386e-06,
+      "loss": 0.3673,
+      "step": 13314
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1878368321104844,
+      "learning_rate": 3.9328335340463655e-06,
+      "loss": 0.6091,
+      "step": 13315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6936864453637621,
+      "learning_rate": 3.932223217663004e-06,
+      "loss": 0.3984,
+      "step": 13316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.838625864241871,
+      "learning_rate": 3.931612934100074e-06,
+      "loss": 0.5405,
+      "step": 13317
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.591960524189726,
+      "learning_rate": 3.931002683370353e-06,
+      "loss": 0.5762,
+      "step": 13318
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7340701928307651,
+      "learning_rate": 3.930392465486615e-06,
+      "loss": 0.5182,
+      "step": 13319
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7775135251069296,
+      "learning_rate": 3.929782280461639e-06,
+      "loss": 0.4575,
+      "step": 13320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5375992414022026,
+      "learning_rate": 3.929172128308194e-06,
+      "loss": 0.3944,
+      "step": 13321
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6535320359268542,
+      "learning_rate": 3.9285620090390555e-06,
+      "loss": 0.4077,
+      "step": 13322
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5775518697076905,
+      "learning_rate": 3.927951922666995e-06,
+      "loss": 0.4776,
+      "step": 13323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4996853427515335,
+      "learning_rate": 3.927341869204785e-06,
+      "loss": 0.4294,
+      "step": 13324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7308755796366926,
+      "learning_rate": 3.926731848665195e-06,
+      "loss": 0.4436,
+      "step": 13325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7755235209833495,
+      "learning_rate": 3.926121861060997e-06,
+      "loss": 0.4028,
+      "step": 13326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7317942593074727,
+      "learning_rate": 3.925511906404962e-06,
+      "loss": 0.3641,
+      "step": 13327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6570679267979285,
+      "learning_rate": 3.924901984709855e-06,
+      "loss": 0.4254,
+      "step": 13328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7638963938144807,
+      "learning_rate": 3.9242920959884464e-06,
+      "loss": 0.6079,
+      "step": 13329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5267464081766702,
+      "learning_rate": 3.923682240253506e-06,
+      "loss": 0.457,
+      "step": 13330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.74561705818966,
+      "learning_rate": 3.923072417517798e-06,
+      "loss": 0.4461,
+      "step": 13331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7513335630747418,
+      "learning_rate": 3.92246262779409e-06,
+      "loss": 0.5886,
+      "step": 13332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.742799966708952,
+      "learning_rate": 3.921852871095147e-06,
+      "loss": 0.4981,
+      "step": 13333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7957450293069948,
+      "learning_rate": 3.921243147433732e-06,
+      "loss": 0.4572,
+      "step": 13334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5522537253644078,
+      "learning_rate": 3.920633456822613e-06,
+      "loss": 0.4045,
+      "step": 13335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7873625732021121,
+      "learning_rate": 3.9200237992745515e-06,
+      "loss": 0.4702,
+      "step": 13336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5444901223242018,
+      "learning_rate": 3.91941417480231e-06,
+      "loss": 0.3883,
+      "step": 13337
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6633717206625367,
+      "learning_rate": 3.918804583418651e-06,
+      "loss": 0.4553,
+      "step": 13338
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.989336698575635,
+      "learning_rate": 3.918195025136335e-06,
+      "loss": 0.7282,
+      "step": 13339
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4985659020760362,
+      "learning_rate": 3.917585499968125e-06,
+      "loss": 0.4605,
+      "step": 13340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8035257989010356,
+      "learning_rate": 3.916976007926779e-06,
+      "loss": 0.5115,
+      "step": 13341
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6080981275741506,
+      "learning_rate": 3.916366549025058e-06,
+      "loss": 0.4455,
+      "step": 13342
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7006823865187577,
+      "learning_rate": 3.915757123275719e-06,
+      "loss": 0.5924,
+      "step": 13343
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7600775871442773,
+      "learning_rate": 3.915147730691521e-06,
+      "loss": 0.4307,
+      "step": 13344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6781119785224778,
+      "learning_rate": 3.914538371285219e-06,
+      "loss": 0.4583,
+      "step": 13345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9657061017375188,
+      "learning_rate": 3.913929045069573e-06,
+      "loss": 0.5386,
+      "step": 13346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9828129141635802,
+      "learning_rate": 3.913319752057337e-06,
+      "loss": 0.5682,
+      "step": 13347
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.514504982283588,
+      "learning_rate": 3.912710492261267e-06,
+      "loss": 0.472,
+      "step": 13348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5038049845392554,
+      "learning_rate": 3.912101265694117e-06,
+      "loss": 0.4306,
+      "step": 13349
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4115428972636375,
+      "learning_rate": 3.91149207236864e-06,
+      "loss": 0.3639,
+      "step": 13350
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5091829895973206,
+      "eval_runtime": 67.9745,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 13350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8011578383030562,
+      "learning_rate": 3.9108829122975914e-06,
+      "loss": 0.5699,
+      "step": 13351
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6197926192039902,
+      "learning_rate": 3.910273785493719e-06,
+      "loss": 0.4601,
+      "step": 13352
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.261461518635062,
+      "learning_rate": 3.90966469196978e-06,
+      "loss": 0.4261,
+      "step": 13353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8073775569841048,
+      "learning_rate": 3.909055631738522e-06,
+      "loss": 0.3667,
+      "step": 13354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.737665049317176,
+      "learning_rate": 3.9084466048126966e-06,
+      "loss": 0.5494,
+      "step": 13355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.645490632207161,
+      "learning_rate": 3.907837611205053e-06,
+      "loss": 0.4158,
+      "step": 13356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.619690113376639,
+      "learning_rate": 3.90722865092834e-06,
+      "loss": 0.4224,
+      "step": 13357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0754353502191094,
+      "learning_rate": 3.906619723995304e-06,
+      "loss": 0.4135,
+      "step": 13358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7882815637455665,
+      "learning_rate": 3.906010830418697e-06,
+      "loss": 0.3678,
+      "step": 13359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.695150085771504,
+      "learning_rate": 3.905401970211262e-06,
+      "loss": 0.4449,
+      "step": 13360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8044056554690546,
+      "learning_rate": 3.904793143385747e-06,
+      "loss": 0.4874,
+      "step": 13361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2588682132402056,
+      "learning_rate": 3.9041843499548955e-06,
+      "loss": 0.662,
+      "step": 13362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.566047882976586,
+      "learning_rate": 3.903575589931455e-06,
+      "loss": 0.4591,
+      "step": 13363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.775249985378024,
+      "learning_rate": 3.9029668633281655e-06,
+      "loss": 0.5493,
+      "step": 13364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921912407164852,
+      "learning_rate": 3.902358170157775e-06,
+      "loss": 0.6655,
+      "step": 13365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7772682296340163,
+      "learning_rate": 3.901749510433024e-06,
+      "loss": 0.7297,
+      "step": 13366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8450513339597565,
+      "learning_rate": 3.901140884166653e-06,
+      "loss": 0.6082,
+      "step": 13367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9289068628083659,
+      "learning_rate": 3.900532291371404e-06,
+      "loss": 0.4897,
+      "step": 13368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.469592204769967,
+      "learning_rate": 3.899923732060019e-06,
+      "loss": 0.387,
+      "step": 13369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7638118513439487,
+      "learning_rate": 3.899315206245237e-06,
+      "loss": 0.5039,
+      "step": 13370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3990744763849827,
+      "learning_rate": 3.898706713939797e-06,
+      "loss": 0.4742,
+      "step": 13371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7081816318469858,
+      "learning_rate": 3.898098255156436e-06,
+      "loss": 0.3734,
+      "step": 13372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.479973277294879,
+      "learning_rate": 3.897489829907895e-06,
+      "loss": 0.3917,
+      "step": 13373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7415727088315773,
+      "learning_rate": 3.896881438206908e-06,
+      "loss": 0.4425,
+      "step": 13374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6512854489925637,
+      "learning_rate": 3.896273080066213e-06,
+      "loss": 0.5212,
+      "step": 13375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6287512570934892,
+      "learning_rate": 3.895664755498545e-06,
+      "loss": 0.3955,
+      "step": 13376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6831160624733132,
+      "learning_rate": 3.895056464516637e-06,
+      "loss": 0.6119,
+      "step": 13377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7224260012770671,
+      "learning_rate": 3.894448207133227e-06,
+      "loss": 0.4449,
+      "step": 13378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8693373729847784,
+      "learning_rate": 3.893839983361045e-06,
+      "loss": 0.4285,
+      "step": 13379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.631687454894412,
+      "learning_rate": 3.893231793212827e-06,
+      "loss": 0.4065,
+      "step": 13380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.550251274829423,
+      "learning_rate": 3.892623636701304e-06,
+      "loss": 0.4477,
+      "step": 13381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.534305392386661,
+      "learning_rate": 3.892015513839205e-06,
+      "loss": 0.7681,
+      "step": 13382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0637864318236483,
+      "learning_rate": 3.891407424639263e-06,
+      "loss": 0.4432,
+      "step": 13383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.218595062482098,
+      "learning_rate": 3.890799369114208e-06,
+      "loss": 0.4726,
+      "step": 13384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4808777060025888,
+      "learning_rate": 3.890191347276767e-06,
+      "loss": 0.4663,
+      "step": 13385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5708394943620074,
+      "learning_rate": 3.889583359139673e-06,
+      "loss": 0.3981,
+      "step": 13386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0082708253306363,
+      "learning_rate": 3.888975404715648e-06,
+      "loss": 0.5107,
+      "step": 13387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.802224593373649,
+      "learning_rate": 3.888367484017424e-06,
+      "loss": 0.6387,
+      "step": 13388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7505917566224878,
+      "learning_rate": 3.8877595970577254e-06,
+      "loss": 0.5632,
+      "step": 13389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7382948027351932,
+      "learning_rate": 3.887151743849278e-06,
+      "loss": 0.521,
+      "step": 13390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.090120304648665,
+      "learning_rate": 3.886543924404808e-06,
+      "loss": 0.5824,
+      "step": 13391
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.786170521637757,
+      "learning_rate": 3.885936138737037e-06,
+      "loss": 0.5434,
+      "step": 13392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.519327283189857,
+      "learning_rate": 3.8853283868586925e-06,
+      "loss": 0.39,
+      "step": 13393
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7284191188068436,
+      "learning_rate": 3.884720668782494e-06,
+      "loss": 0.5015,
+      "step": 13394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5344716711716868,
+      "learning_rate": 3.884112984521165e-06,
+      "loss": 0.4831,
+      "step": 13395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7293513049993332,
+      "learning_rate": 3.883505334087428e-06,
+      "loss": 0.4995,
+      "step": 13396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8344365622143968,
+      "learning_rate": 3.882897717494003e-06,
+      "loss": 0.5595,
+      "step": 13397
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0692377096698578,
+      "learning_rate": 3.882290134753608e-06,
+      "loss": 0.5785,
+      "step": 13398
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.447634892047262,
+      "learning_rate": 3.881682585878966e-06,
+      "loss": 0.3637,
+      "step": 13399
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2617548441251447,
+      "learning_rate": 3.8810750708827925e-06,
+      "loss": 0.5816,
+      "step": 13400
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5090008974075317,
+      "eval_runtime": 67.9539,
+      "eval_samples_per_second": 50.181,
+      "eval_steps_per_second": 1.575,
+      "step": 13400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8951915083191562,
+      "learning_rate": 3.880467589777808e-06,
+      "loss": 0.645,
+      "step": 13401
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6030904640188848,
+      "learning_rate": 3.879860142576728e-06,
+      "loss": 0.5176,
+      "step": 13402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1849310356541674,
+      "learning_rate": 3.879252729292269e-06,
+      "loss": 0.5681,
+      "step": 13403
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0793151948900435,
+      "learning_rate": 3.878645349937148e-06,
+      "loss": 0.7024,
+      "step": 13404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6251962840054428,
+      "learning_rate": 3.87803800452408e-06,
+      "loss": 0.4744,
+      "step": 13405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8434258673799315,
+      "learning_rate": 3.877430693065779e-06,
+      "loss": 0.432,
+      "step": 13406
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7765429919666786,
+      "learning_rate": 3.876823415574958e-06,
+      "loss": 0.4773,
+      "step": 13407
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8720062037901566,
+      "learning_rate": 3.87621617206433e-06,
+      "loss": 0.5631,
+      "step": 13408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.645146006350068,
+      "learning_rate": 3.87560896254661e-06,
+      "loss": 0.4043,
+      "step": 13409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6942424463457038,
+      "learning_rate": 3.8750017870345056e-06,
+      "loss": 0.5234,
+      "step": 13410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.018073554168172,
+      "learning_rate": 3.87439464554073e-06,
+      "loss": 0.5352,
+      "step": 13411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7994315854510823,
+      "learning_rate": 3.8737875380779935e-06,
+      "loss": 0.5422,
+      "step": 13412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4491150256579772,
+      "learning_rate": 3.873180464659002e-06,
+      "loss": 0.4102,
+      "step": 13413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5863081564629873,
+      "learning_rate": 3.87257342529647e-06,
+      "loss": 0.4445,
+      "step": 13414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9577337930327623,
+      "learning_rate": 3.871966420003103e-06,
+      "loss": 0.3265,
+      "step": 13415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.443111788806455,
+      "learning_rate": 3.871359448791605e-06,
+      "loss": 0.3827,
+      "step": 13416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8197221368360057,
+      "learning_rate": 3.870752511674686e-06,
+      "loss": 0.4359,
+      "step": 13417
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7669059061980141,
+      "learning_rate": 3.870145608665051e-06,
+      "loss": 0.4503,
+      "step": 13418
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5582639001706085,
+      "learning_rate": 3.869538739775406e-06,
+      "loss": 0.5007,
+      "step": 13419
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6215480187832418,
+      "learning_rate": 3.868931905018456e-06,
+      "loss": 0.3504,
+      "step": 13420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1983772798331924,
+      "learning_rate": 3.868325104406902e-06,
+      "loss": 0.4591,
+      "step": 13421
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.661009784347571,
+      "learning_rate": 3.867718337953449e-06,
+      "loss": 0.5165,
+      "step": 13422
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7259366067429398,
+      "learning_rate": 3.8671116056707995e-06,
+      "loss": 0.5682,
+      "step": 13423
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8634355269590677,
+      "learning_rate": 3.8665049075716535e-06,
+      "loss": 0.6578,
+      "step": 13424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7322598919973615,
+      "learning_rate": 3.865898243668712e-06,
+      "loss": 0.5237,
+      "step": 13425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.518204678359722,
+      "learning_rate": 3.8652916139746775e-06,
+      "loss": 0.647,
+      "step": 13426
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7463439794584565,
+      "learning_rate": 3.864685018502246e-06,
+      "loss": 0.512,
+      "step": 13427
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.989156217568056,
+      "learning_rate": 3.8640784572641185e-06,
+      "loss": 0.5307,
+      "step": 13428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.255564955851602,
+      "learning_rate": 3.863471930272993e-06,
+      "loss": 0.6373,
+      "step": 13429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1183402847281805,
+      "learning_rate": 3.8628654375415655e-06,
+      "loss": 0.5421,
+      "step": 13430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7197247940643363,
+      "learning_rate": 3.862258979082533e-06,
+      "loss": 0.4424,
+      "step": 13431
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7651248612314852,
+      "learning_rate": 3.861652554908591e-06,
+      "loss": 0.5322,
+      "step": 13432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.761621152880802,
+      "learning_rate": 3.861046165032437e-06,
+      "loss": 0.4559,
+      "step": 13433
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5103189305848261,
+      "learning_rate": 3.860439809466763e-06,
+      "loss": 0.3981,
+      "step": 13434
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.743978421690945,
+      "learning_rate": 3.859833488224262e-06,
+      "loss": 0.5378,
+      "step": 13435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9575017953507297,
+      "learning_rate": 3.85922720131763e-06,
+      "loss": 0.5367,
+      "step": 13436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.762974579869668,
+      "learning_rate": 3.858620948759557e-06,
+      "loss": 0.4865,
+      "step": 13437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4521579988498974,
+      "learning_rate": 3.858014730562734e-06,
+      "loss": 0.3468,
+      "step": 13438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.621879677839954,
+      "learning_rate": 3.857408546739853e-06,
+      "loss": 0.5318,
+      "step": 13439
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6256996837893223,
+      "learning_rate": 3.856802397303603e-06,
+      "loss": 0.429,
+      "step": 13440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8265513864009302,
+      "learning_rate": 3.856196282266675e-06,
+      "loss": 0.4208,
+      "step": 13441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5941064862934706,
+      "learning_rate": 3.855590201641758e-06,
+      "loss": 0.4438,
+      "step": 13442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4661974323395954,
+      "learning_rate": 3.854984155441537e-06,
+      "loss": 0.368,
+      "step": 13443
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9267665651760582,
+      "learning_rate": 3.854378143678702e-06,
+      "loss": 0.6447,
+      "step": 13444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7072003533036917,
+      "learning_rate": 3.853772166365938e-06,
+      "loss": 0.5279,
+      "step": 13445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.810796775440605,
+      "learning_rate": 3.853166223515931e-06,
+      "loss": 0.4961,
+      "step": 13446
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.018537954570743,
+      "learning_rate": 3.852560315141367e-06,
+      "loss": 0.4838,
+      "step": 13447
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5558807287888892,
+      "learning_rate": 3.851954441254928e-06,
+      "loss": 0.4379,
+      "step": 13448
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7736093595545053,
+      "learning_rate": 3.851348601869302e-06,
+      "loss": 0.6127,
+      "step": 13449
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0582423560906693,
+      "learning_rate": 3.850742796997165e-06,
+      "loss": 0.577,
+      "step": 13450
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5082581043243408,
+      "eval_runtime": 67.9737,
+      "eval_samples_per_second": 50.166,
+      "eval_steps_per_second": 1.574,
+      "step": 13450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6342453907671235,
+      "learning_rate": 3.850137026651206e-06,
+      "loss": 0.3947,
+      "step": 13451
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4476125647211602,
+      "learning_rate": 3.849531290844103e-06,
+      "loss": 0.4626,
+      "step": 13452
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9785143305390434,
+      "learning_rate": 3.848925589588537e-06,
+      "loss": 0.5344,
+      "step": 13453
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7339912455506865,
+      "learning_rate": 3.848319922897188e-06,
+      "loss": 0.6159,
+      "step": 13454
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.126688072798982,
+      "learning_rate": 3.847714290782737e-06,
+      "loss": 0.5702,
+      "step": 13455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3738944569840426,
+      "learning_rate": 3.84710869325786e-06,
+      "loss": 0.3536,
+      "step": 13456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.338272849304291,
+      "learning_rate": 3.846503130335235e-06,
+      "loss": 0.69,
+      "step": 13457
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5648925842553312,
+      "learning_rate": 3.845897602027541e-06,
+      "loss": 0.5327,
+      "step": 13458
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5545740214431252,
+      "learning_rate": 3.845292108347453e-06,
+      "loss": 0.4566,
+      "step": 13459
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4726564885678097,
+      "learning_rate": 3.844686649307647e-06,
+      "loss": 0.3923,
+      "step": 13460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.676488236533108,
+      "learning_rate": 3.844081224920798e-06,
+      "loss": 0.4635,
+      "step": 13461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1127757571753714,
+      "learning_rate": 3.843475835199581e-06,
+      "loss": 0.716,
+      "step": 13462
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.823973617219709,
+      "learning_rate": 3.842870480156668e-06,
+      "loss": 0.5719,
+      "step": 13463
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6425852153421758,
+      "learning_rate": 3.84226515980473e-06,
+      "loss": 0.4193,
+      "step": 13464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6453781731103825,
+      "learning_rate": 3.841659874156443e-06,
+      "loss": 0.5418,
+      "step": 13465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4792507760778861,
+      "learning_rate": 3.841054623224475e-06,
+      "loss": 0.3759,
+      "step": 13466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.765357889526238,
+      "learning_rate": 3.840449407021498e-06,
+      "loss": 0.4561,
+      "step": 13467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5438623166351053,
+      "learning_rate": 3.839844225560184e-06,
+      "loss": 0.4409,
+      "step": 13468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5656989603431752,
+      "learning_rate": 3.8392390788531995e-06,
+      "loss": 0.3958,
+      "step": 13469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.050852767137424,
+      "learning_rate": 3.838633966913212e-06,
+      "loss": 0.5139,
+      "step": 13470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6402004065130948,
+      "learning_rate": 3.8380288897528924e-06,
+      "loss": 0.4691,
+      "step": 13471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5604951777675682,
+      "learning_rate": 3.837423847384905e-06,
+      "loss": 0.3434,
+      "step": 13472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7439400962734954,
+      "learning_rate": 3.836818839821916e-06,
+      "loss": 0.4961,
+      "step": 13473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8344148362229769,
+      "learning_rate": 3.836213867076592e-06,
+      "loss": 0.5428,
+      "step": 13474
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0818097664394872,
+      "learning_rate": 3.8356089291615985e-06,
+      "loss": 0.5431,
+      "step": 13475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.073134320920565,
+      "learning_rate": 3.835004026089597e-06,
+      "loss": 0.6246,
+      "step": 13476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8692587451183813,
+      "learning_rate": 3.834399157873252e-06,
+      "loss": 0.5551,
+      "step": 13477
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5631613733326135,
+      "learning_rate": 3.833794324525226e-06,
+      "loss": 0.4366,
+      "step": 13478
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7997189821833586,
+      "learning_rate": 3.8331895260581805e-06,
+      "loss": 0.5343,
+      "step": 13479
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8143159331218002,
+      "learning_rate": 3.832584762484778e-06,
+      "loss": 0.4296,
+      "step": 13480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9055219050632086,
+      "learning_rate": 3.831980033817678e-06,
+      "loss": 0.5545,
+      "step": 13481
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7607715942156128,
+      "learning_rate": 3.83137534006954e-06,
+      "loss": 0.5498,
+      "step": 13482
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.824473519447902,
+      "learning_rate": 3.830770681253021e-06,
+      "loss": 0.5908,
+      "step": 13483
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9166473645762745,
+      "learning_rate": 3.830166057380783e-06,
+      "loss": 0.6373,
+      "step": 13484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8953411489540701,
+      "learning_rate": 3.82956146846548e-06,
+      "loss": 0.4889,
+      "step": 13485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8288875240144535,
+      "learning_rate": 3.828956914519771e-06,
+      "loss": 0.4515,
+      "step": 13486
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7905165251568638,
+      "learning_rate": 3.828352395556312e-06,
+      "loss": 0.4591,
+      "step": 13487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.179331008837777,
+      "learning_rate": 3.827747911587756e-06,
+      "loss": 0.6767,
+      "step": 13488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5140483942027303,
+      "learning_rate": 3.827143462626759e-06,
+      "loss": 0.4801,
+      "step": 13489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7174440265611304,
+      "learning_rate": 3.826539048685975e-06,
+      "loss": 0.4117,
+      "step": 13490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.674299664758448,
+      "learning_rate": 3.825934669778055e-06,
+      "loss": 0.5184,
+      "step": 13491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7110244167846622,
+      "learning_rate": 3.825330325915654e-06,
+      "loss": 0.4918,
+      "step": 13492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9727333934159619,
+      "learning_rate": 3.824726017111422e-06,
+      "loss": 0.47,
+      "step": 13493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.569363132311515,
+      "learning_rate": 3.824121743378011e-06,
+      "loss": 0.3807,
+      "step": 13494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.744405348151891,
+      "learning_rate": 3.8235175047280704e-06,
+      "loss": 0.5749,
+      "step": 13495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.045330105855075,
+      "learning_rate": 3.822913301174249e-06,
+      "loss": 0.5215,
+      "step": 13496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4902805588467374,
+      "learning_rate": 3.822309132729196e-06,
+      "loss": 0.4774,
+      "step": 13497
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6151186311428103,
+      "learning_rate": 3.821704999405559e-06,
+      "loss": 0.389,
+      "step": 13498
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7679118770259463,
+      "learning_rate": 3.821100901215987e-06,
+      "loss": 0.5163,
+      "step": 13499
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0698895850029704,
+      "learning_rate": 3.820496838173123e-06,
+      "loss": 0.516,
+      "step": 13500
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5083070397377014,
+      "eval_runtime": 67.9671,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 13500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7411809057704717,
+      "learning_rate": 3.8198928102896146e-06,
+      "loss": 0.5576,
+      "step": 13501
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7193514494894726,
+      "learning_rate": 3.819288817578108e-06,
+      "loss": 0.5527,
+      "step": 13502
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7115719457551235,
+      "learning_rate": 3.8186848600512464e-06,
+      "loss": 0.3779,
+      "step": 13503
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.249250719391091,
+      "learning_rate": 3.818080937721671e-06,
+      "loss": 0.3776,
+      "step": 13504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8318582693620722,
+      "learning_rate": 3.817477050602028e-06,
+      "loss": 0.5042,
+      "step": 13505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.177723121122112,
+      "learning_rate": 3.816873198704957e-06,
+      "loss": 0.6243,
+      "step": 13506
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.240998610720873,
+      "learning_rate": 3.816269382043101e-06,
+      "loss": 0.6208,
+      "step": 13507
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7909363962933584,
+      "learning_rate": 3.8156656006290985e-06,
+      "loss": 0.4746,
+      "step": 13508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6391954435945015,
+      "learning_rate": 3.815061854475591e-06,
+      "loss": 0.4917,
+      "step": 13509
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7327979823793578,
+      "learning_rate": 3.8144581435952176e-06,
+      "loss": 0.508,
+      "step": 13510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3994462344252496,
+      "learning_rate": 3.813854468000615e-06,
+      "loss": 0.3717,
+      "step": 13511
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.726471076372506,
+      "learning_rate": 3.8132508277044223e-06,
+      "loss": 0.486,
+      "step": 13512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.647036885969757,
+      "learning_rate": 3.8126472227192755e-06,
+      "loss": 0.4524,
+      "step": 13513
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0785429299067113,
+      "learning_rate": 3.812043653057812e-06,
+      "loss": 0.4881,
+      "step": 13514
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4101152998698898,
+      "learning_rate": 3.8114401187326654e-06,
+      "loss": 0.4334,
+      "step": 13515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7134848001645482,
+      "learning_rate": 3.8108366197564707e-06,
+      "loss": 0.4685,
+      "step": 13516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.19190286028776,
+      "learning_rate": 3.8102331561418627e-06,
+      "loss": 0.3971,
+      "step": 13517
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6939987556222076,
+      "learning_rate": 3.8096297279014737e-06,
+      "loss": 0.4748,
+      "step": 13518
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.992764646310981,
+      "learning_rate": 3.809026335047937e-06,
+      "loss": 0.5262,
+      "step": 13519
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5272347125354067,
+      "learning_rate": 3.808422977593883e-06,
+      "loss": 0.6401,
+      "step": 13520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9588553087361893,
+      "learning_rate": 3.807819655551944e-06,
+      "loss": 0.4973,
+      "step": 13521
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6351836146803294,
+      "learning_rate": 3.80721636893475e-06,
+      "loss": 0.4356,
+      "step": 13522
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3007756398612453,
+      "learning_rate": 3.806613117754928e-06,
+      "loss": 0.2893,
+      "step": 13523
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9719695463651503,
+      "learning_rate": 3.8060099020251105e-06,
+      "loss": 0.536,
+      "step": 13524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.176388689351453,
+      "learning_rate": 3.8054067217579227e-06,
+      "loss": 0.5116,
+      "step": 13525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4516792303405668,
+      "learning_rate": 3.804803576965993e-06,
+      "loss": 0.3509,
+      "step": 13526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7629098176797668,
+      "learning_rate": 3.804200467661947e-06,
+      "loss": 0.5891,
+      "step": 13527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6911156399124456,
+      "learning_rate": 3.8035973938584104e-06,
+      "loss": 0.4851,
+      "step": 13528
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5296439226294516,
+      "learning_rate": 3.8029943555680104e-06,
+      "loss": 0.4987,
+      "step": 13529
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.928822123059579,
+      "learning_rate": 3.8023913528033686e-06,
+      "loss": 0.5909,
+      "step": 13530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.42484321251918,
+      "learning_rate": 3.801788385577109e-06,
+      "loss": 0.4528,
+      "step": 13531
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4560376826665031,
+      "learning_rate": 3.8011854539018557e-06,
+      "loss": 0.3411,
+      "step": 13532
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.889423226362998,
+      "learning_rate": 3.8005825577902278e-06,
+      "loss": 0.4796,
+      "step": 13533
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8969942577559873,
+      "learning_rate": 3.799979697254851e-06,
+      "loss": 0.4889,
+      "step": 13534
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7798106348295588,
+      "learning_rate": 3.7993768723083424e-06,
+      "loss": 0.5015,
+      "step": 13535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2028972634349322,
+      "learning_rate": 3.7987740829633223e-06,
+      "loss": 0.4325,
+      "step": 13536
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9299559601563743,
+      "learning_rate": 3.7981713292324097e-06,
+      "loss": 0.6972,
+      "step": 13537
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.271169942681135,
+      "learning_rate": 3.797568611128224e-06,
+      "loss": 0.6635,
+      "step": 13538
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.710664373801328,
+      "learning_rate": 3.796965928663382e-06,
+      "loss": 0.4676,
+      "step": 13539
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0579975337826752,
+      "learning_rate": 3.7963632818505e-06,
+      "loss": 0.3629,
+      "step": 13540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8994512872285627,
+      "learning_rate": 3.7957606707021945e-06,
+      "loss": 0.6963,
+      "step": 13541
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8021143680723453,
+      "learning_rate": 3.795158095231082e-06,
+      "loss": 0.461,
+      "step": 13542
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.170974576214391,
+      "learning_rate": 3.794555555449775e-06,
+      "loss": 0.591,
+      "step": 13543
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6707679021711141,
+      "learning_rate": 3.793953051370888e-06,
+      "loss": 0.4121,
+      "step": 13544
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8521705661230803,
+      "learning_rate": 3.793350583007035e-06,
+      "loss": 0.4852,
+      "step": 13545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4427776649034796,
+      "learning_rate": 3.792748150370825e-06,
+      "loss": 0.411,
+      "step": 13546
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6754478593645892,
+      "learning_rate": 3.792145753474875e-06,
+      "loss": 0.4993,
+      "step": 13547
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.993601625894773,
+      "learning_rate": 3.7915433923317924e-06,
+      "loss": 0.5722,
+      "step": 13548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8651288454375945,
+      "learning_rate": 3.7909410669541857e-06,
+      "loss": 0.4585,
+      "step": 13549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9754327692250164,
+      "learning_rate": 3.7903387773546687e-06,
+      "loss": 0.4831,
+      "step": 13550
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5074734091758728,
+      "eval_runtime": 67.9212,
+      "eval_samples_per_second": 50.205,
+      "eval_steps_per_second": 1.575,
+      "step": 13550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.601605901253713,
+      "learning_rate": 3.789736523545846e-06,
+      "loss": 0.4696,
+      "step": 13551
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5512146124773403,
+      "learning_rate": 3.7891343055403275e-06,
+      "loss": 0.3748,
+      "step": 13552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3757643362655818,
+      "learning_rate": 3.78853212335072e-06,
+      "loss": 0.306,
+      "step": 13553
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0252999628444392,
+      "learning_rate": 3.7879299769896285e-06,
+      "loss": 0.4156,
+      "step": 13554
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8208352906898155,
+      "learning_rate": 3.78732786646966e-06,
+      "loss": 0.5296,
+      "step": 13555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.088706305856973,
+      "learning_rate": 3.7867257918034185e-06,
+      "loss": 0.7565,
+      "step": 13556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4832460281881303,
+      "learning_rate": 3.786123753003509e-06,
+      "loss": 0.479,
+      "step": 13557
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7963116138552506,
+      "learning_rate": 3.7855217500825335e-06,
+      "loss": 0.3874,
+      "step": 13558
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6804648128113617,
+      "learning_rate": 3.784919783053096e-06,
+      "loss": 0.3777,
+      "step": 13559
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.778575228952611,
+      "learning_rate": 3.7843178519277967e-06,
+      "loss": 0.5062,
+      "step": 13560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8741877977713022,
+      "learning_rate": 3.7837159567192383e-06,
+      "loss": 0.4986,
+      "step": 13561
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6562471928788465,
+      "learning_rate": 3.7831140974400205e-06,
+      "loss": 0.4726,
+      "step": 13562
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.910826954198786,
+      "learning_rate": 3.782512274102742e-06,
+      "loss": 0.4818,
+      "step": 13563
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5446811941228051,
+      "learning_rate": 3.781910486720002e-06,
+      "loss": 0.361,
+      "step": 13564
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7917270028117147,
+      "learning_rate": 3.7813087353043975e-06,
+      "loss": 0.4892,
+      "step": 13565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9469967876184695,
+      "learning_rate": 3.7807070198685304e-06,
+      "loss": 0.4993,
+      "step": 13566
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.027546040709806,
+      "learning_rate": 3.780105340424992e-06,
+      "loss": 0.4484,
+      "step": 13567
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6928641093952235,
+      "learning_rate": 3.7795036969863794e-06,
+      "loss": 0.3442,
+      "step": 13568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8546128195275393,
+      "learning_rate": 3.7789020895652905e-06,
+      "loss": 0.4273,
+      "step": 13569
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7826138612131999,
+      "learning_rate": 3.778300518174317e-06,
+      "loss": 0.4814,
+      "step": 13570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.441774983253168,
+      "learning_rate": 3.7776989828260513e-06,
+      "loss": 0.4275,
+      "step": 13571
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8239494885324137,
+      "learning_rate": 3.7770974835330888e-06,
+      "loss": 0.5842,
+      "step": 13572
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7530972489271373,
+      "learning_rate": 3.776496020308019e-06,
+      "loss": 0.5622,
+      "step": 13573
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6258479832374932,
+      "learning_rate": 3.775894593163436e-06,
+      "loss": 0.351,
+      "step": 13574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7001286519664225,
+      "learning_rate": 3.775293202111928e-06,
+      "loss": 0.4909,
+      "step": 13575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.833588942796422,
+      "learning_rate": 3.7746918471660856e-06,
+      "loss": 0.4266,
+      "step": 13576
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6551392537709133,
+      "learning_rate": 3.7740905283384976e-06,
+      "loss": 0.4668,
+      "step": 13577
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.654715578487543,
+      "learning_rate": 3.7734892456417526e-06,
+      "loss": 0.4758,
+      "step": 13578
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8291912723328054,
+      "learning_rate": 3.772887999088438e-06,
+      "loss": 0.4491,
+      "step": 13579
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7158107371270996,
+      "learning_rate": 3.7722867886911406e-06,
+      "loss": 0.4843,
+      "step": 13580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4270193057719451,
+      "learning_rate": 3.771685614462446e-06,
+      "loss": 0.3502,
+      "step": 13581
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7561619186483737,
+      "learning_rate": 3.77108447641494e-06,
+      "loss": 0.5313,
+      "step": 13582
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8542497037496006,
+      "learning_rate": 3.770483374561207e-06,
+      "loss": 0.6267,
+      "step": 13583
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0105914399534637,
+      "learning_rate": 3.769882308913828e-06,
+      "loss": 0.4442,
+      "step": 13584
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8154739477456234,
+      "learning_rate": 3.7692812794853893e-06,
+      "loss": 0.4784,
+      "step": 13585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6597645389329732,
+      "learning_rate": 3.7686802862884718e-06,
+      "loss": 0.4522,
+      "step": 13586
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0729695597774374,
+      "learning_rate": 3.7680793293356587e-06,
+      "loss": 0.6299,
+      "step": 13587
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1964211010039667,
+      "learning_rate": 3.767478408639529e-06,
+      "loss": 0.2168,
+      "step": 13588
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9090332439271358,
+      "learning_rate": 3.766877524212661e-06,
+      "loss": 0.4138,
+      "step": 13589
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5764679024692927,
+      "learning_rate": 3.766276676067636e-06,
+      "loss": 0.4796,
+      "step": 13590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.74317206167676,
+      "learning_rate": 3.765675864217032e-06,
+      "loss": 0.476,
+      "step": 13591
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.034361666194902,
+      "learning_rate": 3.7650750886734268e-06,
+      "loss": 0.533,
+      "step": 13592
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5933784919515819,
+      "learning_rate": 3.7644743494493986e-06,
+      "loss": 0.4505,
+      "step": 13593
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.914356470398811,
+      "learning_rate": 3.76387364655752e-06,
+      "loss": 0.5243,
+      "step": 13594
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7465573484506876,
+      "learning_rate": 3.7632729800103684e-06,
+      "loss": 0.4414,
+      "step": 13595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8205222691855192,
+      "learning_rate": 3.7626723498205174e-06,
+      "loss": 0.3558,
+      "step": 13596
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7758033330254603,
+      "learning_rate": 3.7620717560005438e-06,
+      "loss": 0.5138,
+      "step": 13597
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6525201259529916,
+      "learning_rate": 3.761471198563018e-06,
+      "loss": 0.3575,
+      "step": 13598
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6470081577686402,
+      "learning_rate": 3.7608706775205116e-06,
+      "loss": 0.3673,
+      "step": 13599
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7869265142847641,
+      "learning_rate": 3.7602701928855976e-06,
+      "loss": 0.4679,
+      "step": 13600
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.507072925567627,
+      "eval_runtime": 67.9288,
+      "eval_samples_per_second": 50.2,
+      "eval_steps_per_second": 1.575,
+      "step": 13600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.478892782701891,
+      "learning_rate": 3.7596697446708473e-06,
+      "loss": 0.4686,
+      "step": 13601
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6754085076236085,
+      "learning_rate": 3.7590693328888296e-06,
+      "loss": 0.5021,
+      "step": 13602
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4878139318526946,
+      "learning_rate": 3.758468957552114e-06,
+      "loss": 0.3739,
+      "step": 13603
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1928592879391173,
+      "learning_rate": 3.757868618673269e-06,
+      "loss": 0.4515,
+      "step": 13604
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0144658879133983,
+      "learning_rate": 3.757268316264863e-06,
+      "loss": 0.4638,
+      "step": 13605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.093733282570909,
+      "learning_rate": 3.7566680503394627e-06,
+      "loss": 0.4732,
+      "step": 13606
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1683255672750104,
+      "learning_rate": 3.7560678209096326e-06,
+      "loss": 0.6056,
+      "step": 13607
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8411193649005593,
+      "learning_rate": 3.7554676279879397e-06,
+      "loss": 0.5681,
+      "step": 13608
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.451797479852588,
+      "learning_rate": 3.754867471586948e-06,
+      "loss": 0.7397,
+      "step": 13609
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8879773680858876,
+      "learning_rate": 3.754267351719223e-06,
+      "loss": 0.5218,
+      "step": 13610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7807269566134774,
+      "learning_rate": 3.7536672683973262e-06,
+      "loss": 0.5919,
+      "step": 13611
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9008883354606283,
+      "learning_rate": 3.7530672216338205e-06,
+      "loss": 0.4803,
+      "step": 13612
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8330536041255654,
+      "learning_rate": 3.7524672114412676e-06,
+      "loss": 0.462,
+      "step": 13613
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.829849528747678,
+      "learning_rate": 3.7518672378322273e-06,
+      "loss": 0.5903,
+      "step": 13614
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7743647624653873,
+      "learning_rate": 3.7512673008192612e-06,
+      "loss": 0.5168,
+      "step": 13615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.735695830623404,
+      "learning_rate": 3.750667400414928e-06,
+      "loss": 0.5058,
+      "step": 13616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0057611186702724,
+      "learning_rate": 3.750067536631786e-06,
+      "loss": 0.4947,
+      "step": 13617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3705286821257174,
+      "learning_rate": 3.749467709482394e-06,
+      "loss": 0.4016,
+      "step": 13618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7836887303196542,
+      "learning_rate": 3.7488679189793074e-06,
+      "loss": 0.4978,
+      "step": 13619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.825930290446754,
+      "learning_rate": 3.7482681651350838e-06,
+      "loss": 0.5785,
+      "step": 13620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8282872278991897,
+      "learning_rate": 3.7476684479622783e-06,
+      "loss": 0.5488,
+      "step": 13621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7954700047588281,
+      "learning_rate": 3.7470687674734452e-06,
+      "loss": 0.5351,
+      "step": 13622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6056027291430754,
+      "learning_rate": 3.746469123681139e-06,
+      "loss": 0.421,
+      "step": 13623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3575147883838092,
+      "learning_rate": 3.7458695165979108e-06,
+      "loss": 0.3947,
+      "step": 13624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5081044502410836,
+      "learning_rate": 3.7452699462363163e-06,
+      "loss": 0.4966,
+      "step": 13625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.506697965024366,
+      "learning_rate": 3.744670412608905e-06,
+      "loss": 0.3821,
+      "step": 13626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9054286287716136,
+      "learning_rate": 3.7440709157282297e-06,
+      "loss": 0.5484,
+      "step": 13627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.872515071998152,
+      "learning_rate": 3.7434714556068374e-06,
+      "loss": 0.4637,
+      "step": 13628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6167167389747314,
+      "learning_rate": 3.74287203225728e-06,
+      "loss": 0.4305,
+      "step": 13629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9562762391027682,
+      "learning_rate": 3.742272645692104e-06,
+      "loss": 0.7427,
+      "step": 13630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8689090998481013,
+      "learning_rate": 3.74167329592386e-06,
+      "loss": 0.5109,
+      "step": 13631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8900098886405827,
+      "learning_rate": 3.741073982965093e-06,
+      "loss": 0.6063,
+      "step": 13632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.393222484494158,
+      "learning_rate": 3.74047470682835e-06,
+      "loss": 0.3953,
+      "step": 13633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9901279432066243,
+      "learning_rate": 3.739875467526175e-06,
+      "loss": 0.5508,
+      "step": 13634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5919281124439697,
+      "learning_rate": 3.7392762650711145e-06,
+      "loss": 0.4412,
+      "step": 13635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4633866174104055,
+      "learning_rate": 3.7386770994757114e-06,
+      "loss": 0.3942,
+      "step": 13636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6632367261116212,
+      "learning_rate": 3.7380779707525093e-06,
+      "loss": 0.3693,
+      "step": 13637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0778116121804815,
+      "learning_rate": 3.7374788789140515e-06,
+      "loss": 0.4224,
+      "step": 13638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.040912014875654,
+      "learning_rate": 3.736879823972878e-06,
+      "loss": 0.5677,
+      "step": 13639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6000659174061544,
+      "learning_rate": 3.73628080594153e-06,
+      "loss": 0.508,
+      "step": 13640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.12353443211308,
+      "learning_rate": 3.7356818248325476e-06,
+      "loss": 0.6594,
+      "step": 13641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.740673333165395,
+      "learning_rate": 3.7350828806584705e-06,
+      "loss": 0.4528,
+      "step": 13642
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7892179478589916,
+      "learning_rate": 3.734483973431837e-06,
+      "loss": 0.4062,
+      "step": 13643
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.036419390908011,
+      "learning_rate": 3.7338851031651856e-06,
+      "loss": 0.5044,
+      "step": 13644
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9069228232755038,
+      "learning_rate": 3.7332862698710526e-06,
+      "loss": 0.7169,
+      "step": 13645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5492857658853256,
+      "learning_rate": 3.7326874735619734e-06,
+      "loss": 0.5293,
+      "step": 13646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1218015015517304,
+      "learning_rate": 3.7320887142504847e-06,
+      "loss": 0.5724,
+      "step": 13647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3412959110315623,
+      "learning_rate": 3.73148999194912e-06,
+      "loss": 0.6016,
+      "step": 13648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.490295350650943,
+      "learning_rate": 3.7308913066704145e-06,
+      "loss": 0.4165,
+      "step": 13649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7534195731481625,
+      "learning_rate": 3.7302926584269007e-06,
+      "loss": 0.4164,
+      "step": 13650
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.507564127445221,
+      "eval_runtime": 67.9963,
+      "eval_samples_per_second": 50.15,
+      "eval_steps_per_second": 1.574,
+      "step": 13650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2735636618084778,
+      "learning_rate": 3.7296940472311114e-06,
+      "loss": 0.3402,
+      "step": 13651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8789958718734139,
+      "learning_rate": 3.7290954730955762e-06,
+      "loss": 0.4857,
+      "step": 13652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1592187385094825,
+      "learning_rate": 3.7284969360328283e-06,
+      "loss": 0.5113,
+      "step": 13653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6052966862315283,
+      "learning_rate": 3.727898436055397e-06,
+      "loss": 0.4907,
+      "step": 13654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6345025177207522,
+      "learning_rate": 3.7272999731758107e-06,
+      "loss": 0.4782,
+      "step": 13655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6795011975297467,
+      "learning_rate": 3.726701547406599e-06,
+      "loss": 0.3929,
+      "step": 13656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5816437057942838,
+      "learning_rate": 3.726103158760287e-06,
+      "loss": 0.3633,
+      "step": 13657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7996149118344535,
+      "learning_rate": 3.725504807249406e-06,
+      "loss": 0.6458,
+      "step": 13658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5897059287979127,
+      "learning_rate": 3.724906492886479e-06,
+      "loss": 0.5245,
+      "step": 13659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.016491258170107,
+      "learning_rate": 3.7243082156840315e-06,
+      "loss": 0.4846,
+      "step": 13660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2737125260118372,
+      "learning_rate": 3.723709975654589e-06,
+      "loss": 0.2931,
+      "step": 13661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6542445128109207,
+      "learning_rate": 3.7231117728106735e-06,
+      "loss": 0.4936,
+      "step": 13662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8090973591138666,
+      "learning_rate": 3.7225136071648104e-06,
+      "loss": 0.4879,
+      "step": 13663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5254810754438453,
+      "learning_rate": 3.721915478729521e-06,
+      "loss": 0.4788,
+      "step": 13664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7530005254067174,
+      "learning_rate": 3.721317387517327e-06,
+      "loss": 0.5664,
+      "step": 13665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7375509550949773,
+      "learning_rate": 3.7207193335407475e-06,
+      "loss": 0.4856,
+      "step": 13666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7519130966842529,
+      "learning_rate": 3.720121316812305e-06,
+      "loss": 0.5085,
+      "step": 13667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.38824399145903,
+      "learning_rate": 3.719523337344517e-06,
+      "loss": 0.7581,
+      "step": 13668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4702421490297675,
+      "learning_rate": 3.7189253951499015e-06,
+      "loss": 0.3766,
+      "step": 13669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7150411207628649,
+      "learning_rate": 3.718327490240977e-06,
+      "loss": 0.5136,
+      "step": 13670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.67820834240397,
+      "learning_rate": 3.7177296226302595e-06,
+      "loss": 0.4456,
+      "step": 13671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9997865650704814,
+      "learning_rate": 3.7171317923302654e-06,
+      "loss": 0.5812,
+      "step": 13672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9696114028942604,
+      "learning_rate": 3.716533999353511e-06,
+      "loss": 0.4893,
+      "step": 13673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4619304611382211,
+      "learning_rate": 3.715936243712507e-06,
+      "loss": 0.4051,
+      "step": 13674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.627769966922691,
+      "learning_rate": 3.7153385254197714e-06,
+      "loss": 0.4031,
+      "step": 13675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7353712630904008,
+      "learning_rate": 3.7147408444878153e-06,
+      "loss": 0.4981,
+      "step": 13676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.903408459580641,
+      "learning_rate": 3.7141432009291513e-06,
+      "loss": 0.5561,
+      "step": 13677
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7392630310616302,
+      "learning_rate": 3.713545594756289e-06,
+      "loss": 0.3842,
+      "step": 13678
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6641297549030345,
+      "learning_rate": 3.7129480259817397e-06,
+      "loss": 0.4595,
+      "step": 13679
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8979879826326571,
+      "learning_rate": 3.7123504946180143e-06,
+      "loss": 0.5167,
+      "step": 13680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7376679671302326,
+      "learning_rate": 3.7117530006776217e-06,
+      "loss": 0.4969,
+      "step": 13681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7448312037450018,
+      "learning_rate": 3.7111555441730686e-06,
+      "loss": 0.5079,
+      "step": 13682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0455039954330236,
+      "learning_rate": 3.710558125116862e-06,
+      "loss": 0.571,
+      "step": 13683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5717392068755236,
+      "learning_rate": 3.70996074352151e-06,
+      "loss": 0.4468,
+      "step": 13684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5337637819481245,
+      "learning_rate": 3.709363399399519e-06,
+      "loss": 0.388,
+      "step": 13685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5161147892893705,
+      "learning_rate": 3.7087660927633937e-06,
+      "loss": 0.4843,
+      "step": 13686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.671849699836569,
+      "learning_rate": 3.708168823625635e-06,
+      "loss": 0.4597,
+      "step": 13687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7014028750735657,
+      "learning_rate": 3.707571591998752e-06,
+      "loss": 0.5435,
+      "step": 13688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6362963980663316,
+      "learning_rate": 3.7069743978952423e-06,
+      "loss": 0.4573,
+      "step": 13689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5661719327320063,
+      "learning_rate": 3.7063772413276113e-06,
+      "loss": 0.4235,
+      "step": 13690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7743406595780176,
+      "learning_rate": 3.705780122308358e-06,
+      "loss": 0.5019,
+      "step": 13691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7780854854954498,
+      "learning_rate": 3.7051830408499823e-06,
+      "loss": 0.454,
+      "step": 13692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.866385160394268,
+      "learning_rate": 3.7045859969649866e-06,
+      "loss": 0.4564,
+      "step": 13693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6560977540511677,
+      "learning_rate": 3.7039889906658675e-06,
+      "loss": 0.5203,
+      "step": 13694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9572785024875063,
+      "learning_rate": 3.703392021965122e-06,
+      "loss": 0.5569,
+      "step": 13695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6116541869138623,
+      "learning_rate": 3.7027950908752507e-06,
+      "loss": 0.4358,
+      "step": 13696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4501925477905686,
+      "learning_rate": 3.7021981974087464e-06,
+      "loss": 0.503,
+      "step": 13697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.508003241945523,
+      "learning_rate": 3.701601341578107e-06,
+      "loss": 0.3623,
+      "step": 13698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6379311865609043,
+      "learning_rate": 3.701004523395826e-06,
+      "loss": 0.3589,
+      "step": 13699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8055181828828843,
+      "learning_rate": 3.7004077428743977e-06,
+      "loss": 0.3962,
+      "step": 13700
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.5069215893745422,
+      "eval_runtime": 67.9727,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 13700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6536272359757906,
+      "learning_rate": 3.6998110000263167e-06,
+      "loss": 0.4645,
+      "step": 13701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.313659048964841,
+      "learning_rate": 3.699214294864072e-06,
+      "loss": 0.574,
+      "step": 13702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1477223436584105,
+      "learning_rate": 3.6986176274001597e-06,
+      "loss": 0.4908,
+      "step": 13703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9496975213258836,
+      "learning_rate": 3.698020997647068e-06,
+      "loss": 0.5651,
+      "step": 13704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5112458809360512,
+      "learning_rate": 3.6974244056172857e-06,
+      "loss": 0.5212,
+      "step": 13705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8522154805762827,
+      "learning_rate": 3.696827851323307e-06,
+      "loss": 0.4269,
+      "step": 13706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6536399869108196,
+      "learning_rate": 3.696231334777616e-06,
+      "loss": 0.3438,
+      "step": 13707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5972918403615064,
+      "learning_rate": 3.6956348559927003e-06,
+      "loss": 0.401,
+      "step": 13708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0885194728474166,
+      "learning_rate": 3.6950384149810492e-06,
+      "loss": 0.5557,
+      "step": 13709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.329812173127852,
+      "learning_rate": 3.694442011755147e-06,
+      "loss": 0.428,
+      "step": 13710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2121871339210806,
+      "learning_rate": 3.6938456463274806e-06,
+      "loss": 0.5335,
+      "step": 13711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7565489762254847,
+      "learning_rate": 3.6932493187105345e-06,
+      "loss": 0.5422,
+      "step": 13712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9480799309051013,
+      "learning_rate": 3.6926530289167896e-06,
+      "loss": 0.3762,
+      "step": 13713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7936442430726625,
+      "learning_rate": 3.6920567769587313e-06,
+      "loss": 0.5609,
+      "step": 13714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5440897767742583,
+      "learning_rate": 3.6914605628488422e-06,
+      "loss": 0.5993,
+      "step": 13715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.594588020637801,
+      "learning_rate": 3.690864386599603e-06,
+      "loss": 0.4727,
+      "step": 13716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7176944512226318,
+      "learning_rate": 3.6902682482234944e-06,
+      "loss": 0.4874,
+      "step": 13717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.177649684258606,
+      "learning_rate": 3.6896721477329955e-06,
+      "loss": 0.5303,
+      "step": 13718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.073820969787836,
+      "learning_rate": 3.6890760851405862e-06,
+      "loss": 0.4564,
+      "step": 13719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.581910936468988,
+      "learning_rate": 3.6884800604587446e-06,
+      "loss": 0.4057,
+      "step": 13720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8803464771246436,
+      "learning_rate": 3.6878840736999465e-06,
+      "loss": 0.4808,
+      "step": 13721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7439793648981505,
+      "learning_rate": 3.68728812487667e-06,
+      "loss": 0.4702,
+      "step": 13722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2346631182456047,
+      "learning_rate": 3.686692214001392e-06,
+      "loss": 0.3544,
+      "step": 13723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.563163318661538,
+      "learning_rate": 3.6860963410865845e-06,
+      "loss": 0.4056,
+      "step": 13724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3323455769742851,
+      "learning_rate": 3.6855005061447245e-06,
+      "loss": 0.3624,
+      "step": 13725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7920066153774945,
+      "learning_rate": 3.684904709188285e-06,
+      "loss": 0.4039,
+      "step": 13726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4964378886613428,
+      "learning_rate": 3.6843089502297358e-06,
+      "loss": 0.425,
+      "step": 13727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8361171203376272,
+      "learning_rate": 3.6837132292815538e-06,
+      "loss": 0.4931,
+      "step": 13728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9240301625705583,
+      "learning_rate": 3.6831175463562043e-06,
+      "loss": 0.5312,
+      "step": 13729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.807004318968598,
+      "learning_rate": 3.682521901466163e-06,
+      "loss": 0.4951,
+      "step": 13730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7319824410313127,
+      "learning_rate": 3.6819262946238954e-06,
+      "loss": 0.4421,
+      "step": 13731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4122147693124143,
+      "learning_rate": 3.6813307258418717e-06,
+      "loss": 0.4276,
+      "step": 13732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7573272210216668,
+      "learning_rate": 3.68073519513256e-06,
+      "loss": 0.4384,
+      "step": 13733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4148744597911675,
+      "learning_rate": 3.6801397025084272e-06,
+      "loss": 0.3455,
+      "step": 13734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6264274095242501,
+      "learning_rate": 3.679544247981939e-06,
+      "loss": 0.4372,
+      "step": 13735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8713700825864354,
+      "learning_rate": 3.6789488315655608e-06,
+      "loss": 0.6509,
+      "step": 13736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5071583072777914,
+      "learning_rate": 3.678353453271758e-06,
+      "loss": 0.5189,
+      "step": 13737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5222372907289434,
+      "learning_rate": 3.6777581131129948e-06,
+      "loss": 0.3473,
+      "step": 13738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9721171725030715,
+      "learning_rate": 3.6771628111017333e-06,
+      "loss": 0.4016,
+      "step": 13739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9539599798795835,
+      "learning_rate": 3.6765675472504357e-06,
+      "loss": 0.4972,
+      "step": 13740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1523027433103263,
+      "learning_rate": 3.6759723215715646e-06,
+      "loss": 0.5676,
+      "step": 13741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9493363250000701,
+      "learning_rate": 3.675377134077579e-06,
+      "loss": 0.4324,
+      "step": 13742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8192898442835224,
+      "learning_rate": 3.674781984780941e-06,
+      "loss": 0.5338,
+      "step": 13743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.638446017169365,
+      "learning_rate": 3.674186873694109e-06,
+      "loss": 0.4135,
+      "step": 13744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8120861399601218,
+      "learning_rate": 3.673591800829538e-06,
+      "loss": 0.5061,
+      "step": 13745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5680589868101855,
+      "learning_rate": 3.6729967661996903e-06,
+      "loss": 0.3851,
+      "step": 13746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.648170556610067,
+      "learning_rate": 3.672401769817021e-06,
+      "loss": 0.435,
+      "step": 13747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.307748985616173,
+      "learning_rate": 3.6718068116939842e-06,
+      "loss": 0.3616,
+      "step": 13748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9970409759894232,
+      "learning_rate": 3.6712118918430364e-06,
+      "loss": 0.4467,
+      "step": 13749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8732050583026156,
+      "learning_rate": 3.6706170102766314e-06,
+      "loss": 0.482,
+      "step": 13750
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.5060539841651917,
+      "eval_runtime": 67.9462,
+      "eval_samples_per_second": 50.187,
+      "eval_steps_per_second": 1.575,
+      "step": 13750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.387196243772904,
+      "learning_rate": 3.6700221670072244e-06,
+      "loss": 0.6062,
+      "step": 13751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6096964233014617,
+      "learning_rate": 3.6694273620472665e-06,
+      "loss": 0.4389,
+      "step": 13752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7589838312177188,
+      "learning_rate": 3.6688325954092085e-06,
+      "loss": 0.5299,
+      "step": 13753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5217970244391956,
+      "learning_rate": 3.668237867105504e-06,
+      "loss": 0.3703,
+      "step": 13754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.663382681725485,
+      "learning_rate": 3.6676431771485997e-06,
+      "loss": 0.4327,
+      "step": 13755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.723281823930977,
+      "learning_rate": 3.6670485255509504e-06,
+      "loss": 0.4121,
+      "step": 13756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0359266618497687,
+      "learning_rate": 3.666453912325e-06,
+      "loss": 0.6244,
+      "step": 13757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6265085358838662,
+      "learning_rate": 3.6658593374831985e-06,
+      "loss": 0.4165,
+      "step": 13758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6520728026516365,
+      "learning_rate": 3.6652648010379932e-06,
+      "loss": 0.3779,
+      "step": 13759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.690992266793433,
+      "learning_rate": 3.6646703030018294e-06,
+      "loss": 0.5087,
+      "step": 13760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6720475527133656,
+      "learning_rate": 3.6640758433871516e-06,
+      "loss": 0.3896,
+      "step": 13761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0768444864801983,
+      "learning_rate": 3.6634814222064053e-06,
+      "loss": 0.2446,
+      "step": 13762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6728008808298582,
+      "learning_rate": 3.662887039472036e-06,
+      "loss": 0.4459,
+      "step": 13763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5309087576125902,
+      "learning_rate": 3.6622926951964845e-06,
+      "loss": 0.4212,
+      "step": 13764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.895506417131577,
+      "learning_rate": 3.6616983893921953e-06,
+      "loss": 0.4821,
+      "step": 13765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6131056416182756,
+      "learning_rate": 3.6611041220716093e-06,
+      "loss": 0.5295,
+      "step": 13766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6741376944411863,
+      "learning_rate": 3.660509893247164e-06,
+      "loss": 0.4269,
+      "step": 13767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7227373734204072,
+      "learning_rate": 3.6599157029313037e-06,
+      "loss": 0.5944,
+      "step": 13768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7692149742061924,
+      "learning_rate": 3.659321551136463e-06,
+      "loss": 0.5764,
+      "step": 13769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8798471573093065,
+      "learning_rate": 3.658727437875084e-06,
+      "loss": 0.3812,
+      "step": 13770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0213500122648465,
+      "learning_rate": 3.6581333631596015e-06,
+      "loss": 0.563,
+      "step": 13771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061619893366209,
+      "learning_rate": 3.657539327002454e-06,
+      "loss": 0.6348,
+      "step": 13772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9894460197600816,
+      "learning_rate": 3.6569453294160762e-06,
+      "loss": 0.5875,
+      "step": 13773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8231891802158127,
+      "learning_rate": 3.656351370412903e-06,
+      "loss": 0.7985,
+      "step": 13774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7831321260830895,
+      "learning_rate": 3.6557574500053683e-06,
+      "loss": 0.7264,
+      "step": 13775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6565813811747152,
+      "learning_rate": 3.6551635682059062e-06,
+      "loss": 0.4389,
+      "step": 13776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2519042411153136,
+      "learning_rate": 3.6545697250269487e-06,
+      "loss": 0.666,
+      "step": 13777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7481072576467336,
+      "learning_rate": 3.653975920480928e-06,
+      "loss": 0.5809,
+      "step": 13778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7013784780909345,
+      "learning_rate": 3.6533821545802738e-06,
+      "loss": 0.4727,
+      "step": 13779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.821963640664717,
+      "learning_rate": 3.652788427337418e-06,
+      "loss": 0.5296,
+      "step": 13780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4931675109681588,
+      "learning_rate": 3.6521947387647904e-06,
+      "loss": 0.3749,
+      "step": 13781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5258283297685242,
+      "learning_rate": 3.6516010888748157e-06,
+      "loss": 0.3899,
+      "step": 13782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8098409916230924,
+      "learning_rate": 3.6510074776799254e-06,
+      "loss": 0.4668,
+      "step": 13783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.936678269652148,
+      "learning_rate": 3.6504139051925456e-06,
+      "loss": 0.4974,
+      "step": 13784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0178122415066957,
+      "learning_rate": 3.6498203714251013e-06,
+      "loss": 0.615,
+      "step": 13785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5082693677607495,
+      "learning_rate": 3.6492268763900195e-06,
+      "loss": 0.4241,
+      "step": 13786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6043863162577352,
+      "learning_rate": 3.648633420099722e-06,
+      "loss": 0.4934,
+      "step": 13787
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.105598352523202,
+      "learning_rate": 3.6480400025666342e-06,
+      "loss": 0.5373,
+      "step": 13788
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7283994225121395,
+      "learning_rate": 3.6474466238031793e-06,
+      "loss": 0.5124,
+      "step": 13789
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7158273717868264,
+      "learning_rate": 3.646853283821777e-06,
+      "loss": 0.4575,
+      "step": 13790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7152620817228508,
+      "learning_rate": 3.646259982634852e-06,
+      "loss": 0.5155,
+      "step": 13791
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8524336101703773,
+      "learning_rate": 3.6456667202548236e-06,
+      "loss": 0.4777,
+      "step": 13792
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8524503105748646,
+      "learning_rate": 3.645073496694108e-06,
+      "loss": 0.5746,
+      "step": 13793
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1657205744055177,
+      "learning_rate": 3.6444803119651282e-06,
+      "loss": 0.5149,
+      "step": 13794
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3424573095042818,
+      "learning_rate": 3.6438871660802998e-06,
+      "loss": 0.6651,
+      "step": 13795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8638539164235226,
+      "learning_rate": 3.6432940590520418e-06,
+      "loss": 0.556,
+      "step": 13796
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7850366053004454,
+      "learning_rate": 3.642700990892769e-06,
+      "loss": 0.4284,
+      "step": 13797
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3730292486399287,
+      "learning_rate": 3.642107961614897e-06,
+      "loss": 0.3022,
+      "step": 13798
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5260634963168829,
+      "learning_rate": 3.6415149712308406e-06,
+      "loss": 0.5097,
+      "step": 13799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.142332937907371,
+      "learning_rate": 3.640922019753015e-06,
+      "loss": 0.6159,
+      "step": 13800
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.5055629014968872,
+      "eval_runtime": 68.0012,
+      "eval_samples_per_second": 50.146,
+      "eval_steps_per_second": 1.574,
+      "step": 13800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0915102451638665,
+      "learning_rate": 3.6403291071938307e-06,
+      "loss": 0.3951,
+      "step": 13801
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6606514441901619,
+      "learning_rate": 3.639736233565703e-06,
+      "loss": 0.4368,
+      "step": 13802
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4304310040025048,
+      "learning_rate": 3.639143398881041e-06,
+      "loss": 0.3974,
+      "step": 13803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.856041418544076,
+      "learning_rate": 3.6385506031522566e-06,
+      "loss": 0.5024,
+      "step": 13804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8939953553901923,
+      "learning_rate": 3.6379578463917586e-06,
+      "loss": 0.5385,
+      "step": 13805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9232351583591505,
+      "learning_rate": 3.637365128611956e-06,
+      "loss": 0.5957,
+      "step": 13806
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5876705465989824,
+      "learning_rate": 3.6367724498252587e-06,
+      "loss": 0.4325,
+      "step": 13807
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8598034486669157,
+      "learning_rate": 3.6361798100440714e-06,
+      "loss": 0.4356,
+      "step": 13808
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6566583164156854,
+      "learning_rate": 3.635587209280804e-06,
+      "loss": 0.5047,
+      "step": 13809
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8875206784232623,
+      "learning_rate": 3.6349946475478593e-06,
+      "loss": 0.6883,
+      "step": 13810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8827868562932482,
+      "learning_rate": 3.634402124857641e-06,
+      "loss": 0.4917,
+      "step": 13811
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6996488927586988,
+      "learning_rate": 3.6338096412225575e-06,
+      "loss": 0.5804,
+      "step": 13812
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.674668494571535,
+      "learning_rate": 3.6332171966550113e-06,
+      "loss": 0.4753,
+      "step": 13813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9382860189225808,
+      "learning_rate": 3.632624791167403e-06,
+      "loss": 0.4311,
+      "step": 13814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.228886042945323,
+      "learning_rate": 3.632032424772134e-06,
+      "loss": 0.6962,
+      "step": 13815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3992767690290555,
+      "learning_rate": 3.6314400974816066e-06,
+      "loss": 0.2661,
+      "step": 13816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6972997560542027,
+      "learning_rate": 3.63084780930822e-06,
+      "loss": 0.4891,
+      "step": 13817
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.82830444509972,
+      "learning_rate": 3.630255560264373e-06,
+      "loss": 0.6212,
+      "step": 13818
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7937365608321647,
+      "learning_rate": 3.6296633503624657e-06,
+      "loss": 0.5419,
+      "step": 13819
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.856069422144031,
+      "learning_rate": 3.6290711796148924e-06,
+      "loss": 0.5581,
+      "step": 13820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.959539698328812,
+      "learning_rate": 3.6284790480340536e-06,
+      "loss": 0.4969,
+      "step": 13821
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7547349990700774,
+      "learning_rate": 3.6278869556323425e-06,
+      "loss": 0.5915,
+      "step": 13822
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.94553201738779,
+      "learning_rate": 3.627294902422157e-06,
+      "loss": 0.4841,
+      "step": 13823
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2024813823511327,
+      "learning_rate": 3.6267028884158883e-06,
+      "loss": 0.4902,
+      "step": 13824
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.822697611936721,
+      "learning_rate": 3.62611091362593e-06,
+      "loss": 0.4064,
+      "step": 13825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7431112848653831,
+      "learning_rate": 3.625518978064677e-06,
+      "loss": 0.4575,
+      "step": 13826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.122461242258822,
+      "learning_rate": 3.6249270817445213e-06,
+      "loss": 0.6474,
+      "step": 13827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7344673755576374,
+      "learning_rate": 3.6243352246778496e-06,
+      "loss": 0.4521,
+      "step": 13828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.085049694565667,
+      "learning_rate": 3.623743406877056e-06,
+      "loss": 0.6097,
+      "step": 13829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8847895217335826,
+      "learning_rate": 3.623151628354531e-06,
+      "loss": 0.5443,
+      "step": 13830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5416026256670496,
+      "learning_rate": 3.6225598891226588e-06,
+      "loss": 0.3615,
+      "step": 13831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6024967616056225,
+      "learning_rate": 3.6219681891938286e-06,
+      "loss": 0.6407,
+      "step": 13832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8862428332655796,
+      "learning_rate": 3.621376528580429e-06,
+      "loss": 0.5015,
+      "step": 13833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5750626970975772,
+      "learning_rate": 3.6207849072948455e-06,
+      "loss": 0.3762,
+      "step": 13834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4463691278712447,
+      "learning_rate": 3.6201933253494603e-06,
+      "loss": 0.3674,
+      "step": 13835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7867942474989549,
+      "learning_rate": 3.6196017827566634e-06,
+      "loss": 0.4111,
+      "step": 13836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0144338275863625,
+      "learning_rate": 3.619010279528833e-06,
+      "loss": 0.4752,
+      "step": 13837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.816426749957387,
+      "learning_rate": 3.618418815678354e-06,
+      "loss": 0.5856,
+      "step": 13838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.763984938313847,
+      "learning_rate": 3.6178273912176085e-06,
+      "loss": 0.4341,
+      "step": 13839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8011997465604566,
+      "learning_rate": 3.6172360061589794e-06,
+      "loss": 0.5176,
+      "step": 13840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6877559444386983,
+      "learning_rate": 3.616644660514843e-06,
+      "loss": 0.4507,
+      "step": 13841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3340086823029416,
+      "learning_rate": 3.6160533542975817e-06,
+      "loss": 0.2993,
+      "step": 13842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6299710777496443,
+      "learning_rate": 3.6154620875195744e-06,
+      "loss": 0.48,
+      "step": 13843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.658808776239314,
+      "learning_rate": 3.614870860193196e-06,
+      "loss": 0.3431,
+      "step": 13844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7518713121061955,
+      "learning_rate": 3.614279672330825e-06,
+      "loss": 0.5771,
+      "step": 13845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6601347187367803,
+      "learning_rate": 3.6136885239448405e-06,
+      "loss": 0.4749,
+      "step": 13846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9793413437899054,
+      "learning_rate": 3.613097415047613e-06,
+      "loss": 0.4782,
+      "step": 13847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5728514394187416,
+      "learning_rate": 3.6125063456515184e-06,
+      "loss": 0.4424,
+      "step": 13848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7501321152709703,
+      "learning_rate": 3.6119153157689334e-06,
+      "loss": 0.5514,
+      "step": 13849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6843023469983256,
+      "learning_rate": 3.6113243254122265e-06,
+      "loss": 0.5814,
+      "step": 13850
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.505012571811676,
+      "eval_runtime": 67.9584,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.574,
+      "step": 13850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.409703220497235,
+      "learning_rate": 3.6107333745937724e-06,
+      "loss": 0.4986,
+      "step": 13851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2403668301729014,
+      "learning_rate": 3.6101424633259408e-06,
+      "loss": 0.6477,
+      "step": 13852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7574998105700206,
+      "learning_rate": 3.6095515916211042e-06,
+      "loss": 0.5199,
+      "step": 13853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3938378223407555,
+      "learning_rate": 3.6089607594916308e-06,
+      "loss": 0.4887,
+      "step": 13854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.157094171109226,
+      "learning_rate": 3.6083699669498883e-06,
+      "loss": 0.5524,
+      "step": 13855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8091458041474255,
+      "learning_rate": 3.607779214008247e-06,
+      "loss": 0.5962,
+      "step": 13856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8173710483129124,
+      "learning_rate": 3.6071885006790707e-06,
+      "loss": 0.6884,
+      "step": 13857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0720192241586015,
+      "learning_rate": 3.6065978269747294e-06,
+      "loss": 0.596,
+      "step": 13858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9035580545582436,
+      "learning_rate": 3.6060071929075858e-06,
+      "loss": 0.5335,
+      "step": 13859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8704301141274393,
+      "learning_rate": 3.6054165984900036e-06,
+      "loss": 0.6694,
+      "step": 13860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.564402724744622,
+      "learning_rate": 3.6048260437343487e-06,
+      "loss": 0.4893,
+      "step": 13861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.659548685157608,
+      "learning_rate": 3.604235528652985e-06,
+      "loss": 0.537,
+      "step": 13862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2909555570328712,
+      "learning_rate": 3.603645053258272e-06,
+      "loss": 0.6783,
+      "step": 13863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7648373553060097,
+      "learning_rate": 3.6030546175625725e-06,
+      "loss": 0.4774,
+      "step": 13864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.964090286458108,
+      "learning_rate": 3.602464221578247e-06,
+      "loss": 0.5428,
+      "step": 13865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5394614095910935,
+      "learning_rate": 3.601873865317653e-06,
+      "loss": 0.3598,
+      "step": 13866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.664510337752565,
+      "learning_rate": 3.601283548793151e-06,
+      "loss": 0.4807,
+      "step": 13867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8156547032660688,
+      "learning_rate": 3.6006932720170993e-06,
+      "loss": 0.5685,
+      "step": 13868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1761205692000685,
+      "learning_rate": 3.600103035001854e-06,
+      "loss": 0.6916,
+      "step": 13869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.482276671370745,
+      "learning_rate": 3.599512837759771e-06,
+      "loss": 0.2899,
+      "step": 13870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6082421349842244,
+      "learning_rate": 3.5989226803032053e-06,
+      "loss": 0.6398,
+      "step": 13871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.428667927939651,
+      "learning_rate": 3.5983325626445155e-06,
+      "loss": 0.4605,
+      "step": 13872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6889190767242448,
+      "learning_rate": 3.5977424847960507e-06,
+      "loss": 0.4342,
+      "step": 13873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0762552118998627,
+      "learning_rate": 3.5971524467701644e-06,
+      "loss": 0.5145,
+      "step": 13874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1440590325257936,
+      "learning_rate": 3.5965624485792115e-06,
+      "loss": 0.5162,
+      "step": 13875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9135680452317372,
+      "learning_rate": 3.5959724902355407e-06,
+      "loss": 0.4391,
+      "step": 13876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.561254599306087,
+      "learning_rate": 3.5953825717515024e-06,
+      "loss": 0.4692,
+      "step": 13877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.53719960522883,
+      "learning_rate": 3.5947926931394495e-06,
+      "loss": 0.5109,
+      "step": 13878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6955279363096218,
+      "learning_rate": 3.5942028544117256e-06,
+      "loss": 0.4883,
+      "step": 13879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7912751186409475,
+      "learning_rate": 3.5936130555806816e-06,
+      "loss": 0.4724,
+      "step": 13880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5988021499686391,
+      "learning_rate": 3.5930232966586663e-06,
+      "loss": 0.621,
+      "step": 13881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7388711370019057,
+      "learning_rate": 3.592433577658021e-06,
+      "loss": 0.3559,
+      "step": 13882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.103136349894707,
+      "learning_rate": 3.591843898591094e-06,
+      "loss": 0.634,
+      "step": 13883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6699042892654925,
+      "learning_rate": 3.5912542594702316e-06,
+      "loss": 0.581,
+      "step": 13884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4200871734620786,
+      "learning_rate": 3.5906646603077757e-06,
+      "loss": 0.3486,
+      "step": 13885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4736863615128706,
+      "learning_rate": 3.5900751011160684e-06,
+      "loss": 0.4885,
+      "step": 13886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6875135600652187,
+      "learning_rate": 3.5894855819074516e-06,
+      "loss": 0.4847,
+      "step": 13887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5178963906560268,
+      "learning_rate": 3.5888961026942702e-06,
+      "loss": 0.4425,
+      "step": 13888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.123602685908976,
+      "learning_rate": 3.5883066634888597e-06,
+      "loss": 0.6944,
+      "step": 13889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8397308558397767,
+      "learning_rate": 3.5877172643035617e-06,
+      "loss": 0.5809,
+      "step": 13890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.479759464898651,
+      "learning_rate": 3.587127905150717e-06,
+      "loss": 0.4215,
+      "step": 13891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7003691621542028,
+      "learning_rate": 3.5865385860426594e-06,
+      "loss": 0.4215,
+      "step": 13892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4061539284661613,
+      "learning_rate": 3.5859493069917284e-06,
+      "loss": 0.3727,
+      "step": 13893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7274654281941633,
+      "learning_rate": 3.5853600680102605e-06,
+      "loss": 0.4735,
+      "step": 13894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4720723981221546,
+      "learning_rate": 3.58477086911059e-06,
+      "loss": 0.4122,
+      "step": 13895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.443321687316621,
+      "learning_rate": 3.584181710305051e-06,
+      "loss": 0.3769,
+      "step": 13896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8874650126626327,
+      "learning_rate": 3.5835925916059777e-06,
+      "loss": 0.6037,
+      "step": 13897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8687212312315915,
+      "learning_rate": 3.5830035130257055e-06,
+      "loss": 0.7514,
+      "step": 13898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5329691648506192,
+      "learning_rate": 3.582414474576562e-06,
+      "loss": 0.612,
+      "step": 13899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7464524716536143,
+      "learning_rate": 3.5818254762708805e-06,
+      "loss": 0.4586,
+      "step": 13900
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.5057464241981506,
+      "eval_runtime": 67.941,
+      "eval_samples_per_second": 50.191,
+      "eval_steps_per_second": 1.575,
+      "step": 13900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7063392197621436,
+      "learning_rate": 3.581236518120993e-06,
+      "loss": 0.4327,
+      "step": 13901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2629530564049163,
+      "learning_rate": 3.5806476001392263e-06,
+      "loss": 0.6522,
+      "step": 13902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8581069060548498,
+      "learning_rate": 3.580058722337909e-06,
+      "loss": 0.5977,
+      "step": 13903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7689489404476246,
+      "learning_rate": 3.579469884729372e-06,
+      "loss": 0.6706,
+      "step": 13904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7255209705281984,
+      "learning_rate": 3.5788810873259373e-06,
+      "loss": 0.4585,
+      "step": 13905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5944654420963214,
+      "learning_rate": 3.5782923301399354e-06,
+      "loss": 0.5735,
+      "step": 13906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5617692930102858,
+      "learning_rate": 3.5777036131836905e-06,
+      "loss": 0.489,
+      "step": 13907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.126428437513977,
+      "learning_rate": 3.5771149364695248e-06,
+      "loss": 0.5253,
+      "step": 13908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.632263887893169,
+      "learning_rate": 3.576526300009764e-06,
+      "loss": 0.3467,
+      "step": 13909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8650886673083935,
+      "learning_rate": 3.57593770381673e-06,
+      "loss": 0.5148,
+      "step": 13910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4480126126441704,
+      "learning_rate": 3.575349147902745e-06,
+      "loss": 0.376,
+      "step": 13911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9732022391997799,
+      "learning_rate": 3.5747606322801307e-06,
+      "loss": 0.5864,
+      "step": 13912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.940894008483344,
+      "learning_rate": 3.5741721569612086e-06,
+      "loss": 0.5643,
+      "step": 13913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.936730150045112,
+      "learning_rate": 3.573583721958294e-06,
+      "loss": 0.6489,
+      "step": 13914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3918215153352234,
+      "learning_rate": 3.572995327283707e-06,
+      "loss": 0.3725,
+      "step": 13915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6649958308877353,
+      "learning_rate": 3.572406972949769e-06,
+      "loss": 0.4435,
+      "step": 13916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.528929961900855,
+      "learning_rate": 3.571818658968791e-06,
+      "loss": 0.4767,
+      "step": 13917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.792010362101866,
+      "learning_rate": 3.571230385353092e-06,
+      "loss": 0.4808,
+      "step": 13918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8504712598645086,
+      "learning_rate": 3.570642152114987e-06,
+      "loss": 0.5063,
+      "step": 13919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5366401714535656,
+      "learning_rate": 3.570053959266793e-06,
+      "loss": 0.5263,
+      "step": 13920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0443723344300184,
+      "learning_rate": 3.569465806820818e-06,
+      "loss": 0.6432,
+      "step": 13921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2912422004376432,
+      "learning_rate": 3.5688776947893767e-06,
+      "loss": 0.4213,
+      "step": 13922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0593802950565117,
+      "learning_rate": 3.568289623184783e-06,
+      "loss": 0.5213,
+      "step": 13923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3596876518842862,
+      "learning_rate": 3.5677015920193445e-06,
+      "loss": 0.4373,
+      "step": 13924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8501551318365936,
+      "learning_rate": 3.5671136013053742e-06,
+      "loss": 0.4874,
+      "step": 13925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8578995054969907,
+      "learning_rate": 3.5665256510551804e-06,
+      "loss": 0.5368,
+      "step": 13926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7111291194782616,
+      "learning_rate": 3.56593774128107e-06,
+      "loss": 0.4192,
+      "step": 13927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5261013781644146,
+      "learning_rate": 3.5653498719953515e-06,
+      "loss": 0.4949,
+      "step": 13928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2150360220982916,
+      "learning_rate": 3.5647620432103346e-06,
+      "loss": 0.6266,
+      "step": 13929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9074025234676226,
+      "learning_rate": 3.5641742549383195e-06,
+      "loss": 0.4697,
+      "step": 13930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8820603235149616,
+      "learning_rate": 3.5635865071916153e-06,
+      "loss": 0.5705,
+      "step": 13931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5147980084607873,
+      "learning_rate": 3.5629987999825235e-06,
+      "loss": 0.3319,
+      "step": 13932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5368889216237767,
+      "learning_rate": 3.56241113332335e-06,
+      "loss": 0.4594,
+      "step": 13933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6452271812543235,
+      "learning_rate": 3.561823507226395e-06,
+      "loss": 0.4084,
+      "step": 13934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6832207429685122,
+      "learning_rate": 3.5612359217039615e-06,
+      "loss": 0.4306,
+      "step": 13935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7183294781460332,
+      "learning_rate": 3.5606483767683508e-06,
+      "loss": 0.3795,
+      "step": 13936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8492429480277799,
+      "learning_rate": 3.5600608724318597e-06,
+      "loss": 0.5099,
+      "step": 13937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.356771995212399,
+      "learning_rate": 3.5594734087067906e-06,
+      "loss": 0.409,
+      "step": 13938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.707199712234546,
+      "learning_rate": 3.5588859856054416e-06,
+      "loss": 0.4919,
+      "step": 13939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6383782528561617,
+      "learning_rate": 3.5582986031401056e-06,
+      "loss": 0.4643,
+      "step": 13940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8742121218710655,
+      "learning_rate": 3.5577112613230837e-06,
+      "loss": 0.4567,
+      "step": 13941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7659581114229181,
+      "learning_rate": 3.557123960166672e-06,
+      "loss": 0.4171,
+      "step": 13942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9984618322920973,
+      "learning_rate": 3.5565366996831617e-06,
+      "loss": 0.5794,
+      "step": 13943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.002763034007282,
+      "learning_rate": 3.555949479884848e-06,
+      "loss": 0.628,
+      "step": 13944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6784067394024857,
+      "learning_rate": 3.555362300784024e-06,
+      "loss": 0.4243,
+      "step": 13945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7829655408147222,
+      "learning_rate": 3.554775162392985e-06,
+      "loss": 0.6091,
+      "step": 13946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4204169697442275,
+      "learning_rate": 3.554188064724018e-06,
+      "loss": 0.5482,
+      "step": 13947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4764225861436844,
+      "learning_rate": 3.5536010077894145e-06,
+      "loss": 0.3905,
+      "step": 13948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5328692010068283,
+      "learning_rate": 3.5530139916014672e-06,
+      "loss": 0.4414,
+      "step": 13949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4344036736150356,
+      "learning_rate": 3.5524270161724617e-06,
+      "loss": 0.3026,
+      "step": 13950
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.505784809589386,
+      "eval_runtime": 68.0018,
+      "eval_samples_per_second": 50.146,
+      "eval_steps_per_second": 1.573,
+      "step": 13950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5868237897844495,
+      "learning_rate": 3.5518400815146856e-06,
+      "loss": 0.4112,
+      "step": 13951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.596261257267076,
+      "learning_rate": 3.55125318764043e-06,
+      "loss": 0.4331,
+      "step": 13952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7700408936783467,
+      "learning_rate": 3.550666334561976e-06,
+      "loss": 0.5229,
+      "step": 13953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7436665986391404,
+      "learning_rate": 3.550079522291612e-06,
+      "loss": 0.4829,
+      "step": 13954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.014610158503909,
+      "learning_rate": 3.5494927508416245e-06,
+      "loss": 0.6794,
+      "step": 13955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6858964716141858,
+      "learning_rate": 3.548906020224292e-06,
+      "loss": 0.4153,
+      "step": 13956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7085117666665046,
+      "learning_rate": 3.5483193304519006e-06,
+      "loss": 0.5823,
+      "step": 13957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7612145224330091,
+      "learning_rate": 3.5477326815367307e-06,
+      "loss": 0.3422,
+      "step": 13958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6444186852214016,
+      "learning_rate": 3.5471460734910657e-06,
+      "loss": 0.4372,
+      "step": 13959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8456446261328856,
+      "learning_rate": 3.546559506327184e-06,
+      "loss": 0.4889,
+      "step": 13960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3502393120754257,
+      "learning_rate": 3.5459729800573685e-06,
+      "loss": 0.5885,
+      "step": 13961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.713514358269347,
+      "learning_rate": 3.5453864946938922e-06,
+      "loss": 0.5836,
+      "step": 13962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.404831222645814,
+      "learning_rate": 3.544800050249035e-06,
+      "loss": 0.3796,
+      "step": 13963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.063161292106625,
+      "learning_rate": 3.544213646735075e-06,
+      "loss": 0.5698,
+      "step": 13964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5450405317472717,
+      "learning_rate": 3.5436272841642906e-06,
+      "loss": 0.3194,
+      "step": 13965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9105951109914945,
+      "learning_rate": 3.5430409625489506e-06,
+      "loss": 0.5175,
+      "step": 13966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.758294333128104,
+      "learning_rate": 3.5424546819013326e-06,
+      "loss": 0.5539,
+      "step": 13967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7588127743426185,
+      "learning_rate": 3.5418684422337125e-06,
+      "loss": 0.5521,
+      "step": 13968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8619554191380214,
+      "learning_rate": 3.541282243558358e-06,
+      "loss": 0.5687,
+      "step": 13969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8929391882819873,
+      "learning_rate": 3.5406960858875426e-06,
+      "loss": 0.4907,
+      "step": 13970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125184799735611,
+      "learning_rate": 3.5401099692335404e-06,
+      "loss": 0.533,
+      "step": 13971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7267782876690652,
+      "learning_rate": 3.5395238936086164e-06,
+      "loss": 0.4878,
+      "step": 13972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7174051029214208,
+      "learning_rate": 3.5389378590250424e-06,
+      "loss": 0.4299,
+      "step": 13973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.371071468398453,
+      "learning_rate": 3.538351865495088e-06,
+      "loss": 0.624,
+      "step": 13974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6157475553506877,
+      "learning_rate": 3.537765913031016e-06,
+      "loss": 0.4358,
+      "step": 13975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.829870823266025,
+      "learning_rate": 3.5371800016450973e-06,
+      "loss": 0.4978,
+      "step": 13976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2860702378648434,
+      "learning_rate": 3.5365941313495956e-06,
+      "loss": 0.4708,
+      "step": 13977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9240018588304213,
+      "learning_rate": 3.536008302156778e-06,
+      "loss": 0.5642,
+      "step": 13978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.797186006285749,
+      "learning_rate": 3.5354225140789043e-06,
+      "loss": 0.3963,
+      "step": 13979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8618992946279505,
+      "learning_rate": 3.53483676712824e-06,
+      "loss": 0.604,
+      "step": 13980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8845697798884988,
+      "learning_rate": 3.53425106131705e-06,
+      "loss": 0.5069,
+      "step": 13981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3199613479859227,
+      "learning_rate": 3.533665396657592e-06,
+      "loss": 0.422,
+      "step": 13982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1272762691337146,
+      "learning_rate": 3.533079773162126e-06,
+      "loss": 0.5567,
+      "step": 13983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5388355332336165,
+      "learning_rate": 3.5324941908429155e-06,
+      "loss": 0.3651,
+      "step": 13984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7954614798179735,
+      "learning_rate": 3.5319086497122157e-06,
+      "loss": 0.6532,
+      "step": 13985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.151753420759947,
+      "learning_rate": 3.5313231497822866e-06,
+      "loss": 0.5693,
+      "step": 13986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7276861322922723,
+      "learning_rate": 3.5307376910653845e-06,
+      "loss": 0.4313,
+      "step": 13987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.998529416882668,
+      "learning_rate": 3.530152273573765e-06,
+      "loss": 0.5245,
+      "step": 13988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0564196554214957,
+      "learning_rate": 3.529566897319684e-06,
+      "loss": 0.5931,
+      "step": 13989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8372710279316413,
+      "learning_rate": 3.5289815623153955e-06,
+      "loss": 0.4797,
+      "step": 13990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4664948952382613,
+      "learning_rate": 3.5283962685731567e-06,
+      "loss": 0.5806,
+      "step": 13991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265635906120426,
+      "learning_rate": 3.527811016105214e-06,
+      "loss": 0.7451,
+      "step": 13992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8830731148304956,
+      "learning_rate": 3.527225804923824e-06,
+      "loss": 0.5191,
+      "step": 13993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.696491155182148,
+      "learning_rate": 3.5266406350412373e-06,
+      "loss": 0.4177,
+      "step": 13994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9237925563386125,
+      "learning_rate": 3.5260555064697024e-06,
+      "loss": 0.5135,
+      "step": 13995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9322159639046261,
+      "learning_rate": 3.5254704192214694e-06,
+      "loss": 0.6486,
+      "step": 13996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6693117390669885,
+      "learning_rate": 3.5248853733087872e-06,
+      "loss": 0.486,
+      "step": 13997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9771022038978434,
+      "learning_rate": 3.5243003687439024e-06,
+      "loss": 0.5295,
+      "step": 13998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6792088654484125,
+      "learning_rate": 3.523715405539062e-06,
+      "loss": 0.3937,
+      "step": 13999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.395555383184759,
+      "learning_rate": 3.523130483706514e-06,
+      "loss": 0.3466,
+      "step": 14000
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.5054076910018921,
+      "eval_runtime": 67.9515,
+      "eval_samples_per_second": 50.183,
+      "eval_steps_per_second": 1.575,
+      "step": 14000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.85838133063084,
+      "learning_rate": 3.5225456032584993e-06,
+      "loss": 0.4624,
+      "step": 14001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6988335905769043,
+      "learning_rate": 3.5219607642072638e-06,
+      "loss": 0.4749,
+      "step": 14002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7171685897692346,
+      "learning_rate": 3.521375966565053e-06,
+      "loss": 0.3509,
+      "step": 14003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.098021952479637,
+      "learning_rate": 3.5207912103441066e-06,
+      "loss": 0.5626,
+      "step": 14004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8087062200564163,
+      "learning_rate": 3.520206495556666e-06,
+      "loss": 0.4816,
+      "step": 14005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.669741389638038,
+      "learning_rate": 3.5196218222149735e-06,
+      "loss": 0.4752,
+      "step": 14006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3416277962599545,
+      "learning_rate": 3.519037190331269e-06,
+      "loss": 0.3479,
+      "step": 14007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8298523862972431,
+      "learning_rate": 3.5184525999177877e-06,
+      "loss": 0.4848,
+      "step": 14008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.049964054547157,
+      "learning_rate": 3.5178680509867736e-06,
+      "loss": 0.6151,
+      "step": 14009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3136704601874667,
+      "learning_rate": 3.5172835435504595e-06,
+      "loss": 0.3575,
+      "step": 14010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7087058712735144,
+      "learning_rate": 3.5166990776210806e-06,
+      "loss": 0.5101,
+      "step": 14011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.562997935052242,
+      "learning_rate": 3.516114653210877e-06,
+      "loss": 0.4644,
+      "step": 14012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.841168047984965,
+      "learning_rate": 3.5155302703320826e-06,
+      "loss": 0.5486,
+      "step": 14013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4202849789139798,
+      "learning_rate": 3.5149459289969266e-06,
+      "loss": 0.3608,
+      "step": 14014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6583328541885365,
+      "learning_rate": 3.5143616292176453e-06,
+      "loss": 0.4122,
+      "step": 14015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7078004525624202,
+      "learning_rate": 3.5137773710064717e-06,
+      "loss": 0.5017,
+      "step": 14016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.960655359592902,
+      "learning_rate": 3.5131931543756325e-06,
+      "loss": 0.4898,
+      "step": 14017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5484939995451814,
+      "learning_rate": 3.5126089793373606e-06,
+      "loss": 0.4934,
+      "step": 14018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.765814051693623,
+      "learning_rate": 3.5120248459038886e-06,
+      "loss": 0.5147,
+      "step": 14019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.894629968673126,
+      "learning_rate": 3.5114407540874394e-06,
+      "loss": 0.5934,
+      "step": 14020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.454538507297613,
+      "learning_rate": 3.5108567039002417e-06,
+      "loss": 0.4194,
+      "step": 14021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6153546140629866,
+      "learning_rate": 3.5102726953545263e-06,
+      "loss": 0.399,
+      "step": 14022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7625960371559983,
+      "learning_rate": 3.5096887284625137e-06,
+      "loss": 0.4258,
+      "step": 14023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9786634298280843,
+      "learning_rate": 3.5091048032364316e-06,
+      "loss": 0.6685,
+      "step": 14024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4800369805038258,
+      "learning_rate": 3.508520919688504e-06,
+      "loss": 0.3973,
+      "step": 14025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5391595678059276,
+      "learning_rate": 3.5079370778309563e-06,
+      "loss": 0.3751,
+      "step": 14026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8636840790769036,
+      "learning_rate": 3.5073532776760055e-06,
+      "loss": 0.4694,
+      "step": 14027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.44281679281431,
+      "learning_rate": 3.5067695192358763e-06,
+      "loss": 0.3728,
+      "step": 14028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7527655595318496,
+      "learning_rate": 3.5061858025227925e-06,
+      "loss": 0.5694,
+      "step": 14029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.700203365077964,
+      "learning_rate": 3.5056021275489673e-06,
+      "loss": 0.4641,
+      "step": 14030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6572665533776385,
+      "learning_rate": 3.505018494326623e-06,
+      "loss": 0.4484,
+      "step": 14031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7235932435385435,
+      "learning_rate": 3.5044349028679793e-06,
+      "loss": 0.5226,
+      "step": 14032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6568084273813117,
+      "learning_rate": 3.503851353185249e-06,
+      "loss": 0.5518,
+      "step": 14033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5313498029992205,
+      "learning_rate": 3.5032678452906523e-06,
+      "loss": 0.4525,
+      "step": 14034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7207977858001926,
+      "learning_rate": 3.5026843791964038e-06,
+      "loss": 0.4678,
+      "step": 14035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6919675630599702,
+      "learning_rate": 3.5021009549147157e-06,
+      "loss": 0.5047,
+      "step": 14036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1394921815250787,
+      "learning_rate": 3.501517572457803e-06,
+      "loss": 0.6234,
+      "step": 14037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9304203755789948,
+      "learning_rate": 3.500934231837879e-06,
+      "loss": 0.5569,
+      "step": 14038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5931077582094948,
+      "learning_rate": 3.500350933067157e-06,
+      "loss": 0.5157,
+      "step": 14039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1644553899465997,
+      "learning_rate": 3.4997676761578437e-06,
+      "loss": 0.7253,
+      "step": 14040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2249623937394793,
+      "learning_rate": 3.4991844611221525e-06,
+      "loss": 0.5953,
+      "step": 14041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8402058712669749,
+      "learning_rate": 3.498601287972293e-06,
+      "loss": 0.454,
+      "step": 14042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.518678905728106,
+      "learning_rate": 3.4980181567204706e-06,
+      "loss": 0.444,
+      "step": 14043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7894810088534379,
+      "learning_rate": 3.4974350673788948e-06,
+      "loss": 0.5071,
+      "step": 14044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7117574388978734,
+      "learning_rate": 3.4968520199597743e-06,
+      "loss": 0.4339,
+      "step": 14045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6820392745183657,
+      "learning_rate": 3.4962690144753103e-06,
+      "loss": 0.501,
+      "step": 14046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6170775506808996,
+      "learning_rate": 3.4956860509377098e-06,
+      "loss": 0.4841,
+      "step": 14047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0686857382300343,
+      "learning_rate": 3.4951031293591776e-06,
+      "loss": 0.6861,
+      "step": 14048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0274188099481982,
+      "learning_rate": 3.4945202497519177e-06,
+      "loss": 0.638,
+      "step": 14049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5154014953830761,
+      "learning_rate": 3.4939374121281276e-06,
+      "loss": 0.4782,
+      "step": 14050
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.5046061277389526,
+      "eval_runtime": 68.0243,
+      "eval_samples_per_second": 50.129,
+      "eval_steps_per_second": 1.573,
+      "step": 14050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6993811638726564,
+      "learning_rate": 3.4933546165000133e-06,
+      "loss": 0.361,
+      "step": 14051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6853080834469139,
+      "learning_rate": 3.492771862879776e-06,
+      "loss": 0.5102,
+      "step": 14052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4355473049070917,
+      "learning_rate": 3.4921891512796093e-06,
+      "loss": 0.4277,
+      "step": 14053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7508986082415963,
+      "learning_rate": 3.4916064817117167e-06,
+      "loss": 0.5446,
+      "step": 14054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.92700124262246,
+      "learning_rate": 3.4910238541882967e-06,
+      "loss": 0.7144,
+      "step": 14055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5099068548140215,
+      "learning_rate": 3.4904412687215427e-06,
+      "loss": 0.4197,
+      "step": 14056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.746369222161763,
+      "learning_rate": 3.4898587253236504e-06,
+      "loss": 0.4536,
+      "step": 14057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7664462186169811,
+      "learning_rate": 3.489276224006821e-06,
+      "loss": 0.5917,
+      "step": 14058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.925562601434772,
+      "learning_rate": 3.4886937647832437e-06,
+      "loss": 0.549,
+      "step": 14059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7942833129847018,
+      "learning_rate": 3.4881113476651117e-06,
+      "loss": 0.5032,
+      "step": 14060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.857746379377763,
+      "learning_rate": 3.487528972664622e-06,
+      "loss": 0.5657,
+      "step": 14061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9383853169973262,
+      "learning_rate": 3.4869466397939616e-06,
+      "loss": 0.4831,
+      "step": 14062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7677880446532692,
+      "learning_rate": 3.4863643490653223e-06,
+      "loss": 0.5839,
+      "step": 14063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5656327062237845,
+      "learning_rate": 3.4857821004908967e-06,
+      "loss": 0.3802,
+      "step": 14064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5916829342938394,
+      "learning_rate": 3.4851998940828702e-06,
+      "loss": 0.4815,
+      "step": 14065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8502213562593095,
+      "learning_rate": 3.484617729853432e-06,
+      "loss": 0.5618,
+      "step": 14066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4989356082511582,
+      "learning_rate": 3.4840356078147717e-06,
+      "loss": 0.387,
+      "step": 14067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6414743773086256,
+      "learning_rate": 3.4834535279790717e-06,
+      "loss": 0.4651,
+      "step": 14068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.85552568035413,
+      "learning_rate": 3.482871490358521e-06,
+      "loss": 0.5661,
+      "step": 14069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7653624878623513,
+      "learning_rate": 3.4822894949653015e-06,
+      "loss": 0.5141,
+      "step": 14070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0485855287084513,
+      "learning_rate": 3.4817075418115996e-06,
+      "loss": 0.6553,
+      "step": 14071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4118489073802365,
+      "learning_rate": 3.4811256309095957e-06,
+      "loss": 0.451,
+      "step": 14072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5707094053402932,
+      "learning_rate": 3.480543762271473e-06,
+      "loss": 0.4472,
+      "step": 14073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6086104401594608,
+      "learning_rate": 3.4799619359094133e-06,
+      "loss": 0.5118,
+      "step": 14074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9656298907587446,
+      "learning_rate": 3.479380151835594e-06,
+      "loss": 0.6021,
+      "step": 14075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.818254853609277,
+      "learning_rate": 3.4787984100621976e-06,
+      "loss": 0.5731,
+      "step": 14076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5965344530471233,
+      "learning_rate": 3.478216710601401e-06,
+      "loss": 0.4199,
+      "step": 14077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5256590307248938,
+      "learning_rate": 3.4776350534653815e-06,
+      "loss": 0.3844,
+      "step": 14078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.081036297846742,
+      "learning_rate": 3.4770534386663147e-06,
+      "loss": 0.6411,
+      "step": 14079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7778213060392665,
+      "learning_rate": 3.47647186621638e-06,
+      "loss": 0.4685,
+      "step": 14080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2882885104474753,
+      "learning_rate": 3.475890336127748e-06,
+      "loss": 0.3202,
+      "step": 14081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8151415898811059,
+      "learning_rate": 3.475308848412595e-06,
+      "loss": 0.4777,
+      "step": 14082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.421658788987046,
+      "learning_rate": 3.4747274030830934e-06,
+      "loss": 0.4147,
+      "step": 14083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.095630297893878,
+      "learning_rate": 3.4741460001514176e-06,
+      "loss": 0.655,
+      "step": 14084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3916151334425397,
+      "learning_rate": 3.473564639629735e-06,
+      "loss": 0.4681,
+      "step": 14085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9702616317989523,
+      "learning_rate": 3.4729833215302185e-06,
+      "loss": 0.6883,
+      "step": 14086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4964774242997139,
+      "learning_rate": 3.4724020458650376e-06,
+      "loss": 0.5448,
+      "step": 14087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.81323813966552,
+      "learning_rate": 3.47182081264636e-06,
+      "loss": 0.5231,
+      "step": 14088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7485196168404902,
+      "learning_rate": 3.471239621886354e-06,
+      "loss": 0.4262,
+      "step": 14089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4224347025800763,
+      "learning_rate": 3.470658473597187e-06,
+      "loss": 0.5882,
+      "step": 14090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3801758015713024,
+      "learning_rate": 3.470077367791024e-06,
+      "loss": 0.4609,
+      "step": 14091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6410181969024784,
+      "learning_rate": 3.4694963044800294e-06,
+      "loss": 0.531,
+      "step": 14092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9314322126709564,
+      "learning_rate": 3.468915283676371e-06,
+      "loss": 0.569,
+      "step": 14093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7202472763794712,
+      "learning_rate": 3.4683343053922073e-06,
+      "loss": 0.4094,
+      "step": 14094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.806999410893785,
+      "learning_rate": 3.4677533696397036e-06,
+      "loss": 0.5421,
+      "step": 14095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8163647365545899,
+      "learning_rate": 3.4671724764310205e-06,
+      "loss": 0.567,
+      "step": 14096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.939895400620996,
+      "learning_rate": 3.4665916257783204e-06,
+      "loss": 0.4731,
+      "step": 14097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6349782740532037,
+      "learning_rate": 3.4660108176937603e-06,
+      "loss": 0.4449,
+      "step": 14098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.631779346856254,
+      "learning_rate": 3.4654300521895e-06,
+      "loss": 0.5107,
+      "step": 14099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5507100454262552,
+      "learning_rate": 3.4648493292776997e-06,
+      "loss": 0.4577,
+      "step": 14100
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.5042749643325806,
+      "eval_runtime": 68.1215,
+      "eval_samples_per_second": 50.058,
+      "eval_steps_per_second": 1.571,
+      "step": 14100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.71637021607557,
+      "learning_rate": 3.4642686489705128e-06,
+      "loss": 0.4074,
+      "step": 14101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.559185513098386,
+      "learning_rate": 3.4636880112800975e-06,
+      "loss": 0.4392,
+      "step": 14102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6825279491027867,
+      "learning_rate": 3.463107416218611e-06,
+      "loss": 0.5087,
+      "step": 14103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3558185906203755,
+      "learning_rate": 3.4625268637982036e-06,
+      "loss": 0.3183,
+      "step": 14104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.732579849955325,
+      "learning_rate": 3.4619463540310307e-06,
+      "loss": 0.4879,
+      "step": 14105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.56948012618146,
+      "learning_rate": 3.4613658869292465e-06,
+      "loss": 0.4198,
+      "step": 14106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5000010669610808,
+      "learning_rate": 3.4607854625049976e-06,
+      "loss": 0.4805,
+      "step": 14107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5843855973200698,
+      "learning_rate": 3.46020508077044e-06,
+      "loss": 0.4054,
+      "step": 14108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4621406380992708,
+      "learning_rate": 3.4596247417377248e-06,
+      "loss": 0.4261,
+      "step": 14109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9068011519054613,
+      "learning_rate": 3.459044445418996e-06,
+      "loss": 0.569,
+      "step": 14110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9535877931542158,
+      "learning_rate": 3.458464191826404e-06,
+      "loss": 0.6203,
+      "step": 14111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9902583375188376,
+      "learning_rate": 3.4578839809720977e-06,
+      "loss": 0.5265,
+      "step": 14112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4573633559786627,
+      "learning_rate": 3.457303812868221e-06,
+      "loss": 0.3724,
+      "step": 14113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.622792324967049,
+      "learning_rate": 3.4567236875269196e-06,
+      "loss": 0.4451,
+      "step": 14114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8513423839758003,
+      "learning_rate": 3.4561436049603397e-06,
+      "loss": 0.4351,
+      "step": 14115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8977639435964042,
+      "learning_rate": 3.455563565180623e-06,
+      "loss": 0.569,
+      "step": 14116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3457575105548094,
+      "learning_rate": 3.454983568199913e-06,
+      "loss": 0.3506,
+      "step": 14117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7103554616524177,
+      "learning_rate": 3.454403614030352e-06,
+      "loss": 0.5175,
+      "step": 14118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.106925685053558,
+      "learning_rate": 3.453823702684082e-06,
+      "loss": 0.5741,
+      "step": 14119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6574862844065892,
+      "learning_rate": 3.453243834173241e-06,
+      "loss": 0.3371,
+      "step": 14120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6641453875634882,
+      "learning_rate": 3.4526640085099683e-06,
+      "loss": 0.6066,
+      "step": 14121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.667351058059824,
+      "learning_rate": 3.4520842257064054e-06,
+      "loss": 0.5204,
+      "step": 14122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3403628958500218,
+      "learning_rate": 3.4515044857746844e-06,
+      "loss": 0.3298,
+      "step": 14123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8805237539301067,
+      "learning_rate": 3.450924788726946e-06,
+      "loss": 0.5299,
+      "step": 14124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6957152849025834,
+      "learning_rate": 3.4503451345753245e-06,
+      "loss": 0.5679,
+      "step": 14125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8324512627425071,
+      "learning_rate": 3.449765523331954e-06,
+      "loss": 0.4385,
+      "step": 14126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3072229078988706,
+      "learning_rate": 3.449185955008968e-06,
+      "loss": 0.6146,
+      "step": 14127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.690290601634811,
+      "learning_rate": 3.448606429618501e-06,
+      "loss": 0.4293,
+      "step": 14128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7489531560753937,
+      "learning_rate": 3.4480269471726858e-06,
+      "loss": 0.4745,
+      "step": 14129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7144877737192354,
+      "learning_rate": 3.44744750768365e-06,
+      "loss": 0.4512,
+      "step": 14130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0415261159685048,
+      "learning_rate": 3.446868111163526e-06,
+      "loss": 0.46,
+      "step": 14131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6884350498252987,
+      "learning_rate": 3.4462887576244443e-06,
+      "loss": 0.5069,
+      "step": 14132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5190971134712876,
+      "learning_rate": 3.44570944707853e-06,
+      "loss": 0.3613,
+      "step": 14133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3573287759415686,
+      "learning_rate": 3.4451301795379127e-06,
+      "loss": 0.3613,
+      "step": 14134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6671092205344058,
+      "learning_rate": 3.4445509550147207e-06,
+      "loss": 0.4481,
+      "step": 14135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7743831805300079,
+      "learning_rate": 3.4439717735210756e-06,
+      "loss": 0.4136,
+      "step": 14136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0667376164565887,
+      "learning_rate": 3.4433926350691048e-06,
+      "loss": 0.4749,
+      "step": 14137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8107455303012756,
+      "learning_rate": 3.4428135396709327e-06,
+      "loss": 0.4856,
+      "step": 14138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5376183855394214,
+      "learning_rate": 3.4422344873386807e-06,
+      "loss": 0.4254,
+      "step": 14139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.821354316284062,
+      "learning_rate": 3.4416554780844713e-06,
+      "loss": 0.5506,
+      "step": 14140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.509428607189238,
+      "learning_rate": 3.441076511920426e-06,
+      "loss": 0.4247,
+      "step": 14141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6232084217287508,
+      "learning_rate": 3.4404975888586666e-06,
+      "loss": 0.6336,
+      "step": 14142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6910616011690913,
+      "learning_rate": 3.4399187089113083e-06,
+      "loss": 0.4663,
+      "step": 14143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6272396870288586,
+      "learning_rate": 3.4393398720904738e-06,
+      "loss": 0.4269,
+      "step": 14144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.895337327811185,
+      "learning_rate": 3.4387610784082805e-06,
+      "loss": 0.5361,
+      "step": 14145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.405065138383051,
+      "learning_rate": 3.4381823278768412e-06,
+      "loss": 0.3982,
+      "step": 14146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4706282337072427,
+      "learning_rate": 3.437603620508275e-06,
+      "loss": 0.3238,
+      "step": 14147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.929597210942154,
+      "learning_rate": 3.4370249563146983e-06,
+      "loss": 0.4406,
+      "step": 14148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5734800328353975,
+      "learning_rate": 3.43644633530822e-06,
+      "loss": 0.3774,
+      "step": 14149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5847158750210568,
+      "learning_rate": 3.4358677575009567e-06,
+      "loss": 0.4418,
+      "step": 14150
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.5044123530387878,
+      "eval_runtime": 68.0715,
+      "eval_samples_per_second": 50.094,
+      "eval_steps_per_second": 1.572,
+      "step": 14150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5539545339355518,
+      "learning_rate": 3.4352892229050204e-06,
+      "loss": 0.4448,
+      "step": 14151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.561291134816496,
+      "learning_rate": 3.434710731532521e-06,
+      "loss": 0.3513,
+      "step": 14152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.507988614879797,
+      "learning_rate": 3.4341322833955692e-06,
+      "loss": 0.3889,
+      "step": 14153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6513906383147143,
+      "learning_rate": 3.433553878506275e-06,
+      "loss": 0.3577,
+      "step": 14154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9948270796083765,
+      "learning_rate": 3.4329755168767487e-06,
+      "loss": 0.5377,
+      "step": 14155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5761023896459871,
+      "learning_rate": 3.432397198519093e-06,
+      "loss": 0.4368,
+      "step": 14156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7499180242816113,
+      "learning_rate": 3.4318189234454204e-06,
+      "loss": 0.5685,
+      "step": 14157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5452359184073816,
+      "learning_rate": 3.4312406916678327e-06,
+      "loss": 0.5348,
+      "step": 14158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6153373052402829,
+      "learning_rate": 3.4306625031984366e-06,
+      "loss": 0.4978,
+      "step": 14159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7578027338347164,
+      "learning_rate": 3.4300843580493365e-06,
+      "loss": 0.3593,
+      "step": 14160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.239767960959605,
+      "learning_rate": 3.429506256232633e-06,
+      "loss": 0.6733,
+      "step": 14161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5475582360342546,
+      "learning_rate": 3.428928197760431e-06,
+      "loss": 0.4842,
+      "step": 14162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4245222400547033,
+      "learning_rate": 3.42835018264483e-06,
+      "loss": 0.466,
+      "step": 14163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6232636158318914,
+      "learning_rate": 3.4277722108979327e-06,
+      "loss": 0.5302,
+      "step": 14164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.918714551917863,
+      "learning_rate": 3.427194282531836e-06,
+      "loss": 0.5471,
+      "step": 14165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.060792569694963,
+      "learning_rate": 3.426616397558639e-06,
+      "loss": 0.5935,
+      "step": 14166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.664364730957687,
+      "learning_rate": 3.426038555990443e-06,
+      "loss": 0.4644,
+      "step": 14167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7823350896501502,
+      "learning_rate": 3.4254607578393384e-06,
+      "loss": 0.4758,
+      "step": 14168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.742347881190293,
+      "learning_rate": 3.4248830031174245e-06,
+      "loss": 0.4861,
+      "step": 14169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6435032894690882,
+      "learning_rate": 3.424305291836798e-06,
+      "loss": 0.4073,
+      "step": 14170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0101951015978816,
+      "learning_rate": 3.4237276240095492e-06,
+      "loss": 0.4097,
+      "step": 14171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5748601658201675,
+      "learning_rate": 3.4231499996477724e-06,
+      "loss": 0.4467,
+      "step": 14172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.106492078378564,
+      "learning_rate": 3.4225724187635635e-06,
+      "loss": 0.4181,
+      "step": 14173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6142211480897677,
+      "learning_rate": 3.4219948813690073e-06,
+      "loss": 0.443,
+      "step": 14174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3921311505118048,
+      "learning_rate": 3.4214173874761986e-06,
+      "loss": 0.4037,
+      "step": 14175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1074338055119295,
+      "learning_rate": 3.4208399370972257e-06,
+      "loss": 0.667,
+      "step": 14176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.025161624241632,
+      "learning_rate": 3.4202625302441788e-06,
+      "loss": 0.5724,
+      "step": 14177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5118392796824414,
+      "learning_rate": 3.4196851669291424e-06,
+      "loss": 0.3851,
+      "step": 14178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6009370910322214,
+      "learning_rate": 3.4191078471642047e-06,
+      "loss": 0.5618,
+      "step": 14179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6165597908583076,
+      "learning_rate": 3.4185305709614537e-06,
+      "loss": 0.4842,
+      "step": 14180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5387151811727982,
+      "learning_rate": 3.4179533383329704e-06,
+      "loss": 0.4798,
+      "step": 14181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.958647428453039,
+      "learning_rate": 3.4173761492908413e-06,
+      "loss": 0.4846,
+      "step": 14182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.346662832750773,
+      "learning_rate": 3.416799003847151e-06,
+      "loss": 0.6647,
+      "step": 14183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7615588509969264,
+      "learning_rate": 3.4162219020139765e-06,
+      "loss": 0.481,
+      "step": 14184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8257367723547435,
+      "learning_rate": 3.415644843803403e-06,
+      "loss": 0.4267,
+      "step": 14185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0814436913379923,
+      "learning_rate": 3.415067829227512e-06,
+      "loss": 0.508,
+      "step": 14186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8556016094421501,
+      "learning_rate": 3.4144908582983794e-06,
+      "loss": 0.6264,
+      "step": 14187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8028788113417755,
+      "learning_rate": 3.4139139310280845e-06,
+      "loss": 0.5251,
+      "step": 14188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.858340767105399,
+      "learning_rate": 3.413337047428706e-06,
+      "loss": 0.7838,
+      "step": 14189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.138360255425506,
+      "learning_rate": 3.4127602075123223e-06,
+      "loss": 0.644,
+      "step": 14190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9765630206573932,
+      "learning_rate": 3.4121834112910056e-06,
+      "loss": 0.4961,
+      "step": 14191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6812824746162256,
+      "learning_rate": 3.4116066587768316e-06,
+      "loss": 0.5481,
+      "step": 14192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6355177267903263,
+      "learning_rate": 3.4110299499818778e-06,
+      "loss": 0.4459,
+      "step": 14193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.731388767489945,
+      "learning_rate": 3.4104532849182127e-06,
+      "loss": 0.4528,
+      "step": 14194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.112167183792079,
+      "learning_rate": 3.40987666359791e-06,
+      "loss": 0.5356,
+      "step": 14195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5904588042249908,
+      "learning_rate": 3.4093000860330426e-06,
+      "loss": 0.3823,
+      "step": 14196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6977260160636367,
+      "learning_rate": 3.4087235522356776e-06,
+      "loss": 0.4187,
+      "step": 14197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5265392098263708,
+      "learning_rate": 3.408147062217886e-06,
+      "loss": 0.3942,
+      "step": 14198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0461656928002565,
+      "learning_rate": 3.4075706159917384e-06,
+      "loss": 0.5167,
+      "step": 14199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.439200316604771,
+      "learning_rate": 3.406994213569298e-06,
+      "loss": 0.3597,
+      "step": 14200
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.504010796546936,
+      "eval_runtime": 68.0609,
+      "eval_samples_per_second": 50.102,
+      "eval_steps_per_second": 1.572,
+      "step": 14200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3869797230044814,
+      "learning_rate": 3.4064178549626346e-06,
+      "loss": 0.4866,
+      "step": 14201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7463727265356566,
+      "learning_rate": 3.405841540183813e-06,
+      "loss": 0.5958,
+      "step": 14202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8292332934104083,
+      "learning_rate": 3.4052652692448983e-06,
+      "loss": 0.5382,
+      "step": 14203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5986514223691568,
+      "learning_rate": 3.404689042157954e-06,
+      "loss": 0.5559,
+      "step": 14204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9913764468061812,
+      "learning_rate": 3.4041128589350426e-06,
+      "loss": 0.5064,
+      "step": 14205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.960811061391681,
+      "learning_rate": 3.4035367195882267e-06,
+      "loss": 0.4822,
+      "step": 14206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.834631362095179,
+      "learning_rate": 3.4029606241295674e-06,
+      "loss": 0.4739,
+      "step": 14207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0453972021260833,
+      "learning_rate": 3.4023845725711253e-06,
+      "loss": 0.5235,
+      "step": 14208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9897841525178508,
+      "learning_rate": 3.4018085649249603e-06,
+      "loss": 0.5772,
+      "step": 14209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5482348758976587,
+      "learning_rate": 3.401232601203129e-06,
+      "loss": 0.4143,
+      "step": 14210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2443351736917114,
+      "learning_rate": 3.4006566814176882e-06,
+      "loss": 0.4175,
+      "step": 14211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.718771281280242,
+      "learning_rate": 3.4000808055806987e-06,
+      "loss": 0.5073,
+      "step": 14212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.772534870627097,
+      "learning_rate": 3.399504973704211e-06,
+      "loss": 0.545,
+      "step": 14213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7545062357842633,
+      "learning_rate": 3.398929185800282e-06,
+      "loss": 0.466,
+      "step": 14214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0189749051770303,
+      "learning_rate": 3.398353441880967e-06,
+      "loss": 0.4423,
+      "step": 14215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6053995587264174,
+      "learning_rate": 3.397777741958317e-06,
+      "loss": 0.4129,
+      "step": 14216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6426515834633666,
+      "learning_rate": 3.397202086044383e-06,
+      "loss": 0.3861,
+      "step": 14217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9619576756800576,
+      "learning_rate": 3.396626474151219e-06,
+      "loss": 0.6433,
+      "step": 14218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5988679636636762,
+      "learning_rate": 3.3960509062908715e-06,
+      "loss": 0.3247,
+      "step": 14219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6878010194968276,
+      "learning_rate": 3.3954753824753915e-06,
+      "loss": 0.5223,
+      "step": 14220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7696765248706305,
+      "learning_rate": 3.394899902716827e-06,
+      "loss": 0.5245,
+      "step": 14221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6524951770285097,
+      "learning_rate": 3.394324467027228e-06,
+      "loss": 0.5074,
+      "step": 14222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8669043151818001,
+      "learning_rate": 3.3937490754186363e-06,
+      "loss": 0.6717,
+      "step": 14223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6925703393553226,
+      "learning_rate": 3.3931737279031e-06,
+      "loss": 0.4529,
+      "step": 14224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9078802785355498,
+      "learning_rate": 3.3925984244926637e-06,
+      "loss": 0.4187,
+      "step": 14225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7883145214914244,
+      "learning_rate": 3.392023165199369e-06,
+      "loss": 0.4995,
+      "step": 14226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6871759889744893,
+      "learning_rate": 3.391447950035261e-06,
+      "loss": 0.5119,
+      "step": 14227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7570714018744586,
+      "learning_rate": 3.390872779012381e-06,
+      "loss": 0.4171,
+      "step": 14228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6278884788884833,
+      "learning_rate": 3.390297652142769e-06,
+      "loss": 0.3744,
+      "step": 14229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4320644629757278,
+      "learning_rate": 3.389722569438463e-06,
+      "loss": 0.4001,
+      "step": 14230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6159554721890623,
+      "learning_rate": 3.389147530911508e-06,
+      "loss": 0.4811,
+      "step": 14231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5922244144264384,
+      "learning_rate": 3.3885725365739363e-06,
+      "loss": 0.5249,
+      "step": 14232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1340144117464863,
+      "learning_rate": 3.387997586437787e-06,
+      "loss": 0.5206,
+      "step": 14233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0345982865838237,
+      "learning_rate": 3.387422680515096e-06,
+      "loss": 0.549,
+      "step": 14234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.051712802122901,
+      "learning_rate": 3.386847818817901e-06,
+      "loss": 0.4239,
+      "step": 14235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.369868148463022,
+      "learning_rate": 3.386273001358233e-06,
+      "loss": 0.3761,
+      "step": 14236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.088577370837876,
+      "learning_rate": 3.385698228148126e-06,
+      "loss": 0.5259,
+      "step": 14237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0446979141288537,
+      "learning_rate": 3.385123499199615e-06,
+      "loss": 0.628,
+      "step": 14238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7031180591302668,
+      "learning_rate": 3.3845488145247295e-06,
+      "loss": 0.5488,
+      "step": 14239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7773343418835232,
+      "learning_rate": 3.3839741741354993e-06,
+      "loss": 0.5495,
+      "step": 14240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6019393642199524,
+      "learning_rate": 3.3833995780439577e-06,
+      "loss": 0.3808,
+      "step": 14241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3147532330140326,
+      "learning_rate": 3.38282502626213e-06,
+      "loss": 0.6979,
+      "step": 14242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.777569216742771,
+      "learning_rate": 3.382250518802045e-06,
+      "loss": 0.5117,
+      "step": 14243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6752275419921323,
+      "learning_rate": 3.3816760556757318e-06,
+      "loss": 0.4801,
+      "step": 14244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.033151171262965,
+      "learning_rate": 3.381101636895212e-06,
+      "loss": 0.6541,
+      "step": 14245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8649708554835258,
+      "learning_rate": 3.3805272624725144e-06,
+      "loss": 0.6842,
+      "step": 14246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4092992082128835,
+      "learning_rate": 3.3799529324196622e-06,
+      "loss": 0.3813,
+      "step": 14247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5605326520467455,
+      "learning_rate": 3.37937864674868e-06,
+      "loss": 0.4814,
+      "step": 14248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8137426893264341,
+      "learning_rate": 3.3788044054715864e-06,
+      "loss": 0.4469,
+      "step": 14249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4898871182972655,
+      "learning_rate": 3.3782302086004058e-06,
+      "loss": 0.5407,
+      "step": 14250
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.5042082667350769,
+      "eval_runtime": 68.0785,
+      "eval_samples_per_second": 50.089,
+      "eval_steps_per_second": 1.572,
+      "step": 14250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.766086742456162,
+      "learning_rate": 3.3776560561471594e-06,
+      "loss": 0.4822,
+      "step": 14251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.901403914491155,
+      "learning_rate": 3.3770819481238625e-06,
+      "loss": 0.4712,
+      "step": 14252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1642987335526933,
+      "learning_rate": 3.3765078845425377e-06,
+      "loss": 0.539,
+      "step": 14253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2417181430914312,
+      "learning_rate": 3.375933865415201e-06,
+      "loss": 0.5986,
+      "step": 14254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6933402191672868,
+      "learning_rate": 3.37535989075387e-06,
+      "loss": 0.6407,
+      "step": 14255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5719661639574047,
+      "learning_rate": 3.3747859605705594e-06,
+      "loss": 0.4858,
+      "step": 14256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1263953845923065,
+      "learning_rate": 3.3742120748772866e-06,
+      "loss": 0.4342,
+      "step": 14257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8285204360449503,
+      "learning_rate": 3.373638233686062e-06,
+      "loss": 0.5746,
+      "step": 14258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8345103084929308,
+      "learning_rate": 3.3730644370088998e-06,
+      "loss": 0.5751,
+      "step": 14259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.760051508225432,
+      "learning_rate": 3.372490684857815e-06,
+      "loss": 0.521,
+      "step": 14260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8355856586872323,
+      "learning_rate": 3.3719169772448136e-06,
+      "loss": 0.6551,
+      "step": 14261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8790091448666615,
+      "learning_rate": 3.3713433141819097e-06,
+      "loss": 0.4937,
+      "step": 14262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7775352357187655,
+      "learning_rate": 3.370769695681112e-06,
+      "loss": 0.449,
+      "step": 14263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.816691542296263,
+      "learning_rate": 3.3701961217544266e-06,
+      "loss": 0.5221,
+      "step": 14264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5339548902486058,
+      "learning_rate": 3.369622592413863e-06,
+      "loss": 0.4414,
+      "step": 14265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8832239516976295,
+      "learning_rate": 3.369049107671429e-06,
+      "loss": 0.6869,
+      "step": 14266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6893821556286965,
+      "learning_rate": 3.3684756675391263e-06,
+      "loss": 0.4451,
+      "step": 14267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5055541737007574,
+      "learning_rate": 3.367902272028961e-06,
+      "loss": 0.4327,
+      "step": 14268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9070806729790015,
+      "learning_rate": 3.3673289211529383e-06,
+      "loss": 0.5287,
+      "step": 14269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8407867747750666,
+      "learning_rate": 3.366755614923061e-06,
+      "loss": 0.5029,
+      "step": 14270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.934601884825137,
+      "learning_rate": 3.366182353351329e-06,
+      "loss": 0.6508,
+      "step": 14271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8764756527432045,
+      "learning_rate": 3.3656091364497433e-06,
+      "loss": 0.5681,
+      "step": 14272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8343962717352242,
+      "learning_rate": 3.365035964230307e-06,
+      "loss": 0.4873,
+      "step": 14273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9684212097651332,
+      "learning_rate": 3.3644628367050148e-06,
+      "loss": 0.5027,
+      "step": 14274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6677665019211474,
+      "learning_rate": 3.363889753885866e-06,
+      "loss": 0.3566,
+      "step": 14275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6793000063405508,
+      "learning_rate": 3.3633167157848614e-06,
+      "loss": 0.4059,
+      "step": 14276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9852974473607534,
+      "learning_rate": 3.3627437224139916e-06,
+      "loss": 0.6671,
+      "step": 14277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7620073734253971,
+      "learning_rate": 3.3621707737852537e-06,
+      "loss": 0.4103,
+      "step": 14278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7632240310829144,
+      "learning_rate": 3.361597869910646e-06,
+      "loss": 0.4723,
+      "step": 14279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0726260244205563,
+      "learning_rate": 3.3610250108021563e-06,
+      "loss": 0.4335,
+      "step": 14280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.787879955271473,
+      "learning_rate": 3.36045219647178e-06,
+      "loss": 0.4085,
+      "step": 14281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5241157248001553,
+      "learning_rate": 3.359879426931507e-06,
+      "loss": 0.4682,
+      "step": 14282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.894051662977566,
+      "learning_rate": 3.359306702193331e-06,
+      "loss": 0.463,
+      "step": 14283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.00435389338687,
+      "learning_rate": 3.358734022269239e-06,
+      "loss": 0.6191,
+      "step": 14284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.019129570636047,
+      "learning_rate": 3.358161387171218e-06,
+      "loss": 0.6867,
+      "step": 14285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6580507133147995,
+      "learning_rate": 3.357588796911262e-06,
+      "loss": 0.4678,
+      "step": 14286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9235585348537954,
+      "learning_rate": 3.3570162515013503e-06,
+      "loss": 0.5636,
+      "step": 14287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6127283770115857,
+      "learning_rate": 3.3564437509534743e-06,
+      "loss": 0.4323,
+      "step": 14288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7467891871044794,
+      "learning_rate": 3.3558712952796167e-06,
+      "loss": 0.4343,
+      "step": 14289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9780446292590488,
+      "learning_rate": 3.355298884491761e-06,
+      "loss": 0.5447,
+      "step": 14290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.709970059257823,
+      "learning_rate": 3.354726518601891e-06,
+      "loss": 0.5804,
+      "step": 14291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4736707357958974,
+      "learning_rate": 3.3541541976219905e-06,
+      "loss": 0.4054,
+      "step": 14292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5566398723860422,
+      "learning_rate": 3.3535819215640375e-06,
+      "loss": 0.3669,
+      "step": 14293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1969674131640535,
+      "learning_rate": 3.3530096904400134e-06,
+      "loss": 0.6743,
+      "step": 14294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7815877195172751,
+      "learning_rate": 3.352437504261898e-06,
+      "loss": 0.7107,
+      "step": 14295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7555278256100275,
+      "learning_rate": 3.3518653630416713e-06,
+      "loss": 0.5148,
+      "step": 14296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.847442918692246,
+      "learning_rate": 3.351293266791308e-06,
+      "loss": 0.6524,
+      "step": 14297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8506389260765865,
+      "learning_rate": 3.350721215522784e-06,
+      "loss": 0.5834,
+      "step": 14298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5862436239489965,
+      "learning_rate": 3.350149209248079e-06,
+      "loss": 0.4685,
+      "step": 14299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8432793458246397,
+      "learning_rate": 3.349577247979163e-06,
+      "loss": 0.4622,
+      "step": 14300
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.5045114159584045,
+      "eval_runtime": 68.1405,
+      "eval_samples_per_second": 50.044,
+      "eval_steps_per_second": 1.57,
+      "step": 14300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.478538457211359,
+      "learning_rate": 3.3490053317280125e-06,
+      "loss": 0.4937,
+      "step": 14301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8096622795051258,
+      "learning_rate": 3.3484334605066e-06,
+      "loss": 0.5807,
+      "step": 14302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0891733069950025,
+      "learning_rate": 3.3478616343268955e-06,
+      "loss": 0.688,
+      "step": 14303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7750237658779755,
+      "learning_rate": 3.3472898532008693e-06,
+      "loss": 0.6198,
+      "step": 14304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2311645682929235,
+      "learning_rate": 3.346718117140496e-06,
+      "loss": 0.3095,
+      "step": 14305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7925780604375972,
+      "learning_rate": 3.3461464261577397e-06,
+      "loss": 0.5135,
+      "step": 14306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6447627605596657,
+      "learning_rate": 3.34557478026457e-06,
+      "loss": 0.4042,
+      "step": 14307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4726423215856494,
+      "learning_rate": 3.3450031794729565e-06,
+      "loss": 0.508,
+      "step": 14308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6410203357926112,
+      "learning_rate": 3.344431623794861e-06,
+      "loss": 0.4477,
+      "step": 14309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.977453486697347,
+      "learning_rate": 3.3438601132422505e-06,
+      "loss": 0.609,
+      "step": 14310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6991300086084082,
+      "learning_rate": 3.3432886478270902e-06,
+      "loss": 0.541,
+      "step": 14311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6222316646085617,
+      "learning_rate": 3.3427172275613416e-06,
+      "loss": 0.5614,
+      "step": 14312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7268850299617733,
+      "learning_rate": 3.3421458524569684e-06,
+      "loss": 0.5619,
+      "step": 14313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6700048362420021,
+      "learning_rate": 3.3415745225259304e-06,
+      "loss": 0.5604,
+      "step": 14314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6210564575207207,
+      "learning_rate": 3.341003237780192e-06,
+      "loss": 0.4601,
+      "step": 14315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5069451760567454,
+      "learning_rate": 3.340431998231707e-06,
+      "loss": 0.4654,
+      "step": 14316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4269313237785262,
+      "learning_rate": 3.3398608038924375e-06,
+      "loss": 0.3931,
+      "step": 14317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8053673291822985,
+      "learning_rate": 3.3392896547743413e-06,
+      "loss": 0.4389,
+      "step": 14318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3831865551888947,
+      "learning_rate": 3.338718550889373e-06,
+      "loss": 0.8209,
+      "step": 14319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.503809695122812,
+      "learning_rate": 3.3381474922494895e-06,
+      "loss": 0.3731,
+      "step": 14320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4309167805063296,
+      "learning_rate": 3.3375764788666465e-06,
+      "loss": 0.3543,
+      "step": 14321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7039398063847095,
+      "learning_rate": 3.337005510752796e-06,
+      "loss": 0.4348,
+      "step": 14322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.268834843727017,
+      "learning_rate": 3.336434587919891e-06,
+      "loss": 0.3349,
+      "step": 14323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7831969274705142,
+      "learning_rate": 3.3358637103798858e-06,
+      "loss": 0.454,
+      "step": 14324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9270728638864203,
+      "learning_rate": 3.335292878144728e-06,
+      "loss": 0.6341,
+      "step": 14325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5768494844893441,
+      "learning_rate": 3.33472209122637e-06,
+      "loss": 0.3572,
+      "step": 14326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5380007791034205,
+      "learning_rate": 3.334151349636759e-06,
+      "loss": 0.471,
+      "step": 14327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.242151557340886,
+      "learning_rate": 3.3335806533878464e-06,
+      "loss": 0.673,
+      "step": 14328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5938814695752785,
+      "learning_rate": 3.3330100024915757e-06,
+      "loss": 0.4646,
+      "step": 14329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6744727927995646,
+      "learning_rate": 3.332439396959894e-06,
+      "loss": 0.3781,
+      "step": 14330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5213698045848865,
+      "learning_rate": 3.3318688368047493e-06,
+      "loss": 0.43,
+      "step": 14331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2895803931385614,
+      "learning_rate": 3.3312983220380824e-06,
+      "loss": 0.3459,
+      "step": 14332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.542967847108677,
+      "learning_rate": 3.3307278526718373e-06,
+      "loss": 0.4352,
+      "step": 14333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.486415587700653,
+      "learning_rate": 3.3301574287179596e-06,
+      "loss": 0.4275,
+      "step": 14334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6808571691161518,
+      "learning_rate": 3.3295870501883866e-06,
+      "loss": 0.3942,
+      "step": 14335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9599756578028924,
+      "learning_rate": 3.3290167170950606e-06,
+      "loss": 0.5066,
+      "step": 14336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3064551575196433,
+      "learning_rate": 3.328446429449923e-06,
+      "loss": 0.509,
+      "step": 14337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5862061133431589,
+      "learning_rate": 3.327876187264909e-06,
+      "loss": 0.4918,
+      "step": 14338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.662837128855179,
+      "learning_rate": 3.327305990551958e-06,
+      "loss": 0.521,
+      "step": 14339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5729245100796687,
+      "learning_rate": 3.3267358393230054e-06,
+      "loss": 0.4012,
+      "step": 14340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1706680624880814,
+      "learning_rate": 3.3261657335899903e-06,
+      "loss": 0.6298,
+      "step": 14341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4883559146340757,
+      "learning_rate": 3.325595673364844e-06,
+      "loss": 0.4701,
+      "step": 14342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.573229598907604,
+      "learning_rate": 3.325025658659502e-06,
+      "loss": 0.426,
+      "step": 14343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.960921277853873,
+      "learning_rate": 3.324455689485898e-06,
+      "loss": 0.5705,
+      "step": 14344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.002507951083208,
+      "learning_rate": 3.3238857658559615e-06,
+      "loss": 0.647,
+      "step": 14345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.101370424244631,
+      "learning_rate": 3.3233158877816252e-06,
+      "loss": 0.5932,
+      "step": 14346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6109422733549585,
+      "learning_rate": 3.3227460552748197e-06,
+      "loss": 0.4126,
+      "step": 14347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.069919958283278,
+      "learning_rate": 3.3221762683474724e-06,
+      "loss": 0.6172,
+      "step": 14348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8903896092410453,
+      "learning_rate": 3.3216065270115118e-06,
+      "loss": 0.4073,
+      "step": 14349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8289879652949481,
+      "learning_rate": 3.3210368312788665e-06,
+      "loss": 0.6437,
+      "step": 14350
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.5042588710784912,
+      "eval_runtime": 68.067,
+      "eval_samples_per_second": 50.098,
+      "eval_steps_per_second": 1.572,
+      "step": 14350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8374932334574567,
+      "learning_rate": 3.3204671811614607e-06,
+      "loss": 0.5258,
+      "step": 14351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9928119043854577,
+      "learning_rate": 3.319897576671221e-06,
+      "loss": 0.4815,
+      "step": 14352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.837282647844027,
+      "learning_rate": 3.3193280178200702e-06,
+      "loss": 0.6785,
+      "step": 14353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.381552408973852,
+      "learning_rate": 3.3187585046199342e-06,
+      "loss": 0.5693,
+      "step": 14354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7572446605431395,
+      "learning_rate": 3.318189037082734e-06,
+      "loss": 0.4364,
+      "step": 14355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8879912838564712,
+      "learning_rate": 3.317619615220393e-06,
+      "loss": 0.4079,
+      "step": 14356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5784036218544202,
+      "learning_rate": 3.3170502390448267e-06,
+      "loss": 0.4325,
+      "step": 14357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6007835171922677,
+      "learning_rate": 3.3164809085679594e-06,
+      "loss": 0.5288,
+      "step": 14358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7974012693111856,
+      "learning_rate": 3.3159116238017084e-06,
+      "loss": 0.5349,
+      "step": 14359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4721799947043772,
+      "learning_rate": 3.31534238475799e-06,
+      "loss": 0.4731,
+      "step": 14360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.499525270330977,
+      "learning_rate": 3.314773191448722e-06,
+      "loss": 0.3821,
+      "step": 14361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.779998005711339,
+      "learning_rate": 3.3142040438858193e-06,
+      "loss": 0.5845,
+      "step": 14362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7548088440892882,
+      "learning_rate": 3.3136349420812e-06,
+      "loss": 0.5321,
+      "step": 14363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6611459097439663,
+      "learning_rate": 3.313065886046773e-06,
+      "loss": 0.4177,
+      "step": 14364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7546158161226164,
+      "learning_rate": 3.312496875794453e-06,
+      "loss": 0.6071,
+      "step": 14365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.713351598394215,
+      "learning_rate": 3.3119279113361536e-06,
+      "loss": 0.4352,
+      "step": 14366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2780686261592957,
+      "learning_rate": 3.311358992683783e-06,
+      "loss": 0.3044,
+      "step": 14367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.973058689329714,
+      "learning_rate": 3.3107901198492526e-06,
+      "loss": 0.885,
+      "step": 14368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9842750811992993,
+      "learning_rate": 3.3102212928444726e-06,
+      "loss": 0.5718,
+      "step": 14369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5919076221032997,
+      "learning_rate": 3.3096525116813487e-06,
+      "loss": 0.4094,
+      "step": 14370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5889680617026898,
+      "learning_rate": 3.309083776371789e-06,
+      "loss": 0.4023,
+      "step": 14371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6859390083930483,
+      "learning_rate": 3.3085150869277006e-06,
+      "loss": 0.5953,
+      "step": 14372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9924631065474983,
+      "learning_rate": 3.307946443360986e-06,
+      "loss": 0.6487,
+      "step": 14373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.939673571701221,
+      "learning_rate": 3.3073778456835516e-06,
+      "loss": 0.5016,
+      "step": 14374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6761004152357515,
+      "learning_rate": 3.3068092939073006e-06,
+      "loss": 0.3829,
+      "step": 14375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0022640806798546,
+      "learning_rate": 3.306240788044135e-06,
+      "loss": 0.5385,
+      "step": 14376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6020869778682163,
+      "learning_rate": 3.3056723281059553e-06,
+      "loss": 0.4577,
+      "step": 14377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6758467712301415,
+      "learning_rate": 3.305103914104662e-06,
+      "loss": 0.4749,
+      "step": 14378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.651500215187654,
+      "learning_rate": 3.3045355460521568e-06,
+      "loss": 0.469,
+      "step": 14379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7552105298442935,
+      "learning_rate": 3.303967223960335e-06,
+      "loss": 0.572,
+      "step": 14380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6204698093588468,
+      "learning_rate": 3.3033989478410945e-06,
+      "loss": 0.3515,
+      "step": 14381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4366289800783767,
+      "learning_rate": 3.302830717706335e-06,
+      "loss": 0.291,
+      "step": 14382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1930326976017023,
+      "learning_rate": 3.302262533567948e-06,
+      "loss": 0.6311,
+      "step": 14383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.988223887107557,
+      "learning_rate": 3.3016943954378294e-06,
+      "loss": 0.4971,
+      "step": 14384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5695677385464164,
+      "learning_rate": 3.3011263033278727e-06,
+      "loss": 0.5016,
+      "step": 14385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6744376507708187,
+      "learning_rate": 3.3005582572499745e-06,
+      "loss": 0.5022,
+      "step": 14386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3562392696153027,
+      "learning_rate": 3.2999902572160204e-06,
+      "loss": 0.3004,
+      "step": 14387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.297628842122858,
+      "learning_rate": 3.2994223032379034e-06,
+      "loss": 0.3476,
+      "step": 14388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8111250887095047,
+      "learning_rate": 3.2988543953275164e-06,
+      "loss": 0.5026,
+      "step": 14389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7375593507392737,
+      "learning_rate": 3.298286533496743e-06,
+      "loss": 0.4627,
+      "step": 14390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7176015317855935,
+      "learning_rate": 3.297718717757473e-06,
+      "loss": 0.4944,
+      "step": 14391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.350847334989368,
+      "learning_rate": 3.297150948121596e-06,
+      "loss": 0.3985,
+      "step": 14392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7314491521046893,
+      "learning_rate": 3.2965832246009945e-06,
+      "loss": 0.4426,
+      "step": 14393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.414185779801894,
+      "learning_rate": 3.296015547207554e-06,
+      "loss": 0.4729,
+      "step": 14394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2020005824636906,
+      "learning_rate": 3.2954479159531603e-06,
+      "loss": 0.4263,
+      "step": 14395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.53257021726327,
+      "learning_rate": 3.2948803308496948e-06,
+      "loss": 0.3991,
+      "step": 14396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9704380065982174,
+      "learning_rate": 3.294312791909039e-06,
+      "loss": 0.5712,
+      "step": 14397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6564002851931872,
+      "learning_rate": 3.2937452991430745e-06,
+      "loss": 0.3661,
+      "step": 14398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.010041468700026,
+      "learning_rate": 3.293177852563685e-06,
+      "loss": 0.5171,
+      "step": 14399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.802027009037774,
+      "learning_rate": 3.292610452182743e-06,
+      "loss": 0.4817,
+      "step": 14400
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.5040960311889648,
+      "eval_runtime": 68.0659,
+      "eval_samples_per_second": 50.099,
+      "eval_steps_per_second": 1.572,
+      "step": 14400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2948803950765,
+      "learning_rate": 3.292043098012131e-06,
+      "loss": 0.5871,
+      "step": 14401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.528172467727152,
+      "learning_rate": 3.2914757900637255e-06,
+      "loss": 0.406,
+      "step": 14402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7596700900491835,
+      "learning_rate": 3.2909085283494025e-06,
+      "loss": 0.5334,
+      "step": 14403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6268625206799727,
+      "learning_rate": 3.290341312881039e-06,
+      "loss": 0.33,
+      "step": 14404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.334886602600122,
+      "learning_rate": 3.2897741436705055e-06,
+      "loss": 0.6651,
+      "step": 14405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.121836375730023,
+      "learning_rate": 3.2892070207296776e-06,
+      "loss": 0.5192,
+      "step": 14406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6852474303122684,
+      "learning_rate": 3.2886399440704276e-06,
+      "loss": 0.6037,
+      "step": 14407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7824078987846579,
+      "learning_rate": 3.288072913704629e-06,
+      "loss": 0.4937,
+      "step": 14408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6497389397805857,
+      "learning_rate": 3.287505929644148e-06,
+      "loss": 0.5564,
+      "step": 14409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7520284090361526,
+      "learning_rate": 3.2869389919008567e-06,
+      "loss": 0.5333,
+      "step": 14410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.703990317660139,
+      "learning_rate": 3.2863721004866234e-06,
+      "loss": 0.4846,
+      "step": 14411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6221482389146398,
+      "learning_rate": 3.2858052554133144e-06,
+      "loss": 0.6177,
+      "step": 14412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9538460112399814,
+      "learning_rate": 3.2852384566927976e-06,
+      "loss": 0.4644,
+      "step": 14413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9181024838712721,
+      "learning_rate": 3.284671704336939e-06,
+      "loss": 0.5268,
+      "step": 14414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.515402429332685,
+      "learning_rate": 3.2841049983576005e-06,
+      "loss": 0.4905,
+      "step": 14415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6505313272238007,
+      "learning_rate": 3.2835383387666486e-06,
+      "loss": 0.4551,
+      "step": 14416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6587611421552892,
+      "learning_rate": 3.282971725575946e-06,
+      "loss": 0.5209,
+      "step": 14417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6247294888159785,
+      "learning_rate": 3.282405158797351e-06,
+      "loss": 0.4142,
+      "step": 14418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5899068614496827,
+      "learning_rate": 3.281838638442727e-06,
+      "loss": 0.5139,
+      "step": 14419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7622436975116658,
+      "learning_rate": 3.281272164523933e-06,
+      "loss": 0.6657,
+      "step": 14420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3962973222195556,
+      "learning_rate": 3.2807057370528304e-06,
+      "loss": 0.3966,
+      "step": 14421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6540672432294787,
+      "learning_rate": 3.2801393560412727e-06,
+      "loss": 0.4171,
+      "step": 14422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7962605468817263,
+      "learning_rate": 3.2795730215011184e-06,
+      "loss": 0.5429,
+      "step": 14423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.64402360588607,
+      "learning_rate": 3.2790067334442265e-06,
+      "loss": 0.5457,
+      "step": 14424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9233257171291118,
+      "learning_rate": 3.2784404918824464e-06,
+      "loss": 0.5367,
+      "step": 14425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.472853794881359,
+      "learning_rate": 3.2778742968276354e-06,
+      "loss": 0.6091,
+      "step": 14426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9407261427785416,
+      "learning_rate": 3.277308148291647e-06,
+      "loss": 0.5086,
+      "step": 14427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7547965833731376,
+      "learning_rate": 3.2767420462863308e-06,
+      "loss": 0.5148,
+      "step": 14428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8739603919542056,
+      "learning_rate": 3.2761759908235392e-06,
+      "loss": 0.5671,
+      "step": 14429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0007249424483295,
+      "learning_rate": 3.2756099819151227e-06,
+      "loss": 0.5516,
+      "step": 14430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7381824536542057,
+      "learning_rate": 3.275044019572928e-06,
+      "loss": 0.5307,
+      "step": 14431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6944712732542502,
+      "learning_rate": 3.274478103808805e-06,
+      "loss": 0.3999,
+      "step": 14432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.188683709116792,
+      "learning_rate": 3.2739122346346007e-06,
+      "loss": 0.4612,
+      "step": 14433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5333769598264526,
+      "learning_rate": 3.2733464120621628e-06,
+      "loss": 0.4497,
+      "step": 14434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9656194340239337,
+      "learning_rate": 3.2727806361033337e-06,
+      "loss": 0.4698,
+      "step": 14435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.768960075090523,
+      "learning_rate": 3.272214906769957e-06,
+      "loss": 0.3787,
+      "step": 14436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4138195402174367,
+      "learning_rate": 3.2716492240738807e-06,
+      "loss": 0.3857,
+      "step": 14437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.834598512624227,
+      "learning_rate": 3.271083588026942e-06,
+      "loss": 0.6447,
+      "step": 14438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.338356563376128,
+      "learning_rate": 3.2705179986409834e-06,
+      "loss": 0.6247,
+      "step": 14439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6783384331708182,
+      "learning_rate": 3.269952455927847e-06,
+      "loss": 0.5035,
+      "step": 14440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6247491197213622,
+      "learning_rate": 3.26938695989937e-06,
+      "loss": 0.4379,
+      "step": 14441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7248012486004356,
+      "learning_rate": 3.2688215105673908e-06,
+      "loss": 0.4009,
+      "step": 14442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.894511464250447,
+      "learning_rate": 3.2682561079437495e-06,
+      "loss": 0.5302,
+      "step": 14443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8508980294570334,
+      "learning_rate": 3.2676907520402785e-06,
+      "loss": 0.4647,
+      "step": 14444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8917493774488836,
+      "learning_rate": 3.2671254428688154e-06,
+      "loss": 0.4729,
+      "step": 14445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2612198679907318,
+      "learning_rate": 3.266560180441194e-06,
+      "loss": 0.3226,
+      "step": 14446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1687652372079684,
+      "learning_rate": 3.26599496476925e-06,
+      "loss": 0.5214,
+      "step": 14447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6810804236265586,
+      "learning_rate": 3.2654297958648116e-06,
+      "loss": 0.4067,
+      "step": 14448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6950301056657253,
+      "learning_rate": 3.2648646737397125e-06,
+      "loss": 0.4169,
+      "step": 14449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8980959026781885,
+      "learning_rate": 3.264299598405785e-06,
+      "loss": 0.6556,
+      "step": 14450
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5031856894493103,
+      "eval_runtime": 68.1163,
+      "eval_samples_per_second": 50.061,
+      "eval_steps_per_second": 1.571,
+      "step": 14450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6032684219046285,
+      "learning_rate": 3.263734569874854e-06,
+      "loss": 0.412,
+      "step": 14451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.520292108601187,
+      "learning_rate": 3.263169588158753e-06,
+      "loss": 0.517,
+      "step": 14452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4994704505535228,
+      "learning_rate": 3.2626046532693067e-06,
+      "loss": 0.4706,
+      "step": 14453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9594844262153719,
+      "learning_rate": 3.2620397652183422e-06,
+      "loss": 0.4749,
+      "step": 14454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.014305222701485,
+      "learning_rate": 3.2614749240176853e-06,
+      "loss": 0.4528,
+      "step": 14455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.405299058008849,
+      "learning_rate": 3.2609101296791614e-06,
+      "loss": 0.4492,
+      "step": 14456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9008958809089158,
+      "learning_rate": 3.2603453822145916e-06,
+      "loss": 0.591,
+      "step": 14457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.584929293387504,
+      "learning_rate": 3.2597806816358e-06,
+      "loss": 0.4498,
+      "step": 14458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.731621641257565,
+      "learning_rate": 3.2592160279546102e-06,
+      "loss": 0.4727,
+      "step": 14459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.634645795783399,
+      "learning_rate": 3.25865142118284e-06,
+      "loss": 0.3221,
+      "step": 14460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6312452307083967,
+      "learning_rate": 3.2580868613323095e-06,
+      "loss": 0.4055,
+      "step": 14461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.652115670585536,
+      "learning_rate": 3.25752234841484e-06,
+      "loss": 0.4275,
+      "step": 14462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8524461001237749,
+      "learning_rate": 3.256957882442246e-06,
+      "loss": 0.5188,
+      "step": 14463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9097147268765167,
+      "learning_rate": 3.256393463426345e-06,
+      "loss": 0.6615,
+      "step": 14464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.580479713461842,
+      "learning_rate": 3.2558290913789532e-06,
+      "loss": 0.3594,
+      "step": 14465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6031619301481954,
+      "learning_rate": 3.2552647663118867e-06,
+      "loss": 0.4999,
+      "step": 14466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8121303747975297,
+      "learning_rate": 3.2547004882369575e-06,
+      "loss": 0.4742,
+      "step": 14467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6492047779556076,
+      "learning_rate": 3.2541362571659787e-06,
+      "loss": 0.4528,
+      "step": 14468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.757961214140713,
+      "learning_rate": 3.253572073110764e-06,
+      "loss": 0.5397,
+      "step": 14469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.887215424679411,
+      "learning_rate": 3.2530079360831206e-06,
+      "loss": 0.6409,
+      "step": 14470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7592832976115014,
+      "learning_rate": 3.2524438460948605e-06,
+      "loss": 0.3844,
+      "step": 14471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5349241648113954,
+      "learning_rate": 3.251879803157794e-06,
+      "loss": 0.4354,
+      "step": 14472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.662269875970196,
+      "learning_rate": 3.2513158072837263e-06,
+      "loss": 0.5481,
+      "step": 14473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5303002253089208,
+      "learning_rate": 3.2507518584844655e-06,
+      "loss": 0.3736,
+      "step": 14474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.00720174787893,
+      "learning_rate": 3.2501879567718187e-06,
+      "loss": 0.4916,
+      "step": 14475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8942458437502498,
+      "learning_rate": 3.249624102157588e-06,
+      "loss": 0.5299,
+      "step": 14476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7409331101337044,
+      "learning_rate": 3.249060294653579e-06,
+      "loss": 0.4966,
+      "step": 14477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.314832527332666,
+      "learning_rate": 3.2484965342715946e-06,
+      "loss": 0.6323,
+      "step": 14478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8136769373356028,
+      "learning_rate": 3.2479328210234385e-06,
+      "loss": 0.566,
+      "step": 14479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4696532491050809,
+      "learning_rate": 3.2473691549209084e-06,
+      "loss": 0.459,
+      "step": 14480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5166757842738532,
+      "learning_rate": 3.2468055359758055e-06,
+      "loss": 0.4486,
+      "step": 14481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6681731927845103,
+      "learning_rate": 3.2462419641999317e-06,
+      "loss": 0.5224,
+      "step": 14482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8006930561386862,
+      "learning_rate": 3.24567843960508e-06,
+      "loss": 0.5042,
+      "step": 14483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9567173857778284,
+      "learning_rate": 3.245114962203051e-06,
+      "loss": 0.5711,
+      "step": 14484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6225977922171109,
+      "learning_rate": 3.244551532005641e-06,
+      "loss": 0.4728,
+      "step": 14485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8011667112555338,
+      "learning_rate": 3.243988149024641e-06,
+      "loss": 0.6001,
+      "step": 14486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.431622363793733,
+      "learning_rate": 3.2434248132718487e-06,
+      "loss": 0.3658,
+      "step": 14487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9714211786569908,
+      "learning_rate": 3.2428615247590577e-06,
+      "loss": 0.5246,
+      "step": 14488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4890821759035548,
+      "learning_rate": 3.2422982834980576e-06,
+      "loss": 0.3919,
+      "step": 14489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9045138428705917,
+      "learning_rate": 3.24173508950064e-06,
+      "loss": 0.5877,
+      "step": 14490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.812147075422061,
+      "learning_rate": 3.241171942778596e-06,
+      "loss": 0.4397,
+      "step": 14491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9230112017836354,
+      "learning_rate": 3.240608843343716e-06,
+      "loss": 0.4378,
+      "step": 14492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9035252945733754,
+      "learning_rate": 3.240045791207785e-06,
+      "loss": 0.4297,
+      "step": 14493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6590032699046096,
+      "learning_rate": 3.2394827863825913e-06,
+      "loss": 0.4851,
+      "step": 14494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.935371592751589,
+      "learning_rate": 3.2389198288799228e-06,
+      "loss": 0.5154,
+      "step": 14495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6706733573974684,
+      "learning_rate": 3.2383569187115623e-06,
+      "loss": 0.4792,
+      "step": 14496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0892486188085364,
+      "learning_rate": 3.237794055889295e-06,
+      "loss": 0.5346,
+      "step": 14497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.682751895367872,
+      "learning_rate": 3.237231240424905e-06,
+      "loss": 0.567,
+      "step": 14498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8138832870955577,
+      "learning_rate": 3.2366684723301724e-06,
+      "loss": 0.4978,
+      "step": 14499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.836127952950914,
+      "learning_rate": 3.2361057516168776e-06,
+      "loss": 0.5051,
+      "step": 14500
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5024067759513855,
+      "eval_runtime": 68.054,
+      "eval_samples_per_second": 50.107,
+      "eval_steps_per_second": 1.572,
+      "step": 14500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7217633161672115,
+      "learning_rate": 3.2355430782968062e-06,
+      "loss": 0.4009,
+      "step": 14501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.35882982828647,
+      "learning_rate": 3.2349804523817317e-06,
+      "loss": 0.2946,
+      "step": 14502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8825692777174967,
+      "learning_rate": 3.2344178738834353e-06,
+      "loss": 0.5396,
+      "step": 14503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6028184195201132,
+      "learning_rate": 3.2338553428136955e-06,
+      "loss": 0.512,
+      "step": 14504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4060022084019848,
+      "learning_rate": 3.233292859184284e-06,
+      "loss": 0.4467,
+      "step": 14505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8695319297000381,
+      "learning_rate": 3.232730423006979e-06,
+      "loss": 0.386,
+      "step": 14506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1878444314942422,
+      "learning_rate": 3.232168034293556e-06,
+      "loss": 0.5868,
+      "step": 14507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7805082071622829,
+      "learning_rate": 3.231605693055785e-06,
+      "loss": 0.4785,
+      "step": 14508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6677417870654732,
+      "learning_rate": 3.2310433993054397e-06,
+      "loss": 0.6026,
+      "step": 14509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.035127199377684,
+      "learning_rate": 3.2304811530542935e-06,
+      "loss": 0.4551,
+      "step": 14510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8313922127759052,
+      "learning_rate": 3.2299189543141126e-06,
+      "loss": 0.5641,
+      "step": 14511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.973874348913112,
+      "learning_rate": 3.229356803096668e-06,
+      "loss": 0.474,
+      "step": 14512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6669988298151142,
+      "learning_rate": 3.2287946994137297e-06,
+      "loss": 0.4533,
+      "step": 14513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4169045638066025,
+      "learning_rate": 3.2282326432770643e-06,
+      "loss": 0.6513,
+      "step": 14514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5255774756177922,
+      "learning_rate": 3.2276706346984367e-06,
+      "loss": 0.3642,
+      "step": 14515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9056778654181505,
+      "learning_rate": 3.227108673689612e-06,
+      "loss": 0.5388,
+      "step": 14516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5763963265094934,
+      "learning_rate": 3.226546760262357e-06,
+      "loss": 0.4186,
+      "step": 14517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.650256015012313,
+      "learning_rate": 3.2259848944284332e-06,
+      "loss": 0.5374,
+      "step": 14518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5954604221412618,
+      "learning_rate": 3.225423076199602e-06,
+      "loss": 0.3993,
+      "step": 14519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9292010752195945,
+      "learning_rate": 3.224861305587628e-06,
+      "loss": 0.4268,
+      "step": 14520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.503271168594929,
+      "learning_rate": 3.224299582604268e-06,
+      "loss": 0.4952,
+      "step": 14521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.475548929985385,
+      "learning_rate": 3.223737907261282e-06,
+      "loss": 0.5724,
+      "step": 14522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8638191774096344,
+      "learning_rate": 3.2231762795704313e-06,
+      "loss": 0.4638,
+      "step": 14523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5255063263371453,
+      "learning_rate": 3.2226146995434685e-06,
+      "loss": 0.412,
+      "step": 14524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9473800963649757,
+      "learning_rate": 3.222053167192153e-06,
+      "loss": 0.5261,
+      "step": 14525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7993751060953627,
+      "learning_rate": 3.2214916825282403e-06,
+      "loss": 0.526,
+      "step": 14526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5630834020557374,
+      "learning_rate": 3.220930245563485e-06,
+      "loss": 0.3673,
+      "step": 14527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7025148267010595,
+      "learning_rate": 3.220368856309638e-06,
+      "loss": 0.6126,
+      "step": 14528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6997016796335838,
+      "learning_rate": 3.2198075147784525e-06,
+      "loss": 0.414,
+      "step": 14529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7810620225763778,
+      "learning_rate": 3.2192462209816824e-06,
+      "loss": 0.5493,
+      "step": 14530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7825150148291946,
+      "learning_rate": 3.2186849749310744e-06,
+      "loss": 0.5674,
+      "step": 14531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5286336841459633,
+      "learning_rate": 3.2181237766383795e-06,
+      "loss": 0.479,
+      "step": 14532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4934414435530885,
+      "learning_rate": 3.217562626115347e-06,
+      "loss": 0.3683,
+      "step": 14533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.717689516224214,
+      "learning_rate": 3.2170015233737228e-06,
+      "loss": 0.428,
+      "step": 14534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5751094054363153,
+      "learning_rate": 3.2164404684252525e-06,
+      "loss": 0.583,
+      "step": 14535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.991382788688519,
+      "learning_rate": 3.215879461281685e-06,
+      "loss": 0.5405,
+      "step": 14536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4573005203585152,
+      "learning_rate": 3.2153185019547608e-06,
+      "loss": 0.412,
+      "step": 14537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.973024685608145,
+      "learning_rate": 3.2147575904562245e-06,
+      "loss": 0.5729,
+      "step": 14538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.789978062075768,
+      "learning_rate": 3.214196726797818e-06,
+      "loss": 0.5394,
+      "step": 14539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4375095643554194,
+      "learning_rate": 3.213635910991286e-06,
+      "loss": 0.3655,
+      "step": 14540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0068197126416445,
+      "learning_rate": 3.213075143048363e-06,
+      "loss": 0.6658,
+      "step": 14541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7261296860971531,
+      "learning_rate": 3.2125144229807923e-06,
+      "loss": 0.3727,
+      "step": 14542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.516503684397348,
+      "learning_rate": 3.211953750800314e-06,
+      "loss": 0.3279,
+      "step": 14543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1704362974692617,
+      "learning_rate": 3.211393126518659e-06,
+      "loss": 0.4915,
+      "step": 14544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7086412835712343,
+      "learning_rate": 3.210832550147569e-06,
+      "loss": 0.4477,
+      "step": 14545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8154514361458045,
+      "learning_rate": 3.210272021698779e-06,
+      "loss": 0.4959,
+      "step": 14546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8598094622747938,
+      "learning_rate": 3.2097115411840202e-06,
+      "loss": 0.5999,
+      "step": 14547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5378657376899676,
+      "learning_rate": 3.209151108615028e-06,
+      "loss": 0.376,
+      "step": 14548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6431462904245266,
+      "learning_rate": 3.2085907240035354e-06,
+      "loss": 0.5024,
+      "step": 14549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5176913604261293,
+      "learning_rate": 3.2080303873612695e-06,
+      "loss": 0.3977,
+      "step": 14550
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5022367238998413,
+      "eval_runtime": 68.027,
+      "eval_samples_per_second": 50.127,
+      "eval_steps_per_second": 1.573,
+      "step": 14550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9807655749912354,
+      "learning_rate": 3.2074700986999664e-06,
+      "loss": 0.6372,
+      "step": 14551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.032339846143286,
+      "learning_rate": 3.206909858031354e-06,
+      "loss": 0.5517,
+      "step": 14552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5700952199122968,
+      "learning_rate": 3.2063496653671576e-06,
+      "loss": 0.5569,
+      "step": 14553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7759981855380131,
+      "learning_rate": 3.2057895207191064e-06,
+      "loss": 0.5951,
+      "step": 14554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.532992322377769,
+      "learning_rate": 3.2052294240989286e-06,
+      "loss": 0.4235,
+      "step": 14555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.641835829902551,
+      "learning_rate": 3.2046693755183456e-06,
+      "loss": 0.4098,
+      "step": 14556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5563660042508105,
+      "learning_rate": 3.2041093749890828e-06,
+      "loss": 0.506,
+      "step": 14557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7009546114628635,
+      "learning_rate": 3.2035494225228652e-06,
+      "loss": 0.3972,
+      "step": 14558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4394395417174943,
+      "learning_rate": 3.2029895181314154e-06,
+      "loss": 0.2788,
+      "step": 14559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7049750747669041,
+      "learning_rate": 3.2024296618264517e-06,
+      "loss": 0.4879,
+      "step": 14560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6217028632409844,
+      "learning_rate": 3.2018698536196957e-06,
+      "loss": 0.4494,
+      "step": 14561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6927368273398449,
+      "learning_rate": 3.2013100935228686e-06,
+      "loss": 0.4943,
+      "step": 14562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2926791933560033,
+      "learning_rate": 3.2007503815476854e-06,
+      "loss": 0.7132,
+      "step": 14563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0123259555020376,
+      "learning_rate": 3.200190717705865e-06,
+      "loss": 0.5272,
+      "step": 14564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.008556434115587,
+      "learning_rate": 3.1996311020091253e-06,
+      "loss": 0.4937,
+      "step": 14565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8964489777619402,
+      "learning_rate": 3.199071534469178e-06,
+      "loss": 0.5413,
+      "step": 14566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8363946337724284,
+      "learning_rate": 3.198512015097738e-06,
+      "loss": 0.4218,
+      "step": 14567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.454599123001584,
+      "learning_rate": 3.197952543906522e-06,
+      "loss": 0.3596,
+      "step": 14568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9339079266021881,
+      "learning_rate": 3.197393120907237e-06,
+      "loss": 0.6225,
+      "step": 14569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7795592849739703,
+      "learning_rate": 3.1968337461115986e-06,
+      "loss": 0.5102,
+      "step": 14570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6770815723892512,
+      "learning_rate": 3.1962744195313135e-06,
+      "loss": 0.5261,
+      "step": 14571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5460961393088426,
+      "learning_rate": 3.195715141178095e-06,
+      "loss": 0.5516,
+      "step": 14572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.065517173094266,
+      "learning_rate": 3.1951559110636467e-06,
+      "loss": 0.4944,
+      "step": 14573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6526073929487284,
+      "learning_rate": 3.194596729199678e-06,
+      "loss": 0.5284,
+      "step": 14574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7981332442941533,
+      "learning_rate": 3.1940375955978968e-06,
+      "loss": 0.5233,
+      "step": 14575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9272141955404445,
+      "learning_rate": 3.1934785102700046e-06,
+      "loss": 0.4633,
+      "step": 14576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7358621281333397,
+      "learning_rate": 3.1929194732277068e-06,
+      "loss": 0.4582,
+      "step": 14577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9585582285139693,
+      "learning_rate": 3.1923604844827083e-06,
+      "loss": 0.6134,
+      "step": 14578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9175508789553872,
+      "learning_rate": 3.1918015440467085e-06,
+      "loss": 0.532,
+      "step": 14579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.691352492071988,
+      "learning_rate": 3.19124265193141e-06,
+      "loss": 0.4089,
+      "step": 14580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9205371433827036,
+      "learning_rate": 3.1906838081485136e-06,
+      "loss": 0.5923,
+      "step": 14581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6571062293240464,
+      "learning_rate": 3.1901250127097156e-06,
+      "loss": 0.566,
+      "step": 14582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7635790266151816,
+      "learning_rate": 3.1895662656267157e-06,
+      "loss": 0.6519,
+      "step": 14583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6661477212473503,
+      "learning_rate": 3.189007566911212e-06,
+      "loss": 0.4619,
+      "step": 14584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9091780375003005,
+      "learning_rate": 3.1884489165749e-06,
+      "loss": 0.4981,
+      "step": 14585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6864310806711345,
+      "learning_rate": 3.1878903146294728e-06,
+      "loss": 0.5193,
+      "step": 14586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1108841166716887,
+      "learning_rate": 3.187331761086626e-06,
+      "loss": 0.6834,
+      "step": 14587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6685182391299365,
+      "learning_rate": 3.1867732559580538e-06,
+      "loss": 0.4339,
+      "step": 14588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.345472617542511,
+      "learning_rate": 3.186214799255445e-06,
+      "loss": 0.4887,
+      "step": 14589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8441614053409916,
+      "learning_rate": 3.1856563909904926e-06,
+      "loss": 0.5079,
+      "step": 14590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7743839602899445,
+      "learning_rate": 3.1850980311748873e-06,
+      "loss": 0.4227,
+      "step": 14591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7337545538140748,
+      "learning_rate": 3.1845397198203154e-06,
+      "loss": 0.5017,
+      "step": 14592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9696771351122029,
+      "learning_rate": 3.1839814569384663e-06,
+      "loss": 0.5221,
+      "step": 14593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.013486540348913,
+      "learning_rate": 3.1834232425410284e-06,
+      "loss": 0.537,
+      "step": 14594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6385766160883708,
+      "learning_rate": 3.182865076639685e-06,
+      "loss": 0.3989,
+      "step": 14595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7106074726897527,
+      "learning_rate": 3.1823069592461215e-06,
+      "loss": 0.4766,
+      "step": 14596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6199018811886845,
+      "learning_rate": 3.1817488903720216e-06,
+      "loss": 0.4015,
+      "step": 14597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5987178889573899,
+      "learning_rate": 3.1811908700290717e-06,
+      "loss": 0.3853,
+      "step": 14598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2159242787671642,
+      "learning_rate": 3.1806328982289464e-06,
+      "loss": 0.5531,
+      "step": 14599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.557576288170678,
+      "learning_rate": 3.180074974983335e-06,
+      "loss": 0.4496,
+      "step": 14600
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5016937851905823,
+      "eval_runtime": 67.9921,
+      "eval_samples_per_second": 50.153,
+      "eval_steps_per_second": 1.574,
+      "step": 14600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6862469912787474,
+      "learning_rate": 3.17951710030391e-06,
+      "loss": 0.4169,
+      "step": 14601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.430734786362349,
+      "learning_rate": 3.1789592742023543e-06,
+      "loss": 0.4041,
+      "step": 14602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7260991140726385,
+      "learning_rate": 3.1784014966903453e-06,
+      "loss": 0.5638,
+      "step": 14603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7263341269913375,
+      "learning_rate": 3.1778437677795587e-06,
+      "loss": 0.512,
+      "step": 14604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5563544994263294,
+      "learning_rate": 3.177286087481669e-06,
+      "loss": 0.4561,
+      "step": 14605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9163922848770656,
+      "learning_rate": 3.176728455808353e-06,
+      "loss": 0.5962,
+      "step": 14606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6055655236002417,
+      "learning_rate": 3.1761708727712857e-06,
+      "loss": 0.4372,
+      "step": 14607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7810179053798563,
+      "learning_rate": 3.175613338382136e-06,
+      "loss": 0.3777,
+      "step": 14608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7373128262445885,
+      "learning_rate": 3.1750558526525773e-06,
+      "loss": 0.3179,
+      "step": 14609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125511760634857,
+      "learning_rate": 3.1744984155942814e-06,
+      "loss": 0.599,
+      "step": 14610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.712708573086861,
+      "learning_rate": 3.1739410272189166e-06,
+      "loss": 0.4703,
+      "step": 14611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5278392291103324,
+      "learning_rate": 3.1733836875381495e-06,
+      "loss": 0.4273,
+      "step": 14612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6814341975132139,
+      "learning_rate": 3.172826396563653e-06,
+      "loss": 0.5311,
+      "step": 14613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6369253849715601,
+      "learning_rate": 3.1722691543070882e-06,
+      "loss": 0.3633,
+      "step": 14614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.238206186913203,
+      "learning_rate": 3.1717119607801227e-06,
+      "loss": 0.5903,
+      "step": 14615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5652923492229676,
+      "learning_rate": 3.171154815994423e-06,
+      "loss": 0.4176,
+      "step": 14616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5318690105704604,
+      "learning_rate": 3.170597719961649e-06,
+      "loss": 0.4547,
+      "step": 14617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8126648897014463,
+      "learning_rate": 3.1700406726934653e-06,
+      "loss": 0.4691,
+      "step": 14618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3925326658890878,
+      "learning_rate": 3.169483674201532e-06,
+      "loss": 0.4643,
+      "step": 14619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.100619389663311,
+      "learning_rate": 3.168926724497513e-06,
+      "loss": 0.473,
+      "step": 14620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9832813839187193,
+      "learning_rate": 3.168369823593063e-06,
+      "loss": 0.622,
+      "step": 14621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2426958497478413,
+      "learning_rate": 3.1678129714998427e-06,
+      "loss": 0.5935,
+      "step": 14622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.816868232544495,
+      "learning_rate": 3.167256168229511e-06,
+      "loss": 0.4941,
+      "step": 14623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5767716852672633,
+      "learning_rate": 3.1666994137937207e-06,
+      "loss": 0.4947,
+      "step": 14624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.12791978966324,
+      "learning_rate": 3.166142708204129e-06,
+      "loss": 0.6045,
+      "step": 14625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4207281384939798,
+      "learning_rate": 3.1655860514723925e-06,
+      "loss": 0.455,
+      "step": 14626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8498228554639504,
+      "learning_rate": 3.1650294436101593e-06,
+      "loss": 0.441,
+      "step": 14627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6982886927178216,
+      "learning_rate": 3.164472884629084e-06,
+      "loss": 0.4108,
+      "step": 14628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7686998063038935,
+      "learning_rate": 3.1639163745408206e-06,
+      "loss": 0.3689,
+      "step": 14629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.030931687075638,
+      "learning_rate": 3.1633599133570157e-06,
+      "loss": 0.6371,
+      "step": 14630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7311262853421316,
+      "learning_rate": 3.1628035010893187e-06,
+      "loss": 0.5357,
+      "step": 14631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9564999894159805,
+      "learning_rate": 3.1622471377493796e-06,
+      "loss": 0.5239,
+      "step": 14632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4412520514353941,
+      "learning_rate": 3.1616908233488463e-06,
+      "loss": 0.4296,
+      "step": 14633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.023849860093811,
+      "learning_rate": 3.161134557899361e-06,
+      "loss": 0.587,
+      "step": 14634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9388055413960104,
+      "learning_rate": 3.1605783414125707e-06,
+      "loss": 0.5283,
+      "step": 14635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.412846730056515,
+      "learning_rate": 3.1600221739001217e-06,
+      "loss": 0.4138,
+      "step": 14636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5535055334301382,
+      "learning_rate": 3.1594660553736535e-06,
+      "loss": 0.4729,
+      "step": 14637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5875111864909373,
+      "learning_rate": 3.1589099858448104e-06,
+      "loss": 0.4867,
+      "step": 14638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6508208688506922,
+      "learning_rate": 3.158353965325234e-06,
+      "loss": 0.3731,
+      "step": 14639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3721838050590245,
+      "learning_rate": 3.15779799382656e-06,
+      "loss": 0.397,
+      "step": 14640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5268742027077487,
+      "learning_rate": 3.1572420713604314e-06,
+      "loss": 0.3605,
+      "step": 14641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6165120339076895,
+      "learning_rate": 3.156686197938485e-06,
+      "loss": 0.5347,
+      "step": 14642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.728767387809571,
+      "learning_rate": 3.156130373572358e-06,
+      "loss": 0.4274,
+      "step": 14643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6094023554541173,
+      "learning_rate": 3.155574598273685e-06,
+      "loss": 0.3955,
+      "step": 14644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9891552888297237,
+      "learning_rate": 3.155018872054102e-06,
+      "loss": 0.4821,
+      "step": 14645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8333431939263938,
+      "learning_rate": 3.1544631949252435e-06,
+      "loss": 0.3769,
+      "step": 14646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7245575121549845,
+      "learning_rate": 3.1539075668987398e-06,
+      "loss": 0.6206,
+      "step": 14647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5983521939487042,
+      "learning_rate": 3.153351987986224e-06,
+      "loss": 0.4242,
+      "step": 14648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2834786936959315,
+      "learning_rate": 3.1527964581993265e-06,
+      "loss": 0.3352,
+      "step": 14649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7906662430014813,
+      "learning_rate": 3.152240977549677e-06,
+      "loss": 0.4815,
+      "step": 14650
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5012186765670776,
+      "eval_runtime": 67.8916,
+      "eval_samples_per_second": 50.227,
+      "eval_steps_per_second": 1.576,
+      "step": 14650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6545649645826461,
+      "learning_rate": 3.151685546048905e-06,
+      "loss": 0.6547,
+      "step": 14651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0528937683180146,
+      "learning_rate": 3.151130163708639e-06,
+      "loss": 0.4512,
+      "step": 14652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7718927522370385,
+      "learning_rate": 3.1505748305405022e-06,
+      "loss": 0.4319,
+      "step": 14653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9062174315732312,
+      "learning_rate": 3.1500195465561223e-06,
+      "loss": 0.554,
+      "step": 14654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0445360702514064,
+      "learning_rate": 3.149464311767124e-06,
+      "loss": 0.5458,
+      "step": 14655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8261648653982288,
+      "learning_rate": 3.148909126185131e-06,
+      "loss": 0.4648,
+      "step": 14656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4099754075597934,
+      "learning_rate": 3.148353989821763e-06,
+      "loss": 0.3704,
+      "step": 14657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4866602746438518,
+      "learning_rate": 3.147798902688647e-06,
+      "loss": 0.3955,
+      "step": 14658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5639446885138646,
+      "learning_rate": 3.147243864797397e-06,
+      "loss": 0.3542,
+      "step": 14659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.240115295193601,
+      "learning_rate": 3.146688876159636e-06,
+      "loss": 0.5291,
+      "step": 14660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4997282464283836,
+      "learning_rate": 3.1461339367869827e-06,
+      "loss": 0.3584,
+      "step": 14661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.622284804556663,
+      "learning_rate": 3.145579046691052e-06,
+      "loss": 0.3935,
+      "step": 14662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6565981333381792,
+      "learning_rate": 3.1450242058834613e-06,
+      "loss": 0.619,
+      "step": 14663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6171714258210301,
+      "learning_rate": 3.144469414375827e-06,
+      "loss": 0.4501,
+      "step": 14664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9172672103506163,
+      "learning_rate": 3.143914672179763e-06,
+      "loss": 0.542,
+      "step": 14665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7458410499207324,
+      "learning_rate": 3.1433599793068813e-06,
+      "loss": 0.6362,
+      "step": 14666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6846343998200473,
+      "learning_rate": 3.1428053357687944e-06,
+      "loss": 0.4344,
+      "step": 14667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9084733180718851,
+      "learning_rate": 3.142250741577115e-06,
+      "loss": 0.5364,
+      "step": 14668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.163771969335658,
+      "learning_rate": 3.1416961967434507e-06,
+      "loss": 0.5899,
+      "step": 14669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8468553943126966,
+      "learning_rate": 3.1411417012794115e-06,
+      "loss": 0.428,
+      "step": 14670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8499478373859994,
+      "learning_rate": 3.1405872551966066e-06,
+      "loss": 0.5941,
+      "step": 14671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7203148744478762,
+      "learning_rate": 3.140032858506641e-06,
+      "loss": 0.4393,
+      "step": 14672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8318746106304276,
+      "learning_rate": 3.1394785112211224e-06,
+      "loss": 0.5311,
+      "step": 14673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9028157755841577,
+      "learning_rate": 3.1389242133516557e-06,
+      "loss": 0.5229,
+      "step": 14674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6811780143585504,
+      "learning_rate": 3.1383699649098428e-06,
+      "loss": 0.5952,
+      "step": 14675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6118786801702636,
+      "learning_rate": 3.1378157659072872e-06,
+      "loss": 0.5237,
+      "step": 14676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.964709165875142,
+      "learning_rate": 3.1372616163555926e-06,
+      "loss": 0.546,
+      "step": 14677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8614162769588027,
+      "learning_rate": 3.1367075162663585e-06,
+      "loss": 0.5286,
+      "step": 14678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7574383095699266,
+      "learning_rate": 3.136153465651184e-06,
+      "loss": 0.4639,
+      "step": 14679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6907596309594706,
+      "learning_rate": 3.135599464521669e-06,
+      "loss": 0.3633,
+      "step": 14680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2263577013470583,
+      "learning_rate": 3.135045512889411e-06,
+      "loss": 0.6154,
+      "step": 14681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5846062440294544,
+      "learning_rate": 3.134491610766005e-06,
+      "loss": 0.4937,
+      "step": 14682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7930545844280172,
+      "learning_rate": 3.133937758163048e-06,
+      "loss": 0.5022,
+      "step": 14683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8478034966427899,
+      "learning_rate": 3.1333839550921363e-06,
+      "loss": 0.6102,
+      "step": 14684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6505419908831698,
+      "learning_rate": 3.1328302015648597e-06,
+      "loss": 0.4343,
+      "step": 14685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.301179136120949,
+      "learning_rate": 3.1322764975928128e-06,
+      "loss": 0.4188,
+      "step": 14686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6918958420500225,
+      "learning_rate": 3.1317228431875877e-06,
+      "loss": 0.4509,
+      "step": 14687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7489284608837878,
+      "learning_rate": 3.1311692383607727e-06,
+      "loss": 0.5178,
+      "step": 14688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6737778486629455,
+      "learning_rate": 3.1306156831239586e-06,
+      "loss": 0.4991,
+      "step": 14689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6123703087021348,
+      "learning_rate": 3.130062177488733e-06,
+      "loss": 0.6061,
+      "step": 14690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1326137059939136,
+      "learning_rate": 3.1295087214666864e-06,
+      "loss": 0.5334,
+      "step": 14691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8047432554358025,
+      "learning_rate": 3.1289553150693998e-06,
+      "loss": 0.5057,
+      "step": 14692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7463010220405968,
+      "learning_rate": 3.128401958308461e-06,
+      "loss": 0.3732,
+      "step": 14693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7133054546487425,
+      "learning_rate": 3.1278486511954553e-06,
+      "loss": 0.495,
+      "step": 14694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.438227525076291,
+      "learning_rate": 3.127295393741964e-06,
+      "loss": 0.3661,
+      "step": 14695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5812408033394214,
+      "learning_rate": 3.1267421859595694e-06,
+      "loss": 0.5875,
+      "step": 14696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3672861113646209,
+      "learning_rate": 3.126189027859853e-06,
+      "loss": 0.391,
+      "step": 14697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.028825164740782,
+      "learning_rate": 3.1256359194543952e-06,
+      "loss": 0.5633,
+      "step": 14698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.45549121775091,
+      "learning_rate": 3.1250828607547746e-06,
+      "loss": 0.5962,
+      "step": 14699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.162365296275146,
+      "learning_rate": 3.12452985177257e-06,
+      "loss": 0.5625,
+      "step": 14700
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.5013176798820496,
+      "eval_runtime": 67.989,
+      "eval_samples_per_second": 50.155,
+      "eval_steps_per_second": 1.574,
+      "step": 14700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.421119881499164,
+      "learning_rate": 3.123976892519357e-06,
+      "loss": 0.3836,
+      "step": 14701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6530410095637762,
+      "learning_rate": 3.123423983006712e-06,
+      "loss": 0.4736,
+      "step": 14702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9710662010486442,
+      "learning_rate": 3.1228711232462106e-06,
+      "loss": 0.6043,
+      "step": 14703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4837247684619765,
+      "learning_rate": 3.1223183132494245e-06,
+      "loss": 0.4733,
+      "step": 14704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6676666737267425,
+      "learning_rate": 3.1217655530279277e-06,
+      "loss": 0.5504,
+      "step": 14705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8329306095773634,
+      "learning_rate": 3.121212842593293e-06,
+      "loss": 0.5668,
+      "step": 14706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.794460415280258,
+      "learning_rate": 3.120660181957089e-06,
+      "loss": 0.4731,
+      "step": 14707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.995097856786068,
+      "learning_rate": 3.120107571130886e-06,
+      "loss": 0.684,
+      "step": 14708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5429366335439358,
+      "learning_rate": 3.1195550101262543e-06,
+      "loss": 0.4264,
+      "step": 14709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3308882520242302,
+      "learning_rate": 3.1190024989547587e-06,
+      "loss": 0.3973,
+      "step": 14710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0066333069714797,
+      "learning_rate": 3.118450037627966e-06,
+      "loss": 0.5402,
+      "step": 14711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6938049133539155,
+      "learning_rate": 3.1178976261574434e-06,
+      "loss": 0.5044,
+      "step": 14712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3510180373718492,
+      "learning_rate": 3.117345264554756e-06,
+      "loss": 0.3908,
+      "step": 14713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.485086261012198,
+      "learning_rate": 3.1167929528314634e-06,
+      "loss": 0.3707,
+      "step": 14714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7659849946127149,
+      "learning_rate": 3.1162406909991303e-06,
+      "loss": 0.5818,
+      "step": 14715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8809697664482683,
+      "learning_rate": 3.1156884790693194e-06,
+      "loss": 0.5734,
+      "step": 14716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.879040253523967,
+      "learning_rate": 3.1151363170535875e-06,
+      "loss": 0.3333,
+      "step": 14717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9099957954187299,
+      "learning_rate": 3.1145842049634954e-06,
+      "loss": 0.4423,
+      "step": 14718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8252074262515698,
+      "learning_rate": 3.1140321428106034e-06,
+      "loss": 0.6028,
+      "step": 14719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.796705495803894,
+      "learning_rate": 3.113480130606464e-06,
+      "loss": 0.743,
+      "step": 14720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.969963341295638,
+      "learning_rate": 3.112928168362637e-06,
+      "loss": 0.5901,
+      "step": 14721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6454588860010408,
+      "learning_rate": 3.1123762560906756e-06,
+      "loss": 0.3429,
+      "step": 14722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.683822414474894,
+      "learning_rate": 3.1118243938021354e-06,
+      "loss": 0.4278,
+      "step": 14723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.555080358044956,
+      "learning_rate": 3.1112725815085676e-06,
+      "loss": 0.3889,
+      "step": 14724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.05270366128161,
+      "learning_rate": 3.110720819221523e-06,
+      "loss": 0.4804,
+      "step": 14725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.196858825835615,
+      "learning_rate": 3.1101691069525574e-06,
+      "loss": 0.5013,
+      "step": 14726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0521463959961013,
+      "learning_rate": 3.1096174447132145e-06,
+      "loss": 0.4661,
+      "step": 14727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9398280786895266,
+      "learning_rate": 3.1090658325150457e-06,
+      "loss": 0.4491,
+      "step": 14728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4474066131816414,
+      "learning_rate": 3.1085142703696e-06,
+      "loss": 0.4708,
+      "step": 14729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6967354708606412,
+      "learning_rate": 3.107962758288422e-06,
+      "loss": 0.4309,
+      "step": 14730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.127586668430737,
+      "learning_rate": 3.1074112962830572e-06,
+      "loss": 0.6751,
+      "step": 14731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7349422105261552,
+      "learning_rate": 3.1068598843650526e-06,
+      "loss": 0.4466,
+      "step": 14732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6449851866488239,
+      "learning_rate": 3.1063085225459487e-06,
+      "loss": 0.3785,
+      "step": 14733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6511231155795634,
+      "learning_rate": 3.105757210837289e-06,
+      "loss": 0.5034,
+      "step": 14734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7112566849153537,
+      "learning_rate": 3.1052059492506144e-06,
+      "loss": 0.4861,
+      "step": 14735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.963677770984385,
+      "learning_rate": 3.1046547377974677e-06,
+      "loss": 0.512,
+      "step": 14736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5603925995750036,
+      "learning_rate": 3.1041035764893846e-06,
+      "loss": 0.3471,
+      "step": 14737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6341309097620238,
+      "learning_rate": 3.1035524653379052e-06,
+      "loss": 0.4735,
+      "step": 14738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.701301722810108,
+      "learning_rate": 3.1030014043545673e-06,
+      "loss": 0.4988,
+      "step": 14739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8364750843548105,
+      "learning_rate": 3.1024503935509055e-06,
+      "loss": 0.507,
+      "step": 14740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4653104098830971,
+      "learning_rate": 3.1018994329384557e-06,
+      "loss": 0.3925,
+      "step": 14741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6363246525042208,
+      "learning_rate": 3.101348522528752e-06,
+      "loss": 0.5106,
+      "step": 14742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8484321765582477,
+      "learning_rate": 3.1007976623333263e-06,
+      "loss": 0.4807,
+      "step": 14743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5943056618043572,
+      "learning_rate": 3.100246852363712e-06,
+      "loss": 0.5316,
+      "step": 14744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8872698992987333,
+      "learning_rate": 3.09969609263144e-06,
+      "loss": 0.6043,
+      "step": 14745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4359595565593912,
+      "learning_rate": 3.0991453831480377e-06,
+      "loss": 0.3597,
+      "step": 14746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.505002427961518,
+      "learning_rate": 3.098594723925035e-06,
+      "loss": 0.4281,
+      "step": 14747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5132320694979602,
+      "learning_rate": 3.0980441149739625e-06,
+      "loss": 0.4066,
+      "step": 14748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5785958671299785,
+      "learning_rate": 3.0974935563063433e-06,
+      "loss": 0.381,
+      "step": 14749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5858414030218517,
+      "learning_rate": 3.0969430479337042e-06,
+      "loss": 0.4203,
+      "step": 14750
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.5004783272743225,
+      "eval_runtime": 67.9684,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 14750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8392710809263195,
+      "learning_rate": 3.096392589867571e-06,
+      "loss": 0.5311,
+      "step": 14751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.698079104090396,
+      "learning_rate": 3.095842182119465e-06,
+      "loss": 0.4621,
+      "step": 14752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6187983478835717,
+      "learning_rate": 3.0952918247009094e-06,
+      "loss": 0.3905,
+      "step": 14753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7675905251801052,
+      "learning_rate": 3.0947415176234276e-06,
+      "loss": 0.4518,
+      "step": 14754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9050672355904783,
+      "learning_rate": 3.0941912608985364e-06,
+      "loss": 0.568,
+      "step": 14755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8540177896473575,
+      "learning_rate": 3.093641054537757e-06,
+      "loss": 0.5547,
+      "step": 14756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6904358842817713,
+      "learning_rate": 3.093090898552607e-06,
+      "loss": 0.4831,
+      "step": 14757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0283842432484565,
+      "learning_rate": 3.092540792954606e-06,
+      "loss": 0.6458,
+      "step": 14758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.699429200679183,
+      "learning_rate": 3.091990737755267e-06,
+      "loss": 0.466,
+      "step": 14759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0021277374786077,
+      "learning_rate": 3.091440732966105e-06,
+      "loss": 0.5809,
+      "step": 14760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5363808876348983,
+      "learning_rate": 3.0908907785986377e-06,
+      "loss": 0.3674,
+      "step": 14761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7840751728051039,
+      "learning_rate": 3.090340874664374e-06,
+      "loss": 0.4283,
+      "step": 14762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2128688601624824,
+      "learning_rate": 3.0897910211748276e-06,
+      "loss": 0.5936,
+      "step": 14763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9543270893007727,
+      "learning_rate": 3.0892412181415106e-06,
+      "loss": 0.4935,
+      "step": 14764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9138195055242802,
+      "learning_rate": 3.088691465575929e-06,
+      "loss": 0.5881,
+      "step": 14765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.982657718291287,
+      "learning_rate": 3.0881417634895945e-06,
+      "loss": 0.4465,
+      "step": 14766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.589478957544483,
+      "learning_rate": 3.0875921118940155e-06,
+      "loss": 0.5191,
+      "step": 14767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8405703319423667,
+      "learning_rate": 3.087042510800696e-06,
+      "loss": 0.5039,
+      "step": 14768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8152373698723343,
+      "learning_rate": 3.0864929602211424e-06,
+      "loss": 0.5772,
+      "step": 14769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7898341451512454,
+      "learning_rate": 3.0859434601668605e-06,
+      "loss": 0.5866,
+      "step": 14770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078894589659373,
+      "learning_rate": 3.085394010649353e-06,
+      "loss": 0.6728,
+      "step": 14771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3767211378644273,
+      "learning_rate": 3.084844611680121e-06,
+      "loss": 0.3655,
+      "step": 14772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1433811381970336,
+      "learning_rate": 3.084295263270668e-06,
+      "loss": 0.5993,
+      "step": 14773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.590731792906294,
+      "learning_rate": 3.083745965432494e-06,
+      "loss": 0.5213,
+      "step": 14774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7402109680641091,
+      "learning_rate": 3.083196718177095e-06,
+      "loss": 0.4845,
+      "step": 14775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6009029306995333,
+      "learning_rate": 3.0826475215159725e-06,
+      "loss": 0.5105,
+      "step": 14776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6088050327344507,
+      "learning_rate": 3.082098375460625e-06,
+      "loss": 0.5853,
+      "step": 14777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8365077565464445,
+      "learning_rate": 3.0815492800225435e-06,
+      "loss": 0.4402,
+      "step": 14778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6751013880660628,
+      "learning_rate": 3.0810002352132256e-06,
+      "loss": 0.516,
+      "step": 14779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.94163210707266,
+      "learning_rate": 3.0804512410441673e-06,
+      "loss": 0.4581,
+      "step": 14780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.530980968255496,
+      "learning_rate": 3.0799022975268576e-06,
+      "loss": 0.4721,
+      "step": 14781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5860665531493645,
+      "learning_rate": 3.0793534046727903e-06,
+      "loss": 0.4616,
+      "step": 14782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.65477116936959,
+      "learning_rate": 3.078804562493456e-06,
+      "loss": 0.513,
+      "step": 14783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.064084831368312,
+      "learning_rate": 3.0782557710003456e-06,
+      "loss": 0.5982,
+      "step": 14784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5025418047678671,
+      "learning_rate": 3.0777070302049448e-06,
+      "loss": 0.4106,
+      "step": 14785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4433079213514282,
+      "learning_rate": 3.0771583401187435e-06,
+      "loss": 0.3832,
+      "step": 14786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.853639938560243,
+      "learning_rate": 3.0766097007532286e-06,
+      "loss": 0.4464,
+      "step": 14787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0425273654537257,
+      "learning_rate": 3.0760611121198823e-06,
+      "loss": 0.4267,
+      "step": 14788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9008031407761141,
+      "learning_rate": 3.0755125742301915e-06,
+      "loss": 0.4793,
+      "step": 14789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0822481203968985,
+      "learning_rate": 3.074964087095641e-06,
+      "loss": 0.6753,
+      "step": 14790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5632661796834033,
+      "learning_rate": 3.074415650727709e-06,
+      "loss": 0.428,
+      "step": 14791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8941361477231706,
+      "learning_rate": 3.0738672651378792e-06,
+      "loss": 0.6068,
+      "step": 14792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6867849437974647,
+      "learning_rate": 3.073318930337633e-06,
+      "loss": 0.4262,
+      "step": 14793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.843770763400145,
+      "learning_rate": 3.0727706463384456e-06,
+      "loss": 0.6427,
+      "step": 14794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.017965034865008,
+      "learning_rate": 3.0722224131517975e-06,
+      "loss": 0.7203,
+      "step": 14795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6228079497115715,
+      "learning_rate": 3.0716742307891646e-06,
+      "loss": 0.5272,
+      "step": 14796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6678128543631365,
+      "learning_rate": 3.0711260992620233e-06,
+      "loss": 0.4752,
+      "step": 14797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.804957936329993,
+      "learning_rate": 3.0705780185818498e-06,
+      "loss": 0.5452,
+      "step": 14798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8837831621018621,
+      "learning_rate": 3.070029988760117e-06,
+      "loss": 0.5093,
+      "step": 14799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.707018801759939,
+      "learning_rate": 3.0694820098082954e-06,
+      "loss": 0.4589,
+      "step": 14800
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.499317467212677,
+      "eval_runtime": 67.969,
+      "eval_samples_per_second": 50.17,
+      "eval_steps_per_second": 1.574,
+      "step": 14800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.616170551266822,
+      "learning_rate": 3.0689340817378587e-06,
+      "loss": 0.4563,
+      "step": 14801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5497156760904316,
+      "learning_rate": 3.0683862045602776e-06,
+      "loss": 0.4826,
+      "step": 14802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.639287954991105,
+      "learning_rate": 3.0678383782870215e-06,
+      "loss": 0.5303,
+      "step": 14803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.107580960146617,
+      "learning_rate": 3.0672906029295567e-06,
+      "loss": 0.3914,
+      "step": 14804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7826375963979098,
+      "learning_rate": 3.0667428784993523e-06,
+      "loss": 0.5921,
+      "step": 14805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7545532006271294,
+      "learning_rate": 3.0661952050078764e-06,
+      "loss": 0.5013,
+      "step": 14806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5575424214694062,
+      "learning_rate": 3.065647582466591e-06,
+      "loss": 0.408,
+      "step": 14807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.355329647770258,
+      "learning_rate": 3.0651000108869607e-06,
+      "loss": 0.3497,
+      "step": 14808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4552493011375693,
+      "learning_rate": 3.0645524902804506e-06,
+      "loss": 0.4198,
+      "step": 14809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6773333343606693,
+      "learning_rate": 3.0640050206585205e-06,
+      "loss": 0.5582,
+      "step": 14810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.930232426824478,
+      "learning_rate": 3.0634576020326316e-06,
+      "loss": 0.5145,
+      "step": 14811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0659182803927236,
+      "learning_rate": 3.062910234414247e-06,
+      "loss": 0.6422,
+      "step": 14812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8744431463717128,
+      "learning_rate": 3.0623629178148204e-06,
+      "loss": 0.596,
+      "step": 14813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9694059056494448,
+      "learning_rate": 3.061815652245812e-06,
+      "loss": 0.3947,
+      "step": 14814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9290252055437391,
+      "learning_rate": 3.061268437718679e-06,
+      "loss": 0.4761,
+      "step": 14815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9502111343178934,
+      "learning_rate": 3.0607212742448777e-06,
+      "loss": 0.5596,
+      "step": 14816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.881341591415606,
+      "learning_rate": 3.0601741618358603e-06,
+      "loss": 0.4377,
+      "step": 14817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5455799910146508,
+      "learning_rate": 3.0596271005030808e-06,
+      "loss": 0.4334,
+      "step": 14818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9495808340066987,
+      "learning_rate": 3.059080090257995e-06,
+      "loss": 0.5448,
+      "step": 14819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.902161925995909,
+      "learning_rate": 3.0585331311120485e-06,
+      "loss": 0.5555,
+      "step": 14820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5683643033672687,
+      "learning_rate": 3.057986223076695e-06,
+      "loss": 0.4491,
+      "step": 14821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6309086487775983,
+      "learning_rate": 3.057439366163385e-06,
+      "loss": 0.4348,
+      "step": 14822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5493484097818309,
+      "learning_rate": 3.0568925603835624e-06,
+      "loss": 0.4358,
+      "step": 14823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.899872640633971,
+      "learning_rate": 3.0563458057486762e-06,
+      "loss": 0.5346,
+      "step": 14824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.84488062143379,
+      "learning_rate": 3.0557991022701755e-06,
+      "loss": 0.5122,
+      "step": 14825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.324258119768984,
+      "learning_rate": 3.0552524499595003e-06,
+      "loss": 0.6782,
+      "step": 14826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6625153802748993,
+      "learning_rate": 3.054705848828097e-06,
+      "loss": 0.6038,
+      "step": 14827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7767839494714834,
+      "learning_rate": 3.0541592988874075e-06,
+      "loss": 0.6096,
+      "step": 14828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8256933399725117,
+      "learning_rate": 3.053612800148875e-06,
+      "loss": 0.4178,
+      "step": 14829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.775938304127937,
+      "learning_rate": 3.0530663526239384e-06,
+      "loss": 0.542,
+      "step": 14830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5642232208026723,
+      "learning_rate": 3.0525199563240375e-06,
+      "loss": 0.449,
+      "step": 14831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4893093714321994,
+      "learning_rate": 3.051973611260613e-06,
+      "loss": 0.4346,
+      "step": 14832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9434048750278499,
+      "learning_rate": 3.051427317445099e-06,
+      "loss": 0.5715,
+      "step": 14833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3987998009991083,
+      "learning_rate": 3.0508810748889323e-06,
+      "loss": 0.3698,
+      "step": 14834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.830117022586634,
+      "learning_rate": 3.050334883603551e-06,
+      "loss": 0.5866,
+      "step": 14835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9160340104782816,
+      "learning_rate": 3.0497887436003866e-06,
+      "loss": 0.5119,
+      "step": 14836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.758913930389865,
+      "learning_rate": 3.0492426548908723e-06,
+      "loss": 0.4866,
+      "step": 14837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6709808904016081,
+      "learning_rate": 3.0486966174864426e-06,
+      "loss": 0.402,
+      "step": 14838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.31739868923038,
+      "learning_rate": 3.0481506313985244e-06,
+      "loss": 0.3962,
+      "step": 14839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5785364244196458,
+      "learning_rate": 3.04760469663855e-06,
+      "loss": 0.3605,
+      "step": 14840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6734787977469199,
+      "learning_rate": 3.047058813217949e-06,
+      "loss": 0.5724,
+      "step": 14841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2980395262727584,
+      "learning_rate": 3.0465129811481485e-06,
+      "loss": 0.5583,
+      "step": 14842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.042585533922152,
+      "learning_rate": 3.045967200440574e-06,
+      "loss": 0.4709,
+      "step": 14843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8453563651847855,
+      "learning_rate": 3.045421471106652e-06,
+      "loss": 0.6046,
+      "step": 14844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4823807600216812,
+      "learning_rate": 3.0448757931578065e-06,
+      "loss": 0.354,
+      "step": 14845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.862906323054011,
+      "learning_rate": 3.0443301666054614e-06,
+      "loss": 0.5715,
+      "step": 14846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7958165581610823,
+      "learning_rate": 3.043784591461041e-06,
+      "loss": 0.5144,
+      "step": 14847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9257551298732258,
+      "learning_rate": 3.0432390677359622e-06,
+      "loss": 0.482,
+      "step": 14848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7851439986716386,
+      "learning_rate": 3.042693595441648e-06,
+      "loss": 0.4931,
+      "step": 14849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5090461285364067,
+      "learning_rate": 3.0421481745895177e-06,
+      "loss": 0.4594,
+      "step": 14850
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.5000147819519043,
+      "eval_runtime": 67.9913,
+      "eval_samples_per_second": 50.153,
+      "eval_steps_per_second": 1.574,
+      "step": 14850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.631931807039776,
+      "learning_rate": 3.0416028051909896e-06,
+      "loss": 0.381,
+      "step": 14851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0743366289972145,
+      "learning_rate": 3.041057487257478e-06,
+      "loss": 0.569,
+      "step": 14852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8013672151846207,
+      "learning_rate": 3.040512220800401e-06,
+      "loss": 0.5834,
+      "step": 14853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5169127349072176,
+      "learning_rate": 3.039967005831175e-06,
+      "loss": 0.432,
+      "step": 14854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.211728926775914,
+      "learning_rate": 3.0394218423612095e-06,
+      "loss": 0.7641,
+      "step": 14855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2324137415387635,
+      "learning_rate": 3.0388767304019196e-06,
+      "loss": 0.5717,
+      "step": 14856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3445565355872997,
+      "learning_rate": 3.038331669964718e-06,
+      "loss": 0.5042,
+      "step": 14857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.517882000910134,
+      "learning_rate": 3.037786661061013e-06,
+      "loss": 0.3824,
+      "step": 14858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6931477575043654,
+      "learning_rate": 3.037241703702214e-06,
+      "loss": 0.4678,
+      "step": 14859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8285262747133606,
+      "learning_rate": 3.0366967978997326e-06,
+      "loss": 0.4893,
+      "step": 14860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8898138257667918,
+      "learning_rate": 3.0361519436649718e-06,
+      "loss": 0.3849,
+      "step": 14861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7003977481020354,
+      "learning_rate": 3.03560714100934e-06,
+      "loss": 0.3489,
+      "step": 14862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.602079213575268,
+      "learning_rate": 3.0350623899442416e-06,
+      "loss": 0.4229,
+      "step": 14863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8613873387077116,
+      "learning_rate": 3.034517690481083e-06,
+      "loss": 0.4406,
+      "step": 14864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1208140933355164,
+      "learning_rate": 3.033973042631263e-06,
+      "loss": 0.6305,
+      "step": 14865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7945995039051719,
+      "learning_rate": 3.0334284464061863e-06,
+      "loss": 0.5329,
+      "step": 14866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9772600946358438,
+      "learning_rate": 3.0328839018172544e-06,
+      "loss": 0.5347,
+      "step": 14867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8107072845208376,
+      "learning_rate": 3.032339408875864e-06,
+      "loss": 0.5565,
+      "step": 14868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0515879887701707,
+      "learning_rate": 3.0317949675934156e-06,
+      "loss": 0.7361,
+      "step": 14869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6485804999058224,
+      "learning_rate": 3.031250577981308e-06,
+      "loss": 0.4089,
+      "step": 14870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5339700235487976,
+      "learning_rate": 3.0307062400509345e-06,
+      "loss": 0.3747,
+      "step": 14871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5668295149175655,
+      "learning_rate": 3.0301619538136918e-06,
+      "loss": 0.4281,
+      "step": 14872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6784499696869455,
+      "learning_rate": 3.0296177192809763e-06,
+      "loss": 0.4782,
+      "step": 14873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.539285021146966,
+      "learning_rate": 3.0290735364641784e-06,
+      "loss": 0.3901,
+      "step": 14874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.509937030233168,
+      "learning_rate": 3.0285294053746907e-06,
+      "loss": 0.4165,
+      "step": 14875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.621420940536941,
+      "learning_rate": 3.0279853260239045e-06,
+      "loss": 0.5723,
+      "step": 14876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8024623038560768,
+      "learning_rate": 3.027441298423212e-06,
+      "loss": 0.4522,
+      "step": 14877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6543276377473126,
+      "learning_rate": 3.0268973225839987e-06,
+      "loss": 0.4083,
+      "step": 14878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.191343413657928,
+      "learning_rate": 3.026353398517654e-06,
+      "loss": 0.5369,
+      "step": 14879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8742565432575273,
+      "learning_rate": 3.0258095262355657e-06,
+      "loss": 0.6665,
+      "step": 14880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7679435529565335,
+      "learning_rate": 3.025265705749117e-06,
+      "loss": 0.4614,
+      "step": 14881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.785871268254493,
+      "learning_rate": 3.024721937069693e-06,
+      "loss": 0.4901,
+      "step": 14882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.92170767228058,
+      "learning_rate": 3.02417822020868e-06,
+      "loss": 0.4503,
+      "step": 14883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2624452886538322,
+      "learning_rate": 3.0236345551774566e-06,
+      "loss": 0.3288,
+      "step": 14884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6824546684201374,
+      "learning_rate": 3.0230909419874053e-06,
+      "loss": 0.4971,
+      "step": 14885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.417730949619955,
+      "learning_rate": 3.0225473806499083e-06,
+      "loss": 0.3862,
+      "step": 14886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.689920590210498,
+      "learning_rate": 3.022003871176341e-06,
+      "loss": 0.556,
+      "step": 14887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6485156718611023,
+      "learning_rate": 3.0214604135780844e-06,
+      "loss": 0.4451,
+      "step": 14888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7850262887998833,
+      "learning_rate": 3.0209170078665144e-06,
+      "loss": 0.3896,
+      "step": 14889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.149653461355934,
+      "learning_rate": 3.020373654053008e-06,
+      "loss": 0.7122,
+      "step": 14890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6535679095775406,
+      "learning_rate": 3.019830352148938e-06,
+      "loss": 0.4703,
+      "step": 14891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9372168720880416,
+      "learning_rate": 3.019287102165679e-06,
+      "loss": 0.4174,
+      "step": 14892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.731681927666944,
+      "learning_rate": 3.018743904114605e-06,
+      "loss": 0.4147,
+      "step": 14893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8587290592861776,
+      "learning_rate": 3.018200758007084e-06,
+      "loss": 0.496,
+      "step": 14894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5723448547016947,
+      "learning_rate": 3.017657663854489e-06,
+      "loss": 0.3191,
+      "step": 14895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9006768517934758,
+      "learning_rate": 3.017114621668191e-06,
+      "loss": 0.5349,
+      "step": 14896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.228783880541938,
+      "learning_rate": 3.016571631459555e-06,
+      "loss": 0.2675,
+      "step": 14897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9557135400275032,
+      "learning_rate": 3.0160286932399486e-06,
+      "loss": 0.5026,
+      "step": 14898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4004410414014592,
+      "learning_rate": 3.015485807020741e-06,
+      "loss": 0.3939,
+      "step": 14899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8277706012704558,
+      "learning_rate": 3.014942972813293e-06,
+      "loss": 0.5809,
+      "step": 14900
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.49871689081192017,
+      "eval_runtime": 67.9654,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 14900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5127915799960705,
+      "learning_rate": 3.0144001906289698e-06,
+      "loss": 0.3478,
+      "step": 14901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7344908644467132,
+      "learning_rate": 3.0138574604791372e-06,
+      "loss": 0.4307,
+      "step": 14902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5804925161570074,
+      "learning_rate": 3.0133147823751522e-06,
+      "loss": 0.3806,
+      "step": 14903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8820750166698692,
+      "learning_rate": 3.012772156328378e-06,
+      "loss": 0.4098,
+      "step": 14904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0361577890738762,
+      "learning_rate": 3.012229582350175e-06,
+      "loss": 0.6395,
+      "step": 14905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6365264357323224,
+      "learning_rate": 3.0116870604518992e-06,
+      "loss": 0.465,
+      "step": 14906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8636091014776128,
+      "learning_rate": 3.0111445906449097e-06,
+      "loss": 0.5671,
+      "step": 14907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.028454879736734,
+      "learning_rate": 3.010602172940562e-06,
+      "loss": 0.5324,
+      "step": 14908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.650506095064394,
+      "learning_rate": 3.010059807350213e-06,
+      "loss": 0.4599,
+      "step": 14909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.176282947330331,
+      "learning_rate": 3.0095174938852138e-06,
+      "loss": 0.6486,
+      "step": 14910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5852619882079944,
+      "learning_rate": 3.0089752325569193e-06,
+      "loss": 0.3456,
+      "step": 14911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3419209709758777,
+      "learning_rate": 3.0084330233766822e-06,
+      "loss": 0.3619,
+      "step": 14912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2920037866603873,
+      "learning_rate": 3.007890866355851e-06,
+      "loss": 0.3964,
+      "step": 14913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5117521154163434,
+      "learning_rate": 3.0073487615057762e-06,
+      "loss": 0.4276,
+      "step": 14914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8095433147991593,
+      "learning_rate": 3.0068067088378095e-06,
+      "loss": 0.456,
+      "step": 14915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.552086773833783,
+      "learning_rate": 3.0062647083632922e-06,
+      "loss": 0.46,
+      "step": 14916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7704464998595757,
+      "learning_rate": 3.005722760093576e-06,
+      "loss": 0.5491,
+      "step": 14917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6819951156124557,
+      "learning_rate": 3.005180864040006e-06,
+      "loss": 0.4139,
+      "step": 14918
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7444938791873117,
+      "learning_rate": 3.004639020213923e-06,
+      "loss": 0.4445,
+      "step": 14919
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5772191680044996,
+      "learning_rate": 3.0040972286266717e-06,
+      "loss": 0.3201,
+      "step": 14920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3999104291547297,
+      "learning_rate": 3.0035554892895956e-06,
+      "loss": 0.3572,
+      "step": 14921
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5878436424524842,
+      "learning_rate": 3.0030138022140352e-06,
+      "loss": 0.4776,
+      "step": 14922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9844691403728107,
+      "learning_rate": 3.0024721674113294e-06,
+      "loss": 0.6011,
+      "step": 14923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2659303876398311,
+      "learning_rate": 3.001930584892817e-06,
+      "loss": 0.2968,
+      "step": 14924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2463456048311716,
+      "learning_rate": 3.0013890546698384e-06,
+      "loss": 0.3729,
+      "step": 14925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6447519606895546,
+      "learning_rate": 3.000847576753726e-06,
+      "loss": 0.5901,
+      "step": 14926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.05759742250371,
+      "learning_rate": 3.0003061511558167e-06,
+      "loss": 0.4355,
+      "step": 14927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8917150998926755,
+      "learning_rate": 2.999764777887448e-06,
+      "loss": 0.4214,
+      "step": 14928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.098026513819145,
+      "learning_rate": 2.9992234569599488e-06,
+      "loss": 0.4906,
+      "step": 14929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5508665681909783,
+      "learning_rate": 2.998682188384653e-06,
+      "loss": 0.4409,
+      "step": 14930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.026684830719471,
+      "learning_rate": 2.998140972172894e-06,
+      "loss": 0.5595,
+      "step": 14931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4666074116326346,
+      "learning_rate": 2.997599808335997e-06,
+      "loss": 0.4227,
+      "step": 14932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8525926481370256,
+      "learning_rate": 2.9970586968852953e-06,
+      "loss": 0.5876,
+      "step": 14933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3645693890804078,
+      "learning_rate": 2.996517637832114e-06,
+      "loss": 0.3386,
+      "step": 14934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4175324263736113,
+      "learning_rate": 2.9959766311877836e-06,
+      "loss": 0.3533,
+      "step": 14935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7481498052008517,
+      "learning_rate": 2.995435676963625e-06,
+      "loss": 0.5378,
+      "step": 14936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7415865940906543,
+      "learning_rate": 2.9948947751709644e-06,
+      "loss": 0.4292,
+      "step": 14937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9591540409896404,
+      "learning_rate": 2.994353925821128e-06,
+      "loss": 0.4195,
+      "step": 14938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7908009372393516,
+      "learning_rate": 2.9938131289254337e-06,
+      "loss": 0.4886,
+      "step": 14939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8733251434574971,
+      "learning_rate": 2.9932723844952055e-06,
+      "loss": 0.4494,
+      "step": 14940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8532917028590157,
+      "learning_rate": 2.9927316925417643e-06,
+      "loss": 0.5357,
+      "step": 14941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2083228561568324,
+      "learning_rate": 2.9921910530764265e-06,
+      "loss": 0.6046,
+      "step": 14942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5262029885066208,
+      "learning_rate": 2.9916504661105105e-06,
+      "loss": 0.4399,
+      "step": 14943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3100337048046535,
+      "learning_rate": 2.9911099316553347e-06,
+      "loss": 0.7066,
+      "step": 14944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5750134984557569,
+      "learning_rate": 2.9905694497222136e-06,
+      "loss": 0.4142,
+      "step": 14945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5473385654775886,
+      "learning_rate": 2.9900290203224625e-06,
+      "loss": 0.4198,
+      "step": 14946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6312626383495954,
+      "learning_rate": 2.989488643467397e-06,
+      "loss": 0.3602,
+      "step": 14947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4844848504098067,
+      "learning_rate": 2.9889483191683257e-06,
+      "loss": 0.4064,
+      "step": 14948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2637944406243085,
+      "learning_rate": 2.9884080474365613e-06,
+      "loss": 0.5023,
+      "step": 14949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.452953696531603,
+      "learning_rate": 2.987867828283416e-06,
+      "loss": 0.3623,
+      "step": 14950
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.4988970458507538,
+      "eval_runtime": 68.0051,
+      "eval_samples_per_second": 50.143,
+      "eval_steps_per_second": 1.573,
+      "step": 14950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8894498268333404,
+      "learning_rate": 2.987327661720196e-06,
+      "loss": 0.4664,
+      "step": 14951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.593393154546684,
+      "learning_rate": 2.9867875477582107e-06,
+      "loss": 0.4534,
+      "step": 14952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8290570404312427,
+      "learning_rate": 2.9862474864087688e-06,
+      "loss": 0.535,
+      "step": 14953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.759981819620376,
+      "learning_rate": 2.985707477683173e-06,
+      "loss": 0.4576,
+      "step": 14954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6748732241713982,
+      "learning_rate": 2.985167521592729e-06,
+      "loss": 0.4627,
+      "step": 14955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7449861042285182,
+      "learning_rate": 2.984627618148741e-06,
+      "loss": 0.383,
+      "step": 14956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1810603944556517,
+      "learning_rate": 2.984087767362513e-06,
+      "loss": 0.5167,
+      "step": 14957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7590252507433912,
+      "learning_rate": 2.9835479692453432e-06,
+      "loss": 0.5043,
+      "step": 14958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0167192711519593,
+      "learning_rate": 2.9830082238085344e-06,
+      "loss": 0.4514,
+      "step": 14959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6420703598048672,
+      "learning_rate": 2.982468531063386e-06,
+      "loss": 0.3586,
+      "step": 14960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7807258805600872,
+      "learning_rate": 2.981928891021194e-06,
+      "loss": 0.4961,
+      "step": 14961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8938323088396423,
+      "learning_rate": 2.981389303693256e-06,
+      "loss": 0.5893,
+      "step": 14962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7091741773038793,
+      "learning_rate": 2.980849769090871e-06,
+      "loss": 0.517,
+      "step": 14963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.087097900548385,
+      "learning_rate": 2.980310287225329e-06,
+      "loss": 0.6037,
+      "step": 14964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1133934995361177,
+      "learning_rate": 2.9797708581079266e-06,
+      "loss": 0.5555,
+      "step": 14965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.533174700884242,
+      "learning_rate": 2.9792314817499574e-06,
+      "loss": 0.4067,
+      "step": 14966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5685034879786819,
+      "learning_rate": 2.978692158162709e-06,
+      "loss": 0.4532,
+      "step": 14967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8546499199424042,
+      "learning_rate": 2.9781528873574754e-06,
+      "loss": 0.363,
+      "step": 14968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5174189544075871,
+      "learning_rate": 2.977613669345544e-06,
+      "loss": 0.4175,
+      "step": 14969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4264742991769492,
+      "learning_rate": 2.9770745041382067e-06,
+      "loss": 0.533,
+      "step": 14970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3422128621831997,
+      "learning_rate": 2.976535391746744e-06,
+      "loss": 0.3016,
+      "step": 14971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.720153987902366,
+      "learning_rate": 2.975996332182447e-06,
+      "loss": 0.49,
+      "step": 14972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.732937223566829,
+      "learning_rate": 2.9754573254566e-06,
+      "loss": 0.4342,
+      "step": 14973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.028327832774745,
+      "learning_rate": 2.9749183715804843e-06,
+      "loss": 0.4556,
+      "step": 14974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7855089293678315,
+      "learning_rate": 2.974379470565384e-06,
+      "loss": 0.5323,
+      "step": 14975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8983400884119308,
+      "learning_rate": 2.9738406224225826e-06,
+      "loss": 0.5032,
+      "step": 14976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5950917664641688,
+      "learning_rate": 2.9733018271633574e-06,
+      "loss": 0.4868,
+      "step": 14977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0855635970001636,
+      "learning_rate": 2.9727630847989885e-06,
+      "loss": 0.5411,
+      "step": 14978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8107001685833837,
+      "learning_rate": 2.9722243953407548e-06,
+      "loss": 0.5007,
+      "step": 14979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7371925500749301,
+      "learning_rate": 2.971685758799935e-06,
+      "loss": 0.4655,
+      "step": 14980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7288839291396707,
+      "learning_rate": 2.971147175187802e-06,
+      "loss": 0.3333,
+      "step": 14981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6640606511918534,
+      "learning_rate": 2.9706086445156317e-06,
+      "loss": 0.4737,
+      "step": 14982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.690311026231342,
+      "learning_rate": 2.970070166794701e-06,
+      "loss": 0.4361,
+      "step": 14983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.449639443255329,
+      "learning_rate": 2.969531742036277e-06,
+      "loss": 0.622,
+      "step": 14984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9634579320355157,
+      "learning_rate": 2.968993370251635e-06,
+      "loss": 0.6657,
+      "step": 14985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9839312066420163,
+      "learning_rate": 2.968455051452046e-06,
+      "loss": 0.4203,
+      "step": 14986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7097236955153143,
+      "learning_rate": 2.9679167856487766e-06,
+      "loss": 0.4264,
+      "step": 14987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9030148567889182,
+      "learning_rate": 2.967378572853097e-06,
+      "loss": 0.5127,
+      "step": 14988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0399535405568217,
+      "learning_rate": 2.966840413076275e-06,
+      "loss": 0.5911,
+      "step": 14989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.676215640434352,
+      "learning_rate": 2.9663023063295737e-06,
+      "loss": 0.4708,
+      "step": 14990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.739482207545274,
+      "learning_rate": 2.9657642526242603e-06,
+      "loss": 0.5168,
+      "step": 14991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6067576238493304,
+      "learning_rate": 2.9652262519715985e-06,
+      "loss": 0.3628,
+      "step": 14992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8210954536473674,
+      "learning_rate": 2.9646883043828505e-06,
+      "loss": 0.4674,
+      "step": 14993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0053059896960974,
+      "learning_rate": 2.9641504098692777e-06,
+      "loss": 0.5707,
+      "step": 14994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6497201395379326,
+      "learning_rate": 2.9636125684421433e-06,
+      "loss": 0.4877,
+      "step": 14995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4743699558625527,
+      "learning_rate": 2.9630747801127024e-06,
+      "loss": 0.4292,
+      "step": 14996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7715245879184114,
+      "learning_rate": 2.962537044892215e-06,
+      "loss": 0.5201,
+      "step": 14997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4524744312872904,
+      "learning_rate": 2.9619993627919404e-06,
+      "loss": 0.511,
+      "step": 14998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9563026023176968,
+      "learning_rate": 2.961461733823131e-06,
+      "loss": 0.5654,
+      "step": 14999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1329109276402862,
+      "learning_rate": 2.960924157997044e-06,
+      "loss": 0.5581,
+      "step": 15000
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.49989867210388184,
+      "eval_runtime": 67.9484,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 15000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2039295965189107,
+      "learning_rate": 2.960386635324931e-06,
+      "loss": 0.5717,
+      "step": 15001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4053193963474506,
+      "learning_rate": 2.95984916581805e-06,
+      "loss": 0.3893,
+      "step": 15002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.441395014792041,
+      "learning_rate": 2.959311749487646e-06,
+      "loss": 0.3613,
+      "step": 15003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2540390307326366,
+      "learning_rate": 2.9587743863449732e-06,
+      "loss": 0.4694,
+      "step": 15004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9190823766176406,
+      "learning_rate": 2.958237076401282e-06,
+      "loss": 0.4441,
+      "step": 15005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2965653765856167,
+      "learning_rate": 2.9576998196678164e-06,
+      "loss": 0.6585,
+      "step": 15006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3590537571103798,
+      "learning_rate": 2.9571626161558263e-06,
+      "loss": 0.3396,
+      "step": 15007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2718099390040107,
+      "learning_rate": 2.9566254658765594e-06,
+      "loss": 0.5276,
+      "step": 15008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6070387380356812,
+      "learning_rate": 2.9560883688412567e-06,
+      "loss": 0.4432,
+      "step": 15009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.767642088502666,
+      "learning_rate": 2.9555513250611643e-06,
+      "loss": 0.4349,
+      "step": 15010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.714328641391174,
+      "learning_rate": 2.9550143345475253e-06,
+      "loss": 0.4889,
+      "step": 15011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6790229196379742,
+      "learning_rate": 2.9544773973115793e-06,
+      "loss": 0.4591,
+      "step": 15012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.592988115008711,
+      "learning_rate": 2.953940513364568e-06,
+      "loss": 0.3949,
+      "step": 15013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8016430860974524,
+      "learning_rate": 2.9534036827177304e-06,
+      "loss": 0.4001,
+      "step": 15014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5379103109883374,
+      "learning_rate": 2.9528669053823073e-06,
+      "loss": 0.4269,
+      "step": 15015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9081702495963517,
+      "learning_rate": 2.952330181369531e-06,
+      "loss": 0.4391,
+      "step": 15016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5689776652765481,
+      "learning_rate": 2.95179351069064e-06,
+      "loss": 0.3652,
+      "step": 15017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8628719108263048,
+      "learning_rate": 2.951256893356871e-06,
+      "loss": 0.4948,
+      "step": 15018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1520856844860607,
+      "learning_rate": 2.950720329379454e-06,
+      "loss": 0.6458,
+      "step": 15019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7906705951995605,
+      "learning_rate": 2.9501838187696236e-06,
+      "loss": 0.4056,
+      "step": 15020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5696199492167966,
+      "learning_rate": 2.9496473615386133e-06,
+      "loss": 0.4366,
+      "step": 15021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4868431000712703,
+      "learning_rate": 2.9491109576976495e-06,
+      "loss": 0.3521,
+      "step": 15022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5279549053856645,
+      "learning_rate": 2.9485746072579625e-06,
+      "loss": 0.3749,
+      "step": 15023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6600203100555562,
+      "learning_rate": 2.948038310230784e-06,
+      "loss": 0.3671,
+      "step": 15024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8783374314651586,
+      "learning_rate": 2.947502066627337e-06,
+      "loss": 0.5743,
+      "step": 15025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6751488907897243,
+      "learning_rate": 2.946965876458848e-06,
+      "loss": 0.4675,
+      "step": 15026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0484076510661513,
+      "learning_rate": 2.946429739736543e-06,
+      "loss": 0.5258,
+      "step": 15027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8449305962077882,
+      "learning_rate": 2.945893656471647e-06,
+      "loss": 0.4884,
+      "step": 15028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.803599990184791,
+      "learning_rate": 2.9453576266753788e-06,
+      "loss": 0.4424,
+      "step": 15029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7013737123369201,
+      "learning_rate": 2.944821650358962e-06,
+      "loss": 0.4204,
+      "step": 15030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.629136400546103,
+      "learning_rate": 2.944285727533619e-06,
+      "loss": 0.4962,
+      "step": 15031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.145647000782453,
+      "learning_rate": 2.943749858210565e-06,
+      "loss": 0.7406,
+      "step": 15032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0168386341993125,
+      "learning_rate": 2.9432140424010196e-06,
+      "loss": 0.4594,
+      "step": 15033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.879159709842835,
+      "learning_rate": 2.9426782801162025e-06,
+      "loss": 0.5501,
+      "step": 15034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.770983313410326,
+      "learning_rate": 2.9421425713673257e-06,
+      "loss": 0.4566,
+      "step": 15035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5956506982628982,
+      "learning_rate": 2.9416069161656047e-06,
+      "loss": 0.4425,
+      "step": 15036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6482322054632401,
+      "learning_rate": 2.941071314522256e-06,
+      "loss": 0.3824,
+      "step": 15037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.536955680179321,
+      "learning_rate": 2.940535766448488e-06,
+      "loss": 0.4395,
+      "step": 15038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7982184765168308,
+      "learning_rate": 2.9400002719555133e-06,
+      "loss": 0.5206,
+      "step": 15039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7722240867761871,
+      "learning_rate": 2.939464831054544e-06,
+      "loss": 0.3984,
+      "step": 15040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3785758693909242,
+      "learning_rate": 2.938929443756789e-06,
+      "loss": 0.5025,
+      "step": 15041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6950418189493779,
+      "learning_rate": 2.9383941100734527e-06,
+      "loss": 0.5422,
+      "step": 15042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9355321093685791,
+      "learning_rate": 2.9378588300157474e-06,
+      "loss": 0.5328,
+      "step": 15043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5475487587371508,
+      "learning_rate": 2.937323603594874e-06,
+      "loss": 0.4253,
+      "step": 15044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.351462529748631,
+      "learning_rate": 2.93678843082204e-06,
+      "loss": 0.5187,
+      "step": 15045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.957907446526071,
+      "learning_rate": 2.9362533117084498e-06,
+      "loss": 0.518,
+      "step": 15046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5314429251719721,
+      "learning_rate": 2.935718246265302e-06,
+      "loss": 0.3868,
+      "step": 15047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6402222028570923,
+      "learning_rate": 2.9351832345038e-06,
+      "loss": 0.2939,
+      "step": 15048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5940255090133446,
+      "learning_rate": 2.9346482764351446e-06,
+      "loss": 0.4315,
+      "step": 15049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.620262049254893,
+      "learning_rate": 2.934113372070535e-06,
+      "loss": 0.4835,
+      "step": 15050
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.49961042404174805,
+      "eval_runtime": 67.9939,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 15050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.057477493856286,
+      "learning_rate": 2.9335785214211665e-06,
+      "loss": 0.5483,
+      "step": 15051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0524589649533205,
+      "learning_rate": 2.9330437244982387e-06,
+      "loss": 0.5759,
+      "step": 15052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.966173377316185,
+      "learning_rate": 2.932508981312947e-06,
+      "loss": 0.6121,
+      "step": 15053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3646026299331009,
+      "learning_rate": 2.931974291876484e-06,
+      "loss": 0.304,
+      "step": 15054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.992815163961164,
+      "learning_rate": 2.9314396562000424e-06,
+      "loss": 0.5766,
+      "step": 15055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0661127786144613,
+      "learning_rate": 2.9309050742948197e-06,
+      "loss": 0.4519,
+      "step": 15056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8715717270712493,
+      "learning_rate": 2.930370546172001e-06,
+      "loss": 0.4053,
+      "step": 15057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9920989985032838,
+      "learning_rate": 2.929836071842779e-06,
+      "loss": 0.5825,
+      "step": 15058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8017107607954252,
+      "learning_rate": 2.9293016513183427e-06,
+      "loss": 0.4137,
+      "step": 15059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.030655905991212,
+      "learning_rate": 2.9287672846098812e-06,
+      "loss": 0.6051,
+      "step": 15060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.068390821426595,
+      "learning_rate": 2.928232971728578e-06,
+      "loss": 0.72,
+      "step": 15061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5936045813360997,
+      "learning_rate": 2.9276987126856193e-06,
+      "loss": 0.3756,
+      "step": 15062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9056460430858955,
+      "learning_rate": 2.927164507492193e-06,
+      "loss": 0.3938,
+      "step": 15063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.440961880723237,
+      "learning_rate": 2.9266303561594777e-06,
+      "loss": 0.4621,
+      "step": 15064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8656460820887233,
+      "learning_rate": 2.9260962586986573e-06,
+      "loss": 0.4698,
+      "step": 15065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6031915666789598,
+      "learning_rate": 2.925562215120915e-06,
+      "loss": 0.4479,
+      "step": 15066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0633594161425406,
+      "learning_rate": 2.9250282254374266e-06,
+      "loss": 0.631,
+      "step": 15067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6094346361240013,
+      "learning_rate": 2.9244942896593733e-06,
+      "loss": 0.4817,
+      "step": 15068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8122668185717488,
+      "learning_rate": 2.9239604077979346e-06,
+      "loss": 0.4268,
+      "step": 15069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4449497256163277,
+      "learning_rate": 2.9234265798642826e-06,
+      "loss": 0.3947,
+      "step": 15070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6485475573172317,
+      "learning_rate": 2.9228928058695937e-06,
+      "loss": 0.4215,
+      "step": 15071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5692584764476152,
+      "learning_rate": 2.9223590858250443e-06,
+      "loss": 0.482,
+      "step": 15072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8144137517550956,
+      "learning_rate": 2.9218254197418083e-06,
+      "loss": 0.4577,
+      "step": 15073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6415711828579191,
+      "learning_rate": 2.921291807631053e-06,
+      "loss": 0.5014,
+      "step": 15074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5748371222540092,
+      "learning_rate": 2.9207582495039534e-06,
+      "loss": 0.567,
+      "step": 15075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.472271221744903,
+      "learning_rate": 2.920224745371678e-06,
+      "loss": 0.3565,
+      "step": 15076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7572498430075882,
+      "learning_rate": 2.919691295245395e-06,
+      "loss": 0.4895,
+      "step": 15077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.056105926492725,
+      "learning_rate": 2.919157899136271e-06,
+      "loss": 0.5404,
+      "step": 15078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5598895213565807,
+      "learning_rate": 2.918624557055475e-06,
+      "loss": 0.3969,
+      "step": 15079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7260359228802236,
+      "learning_rate": 2.9180912690141692e-06,
+      "loss": 0.436,
+      "step": 15080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.244998983677264,
+      "learning_rate": 2.917558035023519e-06,
+      "loss": 0.4385,
+      "step": 15081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9015405124328753,
+      "learning_rate": 2.9170248550946896e-06,
+      "loss": 0.6674,
+      "step": 15082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3339294708973364,
+      "learning_rate": 2.9164917292388378e-06,
+      "loss": 0.4074,
+      "step": 15083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5667509019178882,
+      "learning_rate": 2.9159586574671276e-06,
+      "loss": 0.5373,
+      "step": 15084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.681385157446845,
+      "learning_rate": 2.9154256397907175e-06,
+      "loss": 0.4471,
+      "step": 15085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6754300691713653,
+      "learning_rate": 2.9148926762207685e-06,
+      "loss": 0.4797,
+      "step": 15086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5155209017081221,
+      "learning_rate": 2.9143597667684337e-06,
+      "loss": 0.4178,
+      "step": 15087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6091395244927753,
+      "learning_rate": 2.9138269114448713e-06,
+      "loss": 0.3574,
+      "step": 15088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7383173173998594,
+      "learning_rate": 2.913294110261238e-06,
+      "loss": 0.5551,
+      "step": 15089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.778035305716489,
+      "learning_rate": 2.9127613632286832e-06,
+      "loss": 0.4924,
+      "step": 15090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7550093800935078,
+      "learning_rate": 2.912228670358363e-06,
+      "loss": 0.4554,
+      "step": 15091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5917788966254798,
+      "learning_rate": 2.9116960316614282e-06,
+      "loss": 0.4455,
+      "step": 15092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8617032893830119,
+      "learning_rate": 2.911163447149029e-06,
+      "loss": 0.4111,
+      "step": 15093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6940109435519217,
+      "learning_rate": 2.9106309168323154e-06,
+      "loss": 0.3992,
+      "step": 15094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890468647485845,
+      "learning_rate": 2.9100984407224375e-06,
+      "loss": 0.433,
+      "step": 15095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5947011214646296,
+      "learning_rate": 2.9095660188305377e-06,
+      "loss": 0.3943,
+      "step": 15096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9463897362240719,
+      "learning_rate": 2.909033651167764e-06,
+      "loss": 0.5413,
+      "step": 15097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.463205139768196,
+      "learning_rate": 2.9085013377452643e-06,
+      "loss": 0.6699,
+      "step": 15098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8304950190690439,
+      "learning_rate": 2.9079690785741767e-06,
+      "loss": 0.5348,
+      "step": 15099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8662139910806896,
+      "learning_rate": 2.9074368736656465e-06,
+      "loss": 0.5054,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.5001636743545532,
+      "eval_runtime": 67.9776,
+      "eval_samples_per_second": 50.164,
+      "eval_steps_per_second": 1.574,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.578184813737434,
+      "learning_rate": 2.906904723030818e-06,
+      "loss": 0.434,
+      "step": 15101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8290938243841106,
+      "learning_rate": 2.9063726266808254e-06,
+      "loss": 0.3582,
+      "step": 15102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2539877722249386,
+      "learning_rate": 2.905840584626811e-06,
+      "loss": 0.6418,
+      "step": 15103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8856597323967548,
+      "learning_rate": 2.9053085968799154e-06,
+      "loss": 0.4087,
+      "step": 15104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.393661381211557,
+      "learning_rate": 2.9047766634512698e-06,
+      "loss": 0.352,
+      "step": 15105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.901261461860349,
+      "learning_rate": 2.9042447843520137e-06,
+      "loss": 0.4874,
+      "step": 15106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7323024099067779,
+      "learning_rate": 2.9037129595932813e-06,
+      "loss": 0.5174,
+      "step": 15107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1072054639789823,
+      "learning_rate": 2.9031811891862056e-06,
+      "loss": 0.529,
+      "step": 15108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6342317796988792,
+      "learning_rate": 2.9026494731419176e-06,
+      "loss": 0.4714,
+      "step": 15109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8544172285397216,
+      "learning_rate": 2.9021178114715496e-06,
+      "loss": 0.5658,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8380953617363252,
+      "learning_rate": 2.9015862041862336e-06,
+      "loss": 0.466,
+      "step": 15111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7896305164398354,
+      "learning_rate": 2.901054651297095e-06,
+      "loss": 0.5313,
+      "step": 15112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5439718492068022,
+      "learning_rate": 2.900523152815262e-06,
+      "loss": 0.3635,
+      "step": 15113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5496382221022655,
+      "learning_rate": 2.8999917087518647e-06,
+      "loss": 0.4006,
+      "step": 15114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2900396470680444,
+      "learning_rate": 2.899460319118025e-06,
+      "loss": 0.3291,
+      "step": 15115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8088215742794536,
+      "learning_rate": 2.898928983924867e-06,
+      "loss": 0.5439,
+      "step": 15116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9114725514459383,
+      "learning_rate": 2.8983977031835176e-06,
+      "loss": 0.3732,
+      "step": 15117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2036137196035996,
+      "learning_rate": 2.8978664769050947e-06,
+      "loss": 0.5341,
+      "step": 15118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.62337237617788,
+      "learning_rate": 2.89733530510072e-06,
+      "loss": 0.431,
+      "step": 15119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8769747164495931,
+      "learning_rate": 2.8968041877815145e-06,
+      "loss": 0.5266,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2602906105953866,
+      "learning_rate": 2.8962731249585987e-06,
+      "loss": 0.4769,
+      "step": 15121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5368507180279034,
+      "learning_rate": 2.8957421166430856e-06,
+      "loss": 0.4179,
+      "step": 15122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6809169664375194,
+      "learning_rate": 2.8952111628460935e-06,
+      "loss": 0.5769,
+      "step": 15123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1425143522039187,
+      "learning_rate": 2.8946802635787396e-06,
+      "loss": 0.6216,
+      "step": 15124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.07005313276818,
+      "learning_rate": 2.8941494188521358e-06,
+      "loss": 0.6653,
+      "step": 15125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2381262369668686,
+      "learning_rate": 2.8936186286773938e-06,
+      "loss": 0.6468,
+      "step": 15126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6339606191604796,
+      "learning_rate": 2.8930878930656293e-06,
+      "loss": 0.5162,
+      "step": 15127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6895826780700174,
+      "learning_rate": 2.8925572120279477e-06,
+      "loss": 0.4395,
+      "step": 15128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.753372669917979,
+      "learning_rate": 2.8920265855754625e-06,
+      "loss": 0.5008,
+      "step": 15129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7642939585227844,
+      "learning_rate": 2.8914960137192826e-06,
+      "loss": 0.489,
+      "step": 15130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4635717046243357,
+      "learning_rate": 2.8909654964705113e-06,
+      "loss": 0.4353,
+      "step": 15131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5791117739881302,
+      "learning_rate": 2.890435033840257e-06,
+      "loss": 0.3425,
+      "step": 15132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.856549081696374,
+      "learning_rate": 2.8899046258396246e-06,
+      "loss": 0.543,
+      "step": 15133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6771355825352439,
+      "learning_rate": 2.8893742724797193e-06,
+      "loss": 0.5006,
+      "step": 15134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4740174888225754,
+      "learning_rate": 2.88884397377164e-06,
+      "loss": 0.3947,
+      "step": 15135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7044895514060145,
+      "learning_rate": 2.88831372972649e-06,
+      "loss": 0.4186,
+      "step": 15136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6547006762195033,
+      "learning_rate": 2.8877835403553724e-06,
+      "loss": 0.4188,
+      "step": 15137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.048311132397605,
+      "learning_rate": 2.8872534056693823e-06,
+      "loss": 0.4461,
+      "step": 15138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6214450843806347,
+      "learning_rate": 2.8867233256796178e-06,
+      "loss": 0.4544,
+      "step": 15139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8482215412995713,
+      "learning_rate": 2.8861933003971808e-06,
+      "loss": 0.4046,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.83488196190546,
+      "learning_rate": 2.8856633298331594e-06,
+      "loss": 0.6846,
+      "step": 15141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6942589366368839,
+      "learning_rate": 2.885133413998654e-06,
+      "loss": 0.441,
+      "step": 15142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9292994309534943,
+      "learning_rate": 2.884603552904758e-06,
+      "loss": 0.5415,
+      "step": 15143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6283982113901336,
+      "learning_rate": 2.884073746562561e-06,
+      "loss": 0.4057,
+      "step": 15144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4293179831684253,
+      "learning_rate": 2.883543994983155e-06,
+      "loss": 0.4036,
+      "step": 15145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9235373754451421,
+      "learning_rate": 2.8830142981776314e-06,
+      "loss": 0.5238,
+      "step": 15146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2984385526535394,
+      "learning_rate": 2.8824846561570774e-06,
+      "loss": 0.4442,
+      "step": 15147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.473116616994912,
+      "learning_rate": 2.8819550689325805e-06,
+      "loss": 0.3067,
+      "step": 15148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8174157419874113,
+      "learning_rate": 2.8814255365152298e-06,
+      "loss": 0.6235,
+      "step": 15149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7186597755409936,
+      "learning_rate": 2.880896058916108e-06,
+      "loss": 0.5331,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.4997015595436096,
+      "eval_runtime": 67.9569,
+      "eval_samples_per_second": 50.179,
+      "eval_steps_per_second": 1.575,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8827160243380008,
+      "learning_rate": 2.8803666361463e-06,
+      "loss": 0.6028,
+      "step": 15151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6111423973306218,
+      "learning_rate": 2.8798372682168896e-06,
+      "loss": 0.4282,
+      "step": 15152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.563310764124605,
+      "learning_rate": 2.8793079551389603e-06,
+      "loss": 0.3885,
+      "step": 15153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6571483275404115,
+      "learning_rate": 2.8787786969235887e-06,
+      "loss": 0.4667,
+      "step": 15154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6101026120611597,
+      "learning_rate": 2.878249493581857e-06,
+      "loss": 0.3732,
+      "step": 15155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8655414970909376,
+      "learning_rate": 2.877720345124847e-06,
+      "loss": 0.5237,
+      "step": 15156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9339730482142548,
+      "learning_rate": 2.8771912515636296e-06,
+      "loss": 0.48,
+      "step": 15157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9798744367696082,
+      "learning_rate": 2.876662212909285e-06,
+      "loss": 0.4628,
+      "step": 15158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5950275808888095,
+      "learning_rate": 2.8761332291728883e-06,
+      "loss": 0.4299,
+      "step": 15159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7341034201858123,
+      "learning_rate": 2.875604300365512e-06,
+      "loss": 0.6656,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8890048309690024,
+      "learning_rate": 2.8750754264982297e-06,
+      "loss": 0.4539,
+      "step": 15161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.775316442860106,
+      "learning_rate": 2.8745466075821136e-06,
+      "loss": 0.3623,
+      "step": 15162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.514300730641927,
+      "learning_rate": 2.874017843628233e-06,
+      "loss": 0.3619,
+      "step": 15163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4703553574526727,
+      "learning_rate": 2.873489134647657e-06,
+      "loss": 0.3751,
+      "step": 15164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5366567396036603,
+      "learning_rate": 2.872960480651455e-06,
+      "loss": 0.4772,
+      "step": 15165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8515358302339544,
+      "learning_rate": 2.872431881650695e-06,
+      "loss": 0.6411,
+      "step": 15166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.520897972317306,
+      "learning_rate": 2.8719033376564396e-06,
+      "loss": 0.3448,
+      "step": 15167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6792620881222524,
+      "learning_rate": 2.8713748486797566e-06,
+      "loss": 0.4164,
+      "step": 15168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6451557055874002,
+      "learning_rate": 2.8708464147317095e-06,
+      "loss": 0.4631,
+      "step": 15169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8788416257964506,
+      "learning_rate": 2.8703180358233573e-06,
+      "loss": 0.4194,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1013799973698215,
+      "learning_rate": 2.8697897119657646e-06,
+      "loss": 0.6148,
+      "step": 15171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6880140810683308,
+      "learning_rate": 2.8692614431699917e-06,
+      "loss": 0.5724,
+      "step": 15172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9022408520351723,
+      "learning_rate": 2.8687332294470952e-06,
+      "loss": 0.4252,
+      "step": 15173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8338613657676615,
+      "learning_rate": 2.8682050708081345e-06,
+      "loss": 0.5641,
+      "step": 15174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1050463266877526,
+      "learning_rate": 2.8676769672641668e-06,
+      "loss": 0.4933,
+      "step": 15175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.733422732204298,
+      "learning_rate": 2.8671489188262453e-06,
+      "loss": 0.4938,
+      "step": 15176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5867967016293656,
+      "learning_rate": 2.866620925505426e-06,
+      "loss": 0.4799,
+      "step": 15177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7012683509751836,
+      "learning_rate": 2.866092987312762e-06,
+      "loss": 0.5428,
+      "step": 15178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.915444203085859,
+      "learning_rate": 2.865565104259307e-06,
+      "loss": 0.573,
+      "step": 15179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.722924026024472,
+      "learning_rate": 2.8650372763561085e-06,
+      "loss": 0.5291,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.150207686070889,
+      "learning_rate": 2.864509503614218e-06,
+      "loss": 0.6516,
+      "step": 15181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7765027939564393,
+      "learning_rate": 2.8639817860446855e-06,
+      "loss": 0.4192,
+      "step": 15182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7388624974558973,
+      "learning_rate": 2.863454123658556e-06,
+      "loss": 0.5664,
+      "step": 15183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7956656657590326,
+      "learning_rate": 2.8629265164668763e-06,
+      "loss": 0.5084,
+      "step": 15184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.648861177171851,
+      "learning_rate": 2.8623989644806937e-06,
+      "loss": 0.4027,
+      "step": 15185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7653833363406348,
+      "learning_rate": 2.8618714677110494e-06,
+      "loss": 0.4234,
+      "step": 15186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.811968303234606,
+      "learning_rate": 2.8613440261689867e-06,
+      "loss": 0.4164,
+      "step": 15187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6974361997716576,
+      "learning_rate": 2.86081663986555e-06,
+      "loss": 0.374,
+      "step": 15188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0152786147498656,
+      "learning_rate": 2.8602893088117758e-06,
+      "loss": 0.6204,
+      "step": 15189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8022628138371228,
+      "learning_rate": 2.8597620330187037e-06,
+      "loss": 0.4823,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.097964353992617,
+      "learning_rate": 2.8592348124973766e-06,
+      "loss": 0.5656,
+      "step": 15191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.797191372965522,
+      "learning_rate": 2.858707647258827e-06,
+      "loss": 0.4608,
+      "step": 15192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6358474461928239,
+      "learning_rate": 2.8581805373140926e-06,
+      "loss": 0.3875,
+      "step": 15193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9119570774064285,
+      "learning_rate": 2.857653482674209e-06,
+      "loss": 0.6732,
+      "step": 15194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.895732174975029,
+      "learning_rate": 2.857126483350207e-06,
+      "loss": 0.4852,
+      "step": 15195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5464590274046361,
+      "learning_rate": 2.85659953935312e-06,
+      "loss": 0.3984,
+      "step": 15196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.787063879512821,
+      "learning_rate": 2.856072650693982e-06,
+      "loss": 0.4845,
+      "step": 15197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.749968129345515,
+      "learning_rate": 2.855545817383819e-06,
+      "loss": 0.4583,
+      "step": 15198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4316222889549794,
+      "learning_rate": 2.8550190394336617e-06,
+      "loss": 0.4934,
+      "step": 15199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4474958269680243,
+      "learning_rate": 2.854492316854538e-06,
+      "loss": 0.4864,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.4985315799713135,
+      "eval_runtime": 68.0123,
+      "eval_samples_per_second": 50.138,
+      "eval_steps_per_second": 1.573,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5112557810836467,
+      "learning_rate": 2.853965649657476e-06,
+      "loss": 0.3979,
+      "step": 15201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.886981965907725,
+      "learning_rate": 2.853439037853498e-06,
+      "loss": 0.526,
+      "step": 15202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.519940043512179,
+      "learning_rate": 2.8529124814536303e-06,
+      "loss": 0.356,
+      "step": 15203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.074717043002903,
+      "learning_rate": 2.8523859804688975e-06,
+      "loss": 0.556,
+      "step": 15204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.766396138725203,
+      "learning_rate": 2.8518595349103175e-06,
+      "loss": 0.5585,
+      "step": 15205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7235939521213657,
+      "learning_rate": 2.851333144788914e-06,
+      "loss": 0.5591,
+      "step": 15206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6057512349779794,
+      "learning_rate": 2.850806810115706e-06,
+      "loss": 0.3913,
+      "step": 15207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6578267626003977,
+      "learning_rate": 2.850280530901712e-06,
+      "loss": 0.5014,
+      "step": 15208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9469119643505806,
+      "learning_rate": 2.849754307157948e-06,
+      "loss": 0.617,
+      "step": 15209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6451327029746294,
+      "learning_rate": 2.8492281388954345e-06,
+      "loss": 0.498,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0104760750228583,
+      "learning_rate": 2.8487020261251808e-06,
+      "loss": 0.4726,
+      "step": 15211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4892574777925254,
+      "learning_rate": 2.8481759688582034e-06,
+      "loss": 0.4493,
+      "step": 15212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5576349148905297,
+      "learning_rate": 2.847649967105515e-06,
+      "loss": 0.4919,
+      "step": 15213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8081241505119119,
+      "learning_rate": 2.8471240208781293e-06,
+      "loss": 0.5701,
+      "step": 15214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6561975438855583,
+      "learning_rate": 2.8465981301870526e-06,
+      "loss": 0.4642,
+      "step": 15215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.435911230435565,
+      "learning_rate": 2.846072295043295e-06,
+      "loss": 0.5767,
+      "step": 15216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3889425522641847,
+      "learning_rate": 2.8455465154578687e-06,
+      "loss": 0.4676,
+      "step": 15217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8272268314119011,
+      "learning_rate": 2.8450207914417747e-06,
+      "loss": 0.4159,
+      "step": 15218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.231597084676914,
+      "learning_rate": 2.844495123006022e-06,
+      "loss": 0.6487,
+      "step": 15219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6886620273130561,
+      "learning_rate": 2.8439695101616155e-06,
+      "loss": 0.4561,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5349617143361132,
+      "learning_rate": 2.8434439529195567e-06,
+      "loss": 0.3202,
+      "step": 15221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8352434325774651,
+      "learning_rate": 2.8429184512908483e-06,
+      "loss": 0.4796,
+      "step": 15222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9388887697519486,
+      "learning_rate": 2.8423930052864932e-06,
+      "loss": 0.3429,
+      "step": 15223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.489007504529107,
+      "learning_rate": 2.8418676149174878e-06,
+      "loss": 0.499,
+      "step": 15224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9435200973498816,
+      "learning_rate": 2.841342280194833e-06,
+      "loss": 0.3506,
+      "step": 15225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2385095418653553,
+      "learning_rate": 2.840817001129527e-06,
+      "loss": 0.6377,
+      "step": 15226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9427500843230916,
+      "learning_rate": 2.840291777732567e-06,
+      "loss": 0.4525,
+      "step": 15227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6482809059414478,
+      "learning_rate": 2.8397666100149447e-06,
+      "loss": 0.4568,
+      "step": 15228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.663350999072837,
+      "learning_rate": 2.839241497987656e-06,
+      "loss": 0.3715,
+      "step": 15229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7060272591029826,
+      "learning_rate": 2.838716441661695e-06,
+      "loss": 0.5918,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.094541209471363,
+      "learning_rate": 2.8381914410480517e-06,
+      "loss": 0.4883,
+      "step": 15231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7537895334711773,
+      "learning_rate": 2.8376664961577167e-06,
+      "loss": 0.4262,
+      "step": 15232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.485833616509936,
+      "learning_rate": 2.837141607001681e-06,
+      "loss": 0.4419,
+      "step": 15233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9250339633384963,
+      "learning_rate": 2.836616773590931e-06,
+      "loss": 0.5711,
+      "step": 15234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6966078336609318,
+      "learning_rate": 2.8360919959364548e-06,
+      "loss": 0.5181,
+      "step": 15235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6823319361926385,
+      "learning_rate": 2.835567274049238e-06,
+      "loss": 0.4024,
+      "step": 15236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8205923602012866,
+      "learning_rate": 2.835042607940266e-06,
+      "loss": 0.5544,
+      "step": 15237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7898150961444157,
+      "learning_rate": 2.8345179976205208e-06,
+      "loss": 0.3758,
+      "step": 15238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8090947248754163,
+      "learning_rate": 2.8339934431009853e-06,
+      "loss": 0.6048,
+      "step": 15239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.815589725507474,
+      "learning_rate": 2.833468944392642e-06,
+      "loss": 0.484,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8120992573201518,
+      "learning_rate": 2.8329445015064694e-06,
+      "loss": 0.5095,
+      "step": 15241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6607226787717855,
+      "learning_rate": 2.832420114453449e-06,
+      "loss": 0.4987,
+      "step": 15242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.749287012323135,
+      "learning_rate": 2.831895783244555e-06,
+      "loss": 0.5097,
+      "step": 15243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.876662988342701,
+      "learning_rate": 2.8313715078907655e-06,
+      "loss": 0.4224,
+      "step": 15244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8314678794067294,
+      "learning_rate": 2.8308472884030556e-06,
+      "loss": 0.474,
+      "step": 15245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6705923046437525,
+      "learning_rate": 2.830323124792402e-06,
+      "loss": 0.4394,
+      "step": 15246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8977268225848227,
+      "learning_rate": 2.8297990170697736e-06,
+      "loss": 0.5411,
+      "step": 15247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.093058574808394,
+      "learning_rate": 2.8292749652461447e-06,
+      "loss": 0.603,
+      "step": 15248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.492328415785815,
+      "learning_rate": 2.8287509693324865e-06,
+      "loss": 0.5159,
+      "step": 15249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9822250877428853,
+      "learning_rate": 2.828227029339767e-06,
+      "loss": 0.5655,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.49921733140945435,
+      "eval_runtime": 68.0008,
+      "eval_samples_per_second": 50.146,
+      "eval_steps_per_second": 1.574,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6115643694890545,
+      "learning_rate": 2.8277031452789545e-06,
+      "loss": 0.4247,
+      "step": 15251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5082840749964606,
+      "learning_rate": 2.827179317161018e-06,
+      "loss": 0.4745,
+      "step": 15252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.568430103308368,
+      "learning_rate": 2.8266555449969214e-06,
+      "loss": 0.3215,
+      "step": 15253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8773941915003605,
+      "learning_rate": 2.8261318287976304e-06,
+      "loss": 0.5742,
+      "step": 15254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.193016293086604,
+      "learning_rate": 2.8256081685741097e-06,
+      "loss": 0.5353,
+      "step": 15255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.683756917042957,
+      "learning_rate": 2.8250845643373205e-06,
+      "loss": 0.6052,
+      "step": 15256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7669976501363678,
+      "learning_rate": 2.824561016098223e-06,
+      "loss": 0.5389,
+      "step": 15257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1227351022533685,
+      "learning_rate": 2.8240375238677796e-06,
+      "loss": 0.5921,
+      "step": 15258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3266727650162926,
+      "learning_rate": 2.8235140876569504e-06,
+      "loss": 0.3021,
+      "step": 15259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5459775800757383,
+      "learning_rate": 2.822990707476689e-06,
+      "loss": 0.5234,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.595655734064993,
+      "learning_rate": 2.822467383337955e-06,
+      "loss": 0.4291,
+      "step": 15261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3471789510077492,
+      "learning_rate": 2.8219441152517052e-06,
+      "loss": 0.4059,
+      "step": 15262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.68553469417923,
+      "learning_rate": 2.82142090322889e-06,
+      "loss": 0.6765,
+      "step": 15263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.662475305497796,
+      "learning_rate": 2.8208977472804644e-06,
+      "loss": 0.5122,
+      "step": 15264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7450896080442495,
+      "learning_rate": 2.820374647417382e-06,
+      "loss": 0.5334,
+      "step": 15265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9028993686903597,
+      "learning_rate": 2.8198516036505897e-06,
+      "loss": 0.5481,
+      "step": 15266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9830957845027264,
+      "learning_rate": 2.8193286159910404e-06,
+      "loss": 0.5078,
+      "step": 15267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4952626012961088,
+      "learning_rate": 2.8188056844496825e-06,
+      "loss": 0.4195,
+      "step": 15268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5155577900276445,
+      "learning_rate": 2.818282809037461e-06,
+      "loss": 0.3514,
+      "step": 15269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9049201115986396,
+      "learning_rate": 2.817759989765323e-06,
+      "loss": 0.5772,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0284730866362244,
+      "learning_rate": 2.817237226644214e-06,
+      "loss": 0.6856,
+      "step": 15271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1115029002337318,
+      "learning_rate": 2.816714519685079e-06,
+      "loss": 0.6577,
+      "step": 15272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7689494080010686,
+      "learning_rate": 2.8161918688988565e-06,
+      "loss": 0.4306,
+      "step": 15273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8929197180613277,
+      "learning_rate": 2.8156692742964905e-06,
+      "loss": 0.5066,
+      "step": 15274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0211683050836635,
+      "learning_rate": 2.8151467358889227e-06,
+      "loss": 0.4805,
+      "step": 15275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7862175678990606,
+      "learning_rate": 2.814624253687089e-06,
+      "loss": 0.4471,
+      "step": 15276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6510563597912347,
+      "learning_rate": 2.8141018277019284e-06,
+      "loss": 0.3894,
+      "step": 15277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3511774366514016,
+      "learning_rate": 2.8135794579443786e-06,
+      "loss": 0.3812,
+      "step": 15278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.701149726610581,
+      "learning_rate": 2.813057144425374e-06,
+      "loss": 0.4718,
+      "step": 15279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6846229870211196,
+      "learning_rate": 2.8125348871558488e-06,
+      "loss": 0.5081,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6298103947123255,
+      "learning_rate": 2.812012686146738e-06,
+      "loss": 0.4368,
+      "step": 15281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5173249421411563,
+      "learning_rate": 2.81149054140897e-06,
+      "loss": 0.3594,
+      "step": 15282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.883500527749322,
+      "learning_rate": 2.8109684529534776e-06,
+      "loss": 0.463,
+      "step": 15283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5347382796993152,
+      "learning_rate": 2.8104464207911907e-06,
+      "loss": 0.4063,
+      "step": 15284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8296005036809888,
+      "learning_rate": 2.8099244449330387e-06,
+      "loss": 0.6266,
+      "step": 15285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.571427321086816,
+      "learning_rate": 2.809402525389947e-06,
+      "loss": 0.5871,
+      "step": 15286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9483726957847765,
+      "learning_rate": 2.8088806621728416e-06,
+      "loss": 0.4225,
+      "step": 15287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8808373443792843,
+      "learning_rate": 2.8083588552926478e-06,
+      "loss": 0.4348,
+      "step": 15288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.400161203312747,
+      "learning_rate": 2.807837104760289e-06,
+      "loss": 0.4267,
+      "step": 15289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7566893724697932,
+      "learning_rate": 2.8073154105866906e-06,
+      "loss": 0.5079,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7077316618090477,
+      "learning_rate": 2.806793772782771e-06,
+      "loss": 0.4118,
+      "step": 15291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8882803180091585,
+      "learning_rate": 2.806272191359449e-06,
+      "loss": 0.4639,
+      "step": 15292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.328635124578062,
+      "learning_rate": 2.8057506663276466e-06,
+      "loss": 0.3021,
+      "step": 15293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.788802146422989,
+      "learning_rate": 2.805229197698282e-06,
+      "loss": 0.4782,
+      "step": 15294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5216235883551548,
+      "learning_rate": 2.804707785482269e-06,
+      "loss": 0.3539,
+      "step": 15295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9509491258448242,
+      "learning_rate": 2.804186429690524e-06,
+      "loss": 0.479,
+      "step": 15296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8388576087194806,
+      "learning_rate": 2.8036651303339634e-06,
+      "loss": 0.4923,
+      "step": 15297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.206287911720816,
+      "learning_rate": 2.8031438874234974e-06,
+      "loss": 0.4785,
+      "step": 15298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6467981222071955,
+      "learning_rate": 2.802622700970038e-06,
+      "loss": 0.3627,
+      "step": 15299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9450528493561143,
+      "learning_rate": 2.802101570984499e-06,
+      "loss": 0.5624,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.4985363483428955,
+      "eval_runtime": 67.9849,
+      "eval_samples_per_second": 50.158,
+      "eval_steps_per_second": 1.574,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.447278795963755,
+      "learning_rate": 2.801580497477787e-06,
+      "loss": 0.3937,
+      "step": 15301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5113545378580655,
+      "learning_rate": 2.8010594804608106e-06,
+      "loss": 0.4001,
+      "step": 15302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8459130704896702,
+      "learning_rate": 2.800538519944479e-06,
+      "loss": 0.5089,
+      "step": 15303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1336282695655697,
+      "learning_rate": 2.8000176159396953e-06,
+      "loss": 0.5142,
+      "step": 15304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6147432954224417,
+      "learning_rate": 2.7994967684573658e-06,
+      "loss": 0.493,
+      "step": 15305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6181185161184861,
+      "learning_rate": 2.7989759775083945e-06,
+      "loss": 0.4899,
+      "step": 15306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265847624824699,
+      "learning_rate": 2.798455243103685e-06,
+      "loss": 0.752,
+      "step": 15307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4719988796741634,
+      "learning_rate": 2.797934565254135e-06,
+      "loss": 0.5509,
+      "step": 15308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9465563074713672,
+      "learning_rate": 2.797413943970647e-06,
+      "loss": 0.572,
+      "step": 15309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4319009820163942,
+      "learning_rate": 2.7968933792641206e-06,
+      "loss": 0.4374,
+      "step": 15310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0325613826760924,
+      "learning_rate": 2.7963728711454508e-06,
+      "loss": 0.6252,
+      "step": 15311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.69563523552337,
+      "learning_rate": 2.7958524196255365e-06,
+      "loss": 0.5241,
+      "step": 15312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4720433183173238,
+      "learning_rate": 2.7953320247152726e-06,
+      "loss": 0.4057,
+      "step": 15313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3837106366190595,
+      "learning_rate": 2.7948116864255516e-06,
+      "loss": 0.5318,
+      "step": 15314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.210547040680201,
+      "learning_rate": 2.794291404767268e-06,
+      "loss": 0.6466,
+      "step": 15315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8849749424454667,
+      "learning_rate": 2.7937711797513127e-06,
+      "loss": 0.4992,
+      "step": 15316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6170936826814426,
+      "learning_rate": 2.793251011388579e-06,
+      "loss": 0.5396,
+      "step": 15317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.601289123879953,
+      "learning_rate": 2.7927308996899527e-06,
+      "loss": 0.4316,
+      "step": 15318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8135366986405408,
+      "learning_rate": 2.7922108446663228e-06,
+      "loss": 0.4356,
+      "step": 15319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9363688744771654,
+      "learning_rate": 2.7916908463285793e-06,
+      "loss": 0.4424,
+      "step": 15320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.873446642658443,
+      "learning_rate": 2.7911709046876035e-06,
+      "loss": 0.5063,
+      "step": 15321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.659529275855696,
+      "learning_rate": 2.7906510197542825e-06,
+      "loss": 0.5157,
+      "step": 15322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8179922075968984,
+      "learning_rate": 2.7901311915395015e-06,
+      "loss": 0.5713,
+      "step": 15323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0246750238900155,
+      "learning_rate": 2.7896114200541384e-06,
+      "loss": 0.5215,
+      "step": 15324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3810256755880055,
+      "learning_rate": 2.789091705309078e-06,
+      "loss": 0.3596,
+      "step": 15325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6370868277021149,
+      "learning_rate": 2.7885720473151994e-06,
+      "loss": 0.4299,
+      "step": 15326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.735816233590733,
+      "learning_rate": 2.7880524460833803e-06,
+      "loss": 0.5088,
+      "step": 15327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.639863571242201,
+      "learning_rate": 2.7875329016244977e-06,
+      "loss": 0.4727,
+      "step": 15328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8940408364694268,
+      "learning_rate": 2.78701341394943e-06,
+      "loss": 0.5067,
+      "step": 15329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5725728309808633,
+      "learning_rate": 2.786493983069052e-06,
+      "loss": 0.442,
+      "step": 15330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5365775439454228,
+      "learning_rate": 2.7859746089942357e-06,
+      "loss": 0.4746,
+      "step": 15331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.161564660441946,
+      "learning_rate": 2.7854552917358556e-06,
+      "loss": 0.7369,
+      "step": 15332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7616076183058025,
+      "learning_rate": 2.7849360313047846e-06,
+      "loss": 0.4752,
+      "step": 15333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5526301685020023,
+      "learning_rate": 2.78441682771189e-06,
+      "loss": 0.4713,
+      "step": 15334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7015959986860991,
+      "learning_rate": 2.783897680968042e-06,
+      "loss": 0.3766,
+      "step": 15335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7403425233490657,
+      "learning_rate": 2.78337859108411e-06,
+      "loss": 0.4907,
+      "step": 15336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8434441895704519,
+      "learning_rate": 2.7828595580709573e-06,
+      "loss": 0.4083,
+      "step": 15337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7245200890899202,
+      "learning_rate": 2.7823405819394533e-06,
+      "loss": 0.4809,
+      "step": 15338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.034300723429399,
+      "learning_rate": 2.781821662700463e-06,
+      "loss": 0.496,
+      "step": 15339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0792001969721956,
+      "learning_rate": 2.7813028003648467e-06,
+      "loss": 0.586,
+      "step": 15340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0725345636854975,
+      "learning_rate": 2.780783994943468e-06,
+      "loss": 0.564,
+      "step": 15341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7758124885334543,
+      "learning_rate": 2.780265246447188e-06,
+      "loss": 0.5755,
+      "step": 15342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.661650312760901,
+      "learning_rate": 2.779746554886864e-06,
+      "loss": 0.3794,
+      "step": 15343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7971795154388217,
+      "learning_rate": 2.7792279202733567e-06,
+      "loss": 0.5295,
+      "step": 15344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7279357134433042,
+      "learning_rate": 2.7787093426175247e-06,
+      "loss": 0.4435,
+      "step": 15345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8421884154166386,
+      "learning_rate": 2.7781908219302213e-06,
+      "loss": 0.5461,
+      "step": 15346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6961847477695537,
+      "learning_rate": 2.777672358222301e-06,
+      "loss": 0.3585,
+      "step": 15347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.699638799291302,
+      "learning_rate": 2.7771539515046218e-06,
+      "loss": 0.4844,
+      "step": 15348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0493715725592665,
+      "learning_rate": 2.7766356017880313e-06,
+      "loss": 0.5471,
+      "step": 15349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1191263748970273,
+      "learning_rate": 2.7761173090833828e-06,
+      "loss": 0.6157,
+      "step": 15350
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.49835219979286194,
+      "eval_runtime": 67.9478,
+      "eval_samples_per_second": 50.186,
+      "eval_steps_per_second": 1.575,
+      "step": 15350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7884888898294324,
+      "learning_rate": 2.775599073401527e-06,
+      "loss": 0.4553,
+      "step": 15351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4632919188863458,
+      "learning_rate": 2.7750808947533137e-06,
+      "loss": 0.3134,
+      "step": 15352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.79910487039987,
+      "learning_rate": 2.774562773149588e-06,
+      "loss": 0.4423,
+      "step": 15353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9311641941621365,
+      "learning_rate": 2.7740447086011977e-06,
+      "loss": 0.4856,
+      "step": 15354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21998323084874,
+      "learning_rate": 2.773526701118989e-06,
+      "loss": 0.5888,
+      "step": 15355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6227262030552618,
+      "learning_rate": 2.7730087507138054e-06,
+      "loss": 0.4071,
+      "step": 15356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.567463464245594,
+      "learning_rate": 2.7724908573964892e-06,
+      "loss": 0.4268,
+      "step": 15357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9154810134973312,
+      "learning_rate": 2.771973021177884e-06,
+      "loss": 0.5921,
+      "step": 15358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.754352193105281,
+      "learning_rate": 2.771455242068828e-06,
+      "loss": 0.459,
+      "step": 15359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6896133263028303,
+      "learning_rate": 2.770937520080161e-06,
+      "loss": 0.5733,
+      "step": 15360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4850811761808949,
+      "learning_rate": 2.7704198552227247e-06,
+      "loss": 0.2579,
+      "step": 15361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6745349607817415,
+      "learning_rate": 2.76990224750735e-06,
+      "loss": 0.4833,
+      "step": 15362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5101678672240755,
+      "learning_rate": 2.769384696944878e-06,
+      "loss": 0.3677,
+      "step": 15363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6248090392825678,
+      "learning_rate": 2.7688672035461393e-06,
+      "loss": 0.4766,
+      "step": 15364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7343676266203618,
+      "learning_rate": 2.768349767321973e-06,
+      "loss": 0.4542,
+      "step": 15365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7822480139880827,
+      "learning_rate": 2.7678323882832048e-06,
+      "loss": 0.4767,
+      "step": 15366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1998050916175713,
+      "learning_rate": 2.767315066440669e-06,
+      "loss": 0.6821,
+      "step": 15367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3922767331655106,
+      "learning_rate": 2.766797801805196e-06,
+      "loss": 0.5005,
+      "step": 15368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7980877033418114,
+      "learning_rate": 2.766280594387612e-06,
+      "loss": 0.5356,
+      "step": 15369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15097341349683,
+      "learning_rate": 2.765763444198746e-06,
+      "loss": 0.3947,
+      "step": 15370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.83319649656577,
+      "learning_rate": 2.7652463512494253e-06,
+      "loss": 0.5543,
+      "step": 15371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4591974119186748,
+      "learning_rate": 2.7647293155504716e-06,
+      "loss": 0.3562,
+      "step": 15372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8769070386977873,
+      "learning_rate": 2.764212337112712e-06,
+      "loss": 0.4653,
+      "step": 15373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1618099174585663,
+      "learning_rate": 2.7636954159469687e-06,
+      "loss": 0.6556,
+      "step": 15374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7060327659744274,
+      "learning_rate": 2.763178552064061e-06,
+      "loss": 0.5648,
+      "step": 15375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8856943114794087,
+      "learning_rate": 2.7626617454748105e-06,
+      "loss": 0.5158,
+      "step": 15376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5276889167350398,
+      "learning_rate": 2.7621449961900356e-06,
+      "loss": 0.3794,
+      "step": 15377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.913195006465422,
+      "learning_rate": 2.7616283042205565e-06,
+      "loss": 0.5642,
+      "step": 15378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4322699995536852,
+      "learning_rate": 2.761111669577187e-06,
+      "loss": 0.3189,
+      "step": 15379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9679013408151067,
+      "learning_rate": 2.7605950922707425e-06,
+      "loss": 0.5132,
+      "step": 15380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.770674387614859,
+      "learning_rate": 2.76007857231204e-06,
+      "loss": 0.4667,
+      "step": 15381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.007988183031336,
+      "learning_rate": 2.7595621097118896e-06,
+      "loss": 0.5662,
+      "step": 15382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6960302708559822,
+      "learning_rate": 2.7590457044811037e-06,
+      "loss": 0.4411,
+      "step": 15383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7438719683782435,
+      "learning_rate": 2.7585293566304954e-06,
+      "loss": 0.4831,
+      "step": 15384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.975849702207441,
+      "learning_rate": 2.758013066170871e-06,
+      "loss": 0.5512,
+      "step": 15385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6385598503562375,
+      "learning_rate": 2.7574968331130396e-06,
+      "loss": 0.5065,
+      "step": 15386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6321183271210449,
+      "learning_rate": 2.7569806574678083e-06,
+      "loss": 0.3362,
+      "step": 15387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.763998159958472,
+      "learning_rate": 2.756464539245983e-06,
+      "loss": 0.4702,
+      "step": 15388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.83340781259504,
+      "learning_rate": 2.7559484784583694e-06,
+      "loss": 0.6689,
+      "step": 15389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2720823181918197,
+      "learning_rate": 2.7554324751157713e-06,
+      "loss": 0.5949,
+      "step": 15390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0110799717365877,
+      "learning_rate": 2.754916529228988e-06,
+      "loss": 0.5113,
+      "step": 15391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5865840974510341,
+      "learning_rate": 2.754400640808822e-06,
+      "loss": 0.4414,
+      "step": 15392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5633008310270027,
+      "learning_rate": 2.7538848098660754e-06,
+      "loss": 0.3674,
+      "step": 15393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9384491406095914,
+      "learning_rate": 2.7533690364115425e-06,
+      "loss": 0.4513,
+      "step": 15394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7935953636743482,
+      "learning_rate": 2.7528533204560226e-06,
+      "loss": 0.3965,
+      "step": 15395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2257807688621416,
+      "learning_rate": 2.752337662010312e-06,
+      "loss": 0.2658,
+      "step": 15396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0775308593373905,
+      "learning_rate": 2.751822061085208e-06,
+      "loss": 0.5162,
+      "step": 15397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.038837721899298,
+      "learning_rate": 2.7513065176915e-06,
+      "loss": 0.4092,
+      "step": 15398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3943055049679267,
+      "learning_rate": 2.7507910318399837e-06,
+      "loss": 0.6421,
+      "step": 15399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5210656111664993,
+      "learning_rate": 2.75027560354145e-06,
+      "loss": 0.4744,
+      "step": 15400
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.49856796860694885,
+      "eval_runtime": 67.9626,
+      "eval_samples_per_second": 50.175,
+      "eval_steps_per_second": 1.574,
+      "step": 15400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6459166415638726,
+      "learning_rate": 2.749760232806688e-06,
+      "loss": 0.4096,
+      "step": 15401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0101515233604736,
+      "learning_rate": 2.7492449196464867e-06,
+      "loss": 0.4675,
+      "step": 15402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7829060942397392,
+      "learning_rate": 2.7487296640716366e-06,
+      "loss": 0.4785,
+      "step": 15403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6085598399065115,
+      "learning_rate": 2.74821446609292e-06,
+      "loss": 0.5724,
+      "step": 15404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7038734690100088,
+      "learning_rate": 2.747699325721125e-06,
+      "loss": 0.5045,
+      "step": 15405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.191086587628478,
+      "learning_rate": 2.747184242967036e-06,
+      "loss": 0.6443,
+      "step": 15406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8647983179544516,
+      "learning_rate": 2.7466692178414336e-06,
+      "loss": 0.4067,
+      "step": 15407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.097561928333045,
+      "learning_rate": 2.7461542503551007e-06,
+      "loss": 0.6674,
+      "step": 15408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4202867346652805,
+      "learning_rate": 2.7456393405188187e-06,
+      "loss": 0.3546,
+      "step": 15409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6417145379959739,
+      "learning_rate": 2.7451244883433673e-06,
+      "loss": 0.442,
+      "step": 15410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0006168372886073,
+      "learning_rate": 2.744609693839522e-06,
+      "loss": 0.5874,
+      "step": 15411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.848460206785893,
+      "learning_rate": 2.7440949570180625e-06,
+      "loss": 0.5756,
+      "step": 15412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0013263720007295,
+      "learning_rate": 2.7435802778897634e-06,
+      "loss": 0.6277,
+      "step": 15413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.474811738891434,
+      "learning_rate": 2.743065656465399e-06,
+      "loss": 0.3737,
+      "step": 15414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6584916639272815,
+      "learning_rate": 2.742551092755742e-06,
+      "loss": 0.3855,
+      "step": 15415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5050820106582314,
+      "learning_rate": 2.742036586771566e-06,
+      "loss": 0.3463,
+      "step": 15416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3834778023006697,
+      "learning_rate": 2.74152213852364e-06,
+      "loss": 0.3799,
+      "step": 15417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.893358877586104,
+      "learning_rate": 2.741007748022735e-06,
+      "loss": 0.4938,
+      "step": 15418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9568714808290253,
+      "learning_rate": 2.74049341527962e-06,
+      "loss": 0.6639,
+      "step": 15419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4026532613584495,
+      "learning_rate": 2.7399791403050608e-06,
+      "loss": 0.3615,
+      "step": 15420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3540063027653857,
+      "learning_rate": 2.7394649231098234e-06,
+      "loss": 0.412,
+      "step": 15421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7978601061724298,
+      "learning_rate": 2.738950763704673e-06,
+      "loss": 0.5827,
+      "step": 15422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9105555575000774,
+      "learning_rate": 2.7384366621003756e-06,
+      "loss": 0.57,
+      "step": 15423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6121607146054986,
+      "learning_rate": 2.737922618307689e-06,
+      "loss": 0.5272,
+      "step": 15424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.42929867124134,
+      "learning_rate": 2.7374086323373773e-06,
+      "loss": 0.3782,
+      "step": 15425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6057578852780516,
+      "learning_rate": 2.7368947042002016e-06,
+      "loss": 0.5601,
+      "step": 15426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5652300614904124,
+      "learning_rate": 2.7363808339069165e-06,
+      "loss": 0.5,
+      "step": 15427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1791116663236787,
+      "learning_rate": 2.7358670214682835e-06,
+      "loss": 0.4562,
+      "step": 15428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8535687666966085,
+      "learning_rate": 2.735353266895059e-06,
+      "loss": 0.563,
+      "step": 15429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5750683862902726,
+      "learning_rate": 2.7348395701979946e-06,
+      "loss": 0.386,
+      "step": 15430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7818252412880968,
+      "learning_rate": 2.7343259313878458e-06,
+      "loss": 0.481,
+      "step": 15431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5220917562113139,
+      "learning_rate": 2.7338123504753677e-06,
+      "loss": 0.4165,
+      "step": 15432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5497871341789253,
+      "learning_rate": 2.733298827471309e-06,
+      "loss": 0.4242,
+      "step": 15433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5667420806985652,
+      "learning_rate": 2.7327853623864198e-06,
+      "loss": 0.4868,
+      "step": 15434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.690940909469493,
+      "learning_rate": 2.7322719552314514e-06,
+      "loss": 0.4069,
+      "step": 15435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.688500635208722,
+      "learning_rate": 2.7317586060171497e-06,
+      "loss": 0.4908,
+      "step": 15436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8916309549402535,
+      "learning_rate": 2.731245314754263e-06,
+      "loss": 0.6897,
+      "step": 15437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.947140073101482,
+      "learning_rate": 2.7307320814535364e-06,
+      "loss": 0.513,
+      "step": 15438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.299770551453077,
+      "learning_rate": 2.7302189061257127e-06,
+      "loss": 0.3926,
+      "step": 15439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.858220504131006,
+      "learning_rate": 2.729705788781536e-06,
+      "loss": 0.4827,
+      "step": 15440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.716278878206975,
+      "learning_rate": 2.729192729431749e-06,
+      "loss": 0.571,
+      "step": 15441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.461480715888988,
+      "learning_rate": 2.7286797280870898e-06,
+      "loss": 0.333,
+      "step": 15442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8394999186148793,
+      "learning_rate": 2.7281667847582992e-06,
+      "loss": 0.5714,
+      "step": 15443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8461289930373737,
+      "learning_rate": 2.7276538994561153e-06,
+      "loss": 0.4127,
+      "step": 15444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5514871179287024,
+      "learning_rate": 2.727141072191276e-06,
+      "loss": 0.537,
+      "step": 15445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6780029838684998,
+      "learning_rate": 2.7266283029745155e-06,
+      "loss": 0.3615,
+      "step": 15446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0056515449912165,
+      "learning_rate": 2.726115591816569e-06,
+      "loss": 0.6677,
+      "step": 15447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.527934493879015,
+      "learning_rate": 2.72560293872817e-06,
+      "loss": 0.3867,
+      "step": 15448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9591326772011666,
+      "learning_rate": 2.7250903437200493e-06,
+      "loss": 0.4174,
+      "step": 15449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3243610304213176,
+      "learning_rate": 2.7245778068029393e-06,
+      "loss": 0.3343,
+      "step": 15450
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.4980265200138092,
+      "eval_runtime": 67.9867,
+      "eval_samples_per_second": 50.157,
+      "eval_steps_per_second": 1.574,
+      "step": 15450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0240249051512804,
+      "learning_rate": 2.72406532798757e-06,
+      "loss": 0.6814,
+      "step": 15451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.779989335373451,
+      "learning_rate": 2.7235529072846675e-06,
+      "loss": 0.6111,
+      "step": 15452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4954943735691033,
+      "learning_rate": 2.7230405447049602e-06,
+      "loss": 0.477,
+      "step": 15453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7942915018434396,
+      "learning_rate": 2.7225282402591757e-06,
+      "loss": 0.3924,
+      "step": 15454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6047948397448142,
+      "learning_rate": 2.722015993958036e-06,
+      "loss": 0.5099,
+      "step": 15455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1105452503426254,
+      "learning_rate": 2.7215038058122663e-06,
+      "loss": 0.566,
+      "step": 15456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6326259157906153,
+      "learning_rate": 2.7209916758325883e-06,
+      "loss": 0.4655,
+      "step": 15457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.065079275482963,
+      "learning_rate": 2.7204796040297243e-06,
+      "loss": 0.5573,
+      "step": 15458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4995225417416107,
+      "learning_rate": 2.719967590414393e-06,
+      "loss": 0.4706,
+      "step": 15459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5158938504588657,
+      "learning_rate": 2.7194556349973124e-06,
+      "loss": 0.4272,
+      "step": 15460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.939254874980533,
+      "learning_rate": 2.7189437377892035e-06,
+      "loss": 0.5382,
+      "step": 15461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.657466597548457,
+      "learning_rate": 2.7184318988007786e-06,
+      "loss": 0.6508,
+      "step": 15462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8550799450183817,
+      "learning_rate": 2.7179201180427533e-06,
+      "loss": 0.5507,
+      "step": 15463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.142721543774673,
+      "learning_rate": 2.7174083955258445e-06,
+      "loss": 0.5496,
+      "step": 15464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9637513108347406,
+      "learning_rate": 2.7168967312607607e-06,
+      "loss": 0.6281,
+      "step": 15465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9938843948855653,
+      "learning_rate": 2.716385125258215e-06,
+      "loss": 0.5475,
+      "step": 15466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5617590187034671,
+      "learning_rate": 2.7158735775289203e-06,
+      "loss": 0.4904,
+      "step": 15467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8572204475462644,
+      "learning_rate": 2.7153620880835805e-06,
+      "loss": 0.6816,
+      "step": 15468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7289305716715315,
+      "learning_rate": 2.7148506569329058e-06,
+      "loss": 0.5398,
+      "step": 15469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6170718927868262,
+      "learning_rate": 2.7143392840876037e-06,
+      "loss": 0.4767,
+      "step": 15470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6227526065481281,
+      "learning_rate": 2.7138279695583784e-06,
+      "loss": 0.4034,
+      "step": 15471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5741121207859015,
+      "learning_rate": 2.7133167133559333e-06,
+      "loss": 0.3822,
+      "step": 15472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.433738231271735,
+      "learning_rate": 2.712805515490972e-06,
+      "loss": 0.3641,
+      "step": 15473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8921132098074103,
+      "learning_rate": 2.7122943759741983e-06,
+      "loss": 0.6376,
+      "step": 15474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828324266825289,
+      "learning_rate": 2.711783294816307e-06,
+      "loss": 0.4629,
+      "step": 15475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6607166524425148,
+      "learning_rate": 2.7112722720280016e-06,
+      "loss": 0.4064,
+      "step": 15476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.412439187648739,
+      "learning_rate": 2.7107613076199814e-06,
+      "loss": 0.4981,
+      "step": 15477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0701626751274023,
+      "learning_rate": 2.710250401602938e-06,
+      "loss": 0.3951,
+      "step": 15478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.512843903277575,
+      "learning_rate": 2.7097395539875704e-06,
+      "loss": 0.401,
+      "step": 15479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7485387946506676,
+      "learning_rate": 2.7092287647845727e-06,
+      "loss": 0.5457,
+      "step": 15480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6458167517551654,
+      "learning_rate": 2.708718034004637e-06,
+      "loss": 0.5083,
+      "step": 15481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8397938433761387,
+      "learning_rate": 2.7082073616584547e-06,
+      "loss": 0.4476,
+      "step": 15482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.003313615338251,
+      "learning_rate": 2.7076967477567173e-06,
+      "loss": 0.6181,
+      "step": 15483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.631022983957819,
+      "learning_rate": 2.707186192310116e-06,
+      "loss": 0.3492,
+      "step": 15484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8773098265182884,
+      "learning_rate": 2.706675695329334e-06,
+      "loss": 0.4513,
+      "step": 15485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6068578260011543,
+      "learning_rate": 2.706165256825064e-06,
+      "loss": 0.3967,
+      "step": 15486
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0192513742101927,
+      "learning_rate": 2.705654876807988e-06,
+      "loss": 0.3899,
+      "step": 15487
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4636830673505523,
+      "learning_rate": 2.7051445552887917e-06,
+      "loss": 0.432,
+      "step": 15488
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7011824109237905,
+      "learning_rate": 2.7046342922781578e-06,
+      "loss": 0.5407,
+      "step": 15489
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7981484128938623,
+      "learning_rate": 2.7041240877867703e-06,
+      "loss": 0.5912,
+      "step": 15490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7177842884599377,
+      "learning_rate": 2.7036139418253073e-06,
+      "loss": 0.4028,
+      "step": 15491
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5693599150578066,
+      "learning_rate": 2.7031038544044487e-06,
+      "loss": 0.4215,
+      "step": 15492
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8565256018842584,
+      "learning_rate": 2.702593825534876e-06,
+      "loss": 0.5877,
+      "step": 15493
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6101489490216818,
+      "learning_rate": 2.7020838552272622e-06,
+      "loss": 0.4373,
+      "step": 15494
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1272812629080704,
+      "learning_rate": 2.701573943492285e-06,
+      "loss": 0.5915,
+      "step": 15495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5030494701123274,
+      "learning_rate": 2.7010640903406206e-06,
+      "loss": 0.4632,
+      "step": 15496
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.067693265849465,
+      "learning_rate": 2.7005542957829387e-06,
+      "loss": 0.5432,
+      "step": 15497
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1095968943150094,
+      "learning_rate": 2.700044559829915e-06,
+      "loss": 0.4797,
+      "step": 15498
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7917075587706084,
+      "learning_rate": 2.699534882492221e-06,
+      "loss": 0.499,
+      "step": 15499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8455679427194156,
+      "learning_rate": 2.6990252637805224e-06,
+      "loss": 0.4712,
+      "step": 15500
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.49789607524871826,
+      "eval_runtime": 67.9469,
+      "eval_samples_per_second": 50.186,
+      "eval_steps_per_second": 1.575,
+      "step": 15500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8557586948548088,
+      "learning_rate": 2.6985157037054903e-06,
+      "loss": 0.4501,
+      "step": 15501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.548067141460719,
+      "learning_rate": 2.698006202277792e-06,
+      "loss": 0.463,
+      "step": 15502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6607070159725243,
+      "learning_rate": 2.6974967595080946e-06,
+      "loss": 0.4217,
+      "step": 15503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2163982245522478,
+      "learning_rate": 2.696987375407062e-06,
+      "loss": 0.6811,
+      "step": 15504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.864697828326227,
+      "learning_rate": 2.6964780499853562e-06,
+      "loss": 0.4853,
+      "step": 15505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6017481724377831,
+      "learning_rate": 2.695968783253643e-06,
+      "loss": 0.3886,
+      "step": 15506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.608289871556866,
+      "learning_rate": 2.69545957522258e-06,
+      "loss": 0.4079,
+      "step": 15507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4802886871545335,
+      "learning_rate": 2.694950425902829e-06,
+      "loss": 0.3812,
+      "step": 15508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.764385062435949,
+      "learning_rate": 2.6944413353050493e-06,
+      "loss": 0.4876,
+      "step": 15509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9610824933878643,
+      "learning_rate": 2.693932303439897e-06,
+      "loss": 0.4012,
+      "step": 15510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6073901949968512,
+      "learning_rate": 2.6934233303180283e-06,
+      "loss": 0.4847,
+      "step": 15511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5233439565888232,
+      "learning_rate": 2.6929144159501016e-06,
+      "loss": 0.3763,
+      "step": 15512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5329236085173328,
+      "learning_rate": 2.6924055603467656e-06,
+      "loss": 0.3952,
+      "step": 15513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8378603144867172,
+      "learning_rate": 2.6918967635186754e-06,
+      "loss": 0.6287,
+      "step": 15514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8530320947261165,
+      "learning_rate": 2.6913880254764826e-06,
+      "loss": 0.518,
+      "step": 15515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9138546129017064,
+      "learning_rate": 2.6908793462308387e-06,
+      "loss": 0.7513,
+      "step": 15516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.814903625558591,
+      "learning_rate": 2.690370725792389e-06,
+      "loss": 0.4111,
+      "step": 15517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8495627604459977,
+      "learning_rate": 2.689862164171783e-06,
+      "loss": 0.6759,
+      "step": 15518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7248102341817366,
+      "learning_rate": 2.689353661379669e-06,
+      "loss": 0.5688,
+      "step": 15519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7001663071522413,
+      "learning_rate": 2.688845217426688e-06,
+      "loss": 0.4358,
+      "step": 15520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7094455012379541,
+      "learning_rate": 2.688336832323487e-06,
+      "loss": 0.477,
+      "step": 15521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.749262688413012,
+      "learning_rate": 2.6878285060807097e-06,
+      "loss": 0.5598,
+      "step": 15522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.001157027929667,
+      "learning_rate": 2.687320238708994e-06,
+      "loss": 0.712,
+      "step": 15523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3821617232664358,
+      "learning_rate": 2.6868120302189826e-06,
+      "loss": 0.3741,
+      "step": 15524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6586505011209125,
+      "learning_rate": 2.686303880621315e-06,
+      "loss": 0.4389,
+      "step": 15525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.329081275981803,
+      "learning_rate": 2.685795789926627e-06,
+      "loss": 0.4532,
+      "step": 15526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5622869105778594,
+      "learning_rate": 2.6852877581455548e-06,
+      "loss": 0.3957,
+      "step": 15527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9256118069257515,
+      "learning_rate": 2.684779785288737e-06,
+      "loss": 0.5236,
+      "step": 15528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.621483335788959,
+      "learning_rate": 2.684271871366806e-06,
+      "loss": 0.4659,
+      "step": 15529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5511222733063459,
+      "learning_rate": 2.683764016390393e-06,
+      "loss": 0.4803,
+      "step": 15530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5216516304188707,
+      "learning_rate": 2.6832562203701316e-06,
+      "loss": 0.4666,
+      "step": 15531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4130828426239719,
+      "learning_rate": 2.6827484833166535e-06,
+      "loss": 0.4877,
+      "step": 15532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9578036112763284,
+      "learning_rate": 2.682240805240584e-06,
+      "loss": 0.4429,
+      "step": 15533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6541595077575013,
+      "learning_rate": 2.681733186152553e-06,
+      "loss": 0.329,
+      "step": 15534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7745344795712181,
+      "learning_rate": 2.6812256260631876e-06,
+      "loss": 0.6389,
+      "step": 15535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9854340465189273,
+      "learning_rate": 2.6807181249831127e-06,
+      "loss": 0.6121,
+      "step": 15536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9123415251842681,
+      "learning_rate": 2.6802106829229523e-06,
+      "loss": 0.4295,
+      "step": 15537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7504483446463608,
+      "learning_rate": 2.6797032998933323e-06,
+      "loss": 0.4687,
+      "step": 15538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.892432539874238,
+      "learning_rate": 2.6791959759048696e-06,
+      "loss": 0.5914,
+      "step": 15539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6248633451207897,
+      "learning_rate": 2.678688710968188e-06,
+      "loss": 0.3592,
+      "step": 15540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1762543408464405,
+      "learning_rate": 2.6781815050939065e-06,
+      "loss": 0.514,
+      "step": 15541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.639206244679123,
+      "learning_rate": 2.677674358292641e-06,
+      "loss": 0.6425,
+      "step": 15542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9080842712124813,
+      "learning_rate": 2.67716727057501e-06,
+      "loss": 0.426,
+      "step": 15543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8277249126261281,
+      "learning_rate": 2.6766602419516307e-06,
+      "loss": 0.3598,
+      "step": 15544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7885535785156772,
+      "learning_rate": 2.676153272433114e-06,
+      "loss": 0.471,
+      "step": 15545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7642963946800525,
+      "learning_rate": 2.6756463620300744e-06,
+      "loss": 0.4853,
+      "step": 15546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7057598768557567,
+      "learning_rate": 2.6751395107531253e-06,
+      "loss": 0.4557,
+      "step": 15547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.526497190130448,
+      "learning_rate": 2.674632718612874e-06,
+      "loss": 0.5008,
+      "step": 15548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.027533430194497,
+      "learning_rate": 2.674125985619933e-06,
+      "loss": 0.4906,
+      "step": 15549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9569775901036741,
+      "learning_rate": 2.6736193117849086e-06,
+      "loss": 0.6314,
+      "step": 15550
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.4975833296775818,
+      "eval_runtime": 67.9189,
+      "eval_samples_per_second": 50.207,
+      "eval_steps_per_second": 1.575,
+      "step": 15550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0711054850137613,
+      "learning_rate": 2.67311269711841e-06,
+      "loss": 0.5733,
+      "step": 15551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.389819372959665,
+      "learning_rate": 2.67260614163104e-06,
+      "loss": 0.6015,
+      "step": 15552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7697084053224401,
+      "learning_rate": 2.6720996453334037e-06,
+      "loss": 0.5288,
+      "step": 15553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.662736552092786,
+      "learning_rate": 2.671593208236107e-06,
+      "loss": 0.5376,
+      "step": 15554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8544385286001364,
+      "learning_rate": 2.671086830349747e-06,
+      "loss": 0.5925,
+      "step": 15555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4029403911618736,
+      "learning_rate": 2.6705805116849282e-06,
+      "loss": 0.5554,
+      "step": 15556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7230730399815641,
+      "learning_rate": 2.6700742522522505e-06,
+      "loss": 0.5191,
+      "step": 15557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7688418760419369,
+      "learning_rate": 2.669568052062309e-06,
+      "loss": 0.4632,
+      "step": 15558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.435974882322232,
+      "learning_rate": 2.6690619111257022e-06,
+      "loss": 0.4702,
+      "step": 15559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7558055184864887,
+      "learning_rate": 2.6685558294530274e-06,
+      "loss": 0.5032,
+      "step": 15560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6583140687231281,
+      "learning_rate": 2.668049807054876e-06,
+      "loss": 0.4159,
+      "step": 15561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6384247357517234,
+      "learning_rate": 2.667543843941843e-06,
+      "loss": 0.4067,
+      "step": 15562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5192476461040918,
+      "learning_rate": 2.66703794012452e-06,
+      "loss": 0.5174,
+      "step": 15563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6365063894412308,
+      "learning_rate": 2.6665320956135e-06,
+      "loss": 0.4076,
+      "step": 15564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.488129851293008,
+      "learning_rate": 2.666026310419369e-06,
+      "loss": 0.506,
+      "step": 15565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7112223068356516,
+      "learning_rate": 2.6655205845527164e-06,
+      "loss": 0.4738,
+      "step": 15566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3899720422276083,
+      "learning_rate": 2.6650149180241324e-06,
+      "loss": 0.3434,
+      "step": 15567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9312888046973324,
+      "learning_rate": 2.664509310844198e-06,
+      "loss": 0.6264,
+      "step": 15568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7795640699308801,
+      "learning_rate": 2.6640037630235005e-06,
+      "loss": 0.4904,
+      "step": 15569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9048291987077368,
+      "learning_rate": 2.663498274572624e-06,
+      "loss": 0.6106,
+      "step": 15570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7838725264211956,
+      "learning_rate": 2.6629928455021473e-06,
+      "loss": 0.5284,
+      "step": 15571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5047553641314897,
+      "learning_rate": 2.6624874758226544e-06,
+      "loss": 0.3793,
+      "step": 15572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6272730579100807,
+      "learning_rate": 2.6619821655447227e-06,
+      "loss": 0.4795,
+      "step": 15573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5136600100923852,
+      "learning_rate": 2.661476914678933e-06,
+      "loss": 0.6348,
+      "step": 15574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5158847324552824,
+      "learning_rate": 2.66097172323586e-06,
+      "loss": 0.4286,
+      "step": 15575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.679587413639298,
+      "learning_rate": 2.660466591226081e-06,
+      "loss": 0.537,
+      "step": 15576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6369439549198672,
+      "learning_rate": 2.6599615186601714e-06,
+      "loss": 0.4459,
+      "step": 15577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0845917818164383,
+      "learning_rate": 2.659456505548701e-06,
+      "loss": 0.5509,
+      "step": 15578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8653549683394668,
+      "learning_rate": 2.658951551902246e-06,
+      "loss": 0.5809,
+      "step": 15579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1468277985735207,
+      "learning_rate": 2.6584466577313757e-06,
+      "loss": 0.4521,
+      "step": 15580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5913520237605714,
+      "learning_rate": 2.6579418230466586e-06,
+      "loss": 0.3833,
+      "step": 15581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.354048438725766,
+      "learning_rate": 2.6574370478586644e-06,
+      "loss": 0.6546,
+      "step": 15582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1959330119177793,
+      "learning_rate": 2.656932332177961e-06,
+      "loss": 0.4596,
+      "step": 15583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9493557832169117,
+      "learning_rate": 2.6564276760151108e-06,
+      "loss": 0.4772,
+      "step": 15584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.826319700632291,
+      "learning_rate": 2.6559230793806823e-06,
+      "loss": 0.4544,
+      "step": 15585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4060380358865827,
+      "learning_rate": 2.655418542285239e-06,
+      "loss": 0.3542,
+      "step": 15586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.586558813991537,
+      "learning_rate": 2.6549140647393407e-06,
+      "loss": 0.4127,
+      "step": 15587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5167818138054476,
+      "learning_rate": 2.6544096467535485e-06,
+      "loss": 0.4249,
+      "step": 15588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0903503186694707,
+      "learning_rate": 2.6539052883384255e-06,
+      "loss": 0.5633,
+      "step": 15589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4184264392513053,
+      "learning_rate": 2.6534009895045254e-06,
+      "loss": 0.3591,
+      "step": 15590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6547773665645826,
+      "learning_rate": 2.6528967502624077e-06,
+      "loss": 0.4201,
+      "step": 15591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7806376287585322,
+      "learning_rate": 2.652392570622629e-06,
+      "loss": 0.5172,
+      "step": 15592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9989037103807334,
+      "learning_rate": 2.651888450595742e-06,
+      "loss": 0.5144,
+      "step": 15593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.867768419461976,
+      "learning_rate": 2.6513843901923007e-06,
+      "loss": 0.6478,
+      "step": 15594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7058762460572505,
+      "learning_rate": 2.6508803894228586e-06,
+      "loss": 0.627,
+      "step": 15595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2900336353195208,
+      "learning_rate": 2.6503764482979677e-06,
+      "loss": 0.4908,
+      "step": 15596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5466574729896172,
+      "learning_rate": 2.6498725668281734e-06,
+      "loss": 0.3983,
+      "step": 15597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8758000049086632,
+      "learning_rate": 2.649368745024028e-06,
+      "loss": 0.5904,
+      "step": 15598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8766271259624343,
+      "learning_rate": 2.648864982896078e-06,
+      "loss": 0.4893,
+      "step": 15599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6630791967268932,
+      "learning_rate": 2.648361280454867e-06,
+      "loss": 0.5266,
+      "step": 15600
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.49660542607307434,
+      "eval_runtime": 67.9388,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 15600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7273357672283736,
+      "learning_rate": 2.647857637710941e-06,
+      "loss": 0.5461,
+      "step": 15601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7187817630890418,
+      "learning_rate": 2.6473540546748462e-06,
+      "loss": 0.5827,
+      "step": 15602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6227200510820898,
+      "learning_rate": 2.6468505313571203e-06,
+      "loss": 0.5079,
+      "step": 15603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6879534913290217,
+      "learning_rate": 2.646347067768307e-06,
+      "loss": 0.4794,
+      "step": 15604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4599854857654577,
+      "learning_rate": 2.6458436639189467e-06,
+      "loss": 0.3928,
+      "step": 15605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4761758550610615,
+      "learning_rate": 2.6453403198195744e-06,
+      "loss": 0.3284,
+      "step": 15606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9317401128176879,
+      "learning_rate": 2.6448370354807296e-06,
+      "loss": 0.6027,
+      "step": 15607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6309622820855663,
+      "learning_rate": 2.6443338109129485e-06,
+      "loss": 0.5641,
+      "step": 15608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8756078694841323,
+      "learning_rate": 2.6438306461267665e-06,
+      "loss": 0.5722,
+      "step": 15609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7991572413446182,
+      "learning_rate": 2.643327541132714e-06,
+      "loss": 0.5181,
+      "step": 15610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3732763181856957,
+      "learning_rate": 2.6428244959413263e-06,
+      "loss": 0.6029,
+      "step": 15611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5231958600955757,
+      "learning_rate": 2.6423215105631337e-06,
+      "loss": 0.4296,
+      "step": 15612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.687594670617563,
+      "learning_rate": 2.641818585008664e-06,
+      "loss": 0.4597,
+      "step": 15613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5243765984251465,
+      "learning_rate": 2.6413157192884477e-06,
+      "loss": 0.4169,
+      "step": 15614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9096641486237442,
+      "learning_rate": 2.640812913413011e-06,
+      "loss": 0.5152,
+      "step": 15615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1421990444027026,
+      "learning_rate": 2.64031016739288e-06,
+      "loss": 0.502,
+      "step": 15616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.006747722301085,
+      "learning_rate": 2.639807481238579e-06,
+      "loss": 0.5832,
+      "step": 15617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6010341273256306,
+      "learning_rate": 2.6393048549606335e-06,
+      "loss": 0.5525,
+      "step": 15618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.227166528892461,
+      "learning_rate": 2.6388022885695626e-06,
+      "loss": 0.7145,
+      "step": 15619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9178947512690556,
+      "learning_rate": 2.6382997820758883e-06,
+      "loss": 0.4392,
+      "step": 15620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1186144702553977,
+      "learning_rate": 2.63779733549013e-06,
+      "loss": 0.4936,
+      "step": 15621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.42235572846206,
+      "learning_rate": 2.6372949488228095e-06,
+      "loss": 0.2884,
+      "step": 15622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.849834547544323,
+      "learning_rate": 2.6367926220844386e-06,
+      "loss": 0.4809,
+      "step": 15623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7070777904147088,
+      "learning_rate": 2.6362903552855356e-06,
+      "loss": 0.372,
+      "step": 15624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5880848621226187,
+      "learning_rate": 2.6357881484366184e-06,
+      "loss": 0.4715,
+      "step": 15625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6979923499286054,
+      "learning_rate": 2.635286001548194e-06,
+      "loss": 0.4561,
+      "step": 15626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6333556385591002,
+      "learning_rate": 2.634783914630779e-06,
+      "loss": 0.4642,
+      "step": 15627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3564112783147864,
+      "learning_rate": 2.6342818876948838e-06,
+      "loss": 0.5761,
+      "step": 15628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.932934494030656,
+      "learning_rate": 2.6337799207510168e-06,
+      "loss": 0.4325,
+      "step": 15629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8978544843180767,
+      "learning_rate": 2.633278013809686e-06,
+      "loss": 0.4234,
+      "step": 15630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6415439788678452,
+      "learning_rate": 2.632776166881401e-06,
+      "loss": 0.4802,
+      "step": 15631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9880357342922272,
+      "learning_rate": 2.6322743799766647e-06,
+      "loss": 0.7404,
+      "step": 15632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7586075748996473,
+      "learning_rate": 2.6317726531059818e-06,
+      "loss": 0.4339,
+      "step": 15633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8566785045381162,
+      "learning_rate": 2.631270986279859e-06,
+      "loss": 0.4357,
+      "step": 15634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6618571318689546,
+      "learning_rate": 2.6307693795087965e-06,
+      "loss": 0.5481,
+      "step": 15635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6902191094957146,
+      "learning_rate": 2.6302678328032937e-06,
+      "loss": 0.4117,
+      "step": 15636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.679429468054857,
+      "learning_rate": 2.6297663461738526e-06,
+      "loss": 0.5369,
+      "step": 15637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.127209075127932,
+      "learning_rate": 2.6292649196309695e-06,
+      "loss": 0.5875,
+      "step": 15638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.755195894714747,
+      "learning_rate": 2.6287635531851415e-06,
+      "loss": 0.4806,
+      "step": 15639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5011417017637367,
+      "learning_rate": 2.6282622468468665e-06,
+      "loss": 0.447,
+      "step": 15640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.819885476021574,
+      "learning_rate": 2.6277610006266376e-06,
+      "loss": 0.4292,
+      "step": 15641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6497733905155363,
+      "learning_rate": 2.6272598145349466e-06,
+      "loss": 0.4628,
+      "step": 15642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7403262134933157,
+      "learning_rate": 2.6267586885822878e-06,
+      "loss": 0.527,
+      "step": 15643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.118535992565973,
+      "learning_rate": 2.6262576227791525e-06,
+      "loss": 0.4519,
+      "step": 15644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8598839038495916,
+      "learning_rate": 2.6257566171360275e-06,
+      "loss": 0.497,
+      "step": 15645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2000150058766526,
+      "learning_rate": 2.6252556716634025e-06,
+      "loss": 0.4923,
+      "step": 15646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7958690853900163,
+      "learning_rate": 2.624754786371766e-06,
+      "loss": 0.4914,
+      "step": 15647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7406419493776422,
+      "learning_rate": 2.6242539612716e-06,
+      "loss": 0.321,
+      "step": 15648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6795103040991899,
+      "learning_rate": 2.623753196373392e-06,
+      "loss": 0.499,
+      "step": 15649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8680598428034358,
+      "learning_rate": 2.6232524916876257e-06,
+      "loss": 0.6476,
+      "step": 15650
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.4963325262069702,
+      "eval_runtime": 67.9423,
+      "eval_samples_per_second": 50.19,
+      "eval_steps_per_second": 1.575,
+      "step": 15650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9538794992539812,
+      "learning_rate": 2.6227518472247797e-06,
+      "loss": 0.5968,
+      "step": 15651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6787260844025498,
+      "learning_rate": 2.6222512629953373e-06,
+      "loss": 0.4336,
+      "step": 15652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1520487919025366,
+      "learning_rate": 2.6217507390097775e-06,
+      "loss": 0.3855,
+      "step": 15653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4734084229716309,
+      "learning_rate": 2.6212502752785785e-06,
+      "loss": 0.3715,
+      "step": 15654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7368730408808926,
+      "learning_rate": 2.620749871812216e-06,
+      "loss": 0.5797,
+      "step": 15655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.477055417540747,
+      "learning_rate": 2.620249528621167e-06,
+      "loss": 0.3783,
+      "step": 15656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5189662266076764,
+      "learning_rate": 2.6197492457159066e-06,
+      "loss": 0.4636,
+      "step": 15657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6410588735992273,
+      "learning_rate": 2.6192490231069044e-06,
+      "loss": 0.4066,
+      "step": 15658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.698470605966907,
+      "learning_rate": 2.618748860804636e-06,
+      "loss": 0.5364,
+      "step": 15659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7873382008626035,
+      "learning_rate": 2.6182487588195715e-06,
+      "loss": 0.4451,
+      "step": 15660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.977879909096002,
+      "learning_rate": 2.617748717162178e-06,
+      "loss": 0.4632,
+      "step": 15661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7365821820794416,
+      "learning_rate": 2.6172487358429244e-06,
+      "loss": 0.3481,
+      "step": 15662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9619639430392826,
+      "learning_rate": 2.6167488148722797e-06,
+      "loss": 0.5447,
+      "step": 15663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9860732303454882,
+      "learning_rate": 2.616248954260706e-06,
+      "loss": 0.5139,
+      "step": 15664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7675774691066826,
+      "learning_rate": 2.6157491540186707e-06,
+      "loss": 0.534,
+      "step": 15665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5440010601399692,
+      "learning_rate": 2.6152494141566336e-06,
+      "loss": 0.5009,
+      "step": 15666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2772882081702643,
+      "learning_rate": 2.6147497346850603e-06,
+      "loss": 0.2989,
+      "step": 15667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4576360498159113,
+      "learning_rate": 2.6142501156144086e-06,
+      "loss": 0.4847,
+      "step": 15668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8652940826802848,
+      "learning_rate": 2.613750556955137e-06,
+      "loss": 0.6091,
+      "step": 15669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6847831934191229,
+      "learning_rate": 2.6132510587177073e-06,
+      "loss": 0.4384,
+      "step": 15670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7871793091113508,
+      "learning_rate": 2.612751620912571e-06,
+      "loss": 0.506,
+      "step": 15671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8551238660554652,
+      "learning_rate": 2.6122522435501873e-06,
+      "loss": 0.583,
+      "step": 15672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.66152036083885,
+      "learning_rate": 2.61175292664101e-06,
+      "loss": 0.3652,
+      "step": 15673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7087255670122719,
+      "learning_rate": 2.61125367019549e-06,
+      "loss": 0.4055,
+      "step": 15674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9567397145200014,
+      "learning_rate": 2.61075447422408e-06,
+      "loss": 0.6224,
+      "step": 15675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9571051015210992,
+      "learning_rate": 2.610255338737232e-06,
+      "loss": 0.534,
+      "step": 15676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7927018640975743,
+      "learning_rate": 2.609756263745392e-06,
+      "loss": 0.4401,
+      "step": 15677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6249486551435492,
+      "learning_rate": 2.6092572492590095e-06,
+      "loss": 0.4159,
+      "step": 15678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7922421742076033,
+      "learning_rate": 2.6087582952885315e-06,
+      "loss": 0.4658,
+      "step": 15679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9508398848983843,
+      "learning_rate": 2.6082594018444043e-06,
+      "loss": 0.5088,
+      "step": 15680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0189201651177635,
+      "learning_rate": 2.6077605689370687e-06,
+      "loss": 0.5596,
+      "step": 15681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6939839180643987,
+      "learning_rate": 2.6072617965769687e-06,
+      "loss": 0.4579,
+      "step": 15682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7847000384019376,
+      "learning_rate": 2.606763084774546e-06,
+      "loss": 0.5161,
+      "step": 15683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.748903043755019,
+      "learning_rate": 2.6062644335402423e-06,
+      "loss": 0.5954,
+      "step": 15684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8565635791900708,
+      "learning_rate": 2.6057658428844956e-06,
+      "loss": 0.5034,
+      "step": 15685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.706824354995179,
+      "learning_rate": 2.6052673128177424e-06,
+      "loss": 0.4662,
+      "step": 15686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4529039652634677,
+      "learning_rate": 2.60476884335042e-06,
+      "loss": 0.4289,
+      "step": 15687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.272062086350275,
+      "learning_rate": 2.6042704344929633e-06,
+      "loss": 0.511,
+      "step": 15688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6152923833426345,
+      "learning_rate": 2.6037720862558087e-06,
+      "loss": 0.4836,
+      "step": 15689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6892841292009633,
+      "learning_rate": 2.603273798649384e-06,
+      "loss": 0.3829,
+      "step": 15690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.646722679514898,
+      "learning_rate": 2.6027755716841243e-06,
+      "loss": 0.4688,
+      "step": 15691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5379356194594511,
+      "learning_rate": 2.6022774053704593e-06,
+      "loss": 0.4107,
+      "step": 15692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9821769834634804,
+      "learning_rate": 2.6017792997188152e-06,
+      "loss": 0.5332,
+      "step": 15693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7089417891817735,
+      "learning_rate": 2.6012812547396215e-06,
+      "loss": 0.4699,
+      "step": 15694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7963417960251016,
+      "learning_rate": 2.6007832704433055e-06,
+      "loss": 0.4843,
+      "step": 15695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6143859498340576,
+      "learning_rate": 2.60028534684029e-06,
+      "loss": 0.3826,
+      "step": 15696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.57232039962926,
+      "learning_rate": 2.599787483940999e-06,
+      "loss": 0.3667,
+      "step": 15697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4566101217753398,
+      "learning_rate": 2.599289681755858e-06,
+      "loss": 0.4722,
+      "step": 15698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4852872936513006,
+      "learning_rate": 2.5987919402952833e-06,
+      "loss": 0.4176,
+      "step": 15699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6930725356979628,
+      "learning_rate": 2.598294259569698e-06,
+      "loss": 0.4914,
+      "step": 15700
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.49588754773139954,
+      "eval_runtime": 67.905,
+      "eval_samples_per_second": 50.217,
+      "eval_steps_per_second": 1.576,
+      "step": 15700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9125759477182551,
+      "learning_rate": 2.5977966395895202e-06,
+      "loss": 0.502,
+      "step": 15701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7630880854575741,
+      "learning_rate": 2.5972990803651688e-06,
+      "loss": 0.5438,
+      "step": 15702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5404078352808759,
+      "learning_rate": 2.596801581907057e-06,
+      "loss": 0.6797,
+      "step": 15703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7508843055392729,
+      "learning_rate": 2.5963041442256e-06,
+      "loss": 0.531,
+      "step": 15704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6649959493680955,
+      "learning_rate": 2.595806767331214e-06,
+      "loss": 0.5509,
+      "step": 15705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7233920878067766,
+      "learning_rate": 2.5953094512343067e-06,
+      "loss": 0.5105,
+      "step": 15706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.682247468006476,
+      "learning_rate": 2.594812195945294e-06,
+      "loss": 0.4763,
+      "step": 15707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8859230103627644,
+      "learning_rate": 2.5943150014745832e-06,
+      "loss": 0.4689,
+      "step": 15708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.492435876467131,
+      "learning_rate": 2.5938178678325827e-06,
+      "loss": 0.4371,
+      "step": 15709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7437733279761,
+      "learning_rate": 2.5933207950297003e-06,
+      "loss": 0.4831,
+      "step": 15710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7683911498223377,
+      "learning_rate": 2.5928237830763413e-06,
+      "loss": 0.4712,
+      "step": 15711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5926058050696192,
+      "learning_rate": 2.5923268319829108e-06,
+      "loss": 0.4997,
+      "step": 15712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7498294787554778,
+      "learning_rate": 2.591829941759812e-06,
+      "loss": 0.434,
+      "step": 15713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7345655465031493,
+      "learning_rate": 2.591333112417447e-06,
+      "loss": 0.5076,
+      "step": 15714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.038273738592888,
+      "learning_rate": 2.5908363439662177e-06,
+      "loss": 0.5469,
+      "step": 15715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9950092818099543,
+      "learning_rate": 2.590339636416521e-06,
+      "loss": 0.5732,
+      "step": 15716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7361799957041208,
+      "learning_rate": 2.5898429897787567e-06,
+      "loss": 0.5409,
+      "step": 15717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5189568589684943,
+      "learning_rate": 2.5893464040633243e-06,
+      "loss": 0.3054,
+      "step": 15718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8634763967147467,
+      "learning_rate": 2.588849879280615e-06,
+      "loss": 0.4622,
+      "step": 15719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8481707307851232,
+      "learning_rate": 2.5883534154410252e-06,
+      "loss": 0.4929,
+      "step": 15720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.024618798690424,
+      "learning_rate": 2.5878570125549495e-06,
+      "loss": 0.4771,
+      "step": 15721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8823185033008174,
+      "learning_rate": 2.587360670632777e-06,
+      "loss": 0.4736,
+      "step": 15722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.76640167866764,
+      "learning_rate": 2.5868643896849e-06,
+      "loss": 0.4257,
+      "step": 15723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9320262658167406,
+      "learning_rate": 2.5863681697217093e-06,
+      "loss": 0.564,
+      "step": 15724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9552573455369333,
+      "learning_rate": 2.5858720107535895e-06,
+      "loss": 0.4992,
+      "step": 15725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8050939395910943,
+      "learning_rate": 2.585375912790928e-06,
+      "loss": 0.6907,
+      "step": 15726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9584794030480428,
+      "learning_rate": 2.5848798758441128e-06,
+      "loss": 0.4875,
+      "step": 15727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1098080432513595,
+      "learning_rate": 2.5843838999235284e-06,
+      "loss": 0.6624,
+      "step": 15728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7545052511028796,
+      "learning_rate": 2.5838879850395538e-06,
+      "loss": 0.4072,
+      "step": 15729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8508095480194626,
+      "learning_rate": 2.5833921312025735e-06,
+      "loss": 0.524,
+      "step": 15730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9215623286638994,
+      "learning_rate": 2.5828963384229667e-06,
+      "loss": 0.6286,
+      "step": 15731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5675888729182241,
+      "learning_rate": 2.582400606711114e-06,
+      "loss": 0.4574,
+      "step": 15732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9093235436952902,
+      "learning_rate": 2.5819049360773916e-06,
+      "loss": 0.5675,
+      "step": 15733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3589511389093625,
+      "learning_rate": 2.5814093265321795e-06,
+      "loss": 0.3869,
+      "step": 15734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8579030409185837,
+      "learning_rate": 2.580913778085847e-06,
+      "loss": 0.3688,
+      "step": 15735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8442087824572986,
+      "learning_rate": 2.580418290748773e-06,
+      "loss": 0.4636,
+      "step": 15736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0997646727941897,
+      "learning_rate": 2.57992286453133e-06,
+      "loss": 0.5207,
+      "step": 15737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6343447528706077,
+      "learning_rate": 2.579427499443887e-06,
+      "loss": 0.4198,
+      "step": 15738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6690632960273708,
+      "learning_rate": 2.5789321954968145e-06,
+      "loss": 0.4338,
+      "step": 15739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9113020243311847,
+      "learning_rate": 2.578436952700484e-06,
+      "loss": 0.5688,
+      "step": 15740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4667168552840353,
+      "learning_rate": 2.5779417710652587e-06,
+      "loss": 0.3501,
+      "step": 15741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3853806525591534,
+      "learning_rate": 2.5774466506015083e-06,
+      "loss": 0.6453,
+      "step": 15742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5829436368725354,
+      "learning_rate": 2.5769515913195986e-06,
+      "loss": 0.5324,
+      "step": 15743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5707560824030493,
+      "learning_rate": 2.57645659322989e-06,
+      "loss": 0.4068,
+      "step": 15744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4814074548203466,
+      "learning_rate": 2.575961656342747e-06,
+      "loss": 0.3917,
+      "step": 15745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0400131218419544,
+      "learning_rate": 2.575466780668531e-06,
+      "loss": 0.5674,
+      "step": 15746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8008835487510937,
+      "learning_rate": 2.574971966217602e-06,
+      "loss": 0.5131,
+      "step": 15747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6961799715831094,
+      "learning_rate": 2.574477213000317e-06,
+      "loss": 0.5136,
+      "step": 15748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.896910372675409,
+      "learning_rate": 2.573982521027034e-06,
+      "loss": 0.4916,
+      "step": 15749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8870289874178268,
+      "learning_rate": 2.5734878903081115e-06,
+      "loss": 0.4637,
+      "step": 15750
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.495881050825119,
+      "eval_runtime": 67.9374,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 15750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5457776900505287,
+      "learning_rate": 2.5729933208539008e-06,
+      "loss": 0.4411,
+      "step": 15751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6871763552589232,
+      "learning_rate": 2.5724988126747565e-06,
+      "loss": 0.5221,
+      "step": 15752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.618700672774756,
+      "learning_rate": 2.5720043657810336e-06,
+      "loss": 0.4761,
+      "step": 15753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5527528267348747,
+      "learning_rate": 2.5715099801830778e-06,
+      "loss": 0.331,
+      "step": 15754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.528175840316567,
+      "learning_rate": 2.571015655891243e-06,
+      "loss": 0.4636,
+      "step": 15755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2172891492026254,
+      "learning_rate": 2.5705213929158767e-06,
+      "loss": 0.4238,
+      "step": 15756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.056343762434034,
+      "learning_rate": 2.570027191267324e-06,
+      "loss": 0.5492,
+      "step": 15757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7210957321519336,
+      "learning_rate": 2.5695330509559325e-06,
+      "loss": 0.572,
+      "step": 15758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0649582559411908,
+      "learning_rate": 2.569038971992047e-06,
+      "loss": 0.541,
+      "step": 15759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4952864081912087,
+      "learning_rate": 2.568544954386011e-06,
+      "loss": 0.4696,
+      "step": 15760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9147557718456651,
+      "learning_rate": 2.5680509981481638e-06,
+      "loss": 0.4599,
+      "step": 15761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.802935872592185,
+      "learning_rate": 2.5675571032888485e-06,
+      "loss": 0.5909,
+      "step": 15762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5825402911784952,
+      "learning_rate": 2.567063269818405e-06,
+      "loss": 0.5001,
+      "step": 15763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8064108805996748,
+      "learning_rate": 2.566569497747169e-06,
+      "loss": 0.4938,
+      "step": 15764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.713148922532916,
+      "learning_rate": 2.566075787085478e-06,
+      "loss": 0.4917,
+      "step": 15765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5016006474708647,
+      "learning_rate": 2.5655821378436697e-06,
+      "loss": 0.3804,
+      "step": 15766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9051387778030309,
+      "learning_rate": 2.5650885500320754e-06,
+      "loss": 0.5363,
+      "step": 15767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5701616428077476,
+      "learning_rate": 2.5645950236610294e-06,
+      "loss": 0.4729,
+      "step": 15768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.633326730155972,
+      "learning_rate": 2.5641015587408647e-06,
+      "loss": 0.389,
+      "step": 15769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9636076173591244,
+      "learning_rate": 2.5636081552819086e-06,
+      "loss": 0.4479,
+      "step": 15770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9128499005510409,
+      "learning_rate": 2.563114813294492e-06,
+      "loss": 0.5695,
+      "step": 15771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.610505478933371,
+      "learning_rate": 2.562621532788944e-06,
+      "loss": 0.3764,
+      "step": 15772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6317113542885637,
+      "learning_rate": 2.5621283137755892e-06,
+      "loss": 0.4285,
+      "step": 15773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8291379707894464,
+      "learning_rate": 2.5616351562647522e-06,
+      "loss": 0.4744,
+      "step": 15774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6378801688517532,
+      "learning_rate": 2.561142060266759e-06,
+      "loss": 0.4176,
+      "step": 15775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.93603133747622,
+      "learning_rate": 2.5606490257919326e-06,
+      "loss": 0.543,
+      "step": 15776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.768936125731648,
+      "learning_rate": 2.560156052850592e-06,
+      "loss": 0.468,
+      "step": 15777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6849073418847103,
+      "learning_rate": 2.5596631414530583e-06,
+      "loss": 0.4642,
+      "step": 15778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1680851503338197,
+      "learning_rate": 2.5591702916096506e-06,
+      "loss": 0.5657,
+      "step": 15779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.822742200297597,
+      "learning_rate": 2.558677503330686e-06,
+      "loss": 0.4571,
+      "step": 15780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2177944406507284,
+      "learning_rate": 2.558184776626479e-06,
+      "loss": 0.5197,
+      "step": 15781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.681852489659902,
+      "learning_rate": 2.5576921115073507e-06,
+      "loss": 0.5239,
+      "step": 15782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9251550854247266,
+      "learning_rate": 2.5571995079836075e-06,
+      "loss": 0.4984,
+      "step": 15783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8871966712244617,
+      "learning_rate": 2.556706966065565e-06,
+      "loss": 0.4502,
+      "step": 15784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.964097740189231,
+      "learning_rate": 2.5562144857635355e-06,
+      "loss": 0.5158,
+      "step": 15785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.435873215177959,
+      "learning_rate": 2.5557220670878264e-06,
+      "loss": 0.4039,
+      "step": 15786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4768469768648926,
+      "learning_rate": 2.555229710048747e-06,
+      "loss": 0.4224,
+      "step": 15787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6951600592083083,
+      "learning_rate": 2.554737414656605e-06,
+      "loss": 0.4051,
+      "step": 15788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1753697992724548,
+      "learning_rate": 2.554245180921705e-06,
+      "loss": 0.6021,
+      "step": 15789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3924430590342045,
+      "learning_rate": 2.553753008854353e-06,
+      "loss": 0.306,
+      "step": 15790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.567351438662708,
+      "learning_rate": 2.5532608984648525e-06,
+      "loss": 0.4104,
+      "step": 15791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.037364349152539,
+      "learning_rate": 2.552768849763504e-06,
+      "loss": 0.5468,
+      "step": 15792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2951700934960462,
+      "learning_rate": 2.5522768627606087e-06,
+      "loss": 0.3909,
+      "step": 15793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7034914782532882,
+      "learning_rate": 2.551784937466466e-06,
+      "loss": 0.4286,
+      "step": 15794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8233070187162783,
+      "learning_rate": 2.5512930738913766e-06,
+      "loss": 0.4404,
+      "step": 15795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5091869993279814,
+      "learning_rate": 2.5508012720456332e-06,
+      "loss": 0.6116,
+      "step": 15796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5191732920560677,
+      "learning_rate": 2.550309531939533e-06,
+      "loss": 0.4606,
+      "step": 15797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1000798535655534,
+      "learning_rate": 2.5498178535833728e-06,
+      "loss": 0.716,
+      "step": 15798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8894765529126627,
+      "learning_rate": 2.549326236987443e-06,
+      "loss": 0.4999,
+      "step": 15799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6784762574509646,
+      "learning_rate": 2.5488346821620335e-06,
+      "loss": 0.6949,
+      "step": 15800
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.4958854615688324,
+      "eval_runtime": 67.8887,
+      "eval_samples_per_second": 50.229,
+      "eval_steps_per_second": 1.576,
+      "step": 15800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6746511252207883,
+      "learning_rate": 2.548343189117439e-06,
+      "loss": 0.4068,
+      "step": 15801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.854375731146031,
+      "learning_rate": 2.547851757863946e-06,
+      "loss": 0.4511,
+      "step": 15802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3260484949100326,
+      "learning_rate": 2.5473603884118425e-06,
+      "loss": 0.6005,
+      "step": 15803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7746003348563066,
+      "learning_rate": 2.546869080771416e-06,
+      "loss": 0.457,
+      "step": 15804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.643598339496694,
+      "learning_rate": 2.54637783495295e-06,
+      "loss": 0.4973,
+      "step": 15805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7140070023414022,
+      "learning_rate": 2.5458866509667294e-06,
+      "loss": 0.5347,
+      "step": 15806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7971049207488947,
+      "learning_rate": 2.545395528823037e-06,
+      "loss": 0.4683,
+      "step": 15807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9589458521188388,
+      "learning_rate": 2.5449044685321557e-06,
+      "loss": 0.4775,
+      "step": 15808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7343028083491543,
+      "learning_rate": 2.544413470104362e-06,
+      "loss": 0.5415,
+      "step": 15809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7151767441550598,
+      "learning_rate": 2.5439225335499368e-06,
+      "loss": 0.4262,
+      "step": 15810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8507067258159404,
+      "learning_rate": 2.5434316588791587e-06,
+      "loss": 0.4971,
+      "step": 15811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4747816672269163,
+      "learning_rate": 2.542940846102301e-06,
+      "loss": 0.4116,
+      "step": 15812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.687682133419,
+      "learning_rate": 2.5424500952296392e-06,
+      "loss": 0.4766,
+      "step": 15813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9824787941088526,
+      "learning_rate": 2.5419594062714502e-06,
+      "loss": 0.5402,
+      "step": 15814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8126949970288917,
+      "learning_rate": 2.5414687792380024e-06,
+      "loss": 0.5291,
+      "step": 15815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.822861907601944,
+      "learning_rate": 2.540978214139567e-06,
+      "loss": 0.3769,
+      "step": 15816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.676702832617276,
+      "learning_rate": 2.5404877109864166e-06,
+      "loss": 0.5621,
+      "step": 15817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4527622015531045,
+      "learning_rate": 2.5399972697888163e-06,
+      "loss": 0.3457,
+      "step": 15818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6531394264222492,
+      "learning_rate": 2.5395068905570353e-06,
+      "loss": 0.4547,
+      "step": 15819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7449195849972992,
+      "learning_rate": 2.5390165733013384e-06,
+      "loss": 0.4861,
+      "step": 15820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8645232011713317,
+      "learning_rate": 2.538526318031992e-06,
+      "loss": 0.5453,
+      "step": 15821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.468544011918344,
+      "learning_rate": 2.5380361247592556e-06,
+      "loss": 0.4512,
+      "step": 15822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7651977817403035,
+      "learning_rate": 2.5375459934933934e-06,
+      "loss": 0.6142,
+      "step": 15823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9962795772631554,
+      "learning_rate": 2.5370559242446668e-06,
+      "loss": 0.7084,
+      "step": 15824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.562943013883949,
+      "learning_rate": 2.5365659170233336e-06,
+      "loss": 0.3415,
+      "step": 15825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1736666906953688,
+      "learning_rate": 2.5360759718396523e-06,
+      "loss": 0.6157,
+      "step": 15826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.10791302995588,
+      "learning_rate": 2.5355860887038804e-06,
+      "loss": 0.5276,
+      "step": 15827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.263209440647333,
+      "learning_rate": 2.5350962676262714e-06,
+      "loss": 0.5433,
+      "step": 15828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1236767859122176,
+      "learning_rate": 2.5346065086170797e-06,
+      "loss": 0.4524,
+      "step": 15829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6360130178057075,
+      "learning_rate": 2.5341168116865593e-06,
+      "loss": 0.3847,
+      "step": 15830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5694591827912734,
+      "learning_rate": 2.5336271768449616e-06,
+      "loss": 0.4863,
+      "step": 15831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.553486138460443,
+      "learning_rate": 2.5331376041025363e-06,
+      "loss": 0.3458,
+      "step": 15832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.002967171907305,
+      "learning_rate": 2.532648093469534e-06,
+      "loss": 0.4285,
+      "step": 15833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7049512449163435,
+      "learning_rate": 2.532158644956199e-06,
+      "loss": 0.3936,
+      "step": 15834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7406447820328823,
+      "learning_rate": 2.5316692585727797e-06,
+      "loss": 0.479,
+      "step": 15835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8034173169138814,
+      "learning_rate": 2.531179934329522e-06,
+      "loss": 0.5229,
+      "step": 15836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.618970950324607,
+      "learning_rate": 2.530690672236667e-06,
+      "loss": 0.4382,
+      "step": 15837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7941633779615667,
+      "learning_rate": 2.530201472304458e-06,
+      "loss": 0.4764,
+      "step": 15838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.818906445420915,
+      "learning_rate": 2.529712334543137e-06,
+      "loss": 0.4803,
+      "step": 15839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8075780877704353,
+      "learning_rate": 2.5292232589629458e-06,
+      "loss": 0.4504,
+      "step": 15840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6889315774509608,
+      "learning_rate": 2.5287342455741177e-06,
+      "loss": 0.484,
+      "step": 15841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.452780796769099,
+      "learning_rate": 2.528245294386893e-06,
+      "loss": 0.3859,
+      "step": 15842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5118755674465347,
+      "learning_rate": 2.527756405411509e-06,
+      "loss": 0.5188,
+      "step": 15843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2328961599827695,
+      "learning_rate": 2.527267578658196e-06,
+      "loss": 0.6057,
+      "step": 15844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.964527801535623,
+      "learning_rate": 2.5267788141371907e-06,
+      "loss": 0.5764,
+      "step": 15845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5051204190979803,
+      "learning_rate": 2.526290111858726e-06,
+      "loss": 0.3768,
+      "step": 15846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6319768638302279,
+      "learning_rate": 2.525801471833029e-06,
+      "loss": 0.4923,
+      "step": 15847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0369783476002787,
+      "learning_rate": 2.52531289407033e-06,
+      "loss": 0.5624,
+      "step": 15848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7682763110138149,
+      "learning_rate": 2.5248243785808605e-06,
+      "loss": 0.4408,
+      "step": 15849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6801352094718507,
+      "learning_rate": 2.524335925374843e-06,
+      "loss": 0.375,
+      "step": 15850
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.49513956904411316,
+      "eval_runtime": 67.9557,
+      "eval_samples_per_second": 50.18,
+      "eval_steps_per_second": 1.575,
+      "step": 15850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9449843905249864,
+      "learning_rate": 2.5238475344625043e-06,
+      "loss": 0.432,
+      "step": 15851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0812064761701965,
+      "learning_rate": 2.523359205854069e-06,
+      "loss": 0.6433,
+      "step": 15852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6261211331930425,
+      "learning_rate": 2.522870939559762e-06,
+      "loss": 0.4686,
+      "step": 15853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.864919155801231,
+      "learning_rate": 2.5223827355898e-06,
+      "loss": 0.4284,
+      "step": 15854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6701217777039974,
+      "learning_rate": 2.5218945939544064e-06,
+      "loss": 0.4022,
+      "step": 15855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8595556422447765,
+      "learning_rate": 2.5214065146638013e-06,
+      "loss": 0.4256,
+      "step": 15856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0100608667440123,
+      "learning_rate": 2.5209184977281997e-06,
+      "loss": 0.4696,
+      "step": 15857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7484138173158557,
+      "learning_rate": 2.520430543157818e-06,
+      "loss": 0.4584,
+      "step": 15858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8614315457724178,
+      "learning_rate": 2.5199426509628746e-06,
+      "loss": 0.5318,
+      "step": 15859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9783176888642573,
+      "learning_rate": 2.5194548211535786e-06,
+      "loss": 0.5564,
+      "step": 15860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8563710719439115,
+      "learning_rate": 2.5189670537401456e-06,
+      "loss": 0.5149,
+      "step": 15861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9947080849943362,
+      "learning_rate": 2.518479348732786e-06,
+      "loss": 0.5276,
+      "step": 15862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5735075049426532,
+      "learning_rate": 2.517991706141709e-06,
+      "loss": 0.5069,
+      "step": 15863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6656489278259217,
+      "learning_rate": 2.5175041259771233e-06,
+      "loss": 0.4588,
+      "step": 15864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8574658471125132,
+      "learning_rate": 2.517016608249236e-06,
+      "loss": 0.4635,
+      "step": 15865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.072624133063836,
+      "learning_rate": 2.5165291529682546e-06,
+      "loss": 0.602,
+      "step": 15866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7023078188684386,
+      "learning_rate": 2.5160417601443806e-06,
+      "loss": 0.4021,
+      "step": 15867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.865359466473416,
+      "learning_rate": 2.515554429787819e-06,
+      "loss": 0.3448,
+      "step": 15868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.766267623488064,
+      "learning_rate": 2.515067161908773e-06,
+      "loss": 0.5278,
+      "step": 15869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0447080226466223,
+      "learning_rate": 2.514579956517441e-06,
+      "loss": 0.6247,
+      "step": 15870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6777598706699246,
+      "learning_rate": 2.5140928136240226e-06,
+      "loss": 0.3361,
+      "step": 15871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.287438119574504,
+      "learning_rate": 2.5136057332387196e-06,
+      "loss": 0.4712,
+      "step": 15872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.556295963880542,
+      "learning_rate": 2.513118715371723e-06,
+      "loss": 0.3607,
+      "step": 15873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7481390085803559,
+      "learning_rate": 2.5126317600332307e-06,
+      "loss": 0.4996,
+      "step": 15874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5614064315641838,
+      "learning_rate": 2.512144867233439e-06,
+      "loss": 0.4922,
+      "step": 15875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9334547156455102,
+      "learning_rate": 2.5116580369825373e-06,
+      "loss": 0.4483,
+      "step": 15876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.609783556926056,
+      "learning_rate": 2.5111712692907174e-06,
+      "loss": 0.5354,
+      "step": 15877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.094730594568225,
+      "learning_rate": 2.5106845641681715e-06,
+      "loss": 0.5392,
+      "step": 15878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5135766589459712,
+      "learning_rate": 2.510197921625087e-06,
+      "loss": 0.4297,
+      "step": 15879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8804278074456335,
+      "learning_rate": 2.5097113416716525e-06,
+      "loss": 0.4909,
+      "step": 15880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9837040608023742,
+      "learning_rate": 2.5092248243180544e-06,
+      "loss": 0.49,
+      "step": 15881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.50597693376003,
+      "learning_rate": 2.508738369574476e-06,
+      "loss": 0.4186,
+      "step": 15882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3895992111492994,
+      "learning_rate": 2.5082519774511017e-06,
+      "loss": 0.3775,
+      "step": 15883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6398194795571126,
+      "learning_rate": 2.5077656479581158e-06,
+      "loss": 0.4805,
+      "step": 15884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6627862640394735,
+      "learning_rate": 2.5072793811056963e-06,
+      "loss": 0.4502,
+      "step": 15885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8669166148973055,
+      "learning_rate": 2.506793176904024e-06,
+      "loss": 0.5947,
+      "step": 15886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.591773593834507,
+      "learning_rate": 2.5063070353632766e-06,
+      "loss": 0.3695,
+      "step": 15887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6722006135070622,
+      "learning_rate": 2.5058209564936343e-06,
+      "loss": 0.4777,
+      "step": 15888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8553767063756466,
+      "learning_rate": 2.5053349403052686e-06,
+      "loss": 0.5982,
+      "step": 15889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8504562274422864,
+      "learning_rate": 2.504848986808356e-06,
+      "loss": 0.4293,
+      "step": 15890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8715663807568113,
+      "learning_rate": 2.504363096013071e-06,
+      "loss": 0.4651,
+      "step": 15891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7508397041580641,
+      "learning_rate": 2.5038772679295823e-06,
+      "loss": 0.5693,
+      "step": 15892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0388218461182466,
+      "learning_rate": 2.503391502568062e-06,
+      "loss": 0.4354,
+      "step": 15893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.02114031683088,
+      "learning_rate": 2.5029057999386812e-06,
+      "loss": 0.478,
+      "step": 15894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8412361095671879,
+      "learning_rate": 2.502420160051604e-06,
+      "loss": 0.5976,
+      "step": 15895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.595887756690223,
+      "learning_rate": 2.5019345829169996e-06,
+      "loss": 0.5337,
+      "step": 15896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6657379843807323,
+      "learning_rate": 2.501449068545033e-06,
+      "loss": 0.4507,
+      "step": 15897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0329110587351864,
+      "learning_rate": 2.5009636169458673e-06,
+      "loss": 0.5602,
+      "step": 15898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7665567666877007,
+      "learning_rate": 2.5004782281296645e-06,
+      "loss": 0.4836,
+      "step": 15899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.333295624288091,
+      "learning_rate": 2.4999929021065873e-06,
+      "loss": 0.3387,
+      "step": 15900
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.4940439760684967,
+      "eval_runtime": 67.9501,
+      "eval_samples_per_second": 50.184,
+      "eval_steps_per_second": 1.575,
+      "step": 15900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5670214452548872,
+      "learning_rate": 2.499507638886797e-06,
+      "loss": 0.3992,
+      "step": 15901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8805982380204083,
+      "learning_rate": 2.499022438480449e-06,
+      "loss": 0.5981,
+      "step": 15902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6502780040442295,
+      "learning_rate": 2.4985373008977025e-06,
+      "loss": 0.3589,
+      "step": 15903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4601062939659748,
+      "learning_rate": 2.498052226148714e-06,
+      "loss": 0.329,
+      "step": 15904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7373902580097396,
+      "learning_rate": 2.497567214243637e-06,
+      "loss": 0.4523,
+      "step": 15905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3859341268153353,
+      "learning_rate": 2.4970822651926247e-06,
+      "loss": 0.4921,
+      "step": 15906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.647511392400241,
+      "learning_rate": 2.496597379005832e-06,
+      "loss": 0.4454,
+      "step": 15907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.062314075520931,
+      "learning_rate": 2.496112555693406e-06,
+      "loss": 0.7134,
+      "step": 15908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8627065557075548,
+      "learning_rate": 2.4956277952654983e-06,
+      "loss": 0.5454,
+      "step": 15909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0887268531696015,
+      "learning_rate": 2.495143097732256e-06,
+      "loss": 0.4678,
+      "step": 15910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9478832137347173,
+      "learning_rate": 2.4946584631038287e-06,
+      "loss": 0.5074,
+      "step": 15911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7868889145428675,
+      "learning_rate": 2.4941738913903577e-06,
+      "loss": 0.4043,
+      "step": 15912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6535157039799693,
+      "learning_rate": 2.4936893826019895e-06,
+      "loss": 0.4761,
+      "step": 15913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6909178408447172,
+      "learning_rate": 2.493204936748869e-06,
+      "loss": 0.4738,
+      "step": 15914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.914159698482053,
+      "learning_rate": 2.4927205538411327e-06,
+      "loss": 0.495,
+      "step": 15915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.867800278538181,
+      "learning_rate": 2.4922362338889247e-06,
+      "loss": 0.4334,
+      "step": 15916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.97442215285193,
+      "learning_rate": 2.491751976902384e-06,
+      "loss": 0.5149,
+      "step": 15917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7357656768951495,
+      "learning_rate": 2.4912677828916457e-06,
+      "loss": 0.4697,
+      "step": 15918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8995196612625884,
+      "learning_rate": 2.490783651866847e-06,
+      "loss": 0.5231,
+      "step": 15919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4506283639110134,
+      "learning_rate": 2.4902995838381264e-06,
+      "loss": 0.3842,
+      "step": 15920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7390599970587133,
+      "learning_rate": 2.489815578815612e-06,
+      "loss": 0.4459,
+      "step": 15921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6955098507258213,
+      "learning_rate": 2.4893316368094395e-06,
+      "loss": 0.3874,
+      "step": 15922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.077106117627177,
+      "learning_rate": 2.488847757829739e-06,
+      "loss": 0.4352,
+      "step": 15923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5992202489144383,
+      "learning_rate": 2.488363941886641e-06,
+      "loss": 0.4553,
+      "step": 15924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5381738516356493,
+      "learning_rate": 2.487880188990273e-06,
+      "loss": 0.3717,
+      "step": 15925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7610702298128358,
+      "learning_rate": 2.487396499150762e-06,
+      "loss": 0.4597,
+      "step": 15926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6376706354146966,
+      "learning_rate": 2.486912872378235e-06,
+      "loss": 0.4788,
+      "step": 15927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6752141322691692,
+      "learning_rate": 2.486429308682813e-06,
+      "loss": 0.6349,
+      "step": 15928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8682232744320246,
+      "learning_rate": 2.485945808074625e-06,
+      "loss": 0.4896,
+      "step": 15929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.662191995907154,
+      "learning_rate": 2.485462370563787e-06,
+      "loss": 0.4468,
+      "step": 15930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2247960679120051,
+      "learning_rate": 2.4849789961604225e-06,
+      "loss": 0.3171,
+      "step": 15931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.638857085944719,
+      "learning_rate": 2.4844956848746497e-06,
+      "loss": 0.4635,
+      "step": 15932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6155394705614678,
+      "learning_rate": 2.4840124367165887e-06,
+      "loss": 0.3587,
+      "step": 15933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5598892035598473,
+      "learning_rate": 2.483529251696352e-06,
+      "loss": 0.4273,
+      "step": 15934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.418368555635749,
+      "learning_rate": 2.4830461298240573e-06,
+      "loss": 0.6388,
+      "step": 15935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2723488462246015,
+      "learning_rate": 2.4825630711098192e-06,
+      "loss": 0.3356,
+      "step": 15936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8280993758183413,
+      "learning_rate": 2.482080075563747e-06,
+      "loss": 0.4803,
+      "step": 15937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6847571449550203,
+      "learning_rate": 2.481597143195955e-06,
+      "loss": 0.3764,
+      "step": 15938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5574788974277831,
+      "learning_rate": 2.481114274016553e-06,
+      "loss": 0.4414,
+      "step": 15939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8635285288151213,
+      "learning_rate": 2.4806314680356468e-06,
+      "loss": 0.5883,
+      "step": 15940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4765961342373393,
+      "learning_rate": 2.4801487252633453e-06,
+      "loss": 0.4111,
+      "step": 15941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7408702125621947,
+      "learning_rate": 2.479666045709757e-06,
+      "loss": 0.5539,
+      "step": 15942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5661260086062136,
+      "learning_rate": 2.479183429384982e-06,
+      "loss": 0.4723,
+      "step": 15943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4608945142691014,
+      "learning_rate": 2.4787008762991253e-06,
+      "loss": 0.4074,
+      "step": 15944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.479420905102041,
+      "learning_rate": 2.4782183864622892e-06,
+      "loss": 0.4421,
+      "step": 15945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7782146064172053,
+      "learning_rate": 2.477735959884576e-06,
+      "loss": 0.6608,
+      "step": 15946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1314897582195704,
+      "learning_rate": 2.4772535965760817e-06,
+      "loss": 0.4734,
+      "step": 15947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0279267956532294,
+      "learning_rate": 2.476771296546906e-06,
+      "loss": 0.5886,
+      "step": 15948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7179398071124723,
+      "learning_rate": 2.476289059807147e-06,
+      "loss": 0.5884,
+      "step": 15949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5314234736613732,
+      "learning_rate": 2.4758068863668964e-06,
+      "loss": 0.3865,
+      "step": 15950
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.4942709803581238,
+      "eval_runtime": 67.9048,
+      "eval_samples_per_second": 50.217,
+      "eval_steps_per_second": 1.576,
+      "step": 15950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.488942977044147,
+      "learning_rate": 2.475324776236251e-06,
+      "loss": 0.4413,
+      "step": 15951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5804195515503092,
+      "learning_rate": 2.474842729425304e-06,
+      "loss": 0.3844,
+      "step": 15952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8103039359696065,
+      "learning_rate": 2.4743607459441434e-06,
+      "loss": 0.5344,
+      "step": 15953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8043510825806848,
+      "learning_rate": 2.4738788258028618e-06,
+      "loss": 0.5001,
+      "step": 15954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7341328931130415,
+      "learning_rate": 2.473396969011549e-06,
+      "loss": 0.4603,
+      "step": 15955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8763462225996277,
+      "learning_rate": 2.472915175580289e-06,
+      "loss": 0.6802,
+      "step": 15956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0986900392851777,
+      "learning_rate": 2.4724334455191697e-06,
+      "loss": 0.5643,
+      "step": 15957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9681896606709413,
+      "learning_rate": 2.471951778838275e-06,
+      "loss": 0.4553,
+      "step": 15958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15225183154352,
+      "learning_rate": 2.471470175547692e-06,
+      "loss": 0.4687,
+      "step": 15959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6927822994188098,
+      "learning_rate": 2.4709886356574967e-06,
+      "loss": 0.4802,
+      "step": 15960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5380862682647023,
+      "learning_rate": 2.470507159177774e-06,
+      "loss": 0.5234,
+      "step": 15961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3575031037524625,
+      "learning_rate": 2.4700257461186033e-06,
+      "loss": 0.3068,
+      "step": 15962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.023954242049222,
+      "learning_rate": 2.46954439649006e-06,
+      "loss": 0.5683,
+      "step": 15963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7044250501929887,
+      "learning_rate": 2.4690631103022226e-06,
+      "loss": 0.428,
+      "step": 15964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0845382158634824,
+      "learning_rate": 2.4685818875651688e-06,
+      "loss": 0.5276,
+      "step": 15965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.906964158106237,
+      "learning_rate": 2.4681007282889672e-06,
+      "loss": 0.6113,
+      "step": 15966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.107540146508276,
+      "learning_rate": 2.4676196324836947e-06,
+      "loss": 0.7229,
+      "step": 15967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9345888945603238,
+      "learning_rate": 2.467138600159423e-06,
+      "loss": 0.5275,
+      "step": 15968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3571409827625187,
+      "learning_rate": 2.4666576313262197e-06,
+      "loss": 0.3796,
+      "step": 15969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.714593294068616,
+      "learning_rate": 2.4661767259941545e-06,
+      "loss": 0.484,
+      "step": 15970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7520354116735906,
+      "learning_rate": 2.465695884173295e-06,
+      "loss": 0.5082,
+      "step": 15971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7869230253774118,
+      "learning_rate": 2.465215105873709e-06,
+      "loss": 0.4183,
+      "step": 15972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.776237932122093,
+      "learning_rate": 2.464734391105459e-06,
+      "loss": 0.4687,
+      "step": 15973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4280668247172223,
+      "learning_rate": 2.4642537398786085e-06,
+      "loss": 0.5874,
+      "step": 15974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1196654646875563,
+      "learning_rate": 2.4637731522032223e-06,
+      "loss": 0.5796,
+      "step": 15975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2941018370259127,
+      "learning_rate": 2.4632926280893575e-06,
+      "loss": 0.3314,
+      "step": 15976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6709761930385907,
+      "learning_rate": 2.4628121675470754e-06,
+      "loss": 0.4178,
+      "step": 15977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9479870595838698,
+      "learning_rate": 2.462331770586434e-06,
+      "loss": 0.5167,
+      "step": 15978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5520606756644164,
+      "learning_rate": 2.4618514372174905e-06,
+      "loss": 0.4544,
+      "step": 15979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6495748161530532,
+      "learning_rate": 2.4613711674503e-06,
+      "loss": 0.6375,
+      "step": 15980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7930953052591008,
+      "learning_rate": 2.4608909612949184e-06,
+      "loss": 0.4597,
+      "step": 15981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0592393038995986,
+      "learning_rate": 2.4604108187613952e-06,
+      "loss": 0.4143,
+      "step": 15982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5120881512209514,
+      "learning_rate": 2.459930739859783e-06,
+      "loss": 0.4364,
+      "step": 15983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5156671303495008,
+      "learning_rate": 2.459450724600135e-06,
+      "loss": 0.5526,
+      "step": 15984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9013473058529726,
+      "learning_rate": 2.4589707729924956e-06,
+      "loss": 0.5821,
+      "step": 15985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7420756975177025,
+      "learning_rate": 2.4584908850469143e-06,
+      "loss": 0.4467,
+      "step": 15986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7230989141595396,
+      "learning_rate": 2.4580110607734384e-06,
+      "loss": 0.5553,
+      "step": 15987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6184650843511137,
+      "learning_rate": 2.4575313001821098e-06,
+      "loss": 0.421,
+      "step": 15988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8892076766550963,
+      "learning_rate": 2.4570516032829743e-06,
+      "loss": 0.5092,
+      "step": 15989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7544143596344215,
+      "learning_rate": 2.456571970086073e-06,
+      "loss": 0.557,
+      "step": 15990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6663497984046878,
+      "learning_rate": 2.4560924006014485e-06,
+      "loss": 0.4883,
+      "step": 15991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6129264262123946,
+      "learning_rate": 2.4556128948391384e-06,
+      "loss": 0.4783,
+      "step": 15992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9001592807139922,
+      "learning_rate": 2.45513345280918e-06,
+      "loss": 0.6075,
+      "step": 15993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.800704923123985,
+      "learning_rate": 2.4546540745216144e-06,
+      "loss": 0.5364,
+      "step": 15994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8069818407534766,
+      "learning_rate": 2.454174759986472e-06,
+      "loss": 0.4377,
+      "step": 15995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7442880662084979,
+      "learning_rate": 2.4536955092137894e-06,
+      "loss": 0.4281,
+      "step": 15996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6772747970980637,
+      "learning_rate": 2.4532163222136007e-06,
+      "loss": 0.5734,
+      "step": 15997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.71890462215094,
+      "learning_rate": 2.452737198995934e-06,
+      "loss": 0.556,
+      "step": 15998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.680885828282392,
+      "learning_rate": 2.4522581395708215e-06,
+      "loss": 0.3804,
+      "step": 15999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6720769898668404,
+      "learning_rate": 2.451779143948294e-06,
+      "loss": 0.4541,
+      "step": 16000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.49520596861839294,
+      "eval_runtime": 67.9508,
+      "eval_samples_per_second": 50.183,
+      "eval_steps_per_second": 1.575,
+      "step": 16000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7560822514448786,
+      "learning_rate": 2.4513002121383743e-06,
+      "loss": 0.5932,
+      "step": 16001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.144906407860756,
+      "learning_rate": 2.4508213441510914e-06,
+      "loss": 0.5292,
+      "step": 16002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8235689054639093,
+      "learning_rate": 2.4503425399964702e-06,
+      "loss": 0.5033,
+      "step": 16003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9205236658145242,
+      "learning_rate": 2.4498637996845345e-06,
+      "loss": 0.5096,
+      "step": 16004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3077807477354355,
+      "learning_rate": 2.4493851232253046e-06,
+      "loss": 0.4365,
+      "step": 16005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5922307490908554,
+      "learning_rate": 2.448906510628802e-06,
+      "loss": 0.4749,
+      "step": 16006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.524800989994984,
+      "learning_rate": 2.448427961905048e-06,
+      "loss": 0.4084,
+      "step": 16007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3880236253420004,
+      "learning_rate": 2.4479494770640583e-06,
+      "loss": 0.3842,
+      "step": 16008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8275121801359238,
+      "learning_rate": 2.44747105611585e-06,
+      "loss": 0.412,
+      "step": 16009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.78444789371609,
+      "learning_rate": 2.446992699070441e-06,
+      "loss": 0.6376,
+      "step": 16010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.881673732103351,
+      "learning_rate": 2.446514405937841e-06,
+      "loss": 0.5681,
+      "step": 16011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7173011330858237,
+      "learning_rate": 2.4460361767280666e-06,
+      "loss": 0.484,
+      "step": 16012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.887126951432163,
+      "learning_rate": 2.4455580114511295e-06,
+      "loss": 0.5056,
+      "step": 16013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.420529201722664,
+      "learning_rate": 2.4450799101170357e-06,
+      "loss": 0.3828,
+      "step": 16014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6169598870815238,
+      "learning_rate": 2.4446018727357975e-06,
+      "loss": 0.4762,
+      "step": 16015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8413788696511888,
+      "learning_rate": 2.444123899317421e-06,
+      "loss": 0.5525,
+      "step": 16016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5047504623182453,
+      "learning_rate": 2.4436459898719137e-06,
+      "loss": 0.4268,
+      "step": 16017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7078360568535802,
+      "learning_rate": 2.4431681444092785e-06,
+      "loss": 0.4968,
+      "step": 16018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5864489135020388,
+      "learning_rate": 2.4426903629395187e-06,
+      "loss": 0.4888,
+      "step": 16019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8637322764287811,
+      "learning_rate": 2.442212645472639e-06,
+      "loss": 0.5367,
+      "step": 16020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2922916873069374,
+      "learning_rate": 2.441734992018637e-06,
+      "loss": 0.232,
+      "step": 16021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7495710126397004,
+      "learning_rate": 2.441257402587513e-06,
+      "loss": 0.4241,
+      "step": 16022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6685176860892341,
+      "learning_rate": 2.4407798771892676e-06,
+      "loss": 0.3952,
+      "step": 16023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6539190068524154,
+      "learning_rate": 2.440302415833893e-06,
+      "loss": 0.5145,
+      "step": 16024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3279242050426943,
+      "learning_rate": 2.4398250185313872e-06,
+      "loss": 0.3567,
+      "step": 16025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6952783137752268,
+      "learning_rate": 2.439347685291746e-06,
+      "loss": 0.6446,
+      "step": 16026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0225441026432827,
+      "learning_rate": 2.4388704161249563e-06,
+      "loss": 0.5219,
+      "step": 16027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.481198802470452,
+      "learning_rate": 2.438393211041015e-06,
+      "loss": 0.6429,
+      "step": 16028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7761159527786206,
+      "learning_rate": 2.437916070049911e-06,
+      "loss": 0.4476,
+      "step": 16029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5446640806317347,
+      "learning_rate": 2.437438993161631e-06,
+      "loss": 0.4444,
+      "step": 16030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6388252554576672,
+      "learning_rate": 2.436961980386164e-06,
+      "loss": 0.5118,
+      "step": 16031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5671177725840273,
+      "learning_rate": 2.436485031733496e-06,
+      "loss": 0.3552,
+      "step": 16032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.783624134359432,
+      "learning_rate": 2.4360081472136103e-06,
+      "loss": 0.4547,
+      "step": 16033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6219005848685522,
+      "learning_rate": 2.4355313268364905e-06,
+      "loss": 0.4518,
+      "step": 16034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5752160864172617,
+      "learning_rate": 2.43505457061212e-06,
+      "loss": 0.4589,
+      "step": 16035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3904651055097403,
+      "learning_rate": 2.4345778785504774e-06,
+      "loss": 0.4448,
+      "step": 16036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6568081801524623,
+      "learning_rate": 2.434101250661543e-06,
+      "loss": 0.3712,
+      "step": 16037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.698447452927487,
+      "learning_rate": 2.4336246869552946e-06,
+      "loss": 0.5006,
+      "step": 16038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3617166757787087,
+      "learning_rate": 2.43314818744171e-06,
+      "loss": 0.4321,
+      "step": 16039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5984163721780245,
+      "learning_rate": 2.4326717521307624e-06,
+      "loss": 0.3136,
+      "step": 16040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0616109963970817,
+      "learning_rate": 2.432195381032426e-06,
+      "loss": 0.5576,
+      "step": 16041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8800711332146764,
+      "learning_rate": 2.4317190741566756e-06,
+      "loss": 0.4757,
+      "step": 16042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.637064497277078,
+      "learning_rate": 2.4312428315134783e-06,
+      "loss": 0.5162,
+      "step": 16043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7280312982305075,
+      "learning_rate": 2.4307666531128067e-06,
+      "loss": 0.4754,
+      "step": 16044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.933470552210131,
+      "learning_rate": 2.4302905389646304e-06,
+      "loss": 0.5079,
+      "step": 16045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8223843553836738,
+      "learning_rate": 2.4298144890789138e-06,
+      "loss": 0.43,
+      "step": 16046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.877776644954724,
+      "learning_rate": 2.4293385034656235e-06,
+      "loss": 0.5264,
+      "step": 16047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6447536248710464,
+      "learning_rate": 2.4288625821347263e-06,
+      "loss": 0.4129,
+      "step": 16048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2719284384886678,
+      "learning_rate": 2.4283867250961814e-06,
+      "loss": 0.3207,
+      "step": 16049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6595212144160183,
+      "learning_rate": 2.4279109323599525e-06,
+      "loss": 0.42,
+      "step": 16050
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.494769811630249,
+      "eval_runtime": 67.9175,
+      "eval_samples_per_second": 50.208,
+      "eval_steps_per_second": 1.575,
+      "step": 16050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7907248693148186,
+      "learning_rate": 2.4274352039360004e-06,
+      "loss": 0.584,
+      "step": 16051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.867000880983463,
+      "learning_rate": 2.426959539834285e-06,
+      "loss": 0.494,
+      "step": 16052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9112158835503474,
+      "learning_rate": 2.4264839400647618e-06,
+      "loss": 0.717,
+      "step": 16053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6098951395766765,
+      "learning_rate": 2.426008404637388e-06,
+      "loss": 0.3748,
+      "step": 16054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5601019553189286,
+      "learning_rate": 2.425532933562121e-06,
+      "loss": 0.4536,
+      "step": 16055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4888691087741486,
+      "learning_rate": 2.42505752684891e-06,
+      "loss": 0.6617,
+      "step": 16056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.081713011056215,
+      "learning_rate": 2.424582184507711e-06,
+      "loss": 0.7468,
+      "step": 16057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9882817693551942,
+      "learning_rate": 2.4241069065484742e-06,
+      "loss": 0.3972,
+      "step": 16058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6442792071305812,
+      "learning_rate": 2.4236316929811476e-06,
+      "loss": 0.5293,
+      "step": 16059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9606995839925208,
+      "learning_rate": 2.423156543815681e-06,
+      "loss": 0.5573,
+      "step": 16060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6683612589593664,
+      "learning_rate": 2.422681459062022e-06,
+      "loss": 0.3019,
+      "step": 16061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8930606607023164,
+      "learning_rate": 2.4222064387301137e-06,
+      "loss": 0.577,
+      "step": 16062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6541617002114886,
+      "learning_rate": 2.4217314828299027e-06,
+      "loss": 0.3454,
+      "step": 16063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8407746203999509,
+      "learning_rate": 2.4212565913713305e-06,
+      "loss": 0.3814,
+      "step": 16064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.780563825050756,
+      "learning_rate": 2.420781764364341e-06,
+      "loss": 0.5803,
+      "step": 16065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.678842010665217,
+      "learning_rate": 2.420307001818871e-06,
+      "loss": 0.5692,
+      "step": 16066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0983532224274315,
+      "learning_rate": 2.4198323037448614e-06,
+      "loss": 0.6948,
+      "step": 16067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4755844102945104,
+      "learning_rate": 2.4193576701522504e-06,
+      "loss": 0.5146,
+      "step": 16068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8992903450706444,
+      "learning_rate": 2.418883101050972e-06,
+      "loss": 0.5461,
+      "step": 16069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6203783273384094,
+      "learning_rate": 2.4184085964509623e-06,
+      "loss": 0.7242,
+      "step": 16070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5298785117714526,
+      "learning_rate": 2.4179341563621557e-06,
+      "loss": 0.4206,
+      "step": 16071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.031187512441427,
+      "learning_rate": 2.4174597807944817e-06,
+      "loss": 0.5149,
+      "step": 16072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6497004636690467,
+      "learning_rate": 2.416985469757873e-06,
+      "loss": 0.4277,
+      "step": 16073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7822976907401926,
+      "learning_rate": 2.4165112232622594e-06,
+      "loss": 0.5822,
+      "step": 16074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7294833497603401,
+      "learning_rate": 2.416037041317567e-06,
+      "loss": 0.4791,
+      "step": 16075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5809263383842311,
+      "learning_rate": 2.4155629239337218e-06,
+      "loss": 0.4231,
+      "step": 16076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5283566867311955,
+      "learning_rate": 2.4150888711206543e-06,
+      "loss": 0.7383,
+      "step": 16077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.596898749566415,
+      "learning_rate": 2.4146148828882836e-06,
+      "loss": 0.559,
+      "step": 16078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9286490170306636,
+      "learning_rate": 2.4141409592465344e-06,
+      "loss": 0.4631,
+      "step": 16079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8316447021445845,
+      "learning_rate": 2.413667100205328e-06,
+      "loss": 0.7269,
+      "step": 16080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3931268896530624,
+      "learning_rate": 2.413193305774583e-06,
+      "loss": 0.447,
+      "step": 16081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9145387550382031,
+      "learning_rate": 2.412719575964219e-06,
+      "loss": 0.5412,
+      "step": 16082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0435343913001063,
+      "learning_rate": 2.412245910784153e-06,
+      "loss": 0.5438,
+      "step": 16083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3262668216143345,
+      "learning_rate": 2.4117723102443017e-06,
+      "loss": 0.416,
+      "step": 16084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7688307651910145,
+      "learning_rate": 2.4112987743545783e-06,
+      "loss": 0.5478,
+      "step": 16085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5368765680346101,
+      "learning_rate": 2.4108253031248965e-06,
+      "loss": 0.4127,
+      "step": 16086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2164193790861693,
+      "learning_rate": 2.410351896565169e-06,
+      "loss": 0.2387,
+      "step": 16087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5509232517788853,
+      "learning_rate": 2.4098785546853046e-06,
+      "loss": 0.5408,
+      "step": 16088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9449440979972252,
+      "learning_rate": 2.4094052774952137e-06,
+      "loss": 0.3978,
+      "step": 16089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.740735968081041,
+      "learning_rate": 2.4089320650048045e-06,
+      "loss": 0.3998,
+      "step": 16090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.717475244925583,
+      "learning_rate": 2.4084589172239816e-06,
+      "loss": 0.4044,
+      "step": 16091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9408532811767614,
+      "learning_rate": 2.4079858341626504e-06,
+      "loss": 0.6671,
+      "step": 16092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0640376944610628,
+      "learning_rate": 2.4075128158307172e-06,
+      "loss": 0.5635,
+      "step": 16093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.406665027152486,
+      "learning_rate": 2.407039862238081e-06,
+      "loss": 0.3762,
+      "step": 16094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7363188615817693,
+      "learning_rate": 2.406566973394644e-06,
+      "loss": 0.5396,
+      "step": 16095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7229397468826932,
+      "learning_rate": 2.406094149310306e-06,
+      "loss": 0.6011,
+      "step": 16096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8103429053114974,
+      "learning_rate": 2.4056213899949667e-06,
+      "loss": 0.5903,
+      "step": 16097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2607525940318705,
+      "learning_rate": 2.4051486954585203e-06,
+      "loss": 0.5964,
+      "step": 16098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7393085960139236,
+      "learning_rate": 2.404676065710864e-06,
+      "loss": 0.5205,
+      "step": 16099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0388540857621695,
+      "learning_rate": 2.4042035007618924e-06,
+      "loss": 0.5656,
+      "step": 16100
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.4948718249797821,
+      "eval_runtime": 67.8694,
+      "eval_samples_per_second": 50.244,
+      "eval_steps_per_second": 1.577,
+      "step": 16100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.903583769177307,
+      "learning_rate": 2.4037310006214958e-06,
+      "loss": 0.5619,
+      "step": 16101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7715900300510266,
+      "learning_rate": 2.4032585652995687e-06,
+      "loss": 0.4797,
+      "step": 16102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6904594819204324,
+      "learning_rate": 2.4027861948060008e-06,
+      "loss": 0.4551,
+      "step": 16103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.829994218209397,
+      "learning_rate": 2.402313889150679e-06,
+      "loss": 0.4862,
+      "step": 16104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1736202918743492,
+      "learning_rate": 2.401841648343492e-06,
+      "loss": 0.7007,
+      "step": 16105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7525381454667626,
+      "learning_rate": 2.401369472394327e-06,
+      "loss": 0.4668,
+      "step": 16106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.963710580748714,
+      "learning_rate": 2.400897361313066e-06,
+      "loss": 0.5359,
+      "step": 16107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.726667078388121,
+      "learning_rate": 2.400425315109594e-06,
+      "loss": 0.4127,
+      "step": 16108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6038717708240002,
+      "learning_rate": 2.3999533337937926e-06,
+      "loss": 0.5135,
+      "step": 16109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5104162248109256,
+      "learning_rate": 2.3994814173755442e-06,
+      "loss": 0.4265,
+      "step": 16110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3886771217264897,
+      "learning_rate": 2.399009565864725e-06,
+      "loss": 0.2716,
+      "step": 16111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5383405765979665,
+      "learning_rate": 2.398537779271215e-06,
+      "loss": 0.4958,
+      "step": 16112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8299958067387514,
+      "learning_rate": 2.3980660576048908e-06,
+      "loss": 0.4231,
+      "step": 16113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6106839428841127,
+      "learning_rate": 2.397594400875626e-06,
+      "loss": 0.4114,
+      "step": 16114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8443839120403147,
+      "learning_rate": 2.3971228090932955e-06,
+      "loss": 0.5562,
+      "step": 16115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.444876902267862,
+      "learning_rate": 2.396651282267773e-06,
+      "loss": 0.6421,
+      "step": 16116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9019225300204392,
+      "learning_rate": 2.396179820408927e-06,
+      "loss": 0.5202,
+      "step": 16117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4521903987039133,
+      "learning_rate": 2.3957084235266283e-06,
+      "loss": 0.4597,
+      "step": 16118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1194227132781296,
+      "learning_rate": 2.3952370916307472e-06,
+      "loss": 0.4871,
+      "step": 16119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.044994863945938,
+      "learning_rate": 2.394765824731148e-06,
+      "loss": 0.6284,
+      "step": 16120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5512499288849992,
+      "learning_rate": 2.3942946228376977e-06,
+      "loss": 0.3572,
+      "step": 16121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9217014702377375,
+      "learning_rate": 2.3938234859602595e-06,
+      "loss": 0.6284,
+      "step": 16122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.734759381084004,
+      "learning_rate": 2.3933524141086995e-06,
+      "loss": 0.3637,
+      "step": 16123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7615249526314691,
+      "learning_rate": 2.3928814072928756e-06,
+      "loss": 0.4987,
+      "step": 16124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3668325592043429,
+      "learning_rate": 2.3924104655226497e-06,
+      "loss": 0.395,
+      "step": 16125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7228910852344583,
+      "learning_rate": 2.3919395888078796e-06,
+      "loss": 0.445,
+      "step": 16126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.95366228265655,
+      "learning_rate": 2.3914687771584244e-06,
+      "loss": 0.4918,
+      "step": 16127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8299757220584418,
+      "learning_rate": 2.390998030584141e-06,
+      "loss": 0.5496,
+      "step": 16128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5243925712964994,
+      "learning_rate": 2.3905273490948815e-06,
+      "loss": 0.4138,
+      "step": 16129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7870030965814765,
+      "learning_rate": 2.3900567327005004e-06,
+      "loss": 0.4654,
+      "step": 16130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.607014505085809,
+      "learning_rate": 2.3895861814108496e-06,
+      "loss": 0.3752,
+      "step": 16131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5533112141273555,
+      "learning_rate": 2.3891156952357818e-06,
+      "loss": 0.4598,
+      "step": 16132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4833769377585921,
+      "learning_rate": 2.3886452741851436e-06,
+      "loss": 0.4086,
+      "step": 16133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0778424777456164,
+      "learning_rate": 2.3881749182687836e-06,
+      "loss": 0.5949,
+      "step": 16134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9628626850064743,
+      "learning_rate": 2.3877046274965506e-06,
+      "loss": 0.4567,
+      "step": 16135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6882476366466512,
+      "learning_rate": 2.3872344018782863e-06,
+      "loss": 0.4762,
+      "step": 16136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0261355669358005,
+      "learning_rate": 2.3867642414238366e-06,
+      "loss": 0.4284,
+      "step": 16137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5623181391035499,
+      "learning_rate": 2.3862941461430453e-06,
+      "loss": 0.3897,
+      "step": 16138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5168952140978347,
+      "learning_rate": 2.3858241160457507e-06,
+      "loss": 0.4002,
+      "step": 16139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5996803365193402,
+      "learning_rate": 2.385354151141794e-06,
+      "loss": 0.4497,
+      "step": 16140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7571752014369697,
+      "learning_rate": 2.384884251441015e-06,
+      "loss": 0.4936,
+      "step": 16141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6075095059507323,
+      "learning_rate": 2.3844144169532473e-06,
+      "loss": 0.4872,
+      "step": 16142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6992931137637792,
+      "learning_rate": 2.3839446476883295e-06,
+      "loss": 0.5485,
+      "step": 16143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4652082690603931,
+      "learning_rate": 2.3834749436560943e-06,
+      "loss": 0.4795,
+      "step": 16144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5453974846399827,
+      "learning_rate": 2.383005304866377e-06,
+      "loss": 0.4865,
+      "step": 16145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6006031827843956,
+      "learning_rate": 2.3825357313290064e-06,
+      "loss": 0.4087,
+      "step": 16146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3921853381560598,
+      "learning_rate": 2.3820662230538137e-06,
+      "loss": 0.3332,
+      "step": 16147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7552655197737939,
+      "learning_rate": 2.3815967800506297e-06,
+      "loss": 0.6478,
+      "step": 16148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4101193774800382,
+      "learning_rate": 2.3811274023292786e-06,
+      "loss": 0.559,
+      "step": 16149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6740593584632257,
+      "learning_rate": 2.380658089899588e-06,
+      "loss": 0.481,
+      "step": 16150
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.4941338300704956,
+      "eval_runtime": 67.9052,
+      "eval_samples_per_second": 50.217,
+      "eval_steps_per_second": 1.576,
+      "step": 16150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9860743886061956,
+      "learning_rate": 2.3801888427713845e-06,
+      "loss": 0.4594,
+      "step": 16151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.937549495131207,
+      "learning_rate": 2.379719660954488e-06,
+      "loss": 0.4445,
+      "step": 16152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5699883432865982,
+      "learning_rate": 2.379250544458723e-06,
+      "loss": 0.4033,
+      "step": 16153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6511708784589243,
+      "learning_rate": 2.3787814932939104e-06,
+      "loss": 0.4019,
+      "step": 16154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7594424877107133,
+      "learning_rate": 2.3783125074698673e-06,
+      "loss": 0.4189,
+      "step": 16155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6146183719588678,
+      "learning_rate": 2.3778435869964126e-06,
+      "loss": 0.4391,
+      "step": 16156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0520814489712347,
+      "learning_rate": 2.3773747318833637e-06,
+      "loss": 0.6589,
+      "step": 16157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4394226253392375,
+      "learning_rate": 2.376905942140536e-06,
+      "loss": 0.3193,
+      "step": 16158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7329657719647644,
+      "learning_rate": 2.376437217777742e-06,
+      "loss": 0.4796,
+      "step": 16159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5124782496454379,
+      "learning_rate": 2.375968558804794e-06,
+      "loss": 0.3687,
+      "step": 16160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6270657137923674,
+      "learning_rate": 2.3754999652315057e-06,
+      "loss": 0.3186,
+      "step": 16161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0495791115569797,
+      "learning_rate": 2.375031437067683e-06,
+      "loss": 0.4251,
+      "step": 16162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.066267741653617,
+      "learning_rate": 2.374562974323136e-06,
+      "loss": 0.6707,
+      "step": 16163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.46949358094982,
+      "learning_rate": 2.374094577007674e-06,
+      "loss": 0.3073,
+      "step": 16164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.00195673810495,
+      "learning_rate": 2.373626245131099e-06,
+      "loss": 0.4081,
+      "step": 16165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4946717107463527,
+      "learning_rate": 2.3731579787032163e-06,
+      "loss": 0.3416,
+      "step": 16166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.594714716265309,
+      "learning_rate": 2.3726897777338294e-06,
+      "loss": 0.4155,
+      "step": 16167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.362709228457641,
+      "learning_rate": 2.372221642232741e-06,
+      "loss": 0.3725,
+      "step": 16168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8296628644085857,
+      "learning_rate": 2.3717535722097483e-06,
+      "loss": 0.4855,
+      "step": 16169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6185786595123661,
+      "learning_rate": 2.3712855676746515e-06,
+      "loss": 0.405,
+      "step": 16170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0304678681058324,
+      "learning_rate": 2.3708176286372494e-06,
+      "loss": 0.5155,
+      "step": 16171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.574821556004549,
+      "learning_rate": 2.3703497551073353e-06,
+      "loss": 0.3451,
+      "step": 16172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8101120428092021,
+      "learning_rate": 2.3698819470947053e-06,
+      "loss": 0.5123,
+      "step": 16173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.79932417918746,
+      "learning_rate": 2.369414204609152e-06,
+      "loss": 0.4767,
+      "step": 16174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8016068502993101,
+      "learning_rate": 2.368946527660469e-06,
+      "loss": 0.4902,
+      "step": 16175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.655855082211776,
+      "learning_rate": 2.3684789162584446e-06,
+      "loss": 0.437,
+      "step": 16176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6994646261839208,
+      "learning_rate": 2.3680113704128705e-06,
+      "loss": 0.3755,
+      "step": 16177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7011629992303108,
+      "learning_rate": 2.367543890133532e-06,
+      "loss": 0.3781,
+      "step": 16178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.620504565656173,
+      "learning_rate": 2.3670764754302166e-06,
+      "loss": 0.5487,
+      "step": 16179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3912544357985626,
+      "learning_rate": 2.3666091263127104e-06,
+      "loss": 0.4299,
+      "step": 16180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9728228534697319,
+      "learning_rate": 2.3661418427907947e-06,
+      "loss": 0.5859,
+      "step": 16181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7925046392572879,
+      "learning_rate": 2.365674624874253e-06,
+      "loss": 0.4305,
+      "step": 16182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8822331725712977,
+      "learning_rate": 2.3652074725728673e-06,
+      "loss": 0.4952,
+      "step": 16183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.960443192519443,
+      "learning_rate": 2.364740385896415e-06,
+      "loss": 0.4978,
+      "step": 16184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4392402482922344,
+      "learning_rate": 2.364273364854675e-06,
+      "loss": 0.3183,
+      "step": 16185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5067280387260271,
+      "learning_rate": 2.363806409457426e-06,
+      "loss": 0.3515,
+      "step": 16186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.02987115017495,
+      "learning_rate": 2.3633395197144407e-06,
+      "loss": 0.5206,
+      "step": 16187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5639743794853072,
+      "learning_rate": 2.362872695635494e-06,
+      "loss": 0.4621,
+      "step": 16188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.593376503241213,
+      "learning_rate": 2.3624059372303594e-06,
+      "loss": 0.4013,
+      "step": 16189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3371056904788556,
+      "learning_rate": 2.3619392445088083e-06,
+      "loss": 0.3262,
+      "step": 16190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.14381703461806,
+      "learning_rate": 2.361472617480609e-06,
+      "loss": 0.5812,
+      "step": 16191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0125057735013145,
+      "learning_rate": 2.361006056155531e-06,
+      "loss": 0.556,
+      "step": 16192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6385466735717262,
+      "learning_rate": 2.3605395605433433e-06,
+      "loss": 0.3927,
+      "step": 16193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.233701528244034,
+      "learning_rate": 2.3600731306538076e-06,
+      "loss": 0.5719,
+      "step": 16194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7481827193301527,
+      "learning_rate": 2.3596067664966917e-06,
+      "loss": 0.5818,
+      "step": 16195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.149855645918576,
+      "learning_rate": 2.359140468081758e-06,
+      "loss": 0.7436,
+      "step": 16196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.970119858339473,
+      "learning_rate": 2.3586742354187665e-06,
+      "loss": 0.537,
+      "step": 16197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9462733090944302,
+      "learning_rate": 2.3582080685174793e-06,
+      "loss": 0.5511,
+      "step": 16198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6741411582302412,
+      "learning_rate": 2.3577419673876554e-06,
+      "loss": 0.5066,
+      "step": 16199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.69352069553792,
+      "learning_rate": 2.3572759320390514e-06,
+      "loss": 0.4589,
+      "step": 16200
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.4941120743751526,
+      "eval_runtime": 67.9334,
+      "eval_samples_per_second": 50.196,
+      "eval_steps_per_second": 1.575,
+      "step": 16200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5267296486412756,
+      "learning_rate": 2.356809962481423e-06,
+      "loss": 0.4096,
+      "step": 16201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5129240900887027,
+      "learning_rate": 2.3563440587245255e-06,
+      "loss": 0.4274,
+      "step": 16202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.477356493516847,
+      "learning_rate": 2.355878220778114e-06,
+      "loss": 0.489,
+      "step": 16203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6056650769075111,
+      "learning_rate": 2.355412448651938e-06,
+      "loss": 0.3983,
+      "step": 16204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.945101990202539,
+      "learning_rate": 2.354946742355749e-06,
+      "loss": 0.5092,
+      "step": 16205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8750750128964175,
+      "learning_rate": 2.3544811018992983e-06,
+      "loss": 0.744,
+      "step": 16206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6687472112547987,
+      "learning_rate": 2.35401552729233e-06,
+      "loss": 0.4585,
+      "step": 16207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7764690667200336,
+      "learning_rate": 2.353550018544593e-06,
+      "loss": 0.4469,
+      "step": 16208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6191532961107495,
+      "learning_rate": 2.3530845756658336e-06,
+      "loss": 0.4813,
+      "step": 16209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8412126712298669,
+      "learning_rate": 2.352619198665792e-06,
+      "loss": 0.5071,
+      "step": 16210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9172829571103203,
+      "learning_rate": 2.3521538875542126e-06,
+      "loss": 0.4741,
+      "step": 16211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5241885674109719,
+      "learning_rate": 2.351688642340839e-06,
+      "loss": 0.2777,
+      "step": 16212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6209670319990583,
+      "learning_rate": 2.351223463035405e-06,
+      "loss": 0.5588,
+      "step": 16213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6300536047070815,
+      "learning_rate": 2.350758349647653e-06,
+      "loss": 0.4457,
+      "step": 16214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4946580791696855,
+      "learning_rate": 2.350293302187318e-06,
+      "loss": 0.3922,
+      "step": 16215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4693405098926535,
+      "learning_rate": 2.349828320664139e-06,
+      "loss": 0.4225,
+      "step": 16216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5115923959051005,
+      "learning_rate": 2.3493634050878454e-06,
+      "loss": 0.3753,
+      "step": 16217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.477042400444892,
+      "learning_rate": 2.348898555468172e-06,
+      "loss": 0.4576,
+      "step": 16218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7381240477315938,
+      "learning_rate": 2.348433771814852e-06,
+      "loss": 0.5038,
+      "step": 16219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.64975139230848,
+      "learning_rate": 2.347969054137612e-06,
+      "loss": 0.5425,
+      "step": 16220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9528211175513188,
+      "learning_rate": 2.3475044024461814e-06,
+      "loss": 0.4046,
+      "step": 16221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5046571368743156,
+      "learning_rate": 2.3470398167502897e-06,
+      "loss": 0.3413,
+      "step": 16222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7422224650490075,
+      "learning_rate": 2.34657529705966e-06,
+      "loss": 0.4304,
+      "step": 16223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8751433962011006,
+      "learning_rate": 2.346110843384016e-06,
+      "loss": 0.4886,
+      "step": 16224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.218152790523824,
+      "learning_rate": 2.3456464557330855e-06,
+      "loss": 0.6359,
+      "step": 16225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.052918403453959,
+      "learning_rate": 2.345182134116586e-06,
+      "loss": 0.6419,
+      "step": 16226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.688866360725778,
+      "learning_rate": 2.3447178785442386e-06,
+      "loss": 0.3349,
+      "step": 16227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9248330087127403,
+      "learning_rate": 2.344253689025764e-06,
+      "loss": 0.5689,
+      "step": 16228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.720402584342661,
+      "learning_rate": 2.343789565570877e-06,
+      "loss": 0.5457,
+      "step": 16229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7241844645788922,
+      "learning_rate": 2.343325508189295e-06,
+      "loss": 0.4559,
+      "step": 16230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8601853275238762,
+      "learning_rate": 2.342861516890735e-06,
+      "loss": 0.6104,
+      "step": 16231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3734167042336538,
+      "learning_rate": 2.342397591684906e-06,
+      "loss": 0.3393,
+      "step": 16232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0408346011812473,
+      "learning_rate": 2.341933732581522e-06,
+      "loss": 0.4536,
+      "step": 16233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4914687237672741,
+      "learning_rate": 2.341469939590296e-06,
+      "loss": 0.3682,
+      "step": 16234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.709728896638615,
+      "learning_rate": 2.341006212720933e-06,
+      "loss": 0.4832,
+      "step": 16235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.919241189434789,
+      "learning_rate": 2.3405425519831424e-06,
+      "loss": 0.4818,
+      "step": 16236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9390272886642514,
+      "learning_rate": 2.3400789573866316e-06,
+      "loss": 0.6771,
+      "step": 16237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.061200025445726,
+      "learning_rate": 2.3396154289411066e-06,
+      "loss": 0.5272,
+      "step": 16238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6952473788205347,
+      "learning_rate": 2.3391519666562676e-06,
+      "loss": 0.3731,
+      "step": 16239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.550382510462613,
+      "learning_rate": 2.3386885705418194e-06,
+      "loss": 0.4601,
+      "step": 16240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.66258674692601,
+      "learning_rate": 2.3382252406074636e-06,
+      "loss": 0.4978,
+      "step": 16241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8540645340138155,
+      "learning_rate": 2.3377619768628975e-06,
+      "loss": 0.4875,
+      "step": 16242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6220860850645822,
+      "learning_rate": 2.3372987793178203e-06,
+      "loss": 0.4259,
+      "step": 16243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.40114844628429,
+      "learning_rate": 2.3368356479819298e-06,
+      "loss": 0.3618,
+      "step": 16244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.879091650108048,
+      "learning_rate": 2.3363725828649194e-06,
+      "loss": 0.5258,
+      "step": 16245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5690753753361617,
+      "learning_rate": 2.335909583976483e-06,
+      "loss": 0.3733,
+      "step": 16246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.040107697321843,
+      "learning_rate": 2.335446651326315e-06,
+      "loss": 0.6055,
+      "step": 16247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.17724511825253,
+      "learning_rate": 2.3349837849241067e-06,
+      "loss": 0.484,
+      "step": 16248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0307966998174893,
+      "learning_rate": 2.3345209847795457e-06,
+      "loss": 0.6486,
+      "step": 16249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6860864895304386,
+      "learning_rate": 2.3340582509023224e-06,
+      "loss": 0.4374,
+      "step": 16250
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.49419575929641724,
+      "eval_runtime": 67.9376,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 16250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3918817668295187,
+      "learning_rate": 2.3335955833021233e-06,
+      "loss": 0.3844,
+      "step": 16251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.880330945947679,
+      "learning_rate": 2.333132981988634e-06,
+      "loss": 0.5339,
+      "step": 16252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5006797754380494,
+      "learning_rate": 2.3326704469715378e-06,
+      "loss": 0.42,
+      "step": 16253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8993506331263579,
+      "learning_rate": 2.33220797826052e-06,
+      "loss": 0.5645,
+      "step": 16254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8047527085012096,
+      "learning_rate": 2.3317455758652593e-06,
+      "loss": 0.5859,
+      "step": 16255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.551778561424871,
+      "learning_rate": 2.3312832397954367e-06,
+      "loss": 0.4404,
+      "step": 16256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6625782603476638,
+      "learning_rate": 2.3308209700607334e-06,
+      "loss": 0.5247,
+      "step": 16257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8581485219675145,
+      "learning_rate": 2.330358766670823e-06,
+      "loss": 0.4607,
+      "step": 16258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0037401793716,
+      "learning_rate": 2.329896629635383e-06,
+      "loss": 0.6011,
+      "step": 16259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5770546730531887,
+      "learning_rate": 2.329434558964088e-06,
+      "loss": 0.3975,
+      "step": 16260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8768847400358664,
+      "learning_rate": 2.328972554666612e-06,
+      "loss": 0.6225,
+      "step": 16261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6353132234660068,
+      "learning_rate": 2.3285106167526254e-06,
+      "loss": 0.3933,
+      "step": 16262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7561826452915983,
+      "learning_rate": 2.328048745231799e-06,
+      "loss": 0.5136,
+      "step": 16263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8026765833274923,
+      "learning_rate": 2.327586940113803e-06,
+      "loss": 0.6884,
+      "step": 16264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7843986199798927,
+      "learning_rate": 2.3271252014083033e-06,
+      "loss": 0.3786,
+      "step": 16265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9167159886842766,
+      "learning_rate": 2.326663529124966e-06,
+      "loss": 0.556,
+      "step": 16266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5833320379819886,
+      "learning_rate": 2.3262019232734584e-06,
+      "loss": 0.3913,
+      "step": 16267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9201284032172452,
+      "learning_rate": 2.325740383863441e-06,
+      "loss": 0.5117,
+      "step": 16268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.073988974854603,
+      "learning_rate": 2.3252789109045765e-06,
+      "loss": 0.666,
+      "step": 16269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5777561538735896,
+      "learning_rate": 2.324817504406528e-06,
+      "loss": 0.4409,
+      "step": 16270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4331336766149427,
+      "learning_rate": 2.3243561643789515e-06,
+      "loss": 0.4788,
+      "step": 16271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7224223540196875,
+      "learning_rate": 2.323894890831506e-06,
+      "loss": 0.493,
+      "step": 16272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7766954547870446,
+      "learning_rate": 2.3234336837738484e-06,
+      "loss": 0.5655,
+      "step": 16273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7941088677055133,
+      "learning_rate": 2.3229725432156335e-06,
+      "loss": 0.4132,
+      "step": 16274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.702599500635686,
+      "learning_rate": 2.3225114691665146e-06,
+      "loss": 0.5266,
+      "step": 16275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5835362150674164,
+      "learning_rate": 2.3220504616361467e-06,
+      "loss": 0.5164,
+      "step": 16276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8370319753869728,
+      "learning_rate": 2.321589520634177e-06,
+      "loss": 0.5013,
+      "step": 16277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6618992894167588,
+      "learning_rate": 2.3211286461702558e-06,
+      "loss": 0.325,
+      "step": 16278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7528135902005746,
+      "learning_rate": 2.320667838254034e-06,
+      "loss": 0.4977,
+      "step": 16279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9895229014820939,
+      "learning_rate": 2.3202070968951543e-06,
+      "loss": 0.6035,
+      "step": 16280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6828390156471547,
+      "learning_rate": 2.3197464221032644e-06,
+      "loss": 0.5579,
+      "step": 16281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6882894363329695,
+      "learning_rate": 2.3192858138880074e-06,
+      "loss": 0.4059,
+      "step": 16282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.618643721410024,
+      "learning_rate": 2.3188252722590285e-06,
+      "loss": 0.6043,
+      "step": 16283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1445556011462314,
+      "learning_rate": 2.3183647972259645e-06,
+      "loss": 0.495,
+      "step": 16284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.939277663624175,
+      "learning_rate": 2.3179043887984574e-06,
+      "loss": 0.5861,
+      "step": 16285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.678893848634552,
+      "learning_rate": 2.3174440469861466e-06,
+      "loss": 0.4752,
+      "step": 16286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4428136622175012,
+      "learning_rate": 2.3169837717986666e-06,
+      "loss": 0.3643,
+      "step": 16287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0880518443040077,
+      "learning_rate": 2.316523563245655e-06,
+      "loss": 0.6796,
+      "step": 16288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7998256137248683,
+      "learning_rate": 2.3160634213367463e-06,
+      "loss": 0.6516,
+      "step": 16289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9360110987263426,
+      "learning_rate": 2.315603346081571e-06,
+      "loss": 0.5445,
+      "step": 16290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5380462695599266,
+      "learning_rate": 2.3151433374897613e-06,
+      "loss": 0.4038,
+      "step": 16291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4526748801119094,
+      "learning_rate": 2.3146833955709494e-06,
+      "loss": 0.5311,
+      "step": 16292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4238315559815284,
+      "learning_rate": 2.3142235203347605e-06,
+      "loss": 0.3603,
+      "step": 16293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9344684373453445,
+      "learning_rate": 2.3137637117908234e-06,
+      "loss": 0.537,
+      "step": 16294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5979139787394971,
+      "learning_rate": 2.313303969948764e-06,
+      "loss": 0.3785,
+      "step": 16295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5611050923040497,
+      "learning_rate": 2.3128442948182085e-06,
+      "loss": 0.4794,
+      "step": 16296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0311020343296105,
+      "learning_rate": 2.3123846864087758e-06,
+      "loss": 0.5196,
+      "step": 16297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25821520344072,
+      "learning_rate": 2.3119251447300907e-06,
+      "loss": 0.4861,
+      "step": 16298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5013862043610557,
+      "learning_rate": 2.3114656697917735e-06,
+      "loss": 0.392,
+      "step": 16299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.453804869112496,
+      "learning_rate": 2.311006261603441e-06,
+      "loss": 0.3486,
+      "step": 16300
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.4937276542186737,
+      "eval_runtime": 67.9255,
+      "eval_samples_per_second": 50.202,
+      "eval_steps_per_second": 1.575,
+      "step": 16300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5160370555155676,
+      "learning_rate": 2.3105469201747112e-06,
+      "loss": 0.374,
+      "step": 16301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2244176941171208,
+      "learning_rate": 2.3100876455152015e-06,
+      "loss": 0.3451,
+      "step": 16302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7068473448461714,
+      "learning_rate": 2.3096284376345247e-06,
+      "loss": 0.4675,
+      "step": 16303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3547903705393594,
+      "learning_rate": 2.3091692965422953e-06,
+      "loss": 0.2829,
+      "step": 16304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1652476519466903,
+      "learning_rate": 2.308710222248126e-06,
+      "loss": 0.5401,
+      "step": 16305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7230094109644134,
+      "learning_rate": 2.3082512147616248e-06,
+      "loss": 0.5604,
+      "step": 16306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4085614384902159,
+      "learning_rate": 2.3077922740924014e-06,
+      "loss": 0.4379,
+      "step": 16307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0716556686084893,
+      "learning_rate": 2.307333400250064e-06,
+      "loss": 0.6693,
+      "step": 16308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9539458347521776,
+      "learning_rate": 2.3068745932442207e-06,
+      "loss": 0.594,
+      "step": 16309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9470659312853973,
+      "learning_rate": 2.3064158530844728e-06,
+      "loss": 0.5662,
+      "step": 16310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4847819540977067,
+      "learning_rate": 2.305957179780426e-06,
+      "loss": 0.5611,
+      "step": 16311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.968158743159501,
+      "learning_rate": 2.3054985733416825e-06,
+      "loss": 0.5188,
+      "step": 16312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6913936605894986,
+      "learning_rate": 2.3050400337778407e-06,
+      "loss": 0.464,
+      "step": 16313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7552815539616586,
+      "learning_rate": 2.3045815610985017e-06,
+      "loss": 0.5755,
+      "step": 16314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7268379403076204,
+      "learning_rate": 2.304123155313264e-06,
+      "loss": 0.5397,
+      "step": 16315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6666761680993152,
+      "learning_rate": 2.3036648164317216e-06,
+      "loss": 0.4392,
+      "step": 16316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.562829848199195,
+      "learning_rate": 2.3032065444634713e-06,
+      "loss": 0.3504,
+      "step": 16317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.912034082670275,
+      "learning_rate": 2.302748339418108e-06,
+      "loss": 0.5234,
+      "step": 16318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7343497019229719,
+      "learning_rate": 2.30229020130522e-06,
+      "loss": 0.5305,
+      "step": 16319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2638081086892587,
+      "learning_rate": 2.3018321301344017e-06,
+      "loss": 0.3115,
+      "step": 16320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4991745602460471,
+      "learning_rate": 2.3013741259152404e-06,
+      "loss": 0.3489,
+      "step": 16321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8445527294521622,
+      "learning_rate": 2.3009161886573252e-06,
+      "loss": 0.5462,
+      "step": 16322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.783365571103696,
+      "learning_rate": 2.300458318370243e-06,
+      "loss": 0.6292,
+      "step": 16323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.447669978932948,
+      "learning_rate": 2.3000005150635794e-06,
+      "loss": 0.3747,
+      "step": 16324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.017187752054106,
+      "learning_rate": 2.2995427787469163e-06,
+      "loss": 0.4073,
+      "step": 16325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.555132655298082,
+      "learning_rate": 2.2990851094298377e-06,
+      "loss": 0.4105,
+      "step": 16326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7777788397718377,
+      "learning_rate": 2.2986275071219235e-06,
+      "loss": 0.6205,
+      "step": 16327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.460052927830399,
+      "learning_rate": 2.2981699718327552e-06,
+      "loss": 0.3631,
+      "step": 16328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.639209893566234,
+      "learning_rate": 2.2977125035719088e-06,
+      "loss": 0.4718,
+      "step": 16329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8413279723786367,
+      "learning_rate": 2.297255102348962e-06,
+      "loss": 0.4704,
+      "step": 16330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5495019194121775,
+      "learning_rate": 2.296797768173492e-06,
+      "loss": 0.4338,
+      "step": 16331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3231634200303413,
+      "learning_rate": 2.2963405010550696e-06,
+      "loss": 0.5316,
+      "step": 16332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5212975322390603,
+      "learning_rate": 2.295883301003269e-06,
+      "loss": 0.4272,
+      "step": 16333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3735299166963197,
+      "learning_rate": 2.2954261680276625e-06,
+      "loss": 0.3622,
+      "step": 16334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.326026402983545,
+      "learning_rate": 2.2949691021378178e-06,
+      "loss": 0.3046,
+      "step": 16335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8347492415301223,
+      "learning_rate": 2.294512103343304e-06,
+      "loss": 0.3973,
+      "step": 16336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9577862212789003,
+      "learning_rate": 2.2940551716536895e-06,
+      "loss": 0.5048,
+      "step": 16337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7077902673633019,
+      "learning_rate": 2.293598307078538e-06,
+      "loss": 0.4412,
+      "step": 16338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8528374261746734,
+      "learning_rate": 2.2931415096274132e-06,
+      "loss": 0.487,
+      "step": 16339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5186161549880541,
+      "learning_rate": 2.29268477930988e-06,
+      "loss": 0.4605,
+      "step": 16340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9605759544885708,
+      "learning_rate": 2.2922281161354996e-06,
+      "loss": 0.4772,
+      "step": 16341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3879768048306287,
+      "learning_rate": 2.29177152011383e-06,
+      "loss": 0.36,
+      "step": 16342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7459040336338192,
+      "learning_rate": 2.291314991254431e-06,
+      "loss": 0.4872,
+      "step": 16343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.918504939612845,
+      "learning_rate": 2.290858529566861e-06,
+      "loss": 0.5944,
+      "step": 16344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2650192611869369,
+      "learning_rate": 2.290402135060673e-06,
+      "loss": 0.246,
+      "step": 16345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5764111812464734,
+      "learning_rate": 2.289945807745423e-06,
+      "loss": 0.5398,
+      "step": 16346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4825821133148416,
+      "learning_rate": 2.289489547630665e-06,
+      "loss": 0.4858,
+      "step": 16347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6479186131772596,
+      "learning_rate": 2.2890333547259476e-06,
+      "loss": 0.4306,
+      "step": 16348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7926210717838686,
+      "learning_rate": 2.2885772290408226e-06,
+      "loss": 0.4757,
+      "step": 16349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6239530698468334,
+      "learning_rate": 2.28812117058484e-06,
+      "loss": 0.4411,
+      "step": 16350
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.4936005175113678,
+      "eval_runtime": 67.9661,
+      "eval_samples_per_second": 50.172,
+      "eval_steps_per_second": 1.574,
+      "step": 16350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7255765434744699,
+      "learning_rate": 2.2876651793675443e-06,
+      "loss": 0.5057,
+      "step": 16351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5800435781738043,
+      "learning_rate": 2.287209255398483e-06,
+      "loss": 0.4764,
+      "step": 16352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2619336178582805,
+      "learning_rate": 2.2867533986872e-06,
+      "loss": 0.463,
+      "step": 16353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.567038316149474,
+      "learning_rate": 2.2862976092432405e-06,
+      "loss": 0.4474,
+      "step": 16354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6929823896039036,
+      "learning_rate": 2.285841887076143e-06,
+      "loss": 0.4933,
+      "step": 16355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7622025659153997,
+      "learning_rate": 2.285386232195449e-06,
+      "loss": 0.5394,
+      "step": 16356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.057474723618014,
+      "learning_rate": 2.284930644610699e-06,
+      "loss": 0.6483,
+      "step": 16357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9320502190006608,
+      "learning_rate": 2.284475124331428e-06,
+      "loss": 0.4451,
+      "step": 16358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.632737715251063,
+      "learning_rate": 2.284019671367173e-06,
+      "loss": 0.5051,
+      "step": 16359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5761287687993752,
+      "learning_rate": 2.28356428572747e-06,
+      "loss": 0.4554,
+      "step": 16360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.764635411652155,
+      "learning_rate": 2.283108967421849e-06,
+      "loss": 0.5077,
+      "step": 16361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0379984484421096,
+      "learning_rate": 2.282653716459844e-06,
+      "loss": 0.6128,
+      "step": 16362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8749183786095502,
+      "learning_rate": 2.2821985328509866e-06,
+      "loss": 0.5286,
+      "step": 16363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0648537740764055,
+      "learning_rate": 2.2817434166048023e-06,
+      "loss": 0.6826,
+      "step": 16364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4895140825746753,
+      "learning_rate": 2.2812883677308212e-06,
+      "loss": 0.4043,
+      "step": 16365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.006283092940254,
+      "learning_rate": 2.280833386238569e-06,
+      "loss": 0.5056,
+      "step": 16366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6640345609871448,
+      "learning_rate": 2.2803784721375718e-06,
+      "loss": 0.5179,
+      "step": 16367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8611798051300656,
+      "learning_rate": 2.2799236254373494e-06,
+      "loss": 0.506,
+      "step": 16368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8063350977661394,
+      "learning_rate": 2.2794688461474267e-06,
+      "loss": 0.4541,
+      "step": 16369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9771956620891349,
+      "learning_rate": 2.279014134277324e-06,
+      "loss": 0.401,
+      "step": 16370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0144643439875916,
+      "learning_rate": 2.2785594898365577e-06,
+      "loss": 0.5519,
+      "step": 16371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5767841868481283,
+      "learning_rate": 2.2781049128346496e-06,
+      "loss": 0.402,
+      "step": 16372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3377272927235244,
+      "learning_rate": 2.2776504032811133e-06,
+      "loss": 0.3656,
+      "step": 16373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7929213564568112,
+      "learning_rate": 2.277195961185464e-06,
+      "loss": 0.4245,
+      "step": 16374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.40194308739782,
+      "learning_rate": 2.2767415865572156e-06,
+      "loss": 0.4013,
+      "step": 16375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8438527276194099,
+      "learning_rate": 2.2762872794058807e-06,
+      "loss": 0.4783,
+      "step": 16376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0553644958242985,
+      "learning_rate": 2.2758330397409687e-06,
+      "loss": 0.5745,
+      "step": 16377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7323988141167623,
+      "learning_rate": 2.2753788675719894e-06,
+      "loss": 0.5545,
+      "step": 16378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5929151469215523,
+      "learning_rate": 2.274924762908452e-06,
+      "loss": 0.3869,
+      "step": 16379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3275232507373078,
+      "learning_rate": 2.2744707257598593e-06,
+      "loss": 0.2959,
+      "step": 16380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3751488360794035,
+      "learning_rate": 2.2740167561357196e-06,
+      "loss": 0.3226,
+      "step": 16381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5978644861166504,
+      "learning_rate": 2.2735628540455363e-06,
+      "loss": 0.4095,
+      "step": 16382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6591411362231205,
+      "learning_rate": 2.2731090194988093e-06,
+      "loss": 0.5295,
+      "step": 16383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.976738224080863,
+      "learning_rate": 2.2726552525050406e-06,
+      "loss": 0.6546,
+      "step": 16384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4447069912572048,
+      "learning_rate": 2.272201553073731e-06,
+      "loss": 0.2859,
+      "step": 16385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3732443963526957,
+      "learning_rate": 2.271747921214376e-06,
+      "loss": 0.3383,
+      "step": 16386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4341081881899511,
+      "learning_rate": 2.2712943569364725e-06,
+      "loss": 0.3575,
+      "step": 16387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7440311506874304,
+      "learning_rate": 2.2708408602495168e-06,
+      "loss": 0.4886,
+      "step": 16388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.040450751704451,
+      "learning_rate": 2.2703874311630027e-06,
+      "loss": 0.5209,
+      "step": 16389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.520392180903698,
+      "learning_rate": 2.269934069686421e-06,
+      "loss": 0.3456,
+      "step": 16390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3816110034809332,
+      "learning_rate": 2.269480775829263e-06,
+      "loss": 0.4467,
+      "step": 16391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7278193138599482,
+      "learning_rate": 2.26902754960102e-06,
+      "loss": 0.4382,
+      "step": 16392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.635462929247399,
+      "learning_rate": 2.2685743910111767e-06,
+      "loss": 0.3887,
+      "step": 16393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.484807744607308,
+      "learning_rate": 2.2681213000692213e-06,
+      "loss": 0.5131,
+      "step": 16394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8585198504633673,
+      "learning_rate": 2.267668276784641e-06,
+      "loss": 0.5261,
+      "step": 16395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8232599211536955,
+      "learning_rate": 2.2672153211669155e-06,
+      "loss": 0.4571,
+      "step": 16396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1408551020885573,
+      "learning_rate": 2.26676243322553e-06,
+      "loss": 0.4685,
+      "step": 16397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7305419829837163,
+      "learning_rate": 2.266309612969966e-06,
+      "loss": 0.5463,
+      "step": 16398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7761091431325104,
+      "learning_rate": 2.2658568604097005e-06,
+      "loss": 0.5454,
+      "step": 16399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6131975741901565,
+      "learning_rate": 2.2654041755542128e-06,
+      "loss": 0.5398,
+      "step": 16400
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.49320724606513977,
+      "eval_runtime": 67.9251,
+      "eval_samples_per_second": 50.202,
+      "eval_steps_per_second": 1.575,
+      "step": 16400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5236606798992853,
+      "learning_rate": 2.264951558412979e-06,
+      "loss": 0.3573,
+      "step": 16401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7272024292989465,
+      "learning_rate": 2.264499008995477e-06,
+      "loss": 0.4779,
+      "step": 16402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5779262357822472,
+      "learning_rate": 2.2640465273111768e-06,
+      "loss": 0.4373,
+      "step": 16403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.602456462226285,
+      "learning_rate": 2.263594113369553e-06,
+      "loss": 0.3671,
+      "step": 16404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3374375216621928,
+      "learning_rate": 2.2631417671800775e-06,
+      "loss": 0.445,
+      "step": 16405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4633655120233025,
+      "learning_rate": 2.262689488752217e-06,
+      "loss": 0.4446,
+      "step": 16406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3855021917246166,
+      "learning_rate": 2.262237278095441e-06,
+      "loss": 0.4494,
+      "step": 16407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9554950108212954,
+      "learning_rate": 2.261785135219219e-06,
+      "loss": 0.5344,
+      "step": 16408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.628187654904731,
+      "learning_rate": 2.2613330601330116e-06,
+      "loss": 0.572,
+      "step": 16409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7236075456988629,
+      "learning_rate": 2.2608810528462856e-06,
+      "loss": 0.4362,
+      "step": 16410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7398872858981496,
+      "learning_rate": 2.2604291133685034e-06,
+      "loss": 0.4829,
+      "step": 16411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.646671133603511,
+      "learning_rate": 2.2599772417091244e-06,
+      "loss": 0.4568,
+      "step": 16412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5335398778756135,
+      "learning_rate": 2.25952543787761e-06,
+      "loss": 0.4007,
+      "step": 16413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.710427834297959,
+      "learning_rate": 2.259073701883417e-06,
+      "loss": 0.3658,
+      "step": 16414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8958361526094978,
+      "learning_rate": 2.2586220337360048e-06,
+      "loss": 0.4411,
+      "step": 16415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6430757164757446,
+      "learning_rate": 2.2581704334448257e-06,
+      "loss": 0.4085,
+      "step": 16416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.831257186787559,
+      "learning_rate": 2.2577189010193355e-06,
+      "loss": 0.5756,
+      "step": 16417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5078634174604193,
+      "learning_rate": 2.257267436468987e-06,
+      "loss": 0.4271,
+      "step": 16418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3963256640848645,
+      "learning_rate": 2.256816039803229e-06,
+      "loss": 0.6302,
+      "step": 16419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5645390095901586,
+      "learning_rate": 2.2563647110315135e-06,
+      "loss": 0.3806,
+      "step": 16420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7471952967293567,
+      "learning_rate": 2.2559134501632878e-06,
+      "loss": 0.4866,
+      "step": 16421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4878963426583158,
+      "learning_rate": 2.2554622572079987e-06,
+      "loss": 0.3411,
+      "step": 16422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8056846234138253,
+      "learning_rate": 2.2550111321750925e-06,
+      "loss": 0.4263,
+      "step": 16423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9172411550113646,
+      "learning_rate": 2.254560075074014e-06,
+      "loss": 0.4122,
+      "step": 16424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1472675674275017,
+      "learning_rate": 2.254109085914202e-06,
+      "loss": 0.6418,
+      "step": 16425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4585865970331575,
+      "learning_rate": 2.2536581647051015e-06,
+      "loss": 0.406,
+      "step": 16426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5884711250559205,
+      "learning_rate": 2.2532073114561516e-06,
+      "loss": 0.4177,
+      "step": 16427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5185233352263134,
+      "learning_rate": 2.2527565261767897e-06,
+      "loss": 0.3889,
+      "step": 16428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9523405086634702,
+      "learning_rate": 2.252305808876452e-06,
+      "loss": 0.5269,
+      "step": 16429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9455666481288016,
+      "learning_rate": 2.2518551595645767e-06,
+      "loss": 0.4997,
+      "step": 16430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1524650428227274,
+      "learning_rate": 2.2514045782505947e-06,
+      "loss": 0.5215,
+      "step": 16431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8767549240452863,
+      "learning_rate": 2.2509540649439402e-06,
+      "loss": 0.4972,
+      "step": 16432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5125419008779264,
+      "learning_rate": 2.2505036196540437e-06,
+      "loss": 0.427,
+      "step": 16433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6009533616823894,
+      "learning_rate": 2.250053242390337e-06,
+      "loss": 0.4318,
+      "step": 16434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8762818117760656,
+      "learning_rate": 2.249602933162246e-06,
+      "loss": 0.3797,
+      "step": 16435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9751958474661504,
+      "learning_rate": 2.2491526919791993e-06,
+      "loss": 0.5921,
+      "step": 16436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1562479237747594,
+      "learning_rate": 2.248702518850623e-06,
+      "loss": 0.6429,
+      "step": 16437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.820808923938873,
+      "learning_rate": 2.248252413785938e-06,
+      "loss": 0.4199,
+      "step": 16438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8858340188378473,
+      "learning_rate": 2.247802376794569e-06,
+      "loss": 0.5207,
+      "step": 16439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5673213277606024,
+      "learning_rate": 2.247352407885939e-06,
+      "loss": 0.364,
+      "step": 16440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7602918676376393,
+      "learning_rate": 2.2469025070694642e-06,
+      "loss": 0.5615,
+      "step": 16441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7000667580116104,
+      "learning_rate": 2.246452674354566e-06,
+      "loss": 0.4843,
+      "step": 16442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.59391932950583,
+      "learning_rate": 2.2460029097506605e-06,
+      "loss": 0.3411,
+      "step": 16443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7487405331052344,
+      "learning_rate": 2.245553213267162e-06,
+      "loss": 0.5026,
+      "step": 16444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8514861630360564,
+      "learning_rate": 2.2451035849134854e-06,
+      "loss": 0.5083,
+      "step": 16445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7719582321261766,
+      "learning_rate": 2.2446540246990435e-06,
+      "loss": 0.5114,
+      "step": 16446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0596987386133914,
+      "learning_rate": 2.244204532633249e-06,
+      "loss": 0.5468,
+      "step": 16447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.900807453431601,
+      "learning_rate": 2.243755108725509e-06,
+      "loss": 0.4741,
+      "step": 16448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.716807439665009,
+      "learning_rate": 2.243305752985234e-06,
+      "loss": 0.3563,
+      "step": 16449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5880561293448232,
+      "learning_rate": 2.2428564654218306e-06,
+      "loss": 0.41,
+      "step": 16450
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.4934113025665283,
+      "eval_runtime": 67.9464,
+      "eval_samples_per_second": 50.187,
+      "eval_steps_per_second": 1.575,
+      "step": 16450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4560697862253908,
+      "learning_rate": 2.2424072460447027e-06,
+      "loss": 0.4422,
+      "step": 16451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7628506193773708,
+      "learning_rate": 2.2419580948632563e-06,
+      "loss": 0.5483,
+      "step": 16452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7818724842120968,
+      "learning_rate": 2.241509011886894e-06,
+      "loss": 0.4237,
+      "step": 16453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9158868416404573,
+      "learning_rate": 2.241059997125016e-06,
+      "loss": 0.4504,
+      "step": 16454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5983185673545042,
+      "learning_rate": 2.2406110505870233e-06,
+      "loss": 0.3903,
+      "step": 16455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6057668781295067,
+      "learning_rate": 2.240162172282314e-06,
+      "loss": 0.4936,
+      "step": 16456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8402963209500471,
+      "learning_rate": 2.239713362220284e-06,
+      "loss": 0.3805,
+      "step": 16457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0887365307915893,
+      "learning_rate": 2.2392646204103297e-06,
+      "loss": 0.6728,
+      "step": 16458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6058859626434436,
+      "learning_rate": 2.2388159468618455e-06,
+      "loss": 0.4529,
+      "step": 16459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9696941467411204,
+      "learning_rate": 2.238367341584225e-06,
+      "loss": 0.5601,
+      "step": 16460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8634341697534003,
+      "learning_rate": 2.2379188045868564e-06,
+      "loss": 0.4556,
+      "step": 16461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8961891865940277,
+      "learning_rate": 2.2374703358791328e-06,
+      "loss": 0.6081,
+      "step": 16462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9783468481920985,
+      "learning_rate": 2.2370219354704412e-06,
+      "loss": 0.5837,
+      "step": 16463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.713692754839265,
+      "learning_rate": 2.2365736033701686e-06,
+      "loss": 0.3757,
+      "step": 16464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7980658895978403,
+      "learning_rate": 2.2361253395876996e-06,
+      "loss": 0.5664,
+      "step": 16465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5715338747317844,
+      "learning_rate": 2.235677144132421e-06,
+      "loss": 0.4858,
+      "step": 16466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1016443449482543,
+      "learning_rate": 2.2352290170137125e-06,
+      "loss": 0.515,
+      "step": 16467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9184037166286212,
+      "learning_rate": 2.234780958240957e-06,
+      "loss": 0.4811,
+      "step": 16468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7022335646464142,
+      "learning_rate": 2.234332967823535e-06,
+      "loss": 0.4,
+      "step": 16469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5127510791206022,
+      "learning_rate": 2.2338850457708218e-06,
+      "loss": 0.6006,
+      "step": 16470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.010088076403467,
+      "learning_rate": 2.2334371920921977e-06,
+      "loss": 0.6029,
+      "step": 16471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5277717112556561,
+      "learning_rate": 2.232989406797038e-06,
+      "loss": 0.3871,
+      "step": 16472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6335496120907895,
+      "learning_rate": 2.2325416898947143e-06,
+      "loss": 0.5759,
+      "step": 16473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.737018571175296,
+      "learning_rate": 2.2320940413946014e-06,
+      "loss": 0.5293,
+      "step": 16474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8818099683001732,
+      "learning_rate": 2.2316464613060714e-06,
+      "loss": 0.6047,
+      "step": 16475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0945626316176393,
+      "learning_rate": 2.2311989496384907e-06,
+      "loss": 0.5417,
+      "step": 16476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9908527343988978,
+      "learning_rate": 2.23075150640123e-06,
+      "loss": 0.6322,
+      "step": 16477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7446456721403911,
+      "learning_rate": 2.2303041316036574e-06,
+      "loss": 0.4032,
+      "step": 16478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7480842896263724,
+      "learning_rate": 2.2298568252551352e-06,
+      "loss": 0.5833,
+      "step": 16479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7334126055387427,
+      "learning_rate": 2.2294095873650296e-06,
+      "loss": 0.4874,
+      "step": 16480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5881585093082675,
+      "learning_rate": 2.228962417942703e-06,
+      "loss": 0.4022,
+      "step": 16481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9484878702533734,
+      "learning_rate": 2.2285153169975172e-06,
+      "loss": 0.3826,
+      "step": 16482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6172225521838788,
+      "learning_rate": 2.2280682845388303e-06,
+      "loss": 0.3612,
+      "step": 16483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3903105007032022,
+      "learning_rate": 2.227621320576002e-06,
+      "loss": 0.319,
+      "step": 16484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2783883392521846,
+      "learning_rate": 2.2271744251183893e-06,
+      "loss": 0.7139,
+      "step": 16485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0509492439587866,
+      "learning_rate": 2.226727598175346e-06,
+      "loss": 0.5006,
+      "step": 16486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4485448507666192,
+      "learning_rate": 2.226280839756228e-06,
+      "loss": 0.3797,
+      "step": 16487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7798369929151947,
+      "learning_rate": 2.225834149870388e-06,
+      "loss": 0.4181,
+      "step": 16488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4983921564858638,
+      "learning_rate": 2.225387528527175e-06,
+      "loss": 0.4335,
+      "step": 16489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7252669223533028,
+      "learning_rate": 2.2249409757359405e-06,
+      "loss": 0.5486,
+      "step": 16490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.540571211972362,
+      "learning_rate": 2.2244944915060337e-06,
+      "loss": 0.4018,
+      "step": 16491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7439211272462416,
+      "learning_rate": 2.2240480758467985e-06,
+      "loss": 0.4169,
+      "step": 16492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3168274041041528,
+      "learning_rate": 2.223601728767582e-06,
+      "loss": 0.443,
+      "step": 16493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1900567439285226,
+      "learning_rate": 2.2231554502777287e-06,
+      "loss": 0.5115,
+      "step": 16494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7515042600813628,
+      "learning_rate": 2.2227092403865824e-06,
+      "loss": 0.5197,
+      "step": 16495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8727689315289162,
+      "learning_rate": 2.2222630991034807e-06,
+      "loss": 0.5338,
+      "step": 16496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5697775131856009,
+      "learning_rate": 2.2218170264377655e-06,
+      "loss": 0.4348,
+      "step": 16497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6898926008200734,
+      "learning_rate": 2.2213710223987756e-06,
+      "loss": 0.5685,
+      "step": 16498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7048060401511997,
+      "learning_rate": 2.2209250869958456e-06,
+      "loss": 0.4751,
+      "step": 16499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.764393026739503,
+      "learning_rate": 2.2204792202383125e-06,
+      "loss": 0.5337,
+      "step": 16500
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.4924033582210541,
+      "eval_runtime": 67.9312,
+      "eval_samples_per_second": 50.198,
+      "eval_steps_per_second": 1.575,
+      "step": 16500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7081837912779176,
+      "learning_rate": 2.220033422135511e-06,
+      "loss": 0.4024,
+      "step": 16501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0485179531715656,
+      "learning_rate": 2.219587692696771e-06,
+      "loss": 0.639,
+      "step": 16502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7278426074210353,
+      "learning_rate": 2.219142031931426e-06,
+      "loss": 0.5562,
+      "step": 16503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6810837912403456,
+      "learning_rate": 2.218696439848804e-06,
+      "loss": 0.5237,
+      "step": 16504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.073099234500854,
+      "learning_rate": 2.218250916458236e-06,
+      "loss": 0.5665,
+      "step": 16505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6923127414604908,
+      "learning_rate": 2.217805461769045e-06,
+      "loss": 0.5433,
+      "step": 16506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7708167909581913,
+      "learning_rate": 2.2173600757905577e-06,
+      "loss": 0.4517,
+      "step": 16507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5380919970425868,
+      "learning_rate": 2.2169147585321e-06,
+      "loss": 0.3083,
+      "step": 16508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6705870369931037,
+      "learning_rate": 2.2164695100029905e-06,
+      "loss": 0.4013,
+      "step": 16509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5127988575089857,
+      "learning_rate": 2.216024330212553e-06,
+      "loss": 0.4842,
+      "step": 16510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9532094848238133,
+      "learning_rate": 2.215579219170108e-06,
+      "loss": 0.4603,
+      "step": 16511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9574427021971956,
+      "learning_rate": 2.21513417688497e-06,
+      "loss": 0.5246,
+      "step": 16512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.408820732412592,
+      "learning_rate": 2.214689203366458e-06,
+      "loss": 0.7004,
+      "step": 16513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5938201220840074,
+      "learning_rate": 2.2142442986238883e-06,
+      "loss": 0.4311,
+      "step": 16514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.56906353169048,
+      "learning_rate": 2.213799462666572e-06,
+      "loss": 0.4436,
+      "step": 16515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3229900515330253,
+      "learning_rate": 2.2133546955038228e-06,
+      "loss": 0.5557,
+      "step": 16516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8530445974133276,
+      "learning_rate": 2.212909997144952e-06,
+      "loss": 0.6437,
+      "step": 16517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6639630269271353,
+      "learning_rate": 2.2124653675992693e-06,
+      "loss": 0.4523,
+      "step": 16518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6499590898534056,
+      "learning_rate": 2.2120208068760803e-06,
+      "loss": 0.4062,
+      "step": 16519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4690187787288862,
+      "learning_rate": 2.2115763149846957e-06,
+      "loss": 0.3101,
+      "step": 16520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.852567575902064,
+      "learning_rate": 2.211131891934417e-06,
+      "loss": 0.4674,
+      "step": 16521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6878165109619052,
+      "learning_rate": 2.2106875377345494e-06,
+      "loss": 0.4947,
+      "step": 16522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9087377586516954,
+      "learning_rate": 2.2102432523943966e-06,
+      "loss": 0.6833,
+      "step": 16523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.977027670054712,
+      "learning_rate": 2.209799035923257e-06,
+      "loss": 0.4347,
+      "step": 16524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7449119385826122,
+      "learning_rate": 2.2093548883304303e-06,
+      "loss": 0.445,
+      "step": 16525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5049670925854643,
+      "learning_rate": 2.2089108096252156e-06,
+      "loss": 0.3975,
+      "step": 16526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.429216999163451,
+      "learning_rate": 2.20846679981691e-06,
+      "loss": 0.2878,
+      "step": 16527
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8388066095974234,
+      "learning_rate": 2.208022858914806e-06,
+      "loss": 0.557,
+      "step": 16528
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4873934912353055,
+      "learning_rate": 2.2075789869281987e-06,
+      "loss": 0.4931,
+      "step": 16529
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9584991471810216,
+      "learning_rate": 2.207135183866382e-06,
+      "loss": 0.5333,
+      "step": 16530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4763056353768558,
+      "learning_rate": 2.2066914497386434e-06,
+      "loss": 0.3429,
+      "step": 16531
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09845849279973,
+      "learning_rate": 2.206247784554274e-06,
+      "loss": 0.5514,
+      "step": 16532
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9436595673461008,
+      "learning_rate": 2.205804188322563e-06,
+      "loss": 0.4118,
+      "step": 16533
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7566112798161988,
+      "learning_rate": 2.2053606610527935e-06,
+      "loss": 0.4524,
+      "step": 16534
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.812339429636297,
+      "learning_rate": 2.2049172027542517e-06,
+      "loss": 0.4987,
+      "step": 16535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7209360810076046,
+      "learning_rate": 2.2044738134362237e-06,
+      "loss": 0.4452,
+      "step": 16536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6194464205171464,
+      "learning_rate": 2.2040304931079878e-06,
+      "loss": 0.5675,
+      "step": 16537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7596853467571987,
+      "learning_rate": 2.203587241778826e-06,
+      "loss": 0.4039,
+      "step": 16538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5858752168319152,
+      "learning_rate": 2.2031440594580182e-06,
+      "loss": 0.4425,
+      "step": 16539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5108503306812464,
+      "learning_rate": 2.202700946154843e-06,
+      "loss": 0.3899,
+      "step": 16540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.197191364080968,
+      "learning_rate": 2.202257901878575e-06,
+      "loss": 0.6188,
+      "step": 16541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.377337569935454,
+      "learning_rate": 2.201814926638488e-06,
+      "loss": 0.4794,
+      "step": 16542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6325368793720154,
+      "learning_rate": 2.201372020443859e-06,
+      "loss": 0.4567,
+      "step": 16543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3119237961438595,
+      "learning_rate": 2.200929183303957e-06,
+      "loss": 0.4408,
+      "step": 16544
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3010225116287235,
+      "learning_rate": 2.200486415228053e-06,
+      "loss": 0.4449,
+      "step": 16545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4020474817707984,
+      "learning_rate": 2.2000437162254178e-06,
+      "loss": 0.4077,
+      "step": 16546
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8910890383594228,
+      "learning_rate": 2.199601086305316e-06,
+      "loss": 0.4555,
+      "step": 16547
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6070044944853061,
+      "learning_rate": 2.1991585254770166e-06,
+      "loss": 0.5393,
+      "step": 16548
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0480752237746387,
+      "learning_rate": 2.1987160337497832e-06,
+      "loss": 0.6056,
+      "step": 16549
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.068438201986122,
+      "learning_rate": 2.198273611132879e-06,
+      "loss": 0.6096,
+      "step": 16550
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.4927782714366913,
+      "eval_runtime": 68.0371,
+      "eval_samples_per_second": 50.12,
+      "eval_steps_per_second": 1.573,
+      "step": 16550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.022470463364116,
+      "learning_rate": 2.1978312576355654e-06,
+      "loss": 0.426,
+      "step": 16551
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1525851598739325,
+      "learning_rate": 2.1973889732671034e-06,
+      "loss": 0.4947,
+      "step": 16552
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9857315777559985,
+      "learning_rate": 2.1969467580367533e-06,
+      "loss": 0.485,
+      "step": 16553
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.580364906985609,
+      "learning_rate": 2.1965046119537696e-06,
+      "loss": 0.3461,
+      "step": 16554
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.801499857122631,
+      "learning_rate": 2.19606253502741e-06,
+      "loss": 0.6125,
+      "step": 16555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3997844100849546,
+      "learning_rate": 2.195620527266931e-06,
+      "loss": 0.3327,
+      "step": 16556
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2471039150872456,
+      "learning_rate": 2.1951785886815824e-06,
+      "loss": 0.5362,
+      "step": 16557
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1351539754929076,
+      "learning_rate": 2.1947367192806168e-06,
+      "loss": 0.6707,
+      "step": 16558
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.661331208018179,
+      "learning_rate": 2.1942949190732868e-06,
+      "loss": 0.4577,
+      "step": 16559
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9212632923623827,
+      "learning_rate": 2.1938531880688374e-06,
+      "loss": 0.5548,
+      "step": 16560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0062505005412783,
+      "learning_rate": 2.1934115262765187e-06,
+      "loss": 0.5732,
+      "step": 16561
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9176600562780144,
+      "learning_rate": 2.1929699337055777e-06,
+      "loss": 0.466,
+      "step": 16562
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9066179051823806,
+      "learning_rate": 2.1925284103652547e-06,
+      "loss": 0.4794,
+      "step": 16563
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3450228447974637,
+      "learning_rate": 2.192086956264795e-06,
+      "loss": 0.3412,
+      "step": 16564
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.370967913016982,
+      "learning_rate": 2.1916455714134407e-06,
+      "loss": 0.3356,
+      "step": 16565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9326902501550973,
+      "learning_rate": 2.1912042558204333e-06,
+      "loss": 0.6219,
+      "step": 16566
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.157271891839123,
+      "learning_rate": 2.1907630094950076e-06,
+      "loss": 0.7467,
+      "step": 16567
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.801808467224235,
+      "learning_rate": 2.1903218324464033e-06,
+      "loss": 0.5365,
+      "step": 16568
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4936541187843106,
+      "learning_rate": 2.189880724683855e-06,
+      "loss": 0.4121,
+      "step": 16569
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9077602265315317,
+      "learning_rate": 2.1894396862165985e-06,
+      "loss": 0.5996,
+      "step": 16570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9634306934167949,
+      "learning_rate": 2.188998717053867e-06,
+      "loss": 0.4772,
+      "step": 16571
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7642799290048123,
+      "learning_rate": 2.1885578172048895e-06,
+      "loss": 0.6071,
+      "step": 16572
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0114102952922317,
+      "learning_rate": 2.188116986678897e-06,
+      "loss": 0.5902,
+      "step": 16573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5309688240120014,
+      "learning_rate": 2.1876762254851185e-06,
+      "loss": 0.3618,
+      "step": 16574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.095476034810948,
+      "learning_rate": 2.1872355336327824e-06,
+      "loss": 0.5038,
+      "step": 16575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8049096771100521,
+      "learning_rate": 2.1867949111311108e-06,
+      "loss": 0.4936,
+      "step": 16576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6446590888426205,
+      "learning_rate": 2.1863543579893305e-06,
+      "loss": 0.3941,
+      "step": 16577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6834214269304753,
+      "learning_rate": 2.1859138742166645e-06,
+      "loss": 0.4355,
+      "step": 16578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7945817164279911,
+      "learning_rate": 2.1854734598223322e-06,
+      "loss": 0.4656,
+      "step": 16579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8629133347486355,
+      "learning_rate": 2.185033114815554e-06,
+      "loss": 0.4257,
+      "step": 16580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.568971237073436,
+      "learning_rate": 2.1845928392055497e-06,
+      "loss": 0.4728,
+      "step": 16581
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.685919198928921,
+      "learning_rate": 2.1841526330015338e-06,
+      "loss": 0.482,
+      "step": 16582
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5530789525361226,
+      "learning_rate": 2.1837124962127234e-06,
+      "loss": 0.3964,
+      "step": 16583
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7209311993693377,
+      "learning_rate": 2.1832724288483313e-06,
+      "loss": 0.4108,
+      "step": 16584
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4165196930525286,
+      "learning_rate": 2.1828324309175724e-06,
+      "loss": 0.3878,
+      "step": 16585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.769893126636981,
+      "learning_rate": 2.1823925024296555e-06,
+      "loss": 0.5496,
+      "step": 16586
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5965890942249985,
+      "learning_rate": 2.1819526433937906e-06,
+      "loss": 0.4704,
+      "step": 16587
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7828146918749495,
+      "learning_rate": 2.1815128538191873e-06,
+      "loss": 0.5441,
+      "step": 16588
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.510134297354676,
+      "learning_rate": 2.18107313371505e-06,
+      "loss": 0.3765,
+      "step": 16589
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9953048304930134,
+      "learning_rate": 2.180633483090585e-06,
+      "loss": 0.4292,
+      "step": 16590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5898738677240252,
+      "learning_rate": 2.180193901954998e-06,
+      "loss": 0.4802,
+      "step": 16591
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6908563394783098,
+      "learning_rate": 2.1797543903174877e-06,
+      "loss": 0.4633,
+      "step": 16592
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.18282847914123,
+      "learning_rate": 2.179314948187257e-06,
+      "loss": 0.6708,
+      "step": 16593
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7217397526754958,
+      "learning_rate": 2.1788755755735074e-06,
+      "loss": 0.6189,
+      "step": 16594
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0533289769537473,
+      "learning_rate": 2.1784362724854325e-06,
+      "loss": 0.5934,
+      "step": 16595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5976409377864276,
+      "learning_rate": 2.177997038932232e-06,
+      "loss": 0.3743,
+      "step": 16596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.683431221939351,
+      "learning_rate": 2.1775578749231e-06,
+      "loss": 0.4995,
+      "step": 16597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7851398342319142,
+      "learning_rate": 2.1771187804672312e-06,
+      "loss": 0.4798,
+      "step": 16598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2321254704118596,
+      "learning_rate": 2.1766797555738155e-06,
+      "loss": 0.4263,
+      "step": 16599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7145271325115323,
+      "learning_rate": 2.176240800252045e-06,
+      "loss": 0.3959,
+      "step": 16600
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.4924680292606354,
+      "eval_runtime": 68.0527,
+      "eval_samples_per_second": 50.108,
+      "eval_steps_per_second": 1.572,
+      "step": 16600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0382655687079074,
+      "learning_rate": 2.17580191451111e-06,
+      "loss": 0.5293,
+      "step": 16601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3532869521477195,
+      "learning_rate": 2.175363098360197e-06,
+      "loss": 0.4999,
+      "step": 16602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5463656045695529,
+      "learning_rate": 2.1749243518084913e-06,
+      "loss": 0.481,
+      "step": 16603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6210785478850827,
+      "learning_rate": 2.174485674865181e-06,
+      "loss": 0.4381,
+      "step": 16604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6995436819257432,
+      "learning_rate": 2.1740470675394452e-06,
+      "loss": 0.5491,
+      "step": 16605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.39832220441225,
+      "learning_rate": 2.1736085298404695e-06,
+      "loss": 0.3228,
+      "step": 16606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7222413442176312,
+      "learning_rate": 2.173170061777434e-06,
+      "loss": 0.5285,
+      "step": 16607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.379332364941836,
+      "learning_rate": 2.1727316633595152e-06,
+      "loss": 0.4549,
+      "step": 16608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7462195374292535,
+      "learning_rate": 2.172293334595893e-06,
+      "loss": 0.5918,
+      "step": 16609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.575563111236623,
+      "learning_rate": 2.1718550754957422e-06,
+      "loss": 0.451,
+      "step": 16610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.087566661803333,
+      "learning_rate": 2.17141688606824e-06,
+      "loss": 0.5061,
+      "step": 16611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.70953082249077,
+      "learning_rate": 2.1709787663225564e-06,
+      "loss": 0.366,
+      "step": 16612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3866392999472186,
+      "learning_rate": 2.1705407162678645e-06,
+      "loss": 0.4228,
+      "step": 16613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6070527206509893,
+      "learning_rate": 2.1701027359133366e-06,
+      "loss": 0.5308,
+      "step": 16614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0499729768059387,
+      "learning_rate": 2.1696648252681375e-06,
+      "loss": 0.768,
+      "step": 16615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7521743022430403,
+      "learning_rate": 2.1692269843414373e-06,
+      "loss": 0.4112,
+      "step": 16616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4055633344885377,
+      "learning_rate": 2.168789213142402e-06,
+      "loss": 0.3657,
+      "step": 16617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7438914804541819,
+      "learning_rate": 2.1683515116801935e-06,
+      "loss": 0.4213,
+      "step": 16618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9716424874470533,
+      "learning_rate": 2.1679138799639778e-06,
+      "loss": 0.5419,
+      "step": 16619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2701868122569917,
+      "learning_rate": 2.1674763180029166e-06,
+      "loss": 0.3545,
+      "step": 16620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2653288788024435,
+      "learning_rate": 2.1670388258061677e-06,
+      "loss": 0.5879,
+      "step": 16621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6965863555648337,
+      "learning_rate": 2.16660140338289e-06,
+      "loss": 0.3943,
+      "step": 16622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6723029978389183,
+      "learning_rate": 2.1661640507422443e-06,
+      "loss": 0.4243,
+      "step": 16623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8007542516408188,
+      "learning_rate": 2.1657267678933806e-06,
+      "loss": 0.4284,
+      "step": 16624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5453587094004384,
+      "learning_rate": 2.1652895548454573e-06,
+      "loss": 0.3332,
+      "step": 16625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4715983546884506,
+      "learning_rate": 2.1648524116076268e-06,
+      "loss": 0.4042,
+      "step": 16626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0426842335209616,
+      "learning_rate": 2.164415338189038e-06,
+      "loss": 0.3918,
+      "step": 16627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9358205719443933,
+      "learning_rate": 2.1639783345988425e-06,
+      "loss": 0.488,
+      "step": 16628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.692463951803529,
+      "learning_rate": 2.1635414008461897e-06,
+      "loss": 0.4448,
+      "step": 16629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.247068986998109,
+      "learning_rate": 2.1631045369402235e-06,
+      "loss": 0.6303,
+      "step": 16630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4907814465594718,
+      "learning_rate": 2.1626677428900916e-06,
+      "loss": 0.3184,
+      "step": 16631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0817782484587157,
+      "learning_rate": 2.162231018704938e-06,
+      "loss": 0.5377,
+      "step": 16632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9226624196220437,
+      "learning_rate": 2.1617943643939056e-06,
+      "loss": 0.5354,
+      "step": 16633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8049457690762571,
+      "learning_rate": 2.161357779966134e-06,
+      "loss": 0.4061,
+      "step": 16634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6861368510742836,
+      "learning_rate": 2.1609212654307627e-06,
+      "loss": 0.5713,
+      "step": 16635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5537438455103534,
+      "learning_rate": 2.160484820796933e-06,
+      "loss": 0.4114,
+      "step": 16636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.699257558835013,
+      "learning_rate": 2.160048446073778e-06,
+      "loss": 0.4604,
+      "step": 16637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1695991017502636,
+      "learning_rate": 2.1596121412704334e-06,
+      "loss": 0.5932,
+      "step": 16638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5625394545747517,
+      "learning_rate": 2.159175906396036e-06,
+      "loss": 0.4659,
+      "step": 16639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0114902244088015,
+      "learning_rate": 2.158739741459714e-06,
+      "loss": 0.4121,
+      "step": 16640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1205132232070363,
+      "learning_rate": 2.1583036464706005e-06,
+      "loss": 0.6666,
+      "step": 16641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.061496822297282,
+      "learning_rate": 2.1578676214378257e-06,
+      "loss": 0.5747,
+      "step": 16642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9332239193691023,
+      "learning_rate": 2.157431666370515e-06,
+      "loss": 0.6634,
+      "step": 16643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.600546492583941,
+      "learning_rate": 2.1569957812777964e-06,
+      "loss": 0.4276,
+      "step": 16644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9974706730541107,
+      "learning_rate": 2.1565599661687947e-06,
+      "loss": 0.5003,
+      "step": 16645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1962421316946505,
+      "learning_rate": 2.1561242210526347e-06,
+      "loss": 0.4529,
+      "step": 16646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7362328785597942,
+      "learning_rate": 2.1556885459384357e-06,
+      "loss": 0.5307,
+      "step": 16647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4528755282377526,
+      "learning_rate": 2.1552529408353205e-06,
+      "loss": 0.3259,
+      "step": 16648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9157069708306012,
+      "learning_rate": 2.154817405752408e-06,
+      "loss": 0.8104,
+      "step": 16649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6011431679645947,
+      "learning_rate": 2.1543819406988142e-06,
+      "loss": 0.4879,
+      "step": 16650
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.4918224513530731,
+      "eval_runtime": 68.0375,
+      "eval_samples_per_second": 50.119,
+      "eval_steps_per_second": 1.573,
+      "step": 16650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.882654213264473,
+      "learning_rate": 2.153946545683657e-06,
+      "loss": 0.5049,
+      "step": 16651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4195629072655667,
+      "learning_rate": 2.153511220716051e-06,
+      "loss": 0.4291,
+      "step": 16652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8034665758700956,
+      "learning_rate": 2.1530759658051075e-06,
+      "loss": 0.5594,
+      "step": 16653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8740211197340093,
+      "learning_rate": 2.152640780959941e-06,
+      "loss": 0.5669,
+      "step": 16654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7337500375564305,
+      "learning_rate": 2.152205666189661e-06,
+      "loss": 0.4466,
+      "step": 16655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.760722719814784,
+      "learning_rate": 2.151770621503375e-06,
+      "loss": 0.5783,
+      "step": 16656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8422892906503658,
+      "learning_rate": 2.1513356469101915e-06,
+      "loss": 0.3181,
+      "step": 16657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7649146538493918,
+      "learning_rate": 2.1509007424192165e-06,
+      "loss": 0.4559,
+      "step": 16658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.833558147756404,
+      "learning_rate": 2.1504659080395547e-06,
+      "loss": 0.4916,
+      "step": 16659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.884259594303208,
+      "learning_rate": 2.1500311437803077e-06,
+      "loss": 0.3552,
+      "step": 16660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9065210037803815,
+      "learning_rate": 2.149596449650578e-06,
+      "loss": 0.5283,
+      "step": 16661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7621767748910582,
+      "learning_rate": 2.1491618256594673e-06,
+      "loss": 0.5184,
+      "step": 16662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.77782027824229,
+      "learning_rate": 2.148727271816071e-06,
+      "loss": 0.4136,
+      "step": 16663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6391027185499056,
+      "learning_rate": 2.148292788129488e-06,
+      "loss": 0.3702,
+      "step": 16664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9008171159778857,
+      "learning_rate": 2.147858374608814e-06,
+      "loss": 0.4804,
+      "step": 16665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.085727828830995,
+      "learning_rate": 2.1474240312631422e-06,
+      "loss": 0.5883,
+      "step": 16666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7073669202334298,
+      "learning_rate": 2.1469897581015646e-06,
+      "loss": 0.4557,
+      "step": 16667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.932826333526124,
+      "learning_rate": 2.1465555551331766e-06,
+      "loss": 0.5306,
+      "step": 16668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9220590220247018,
+      "learning_rate": 2.1461214223670635e-06,
+      "loss": 0.4637,
+      "step": 16669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.68454981812957,
+      "learning_rate": 2.1456873598123152e-06,
+      "loss": 0.4444,
+      "step": 16670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9806254992218815,
+      "learning_rate": 2.14525336747802e-06,
+      "loss": 0.5343,
+      "step": 16671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0281016820986064,
+      "learning_rate": 2.1448194453732606e-06,
+      "loss": 0.5567,
+      "step": 16672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4502546191917856,
+      "learning_rate": 2.1443855935071223e-06,
+      "loss": 0.4271,
+      "step": 16673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6835474455430397,
+      "learning_rate": 2.1439518118886883e-06,
+      "loss": 0.4853,
+      "step": 16674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7624624873009709,
+      "learning_rate": 2.1435181005270374e-06,
+      "loss": 0.3468,
+      "step": 16675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1657881547542224,
+      "learning_rate": 2.1430844594312507e-06,
+      "loss": 0.4849,
+      "step": 16676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.746367036603448,
+      "learning_rate": 2.1426508886104054e-06,
+      "loss": 0.4071,
+      "step": 16677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9117225038479755,
+      "learning_rate": 2.1422173880735796e-06,
+      "loss": 0.5066,
+      "step": 16678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7645628868904684,
+      "learning_rate": 2.1417839578298455e-06,
+      "loss": 0.514,
+      "step": 16679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.994309986548391,
+      "learning_rate": 2.1413505978882794e-06,
+      "loss": 0.5245,
+      "step": 16680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7005134984729326,
+      "learning_rate": 2.140917308257953e-06,
+      "loss": 0.4184,
+      "step": 16681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6672240446109368,
+      "learning_rate": 2.1404840889479357e-06,
+      "loss": 0.5141,
+      "step": 16682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6543031944093296,
+      "learning_rate": 2.1400509399672965e-06,
+      "loss": 0.4314,
+      "step": 16683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7716882853410216,
+      "learning_rate": 2.1396178613251058e-06,
+      "loss": 0.4704,
+      "step": 16684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6761025507803995,
+      "learning_rate": 2.1391848530304265e-06,
+      "loss": 0.4515,
+      "step": 16685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6443336410719886,
+      "learning_rate": 2.138751915092325e-06,
+      "loss": 0.4771,
+      "step": 16686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.558155851717589,
+      "learning_rate": 2.1383190475198653e-06,
+      "loss": 0.5037,
+      "step": 16687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.846233943008376,
+      "learning_rate": 2.137886250322107e-06,
+      "loss": 0.503,
+      "step": 16688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.229138526760543,
+      "learning_rate": 2.137453523508112e-06,
+      "loss": 0.6126,
+      "step": 16689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4902909087113172,
+      "learning_rate": 2.1370208670869386e-06,
+      "loss": 0.3594,
+      "step": 16690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.798646349315153,
+      "learning_rate": 2.1365882810676463e-06,
+      "loss": 0.5092,
+      "step": 16691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4770155212682785,
+      "learning_rate": 2.136155765459288e-06,
+      "loss": 0.4459,
+      "step": 16692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8761264993800821,
+      "learning_rate": 2.1357233202709193e-06,
+      "loss": 0.5434,
+      "step": 16693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8425965241566866,
+      "learning_rate": 2.1352909455115946e-06,
+      "loss": 0.4075,
+      "step": 16694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5504376055473905,
+      "learning_rate": 2.134858641190362e-06,
+      "loss": 0.3621,
+      "step": 16695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6263657226063142,
+      "learning_rate": 2.1344264073162745e-06,
+      "loss": 0.4232,
+      "step": 16696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5750348492695863,
+      "learning_rate": 2.1339942438983803e-06,
+      "loss": 0.3692,
+      "step": 16697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4215905601002183,
+      "learning_rate": 2.1335621509457253e-06,
+      "loss": 0.5896,
+      "step": 16698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8609166539092528,
+      "learning_rate": 2.1331301284673555e-06,
+      "loss": 0.5411,
+      "step": 16699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4501787534723851,
+      "learning_rate": 2.1326981764723167e-06,
+      "loss": 0.4024,
+      "step": 16700
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.49206414818763733,
+      "eval_runtime": 68.0674,
+      "eval_samples_per_second": 50.097,
+      "eval_steps_per_second": 1.572,
+      "step": 16700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.613206389039781,
+      "learning_rate": 2.132266294969648e-06,
+      "loss": 0.5068,
+      "step": 16701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.039376457945496,
+      "learning_rate": 2.131834483968393e-06,
+      "loss": 0.6281,
+      "step": 16702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8499419541426743,
+      "learning_rate": 2.1314027434775914e-06,
+      "loss": 0.5051,
+      "step": 16703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.559638767175366,
+      "learning_rate": 2.130971073506282e-06,
+      "loss": 0.5819,
+      "step": 16704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.906620017447916,
+      "learning_rate": 2.1305394740635e-06,
+      "loss": 0.4029,
+      "step": 16705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1966899585157766,
+      "learning_rate": 2.13010794515828e-06,
+      "loss": 0.7307,
+      "step": 16706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8864164039320215,
+      "learning_rate": 2.1296764867996594e-06,
+      "loss": 0.5062,
+      "step": 16707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8902799245233737,
+      "learning_rate": 2.1292450989966666e-06,
+      "loss": 0.5106,
+      "step": 16708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.818961365746512,
+      "learning_rate": 2.128813781758334e-06,
+      "loss": 0.4003,
+      "step": 16709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7382351019348743,
+      "learning_rate": 2.128382535093692e-06,
+      "loss": 0.4373,
+      "step": 16710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1563005166072498,
+      "learning_rate": 2.127951359011766e-06,
+      "loss": 0.6017,
+      "step": 16711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6970201060800136,
+      "learning_rate": 2.127520253521584e-06,
+      "loss": 0.4364,
+      "step": 16712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8407313624063404,
+      "learning_rate": 2.1270892186321724e-06,
+      "loss": 0.4926,
+      "step": 16713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4989534613009674,
+      "learning_rate": 2.1266582543525515e-06,
+      "loss": 0.4962,
+      "step": 16714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9413010592234863,
+      "learning_rate": 2.126227360691745e-06,
+      "loss": 0.4712,
+      "step": 16715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8348499810072945,
+      "learning_rate": 2.1257965376587728e-06,
+      "loss": 0.4236,
+      "step": 16716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8582212727909602,
+      "learning_rate": 2.1253657852626542e-06,
+      "loss": 0.4614,
+      "step": 16717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0041249836428383,
+      "learning_rate": 2.1249351035124072e-06,
+      "loss": 0.6682,
+      "step": 16718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.840672031198475,
+      "learning_rate": 2.124504492417049e-06,
+      "loss": 0.6955,
+      "step": 16719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5534854406324579,
+      "learning_rate": 2.1240739519855904e-06,
+      "loss": 0.4023,
+      "step": 16720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7291522569426836,
+      "learning_rate": 2.1236434822270475e-06,
+      "loss": 0.5481,
+      "step": 16721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.523678423020937,
+      "learning_rate": 2.1232130831504326e-06,
+      "loss": 0.4724,
+      "step": 16722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3138132485173366,
+      "learning_rate": 2.1227827547647527e-06,
+      "loss": 0.6133,
+      "step": 16723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6837679582106861,
+      "learning_rate": 2.122352497079018e-06,
+      "loss": 0.387,
+      "step": 16724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6470849944895276,
+      "learning_rate": 2.1219223101022362e-06,
+      "loss": 0.4452,
+      "step": 16725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8583913154134737,
+      "learning_rate": 2.1214921938434143e-06,
+      "loss": 0.5329,
+      "step": 16726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6769336655043325,
+      "learning_rate": 2.1210621483115525e-06,
+      "loss": 0.4213,
+      "step": 16727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.430374469123621,
+      "learning_rate": 2.1206321735156565e-06,
+      "loss": 0.781,
+      "step": 16728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.616315628411636,
+      "learning_rate": 2.1202022694647286e-06,
+      "loss": 0.4065,
+      "step": 16729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.849109456746918,
+      "learning_rate": 2.119772436167765e-06,
+      "loss": 0.4224,
+      "step": 16730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9756136090734628,
+      "learning_rate": 2.1193426736337664e-06,
+      "loss": 0.4085,
+      "step": 16731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8360057887351056,
+      "learning_rate": 2.11891298187173e-06,
+      "loss": 0.5026,
+      "step": 16732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8486939511456975,
+      "learning_rate": 2.118483360890649e-06,
+      "loss": 0.5054,
+      "step": 16733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.000849485863303,
+      "learning_rate": 2.1180538106995192e-06,
+      "loss": 0.4424,
+      "step": 16734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5320781171466795,
+      "learning_rate": 2.1176243313073325e-06,
+      "loss": 0.4065,
+      "step": 16735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8058317536624406,
+      "learning_rate": 2.1171949227230792e-06,
+      "loss": 0.5598,
+      "step": 16736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8116359977505372,
+      "learning_rate": 2.1167655849557487e-06,
+      "loss": 0.4517,
+      "step": 16737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9733096156010128,
+      "learning_rate": 2.1163363180143295e-06,
+      "loss": 0.5692,
+      "step": 16738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7470449912173582,
+      "learning_rate": 2.1159071219078087e-06,
+      "loss": 0.5596,
+      "step": 16739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.814379293691226,
+      "learning_rate": 2.115477996645169e-06,
+      "loss": 0.5551,
+      "step": 16740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8744861907397807,
+      "learning_rate": 2.115048942235395e-06,
+      "loss": 0.5407,
+      "step": 16741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7612918949163066,
+      "learning_rate": 2.114619958687471e-06,
+      "loss": 0.4825,
+      "step": 16742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8239151556546087,
+      "learning_rate": 2.1141910460103735e-06,
+      "loss": 0.5095,
+      "step": 16743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4718751340766785,
+      "learning_rate": 2.113762204213084e-06,
+      "loss": 0.3588,
+      "step": 16744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6379602078392845,
+      "learning_rate": 2.11333343330458e-06,
+      "loss": 0.4245,
+      "step": 16745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8339247282108035,
+      "learning_rate": 2.112904733293837e-06,
+      "loss": 0.5317,
+      "step": 16746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7241281493700362,
+      "learning_rate": 2.112476104189829e-06,
+      "loss": 0.5005,
+      "step": 16747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9648169946039622,
+      "learning_rate": 2.1120475460015303e-06,
+      "loss": 0.6438,
+      "step": 16748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9648829509683903,
+      "learning_rate": 2.111619058737911e-06,
+      "loss": 0.5514,
+      "step": 16749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7152305072821497,
+      "learning_rate": 2.1111906424079424e-06,
+      "loss": 0.4298,
+      "step": 16750
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.49209895730018616,
+      "eval_runtime": 68.0655,
+      "eval_samples_per_second": 50.099,
+      "eval_steps_per_second": 1.572,
+      "step": 16750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8348170010616103,
+      "learning_rate": 2.1107622970205928e-06,
+      "loss": 0.5657,
+      "step": 16751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.390526906356096,
+      "learning_rate": 2.11033402258483e-06,
+      "loss": 0.3178,
+      "step": 16752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5284813537387498,
+      "learning_rate": 2.1099058191096187e-06,
+      "loss": 0.4153,
+      "step": 16753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9770672653911634,
+      "learning_rate": 2.109477686603923e-06,
+      "loss": 0.5397,
+      "step": 16754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9758919109738122,
+      "learning_rate": 2.1090496250767073e-06,
+      "loss": 0.4413,
+      "step": 16755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7228357303657262,
+      "learning_rate": 2.1086216345369303e-06,
+      "loss": 0.4896,
+      "step": 16756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.698661366919792,
+      "learning_rate": 2.1081937149935526e-06,
+      "loss": 0.4835,
+      "step": 16757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4218210178684956,
+      "learning_rate": 2.1077658664555343e-06,
+      "loss": 0.363,
+      "step": 16758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7779128768614283,
+      "learning_rate": 2.1073380889318292e-06,
+      "loss": 0.4552,
+      "step": 16759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5345393420678797,
+      "learning_rate": 2.106910382431394e-06,
+      "loss": 0.4146,
+      "step": 16760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4867215155391953,
+      "learning_rate": 2.1064827469631827e-06,
+      "loss": 0.3602,
+      "step": 16761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.79141082576102,
+      "learning_rate": 2.106055182536148e-06,
+      "loss": 0.499,
+      "step": 16762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6995016983433715,
+      "learning_rate": 2.1056276891592394e-06,
+      "loss": 0.441,
+      "step": 16763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1499803210656743,
+      "learning_rate": 2.105200266841406e-06,
+      "loss": 0.7774,
+      "step": 16764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.129097953096552,
+      "learning_rate": 2.104772915591597e-06,
+      "loss": 0.6589,
+      "step": 16765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1246813209850193,
+      "learning_rate": 2.1043456354187582e-06,
+      "loss": 0.6607,
+      "step": 16766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.146828878625548,
+      "learning_rate": 2.1039184263318353e-06,
+      "loss": 0.4925,
+      "step": 16767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5721786636967263,
+      "learning_rate": 2.1034912883397695e-06,
+      "loss": 0.407,
+      "step": 16768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9307790126785542,
+      "learning_rate": 2.1030642214515045e-06,
+      "loss": 0.6595,
+      "step": 16769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9081628802277486,
+      "learning_rate": 2.1026372256759795e-06,
+      "loss": 0.4276,
+      "step": 16770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4563947510218351,
+      "learning_rate": 2.1022103010221355e-06,
+      "loss": 0.3748,
+      "step": 16771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5376861695957393,
+      "learning_rate": 2.1017834474989068e-06,
+      "loss": 0.3913,
+      "step": 16772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9133761269326903,
+      "learning_rate": 2.1013566651152314e-06,
+      "loss": 0.3844,
+      "step": 16773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0907488453492964,
+      "learning_rate": 2.1009299538800446e-06,
+      "loss": 0.6412,
+      "step": 16774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.560541040177774,
+      "learning_rate": 2.1005033138022763e-06,
+      "loss": 0.5032,
+      "step": 16775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0416447479907593,
+      "learning_rate": 2.1000767448908596e-06,
+      "loss": 0.5017,
+      "step": 16776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5599379996153961,
+      "learning_rate": 2.0996502471547263e-06,
+      "loss": 0.4577,
+      "step": 16777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7003759919352903,
+      "learning_rate": 2.0992238206028013e-06,
+      "loss": 0.5025,
+      "step": 16778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6151903011935962,
+      "learning_rate": 2.0987974652440134e-06,
+      "loss": 0.4632,
+      "step": 16779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7720979211344277,
+      "learning_rate": 2.0983711810872894e-06,
+      "loss": 0.5322,
+      "step": 16780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6484322155681232,
+      "learning_rate": 2.0979449681415506e-06,
+      "loss": 0.5478,
+      "step": 16781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.563410942748733,
+      "learning_rate": 2.0975188264157205e-06,
+      "loss": 0.5243,
+      "step": 16782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7905576406364188,
+      "learning_rate": 2.0970927559187206e-06,
+      "loss": 0.4502,
+      "step": 16783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6869539066148103,
+      "learning_rate": 2.0966667566594713e-06,
+      "loss": 0.4296,
+      "step": 16784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7891045015058962,
+      "learning_rate": 2.096240828646889e-06,
+      "loss": 0.6415,
+      "step": 16785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9241458428637834,
+      "learning_rate": 2.0958149718898895e-06,
+      "loss": 0.4796,
+      "step": 16786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5295405161645708,
+      "learning_rate": 2.0953891863973913e-06,
+      "loss": 0.3984,
+      "step": 16787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.814435336720003,
+      "learning_rate": 2.0949634721783046e-06,
+      "loss": 0.6041,
+      "step": 16788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7101097694797394,
+      "learning_rate": 2.0945378292415424e-06,
+      "loss": 0.4076,
+      "step": 16789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5890154377095496,
+      "learning_rate": 2.0941122575960176e-06,
+      "loss": 0.4137,
+      "step": 16790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6347324688027485,
+      "learning_rate": 2.0936867572506346e-06,
+      "loss": 0.4381,
+      "step": 16791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5520257930804442,
+      "learning_rate": 2.0932613282143043e-06,
+      "loss": 0.3599,
+      "step": 16792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8067407591216427,
+      "learning_rate": 2.0928359704959335e-06,
+      "loss": 0.3586,
+      "step": 16793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.833299196024839,
+      "learning_rate": 2.0924106841044238e-06,
+      "loss": 0.3934,
+      "step": 16794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4317565522645497,
+      "learning_rate": 2.09198546904868e-06,
+      "loss": 0.4611,
+      "step": 16795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.087301892486748,
+      "learning_rate": 2.0915603253376036e-06,
+      "loss": 0.5257,
+      "step": 16796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5799306848763144,
+      "learning_rate": 2.091135252980096e-06,
+      "loss": 0.4951,
+      "step": 16797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6594527950606788,
+      "learning_rate": 2.090710251985053e-06,
+      "loss": 0.5435,
+      "step": 16798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7307588167669326,
+      "learning_rate": 2.0902853223613744e-06,
+      "loss": 0.3361,
+      "step": 16799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4070957083721616,
+      "learning_rate": 2.0898604641179547e-06,
+      "loss": 0.2748,
+      "step": 16800
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.49124929308891296,
+      "eval_runtime": 68.039,
+      "eval_samples_per_second": 50.118,
+      "eval_steps_per_second": 1.573,
+      "step": 16800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4285074160830515,
+      "learning_rate": 2.089435677263687e-06,
+      "loss": 0.3667,
+      "step": 16801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8287188723461287,
+      "learning_rate": 2.089010961807466e-06,
+      "loss": 0.5229,
+      "step": 16802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8226524657360939,
+      "learning_rate": 2.088586317758182e-06,
+      "loss": 0.3914,
+      "step": 16803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6117975687575994,
+      "learning_rate": 2.0881617451247237e-06,
+      "loss": 0.5158,
+      "step": 16804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7381685191403353,
+      "learning_rate": 2.0877372439159806e-06,
+      "loss": 0.422,
+      "step": 16805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.587285755885332,
+      "learning_rate": 2.0873128141408396e-06,
+      "loss": 0.3974,
+      "step": 16806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8264199914421417,
+      "learning_rate": 2.086888455808184e-06,
+      "loss": 0.4925,
+      "step": 16807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4460945769727334,
+      "learning_rate": 2.0864641689268983e-06,
+      "loss": 0.4632,
+      "step": 16808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.528757478883879,
+      "learning_rate": 2.086039953505865e-06,
+      "loss": 0.3559,
+      "step": 16809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7139573439219782,
+      "learning_rate": 2.0856158095539664e-06,
+      "loss": 0.3252,
+      "step": 16810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5025887207624897,
+      "learning_rate": 2.0851917370800784e-06,
+      "loss": 0.4352,
+      "step": 16811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8012629260974122,
+      "learning_rate": 2.084767736093081e-06,
+      "loss": 0.5425,
+      "step": 16812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5130438518083404,
+      "learning_rate": 2.0843438066018503e-06,
+      "loss": 0.3577,
+      "step": 16813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8805483702843775,
+      "learning_rate": 2.08391994861526e-06,
+      "loss": 0.4529,
+      "step": 16814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7953138397177872,
+      "learning_rate": 2.0834961621421833e-06,
+      "loss": 0.4432,
+      "step": 16815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8452188856027134,
+      "learning_rate": 2.0830724471914922e-06,
+      "loss": 0.4614,
+      "step": 16816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5986182508862041,
+      "learning_rate": 2.0826488037720573e-06,
+      "loss": 0.4492,
+      "step": 16817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4691415966381849,
+      "learning_rate": 2.082225231892747e-06,
+      "loss": 0.5453,
+      "step": 16818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6682939623600053,
+      "learning_rate": 2.0818017315624305e-06,
+      "loss": 0.5332,
+      "step": 16819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5772012826369135,
+      "learning_rate": 2.08137830278997e-06,
+      "loss": 0.4281,
+      "step": 16820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.635325015642826,
+      "learning_rate": 2.0809549455842304e-06,
+      "loss": 0.4036,
+      "step": 16821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.637848819025558,
+      "learning_rate": 2.080531659954078e-06,
+      "loss": 0.5492,
+      "step": 16822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.375199136938388,
+      "learning_rate": 2.0801084459083694e-06,
+      "loss": 0.3935,
+      "step": 16823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4544047813029646,
+      "learning_rate": 2.0796853034559666e-06,
+      "loss": 0.3723,
+      "step": 16824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.685064746426137,
+      "learning_rate": 2.079262232605729e-06,
+      "loss": 0.5431,
+      "step": 16825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7809142183712097,
+      "learning_rate": 2.0788392333665097e-06,
+      "loss": 0.4292,
+      "step": 16826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7966663085719412,
+      "learning_rate": 2.078416305747167e-06,
+      "loss": 0.5261,
+      "step": 16827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5494824800736247,
+      "learning_rate": 2.0779934497565547e-06,
+      "loss": 0.3437,
+      "step": 16828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.754043530804374,
+      "learning_rate": 2.0775706654035227e-06,
+      "loss": 0.5162,
+      "step": 16829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.260556639721278,
+      "learning_rate": 2.077147952696923e-06,
+      "loss": 0.5003,
+      "step": 16830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.479723189119573,
+      "learning_rate": 2.0767253116456047e-06,
+      "loss": 0.336,
+      "step": 16831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9401570187907247,
+      "learning_rate": 2.0763027422584174e-06,
+      "loss": 0.6035,
+      "step": 16832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9271003451800337,
+      "learning_rate": 2.075880244544204e-06,
+      "loss": 0.49,
+      "step": 16833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6660469272696103,
+      "learning_rate": 2.075457818511811e-06,
+      "loss": 0.545,
+      "step": 16834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.836618385850465,
+      "learning_rate": 2.0750354641700833e-06,
+      "loss": 0.4821,
+      "step": 16835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9157486903426826,
+      "learning_rate": 2.0746131815278594e-06,
+      "loss": 0.6226,
+      "step": 16836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.511714520025309,
+      "learning_rate": 2.07419097059398e-06,
+      "loss": 0.4602,
+      "step": 16837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0637191943451545,
+      "learning_rate": 2.0737688313772876e-06,
+      "loss": 0.5465,
+      "step": 16838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.501640922208207,
+      "learning_rate": 2.0733467638866144e-06,
+      "loss": 0.7584,
+      "step": 16839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7583036482349734,
+      "learning_rate": 2.0729247681307977e-06,
+      "loss": 0.5171,
+      "step": 16840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7875476717781444,
+      "learning_rate": 2.0725028441186733e-06,
+      "loss": 0.4808,
+      "step": 16841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.315187783506495,
+      "learning_rate": 2.072080991859074e-06,
+      "loss": 0.5735,
+      "step": 16842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0337614320096926,
+      "learning_rate": 2.0716592113608292e-06,
+      "loss": 0.5132,
+      "step": 16843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.875492502749162,
+      "learning_rate": 2.0712375026327684e-06,
+      "loss": 0.4315,
+      "step": 16844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7551071382877048,
+      "learning_rate": 2.070815865683723e-06,
+      "loss": 0.5333,
+      "step": 16845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7338529932812325,
+      "learning_rate": 2.0703943005225158e-06,
+      "loss": 0.5329,
+      "step": 16846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.842789197404932,
+      "learning_rate": 2.069972807157973e-06,
+      "loss": 0.4465,
+      "step": 16847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.990641770134104,
+      "learning_rate": 2.069551385598922e-06,
+      "loss": 0.6501,
+      "step": 16848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5487892624871948,
+      "learning_rate": 2.0691300358541794e-06,
+      "loss": 0.3422,
+      "step": 16849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6908427326967372,
+      "learning_rate": 2.0687087579325685e-06,
+      "loss": 0.3766,
+      "step": 16850
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.49159419536590576,
+      "eval_runtime": 68.0023,
+      "eval_samples_per_second": 50.145,
+      "eval_steps_per_second": 1.573,
+      "step": 16850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5235380831806438,
+      "learning_rate": 2.06828755184291e-06,
+      "loss": 0.4572,
+      "step": 16851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5477157018956336,
+      "learning_rate": 2.0678664175940193e-06,
+      "loss": 0.4199,
+      "step": 16852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6716573063154208,
+      "learning_rate": 2.067445355194713e-06,
+      "loss": 0.4789,
+      "step": 16853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7665328629085904,
+      "learning_rate": 2.0670243646538062e-06,
+      "loss": 0.3688,
+      "step": 16854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7233221988074596,
+      "learning_rate": 2.0666034459801136e-06,
+      "loss": 0.4139,
+      "step": 16855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.794567679107905,
+      "learning_rate": 2.0661825991824436e-06,
+      "loss": 0.3288,
+      "step": 16856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.890724504034207,
+      "learning_rate": 2.0657618242696082e-06,
+      "loss": 0.5681,
+      "step": 16857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1454876960786784,
+      "learning_rate": 2.065341121250417e-06,
+      "loss": 0.5281,
+      "step": 16858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5995506271967104,
+      "learning_rate": 2.064920490133675e-06,
+      "loss": 0.3814,
+      "step": 16859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8078419710404403,
+      "learning_rate": 2.0644999309281887e-06,
+      "loss": 0.52,
+      "step": 16860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2643153967059546,
+      "learning_rate": 2.064079443642764e-06,
+      "loss": 0.5218,
+      "step": 16861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.282989169447851,
+      "learning_rate": 2.063659028286201e-06,
+      "loss": 0.4535,
+      "step": 16862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.090815827498234,
+      "learning_rate": 2.0632386848673013e-06,
+      "loss": 0.5306,
+      "step": 16863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7287454330842666,
+      "learning_rate": 2.0628184133948653e-06,
+      "loss": 0.5866,
+      "step": 16864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.657875060388487,
+      "learning_rate": 2.0623982138776914e-06,
+      "loss": 0.5046,
+      "step": 16865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.053935983231554,
+      "learning_rate": 2.061978086324575e-06,
+      "loss": 0.643,
+      "step": 16866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9110689267214978,
+      "learning_rate": 2.0615580307443136e-06,
+      "loss": 0.4908,
+      "step": 16867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6385007748803861,
+      "learning_rate": 2.061138047145697e-06,
+      "loss": 0.6215,
+      "step": 16868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9595226289391792,
+      "learning_rate": 2.0607181355375206e-06,
+      "loss": 0.4876,
+      "step": 16869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.801446083024926,
+      "learning_rate": 2.0602982959285745e-06,
+      "loss": 0.5863,
+      "step": 16870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.710109313542989,
+      "learning_rate": 2.0598785283276455e-06,
+      "loss": 0.484,
+      "step": 16871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7577682290822998,
+      "learning_rate": 2.0594588327435235e-06,
+      "loss": 0.3999,
+      "step": 16872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6968482835727177,
+      "learning_rate": 2.0590392091849953e-06,
+      "loss": 0.5014,
+      "step": 16873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5777486792602142,
+      "learning_rate": 2.058619657660842e-06,
+      "loss": 0.4632,
+      "step": 16874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8532891118787407,
+      "learning_rate": 2.0582001781798493e-06,
+      "loss": 0.4333,
+      "step": 16875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6553020537616485,
+      "learning_rate": 2.0577807707507972e-06,
+      "loss": 0.4736,
+      "step": 16876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.576366475494919,
+      "learning_rate": 2.057361435382469e-06,
+      "loss": 0.4055,
+      "step": 16877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3720173962542814,
+      "learning_rate": 2.0569421720836384e-06,
+      "loss": 0.6153,
+      "step": 16878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0139717234491608,
+      "learning_rate": 2.0565229808630856e-06,
+      "loss": 0.5548,
+      "step": 16879
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.787424054902018,
+      "learning_rate": 2.0561038617295863e-06,
+      "loss": 0.4708,
+      "step": 16880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9043098347365675,
+      "learning_rate": 2.0556848146919122e-06,
+      "loss": 0.477,
+      "step": 16881
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6317623964017598,
+      "learning_rate": 2.0552658397588377e-06,
+      "loss": 0.3671,
+      "step": 16882
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6790834923409428,
+      "learning_rate": 2.0548469369391344e-06,
+      "loss": 0.4409,
+      "step": 16883
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6263570342969589,
+      "learning_rate": 2.054428106241569e-06,
+      "loss": 0.5642,
+      "step": 16884
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4994152491493498,
+      "learning_rate": 2.054009347674911e-06,
+      "loss": 0.4152,
+      "step": 16885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4776476795865203,
+      "learning_rate": 2.053590661247929e-06,
+      "loss": 0.3246,
+      "step": 16886
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.842293145927848,
+      "learning_rate": 2.053172046969384e-06,
+      "loss": 0.5166,
+      "step": 16887
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5843250181141588,
+      "learning_rate": 2.0527535048480413e-06,
+      "loss": 0.4655,
+      "step": 16888
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.640025972065472,
+      "learning_rate": 2.0523350348926622e-06,
+      "loss": 0.4605,
+      "step": 16889
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4543033857975796,
+      "learning_rate": 2.05191663711201e-06,
+      "loss": 0.5131,
+      "step": 16890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8230130499306607,
+      "learning_rate": 2.0514983115148394e-06,
+      "loss": 0.4018,
+      "step": 16891
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.475023532960377,
+      "learning_rate": 2.0510800581099106e-06,
+      "loss": 0.3413,
+      "step": 16892
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9514179264976457,
+      "learning_rate": 2.0506618769059787e-06,
+      "loss": 0.4655,
+      "step": 16893
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9986759524239708,
+      "learning_rate": 2.0502437679117974e-06,
+      "loss": 0.569,
+      "step": 16894
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4530847913584795,
+      "learning_rate": 2.04982573113612e-06,
+      "loss": 0.4536,
+      "step": 16895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.596224740477969,
+      "learning_rate": 2.0494077665876992e-06,
+      "loss": 0.5492,
+      "step": 16896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6581947559343524,
+      "learning_rate": 2.0489898742752824e-06,
+      "loss": 0.3663,
+      "step": 16897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2421656430548866,
+      "learning_rate": 2.0485720542076186e-06,
+      "loss": 0.574,
+      "step": 16898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6928945485074565,
+      "learning_rate": 2.0481543063934567e-06,
+      "loss": 0.4117,
+      "step": 16899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1522374648300455,
+      "learning_rate": 2.047736630841539e-06,
+      "loss": 0.4831,
+      "step": 16900
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.491344153881073,
+      "eval_runtime": 68.0721,
+      "eval_samples_per_second": 50.094,
+      "eval_steps_per_second": 1.572,
+      "step": 16900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.500978168427833,
+      "learning_rate": 2.047319027560611e-06,
+      "loss": 0.3963,
+      "step": 16901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3942076641734835,
+      "learning_rate": 2.046901496559414e-06,
+      "loss": 0.6405,
+      "step": 16902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8853942718810055,
+      "learning_rate": 2.046484037846691e-06,
+      "loss": 0.5778,
+      "step": 16903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6612348090651656,
+      "learning_rate": 2.0460666514311783e-06,
+      "loss": 0.3602,
+      "step": 16904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7200879583307878,
+      "learning_rate": 2.0456493373216147e-06,
+      "loss": 0.4549,
+      "step": 16905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9970970765788063,
+      "learning_rate": 2.0452320955267384e-06,
+      "loss": 0.6361,
+      "step": 16906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8992903504289835,
+      "learning_rate": 2.044814926055281e-06,
+      "loss": 0.5173,
+      "step": 16907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8409845944914123,
+      "learning_rate": 2.044397828915977e-06,
+      "loss": 0.5109,
+      "step": 16908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6383672839199184,
+      "learning_rate": 2.043980804117559e-06,
+      "loss": 0.4777,
+      "step": 16909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3992555640554738,
+      "learning_rate": 2.043563851668755e-06,
+      "loss": 0.3839,
+      "step": 16910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4965904540586872,
+      "learning_rate": 2.0431469715782954e-06,
+      "loss": 0.3756,
+      "step": 16911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.685300699859227,
+      "learning_rate": 2.0427301638549076e-06,
+      "loss": 0.4137,
+      "step": 16912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6607877495105716,
+      "learning_rate": 2.042313428507314e-06,
+      "loss": 0.4106,
+      "step": 16913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8457970942659623,
+      "learning_rate": 2.0418967655442427e-06,
+      "loss": 0.638,
+      "step": 16914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8103172935985674,
+      "learning_rate": 2.041480174974416e-06,
+      "loss": 0.3486,
+      "step": 16915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.673170931957288,
+      "learning_rate": 2.0410636568065512e-06,
+      "loss": 0.5108,
+      "step": 16916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8139695595423424,
+      "learning_rate": 2.040647211049372e-06,
+      "loss": 0.3806,
+      "step": 16917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.010362349804577,
+      "learning_rate": 2.0402308377115953e-06,
+      "loss": 0.5463,
+      "step": 16918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9896625417389386,
+      "learning_rate": 2.0398145368019354e-06,
+      "loss": 0.6877,
+      "step": 16919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.186145128200562,
+      "learning_rate": 2.0393983083291097e-06,
+      "loss": 0.4967,
+      "step": 16920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6429031504597176,
+      "learning_rate": 2.0389821523018304e-06,
+      "loss": 0.4387,
+      "step": 16921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7877072110344039,
+      "learning_rate": 2.0385660687288117e-06,
+      "loss": 0.5415,
+      "step": 16922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.587160818453387,
+      "learning_rate": 2.0381500576187608e-06,
+      "loss": 0.5187,
+      "step": 16923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8185198428041995,
+      "learning_rate": 2.0377341189803884e-06,
+      "loss": 0.5971,
+      "step": 16924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.558986100679423,
+      "learning_rate": 2.0373182528224026e-06,
+      "loss": 0.4221,
+      "step": 16925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7016601223646546,
+      "learning_rate": 2.0369024591535074e-06,
+      "loss": 0.3907,
+      "step": 16926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.193182549180078,
+      "learning_rate": 2.0364867379824083e-06,
+      "loss": 0.6665,
+      "step": 16927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4622345871910871,
+      "learning_rate": 2.0360710893178096e-06,
+      "loss": 0.3985,
+      "step": 16928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.817221735687574,
+      "learning_rate": 2.0356555131684098e-06,
+      "loss": 0.4824,
+      "step": 16929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8244583092989128,
+      "learning_rate": 2.0352400095429098e-06,
+      "loss": 0.4463,
+      "step": 16930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6772940527083084,
+      "learning_rate": 2.034824578450009e-06,
+      "loss": 0.4409,
+      "step": 16931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3864412971718425,
+      "learning_rate": 2.034409219898403e-06,
+      "loss": 0.3507,
+      "step": 16932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0944864367778204,
+      "learning_rate": 2.0339939338967866e-06,
+      "loss": 0.6217,
+      "step": 16933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.88175228962291,
+      "learning_rate": 2.033578720453855e-06,
+      "loss": 0.4287,
+      "step": 16934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.544562277702101,
+      "learning_rate": 2.0331635795783007e-06,
+      "loss": 0.5818,
+      "step": 16935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6817172113938785,
+      "learning_rate": 2.0327485112788127e-06,
+      "loss": 0.4301,
+      "step": 16936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6975122959512878,
+      "learning_rate": 2.0323335155640805e-06,
+      "loss": 0.5682,
+      "step": 16937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8790376207662196,
+      "learning_rate": 2.0319185924427938e-06,
+      "loss": 0.4966,
+      "step": 16938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6611686207948988,
+      "learning_rate": 2.0315037419236355e-06,
+      "loss": 0.3823,
+      "step": 16939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.669995737758737,
+      "learning_rate": 2.0310889640152933e-06,
+      "loss": 0.3503,
+      "step": 16940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9696614102024232,
+      "learning_rate": 2.030674258726449e-06,
+      "loss": 0.5712,
+      "step": 16941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8171676929517604,
+      "learning_rate": 2.0302596260657837e-06,
+      "loss": 0.4906,
+      "step": 16942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9436494972930247,
+      "learning_rate": 2.029845066041977e-06,
+      "loss": 0.3843,
+      "step": 16943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7061599330150157,
+      "learning_rate": 2.0294305786637106e-06,
+      "loss": 0.4322,
+      "step": 16944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5238448662277926,
+      "learning_rate": 2.0290161639396584e-06,
+      "loss": 0.4422,
+      "step": 16945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9108626320039286,
+      "learning_rate": 2.0286018218784965e-06,
+      "loss": 0.4665,
+      "step": 16946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.155959340139782,
+      "learning_rate": 2.028187552488899e-06,
+      "loss": 0.4465,
+      "step": 16947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.158350168861065,
+      "learning_rate": 2.02777335577954e-06,
+      "loss": 0.4068,
+      "step": 16948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6774778166033673,
+      "learning_rate": 2.0273592317590884e-06,
+      "loss": 0.5002,
+      "step": 16949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8537377987416988,
+      "learning_rate": 2.0269451804362135e-06,
+      "loss": 0.4415,
+      "step": 16950
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.491118460893631,
+      "eval_runtime": 68.0341,
+      "eval_samples_per_second": 50.122,
+      "eval_steps_per_second": 1.573,
+      "step": 16950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7233418513749084,
+      "learning_rate": 2.026531201819586e-06,
+      "loss": 0.5531,
+      "step": 16951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8212632182931978,
+      "learning_rate": 2.0261172959178687e-06,
+      "loss": 0.4673,
+      "step": 16952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7023210998949772,
+      "learning_rate": 2.0257034627397283e-06,
+      "loss": 0.3417,
+      "step": 16953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6379987537836749,
+      "learning_rate": 2.0252897022938294e-06,
+      "loss": 0.5259,
+      "step": 16954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9970067588640465,
+      "learning_rate": 2.024876014588831e-06,
+      "loss": 0.6151,
+      "step": 16955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9331448158171913,
+      "learning_rate": 2.024462399633394e-06,
+      "loss": 0.6127,
+      "step": 16956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5783737085955956,
+      "learning_rate": 2.02404885743618e-06,
+      "loss": 0.3842,
+      "step": 16957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6307618447069212,
+      "learning_rate": 2.0236353880058424e-06,
+      "loss": 0.3833,
+      "step": 16958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8010523673549974,
+      "learning_rate": 2.0232219913510395e-06,
+      "loss": 0.3851,
+      "step": 16959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6924531893633428,
+      "learning_rate": 2.022808667480424e-06,
+      "loss": 0.4564,
+      "step": 16960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8830654387177495,
+      "learning_rate": 2.022395416402651e-06,
+      "loss": 0.4921,
+      "step": 16961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9175976917424409,
+      "learning_rate": 2.021982238126367e-06,
+      "loss": 0.5673,
+      "step": 16962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8207606189645607,
+      "learning_rate": 2.021569132660227e-06,
+      "loss": 0.418,
+      "step": 16963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5611752263868734,
+      "learning_rate": 2.0211561000128756e-06,
+      "loss": 0.3886,
+      "step": 16964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9926656817599702,
+      "learning_rate": 2.0207431401929607e-06,
+      "loss": 0.4198,
+      "step": 16965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8458652793801387,
+      "learning_rate": 2.020330253209128e-06,
+      "loss": 0.465,
+      "step": 16966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.663200875146091,
+      "learning_rate": 2.0199174390700193e-06,
+      "loss": 0.5505,
+      "step": 16967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.172564937953964,
+      "learning_rate": 2.0195046977842774e-06,
+      "loss": 0.5269,
+      "step": 16968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9547163819489692,
+      "learning_rate": 2.0190920293605426e-06,
+      "loss": 0.5515,
+      "step": 16969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6057022183960192,
+      "learning_rate": 2.0186794338074554e-06,
+      "loss": 0.3989,
+      "step": 16970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.938580650098847,
+      "learning_rate": 2.0182669111336505e-06,
+      "loss": 0.6054,
+      "step": 16971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.626289869539066,
+      "learning_rate": 2.0178544613477655e-06,
+      "loss": 0.4047,
+      "step": 16972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.630202248015343,
+      "learning_rate": 2.017442084458436e-06,
+      "loss": 0.4251,
+      "step": 16973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8165970310325568,
+      "learning_rate": 2.017029780474292e-06,
+      "loss": 0.4755,
+      "step": 16974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2211057726723826,
+      "learning_rate": 2.016617549403966e-06,
+      "loss": 0.6088,
+      "step": 16975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.679369475185129,
+      "learning_rate": 2.016205391256089e-06,
+      "loss": 0.431,
+      "step": 16976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.187624738202173,
+      "learning_rate": 2.0157933060392874e-06,
+      "loss": 0.6716,
+      "step": 16977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8953342804989044,
+      "learning_rate": 2.0153812937621886e-06,
+      "loss": 0.489,
+      "step": 16978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7690562171778115,
+      "learning_rate": 2.0149693544334198e-06,
+      "loss": 0.4863,
+      "step": 16979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8591249525446927,
+      "learning_rate": 2.014557488061601e-06,
+      "loss": 0.632,
+      "step": 16980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5253728576594876,
+      "learning_rate": 2.014145694655357e-06,
+      "loss": 0.4817,
+      "step": 16981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4420383093085918,
+      "learning_rate": 2.0137339742233068e-06,
+      "loss": 0.4926,
+      "step": 16982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2692834929356405,
+      "learning_rate": 2.0133223267740722e-06,
+      "loss": 0.3309,
+      "step": 16983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6756965780466089,
+      "learning_rate": 2.0129107523162676e-06,
+      "loss": 0.5413,
+      "step": 16984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5586351208059028,
+      "learning_rate": 2.0124992508585105e-06,
+      "loss": 0.4203,
+      "step": 16985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9314198578413708,
+      "learning_rate": 2.012087822409417e-06,
+      "loss": 0.4988,
+      "step": 16986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5489020050472808,
+      "learning_rate": 2.0116764669775973e-06,
+      "loss": 0.3938,
+      "step": 16987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5141588469849245,
+      "learning_rate": 2.0112651845716643e-06,
+      "loss": 0.3953,
+      "step": 16988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0002592801924775,
+      "learning_rate": 2.010853975200228e-06,
+      "loss": 0.5483,
+      "step": 16989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9617776257616109,
+      "learning_rate": 2.010442838871896e-06,
+      "loss": 0.5348,
+      "step": 16990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8294507053384423,
+      "learning_rate": 2.0100317755952757e-06,
+      "loss": 0.4296,
+      "step": 16991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9057426082722289,
+      "learning_rate": 2.0096207853789732e-06,
+      "loss": 0.5364,
+      "step": 16992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.703928052337348,
+      "learning_rate": 2.0092098682315906e-06,
+      "loss": 0.4082,
+      "step": 16993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8714758676422956,
+      "learning_rate": 2.0087990241617316e-06,
+      "loss": 0.4715,
+      "step": 16994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7277951248497285,
+      "learning_rate": 2.0083882531779965e-06,
+      "loss": 0.378,
+      "step": 16995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2015074449528207,
+      "learning_rate": 2.0079775552889857e-06,
+      "loss": 0.4993,
+      "step": 16996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8839130477579533,
+      "learning_rate": 2.007566930503295e-06,
+      "loss": 0.6015,
+      "step": 16997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4928893683937217,
+      "learning_rate": 2.00715637882952e-06,
+      "loss": 0.3713,
+      "step": 16998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.354855774016581,
+      "learning_rate": 2.0067459002762596e-06,
+      "loss": 0.2792,
+      "step": 16999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9517369210210709,
+      "learning_rate": 2.006335494852101e-06,
+      "loss": 0.5987,
+      "step": 17000
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.49126937985420227,
+      "eval_runtime": 68.0843,
+      "eval_samples_per_second": 50.085,
+      "eval_steps_per_second": 1.572,
+      "step": 17000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9292674905601508,
+      "learning_rate": 2.0059251625656407e-06,
+      "loss": 0.4288,
+      "step": 17001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7990351974784482,
+      "learning_rate": 2.0055149034254673e-06,
+      "loss": 0.4304,
+      "step": 17002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.023823005448838,
+      "learning_rate": 2.0051047174401676e-06,
+      "loss": 0.6422,
+      "step": 17003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5736238803269147,
+      "learning_rate": 2.00469460461833e-06,
+      "loss": 0.4298,
+      "step": 17004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9107965751267515,
+      "learning_rate": 2.004284564968541e-06,
+      "loss": 0.5082,
+      "step": 17005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9689198554856466,
+      "learning_rate": 2.003874598499382e-06,
+      "loss": 0.4497,
+      "step": 17006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7749873521605768,
+      "learning_rate": 2.003464705219437e-06,
+      "loss": 0.3745,
+      "step": 17007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2420483175216823,
+      "learning_rate": 2.0030548851372863e-06,
+      "loss": 0.5611,
+      "step": 17008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8909731802705956,
+      "learning_rate": 2.0026451382615114e-06,
+      "loss": 0.3975,
+      "step": 17009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.812688849259487,
+      "learning_rate": 2.0022354646006865e-06,
+      "loss": 0.452,
+      "step": 17010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8151496967796115,
+      "learning_rate": 2.0018258641633895e-06,
+      "loss": 0.5045,
+      "step": 17011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4240761463872864,
+      "learning_rate": 2.001416336958196e-06,
+      "loss": 0.3373,
+      "step": 17012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5510272998809451,
+      "learning_rate": 2.0010068829936774e-06,
+      "loss": 0.3545,
+      "step": 17013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8268940045946696,
+      "learning_rate": 2.000597502278407e-06,
+      "loss": 0.4208,
+      "step": 17014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4937620993324168,
+      "learning_rate": 2.000188194820955e-06,
+      "loss": 0.3376,
+      "step": 17015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4893119401942176,
+      "learning_rate": 1.999778960629889e-06,
+      "loss": 0.4382,
+      "step": 17016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8914607908658352,
+      "learning_rate": 1.999369799713776e-06,
+      "loss": 0.639,
+      "step": 17017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7684313201186557,
+      "learning_rate": 1.9989607120811826e-06,
+      "loss": 0.4139,
+      "step": 17018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8273132937729681,
+      "learning_rate": 1.9985516977406717e-06,
+      "loss": 0.4428,
+      "step": 17019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.18645892475102,
+      "learning_rate": 1.9981427567008055e-06,
+      "loss": 0.5413,
+      "step": 17020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6076847295643166,
+      "learning_rate": 1.9977338889701475e-06,
+      "loss": 0.4642,
+      "step": 17021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8546716264949439,
+      "learning_rate": 1.9973250945572538e-06,
+      "loss": 0.4504,
+      "step": 17022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6014176725383455,
+      "learning_rate": 1.996916373470683e-06,
+      "loss": 0.4779,
+      "step": 17023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.443502999091037,
+      "learning_rate": 1.996507725718994e-06,
+      "loss": 0.6042,
+      "step": 17024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.605971183956084,
+      "learning_rate": 1.996099151310738e-06,
+      "loss": 0.3832,
+      "step": 17025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.821793176427372,
+      "learning_rate": 1.9956906502544707e-06,
+      "loss": 0.5757,
+      "step": 17026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5306726804614192,
+      "learning_rate": 1.995282222558742e-06,
+      "loss": 0.4983,
+      "step": 17027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8289157096582351,
+      "learning_rate": 1.994873868232105e-06,
+      "loss": 0.4414,
+      "step": 17028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5987822132490552,
+      "learning_rate": 1.994465587283105e-06,
+      "loss": 0.3711,
+      "step": 17029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.144934870101675,
+      "learning_rate": 1.994057379720291e-06,
+      "loss": 0.3834,
+      "step": 17030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5502988462798746,
+      "learning_rate": 1.993649245552209e-06,
+      "loss": 0.3093,
+      "step": 17031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.881760460513146,
+      "learning_rate": 1.9932411847874007e-06,
+      "loss": 0.4662,
+      "step": 17032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8233173431174332,
+      "learning_rate": 1.9928331974344104e-06,
+      "loss": 0.5367,
+      "step": 17033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7323243432492565,
+      "learning_rate": 1.9924252835017795e-06,
+      "loss": 0.4567,
+      "step": 17034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8356089639294002,
+      "learning_rate": 1.992017442998046e-06,
+      "loss": 0.4603,
+      "step": 17035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7633834556313597,
+      "learning_rate": 1.9916096759317483e-06,
+      "loss": 0.4382,
+      "step": 17036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7591774034266985,
+      "learning_rate": 1.991201982311424e-06,
+      "loss": 0.4877,
+      "step": 17037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0446034775571476,
+      "learning_rate": 1.9907943621456055e-06,
+      "loss": 0.582,
+      "step": 17038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.000072737028108,
+      "learning_rate": 1.9903868154428277e-06,
+      "loss": 0.7732,
+      "step": 17039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7384222142950627,
+      "learning_rate": 1.9899793422116217e-06,
+      "loss": 0.4603,
+      "step": 17040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3718667067511854,
+      "learning_rate": 1.989571942460519e-06,
+      "loss": 0.4317,
+      "step": 17041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9019381997674936,
+      "learning_rate": 1.9891646161980472e-06,
+      "loss": 0.4696,
+      "step": 17042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5631909813791278,
+      "learning_rate": 1.988757363432732e-06,
+      "loss": 0.4209,
+      "step": 17043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2528727122575676,
+      "learning_rate": 1.9883501841731026e-06,
+      "loss": 0.6105,
+      "step": 17044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.182902249506562,
+      "learning_rate": 1.9879430784276795e-06,
+      "loss": 0.64,
+      "step": 17045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7753674431373194,
+      "learning_rate": 1.9875360462049863e-06,
+      "loss": 0.4495,
+      "step": 17046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.821147528574686,
+      "learning_rate": 1.987129087513546e-06,
+      "loss": 0.4449,
+      "step": 17047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4917247714216708,
+      "learning_rate": 1.9867222023618756e-06,
+      "loss": 0.38,
+      "step": 17048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8886800771083223,
+      "learning_rate": 1.9863153907584927e-06,
+      "loss": 0.445,
+      "step": 17049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9883652687428832,
+      "learning_rate": 1.9859086527119163e-06,
+      "loss": 0.5006,
+      "step": 17050
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.49126219749450684,
+      "eval_runtime": 67.9629,
+      "eval_samples_per_second": 50.174,
+      "eval_steps_per_second": 1.574,
+      "step": 17050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.09032892692906,
+      "learning_rate": 1.985501988230658e-06,
+      "loss": 0.5752,
+      "step": 17051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7127376509072973,
+      "learning_rate": 1.9850953973232333e-06,
+      "loss": 0.4501,
+      "step": 17052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.744320278418118,
+      "learning_rate": 1.9846888799981528e-06,
+      "loss": 0.4677,
+      "step": 17053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.815808573821082,
+      "learning_rate": 1.9842824362639287e-06,
+      "loss": 0.4708,
+      "step": 17054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8861417435207568,
+      "learning_rate": 1.9838760661290666e-06,
+      "loss": 0.4808,
+      "step": 17055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.672603612333792,
+      "learning_rate": 1.983469769602075e-06,
+      "loss": 0.4259,
+      "step": 17056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.92836016562937,
+      "learning_rate": 1.983063546691461e-06,
+      "loss": 0.5976,
+      "step": 17057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1564918720369273,
+      "learning_rate": 1.9826573974057263e-06,
+      "loss": 0.3305,
+      "step": 17058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9782065007860479,
+      "learning_rate": 1.982251321753374e-06,
+      "loss": 0.5395,
+      "step": 17059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6064080303319959,
+      "learning_rate": 1.9818453197429063e-06,
+      "loss": 0.5489,
+      "step": 17060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8300164941549002,
+      "learning_rate": 1.9814393913828196e-06,
+      "loss": 0.5459,
+      "step": 17061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.101908487935318,
+      "learning_rate": 1.981033536681615e-06,
+      "loss": 0.5727,
+      "step": 17062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.867283597601236,
+      "learning_rate": 1.980627755647788e-06,
+      "loss": 0.4504,
+      "step": 17063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1763985171768883,
+      "learning_rate": 1.9802220482898327e-06,
+      "loss": 0.5343,
+      "step": 17064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6269832512122029,
+      "learning_rate": 1.9798164146162424e-06,
+      "loss": 0.4259,
+      "step": 17065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4720061676866698,
+      "learning_rate": 1.97941085463551e-06,
+      "loss": 0.4088,
+      "step": 17066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.651618851319702,
+      "learning_rate": 1.9790053683561236e-06,
+      "loss": 0.4239,
+      "step": 17067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7828009148880317,
+      "learning_rate": 1.9785999557865725e-06,
+      "loss": 0.4587,
+      "step": 17068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.702243133732288,
+      "learning_rate": 1.9781946169353454e-06,
+      "loss": 0.5766,
+      "step": 17069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5245453705667262,
+      "learning_rate": 1.9777893518109256e-06,
+      "loss": 0.4105,
+      "step": 17070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7429053644895367,
+      "learning_rate": 1.977384160421798e-06,
+      "loss": 0.3911,
+      "step": 17071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8051982395210613,
+      "learning_rate": 1.976979042776446e-06,
+      "loss": 0.5098,
+      "step": 17072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8548779182818833,
+      "learning_rate": 1.9765739988833484e-06,
+      "loss": 0.4671,
+      "step": 17073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8136432696470628,
+      "learning_rate": 1.976169028750986e-06,
+      "loss": 0.3619,
+      "step": 17074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7220680475934378,
+      "learning_rate": 1.9757641323878358e-06,
+      "loss": 0.4879,
+      "step": 17075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8287040484463588,
+      "learning_rate": 1.9753593098023755e-06,
+      "loss": 0.4884,
+      "step": 17076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9348278763803388,
+      "learning_rate": 1.974954561003078e-06,
+      "loss": 0.7233,
+      "step": 17077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5855101918947152,
+      "learning_rate": 1.974549885998417e-06,
+      "loss": 0.5686,
+      "step": 17078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2378256785290134,
+      "learning_rate": 1.9741452847968664e-06,
+      "loss": 0.4623,
+      "step": 17079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6339209195334972,
+      "learning_rate": 1.9737407574068924e-06,
+      "loss": 0.4181,
+      "step": 17080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5160916850940607,
+      "learning_rate": 1.973336303836965e-06,
+      "loss": 0.4031,
+      "step": 17081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7689441177224268,
+      "learning_rate": 1.9729319240955538e-06,
+      "loss": 0.5014,
+      "step": 17082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5747441490903968,
+      "learning_rate": 1.97252761819112e-06,
+      "loss": 0.5381,
+      "step": 17083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7487569280176567,
+      "learning_rate": 1.97212338613213e-06,
+      "loss": 0.457,
+      "step": 17084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2099995884439236,
+      "learning_rate": 1.9717192279270467e-06,
+      "loss": 0.5804,
+      "step": 17085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9290651323227885,
+      "learning_rate": 1.9713151435843287e-06,
+      "loss": 0.5598,
+      "step": 17086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8492685143357663,
+      "learning_rate": 1.9709111331124365e-06,
+      "loss": 0.3764,
+      "step": 17087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1359288779192993,
+      "learning_rate": 1.9705071965198274e-06,
+      "loss": 0.532,
+      "step": 17088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3508049450008963,
+      "learning_rate": 1.9701033338149593e-06,
+      "loss": 0.5775,
+      "step": 17089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1542180130001136,
+      "learning_rate": 1.969699545006284e-06,
+      "loss": 0.4162,
+      "step": 17090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6795545162704697,
+      "learning_rate": 1.969295830102256e-06,
+      "loss": 0.5325,
+      "step": 17091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8302776460543813,
+      "learning_rate": 1.9688921891113283e-06,
+      "loss": 0.5452,
+      "step": 17092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.760403484718279,
+      "learning_rate": 1.968488622041948e-06,
+      "loss": 0.5463,
+      "step": 17093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8168325390774058,
+      "learning_rate": 1.968085128902565e-06,
+      "loss": 0.4477,
+      "step": 17094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.942297043121562,
+      "learning_rate": 1.967681709701627e-06,
+      "loss": 0.441,
+      "step": 17095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.44295878890837,
+      "learning_rate": 1.9672783644475778e-06,
+      "loss": 0.6609,
+      "step": 17096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6077600125823686,
+      "learning_rate": 1.9668750931488613e-06,
+      "loss": 0.4692,
+      "step": 17097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8803691835384113,
+      "learning_rate": 1.9664718958139203e-06,
+      "loss": 0.4496,
+      "step": 17098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.907472370411501,
+      "learning_rate": 1.9660687724511966e-06,
+      "loss": 0.5437,
+      "step": 17099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7178419644399747,
+      "learning_rate": 1.965665723069127e-06,
+      "loss": 0.5315,
+      "step": 17100
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.4908825755119324,
+      "eval_runtime": 67.9232,
+      "eval_samples_per_second": 50.204,
+      "eval_steps_per_second": 1.575,
+      "step": 17100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.741880473921168,
+      "learning_rate": 1.9652627476761497e-06,
+      "loss": 0.439,
+      "step": 17101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6251309990130118,
+      "learning_rate": 1.964859846280703e-06,
+      "loss": 0.4071,
+      "step": 17102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.720972556144955,
+      "learning_rate": 1.9644570188912186e-06,
+      "loss": 0.4213,
+      "step": 17103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.681095355218643,
+      "learning_rate": 1.9640542655161306e-06,
+      "loss": 0.4204,
+      "step": 17104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.748943442329934,
+      "learning_rate": 1.963651586163871e-06,
+      "loss": 0.4314,
+      "step": 17105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8022478964470376,
+      "learning_rate": 1.963248980842868e-06,
+      "loss": 0.6212,
+      "step": 17106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8697931240232628,
+      "learning_rate": 1.9628464495615506e-06,
+      "loss": 0.6589,
+      "step": 17107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7815505307734743,
+      "learning_rate": 1.962443992328347e-06,
+      "loss": 0.4319,
+      "step": 17108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.903973547749835,
+      "learning_rate": 1.962041609151681e-06,
+      "loss": 0.4898,
+      "step": 17109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0409596421955953,
+      "learning_rate": 1.9616393000399734e-06,
+      "loss": 0.558,
+      "step": 17110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.09008780958771,
+      "learning_rate": 1.961237065001653e-06,
+      "loss": 0.4717,
+      "step": 17111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8487718764221033,
+      "learning_rate": 1.9608349040451353e-06,
+      "loss": 0.4869,
+      "step": 17112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7049050302923627,
+      "learning_rate": 1.9604328171788407e-06,
+      "loss": 0.4951,
+      "step": 17113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.84273844558538,
+      "learning_rate": 1.960030804411187e-06,
+      "loss": 0.4601,
+      "step": 17114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7969417843545197,
+      "learning_rate": 1.959628865750589e-06,
+      "loss": 0.4862,
+      "step": 17115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.122727749078996,
+      "learning_rate": 1.9592270012054616e-06,
+      "loss": 0.486,
+      "step": 17116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.635325446131076,
+      "learning_rate": 1.9588252107842186e-06,
+      "loss": 0.4761,
+      "step": 17117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0483080340858266,
+      "learning_rate": 1.9584234944952695e-06,
+      "loss": 0.4192,
+      "step": 17118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9752432954625545,
+      "learning_rate": 1.9580218523470245e-06,
+      "loss": 0.4131,
+      "step": 17119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.945873726616559,
+      "learning_rate": 1.9576202843478925e-06,
+      "loss": 0.4764,
+      "step": 17120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5861932534422434,
+      "learning_rate": 1.9572187905062808e-06,
+      "loss": 0.4745,
+      "step": 17121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8274753811835998,
+      "learning_rate": 1.9568173708305914e-06,
+      "loss": 0.5105,
+      "step": 17122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.143640631237111,
+      "learning_rate": 1.95641602532923e-06,
+      "loss": 0.6645,
+      "step": 17123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2002720535240203,
+      "learning_rate": 1.956014754010599e-06,
+      "loss": 0.7426,
+      "step": 17124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.871550652110205,
+      "learning_rate": 1.9556135568830967e-06,
+      "loss": 0.5684,
+      "step": 17125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9891211032549443,
+      "learning_rate": 1.9552124339551233e-06,
+      "loss": 0.4633,
+      "step": 17126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.838393375346133,
+      "learning_rate": 1.954811385235077e-06,
+      "loss": 0.5426,
+      "step": 17127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5054636178184855,
+      "learning_rate": 1.9544104107313513e-06,
+      "loss": 0.5858,
+      "step": 17128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4741221484730955,
+      "learning_rate": 1.9540095104523413e-06,
+      "loss": 0.4025,
+      "step": 17129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.331210830937696,
+      "learning_rate": 1.953608684406441e-06,
+      "loss": 0.6043,
+      "step": 17130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7046897110923938,
+      "learning_rate": 1.9532079326020386e-06,
+      "loss": 0.4188,
+      "step": 17131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2591088994452373,
+      "learning_rate": 1.9528072550475257e-06,
+      "loss": 0.5945,
+      "step": 17132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9695105366389536,
+      "learning_rate": 1.9524066517512892e-06,
+      "loss": 0.5549,
+      "step": 17133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4830496892802991,
+      "learning_rate": 1.952006122721717e-06,
+      "loss": 0.4577,
+      "step": 17134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8644421806453562,
+      "learning_rate": 1.9516056679671918e-06,
+      "loss": 0.4736,
+      "step": 17135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5169860692938595,
+      "learning_rate": 1.951205287496098e-06,
+      "loss": 0.4117,
+      "step": 17136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.454218070748855,
+      "learning_rate": 1.9508049813168185e-06,
+      "loss": 0.3695,
+      "step": 17137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5992769488806666,
+      "learning_rate": 1.950404749437731e-06,
+      "loss": 0.357,
+      "step": 17138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1179966421687233,
+      "learning_rate": 1.950004591867216e-06,
+      "loss": 0.5726,
+      "step": 17139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8825931137224174,
+      "learning_rate": 1.94960450861365e-06,
+      "loss": 0.4879,
+      "step": 17140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.743893773787922,
+      "learning_rate": 1.949204499685408e-06,
+      "loss": 0.4053,
+      "step": 17141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6724692294141565,
+      "learning_rate": 1.948804565090864e-06,
+      "loss": 0.3561,
+      "step": 17142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5506872241587109,
+      "learning_rate": 1.948404704838392e-06,
+      "loss": 0.3784,
+      "step": 17143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4806644975094203,
+      "learning_rate": 1.94800491893636e-06,
+      "loss": 0.4934,
+      "step": 17144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.850668806112288,
+      "learning_rate": 1.9476052073931395e-06,
+      "loss": 0.4329,
+      "step": 17145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4819043148402742,
+      "learning_rate": 1.9472055702170965e-06,
+      "loss": 0.4104,
+      "step": 17146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9574538531850938,
+      "learning_rate": 1.9468060074166e-06,
+      "loss": 0.5684,
+      "step": 17147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4006972879015835,
+      "learning_rate": 1.9464065190000116e-06,
+      "loss": 0.3697,
+      "step": 17148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7416582463065502,
+      "learning_rate": 1.9460071049756956e-06,
+      "loss": 0.5816,
+      "step": 17149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2746211671433032,
+      "learning_rate": 1.9456077653520144e-06,
+      "loss": 0.429,
+      "step": 17150
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.49062973260879517,
+      "eval_runtime": 67.9647,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 17150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.882074098542107,
+      "learning_rate": 1.9452085001373256e-06,
+      "loss": 0.5585,
+      "step": 17151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2155497713062897,
+      "learning_rate": 1.9448093093399886e-06,
+      "loss": 0.5504,
+      "step": 17152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8020803005472001,
+      "learning_rate": 1.9444101929683625e-06,
+      "loss": 0.439,
+      "step": 17153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.245699018580629,
+      "learning_rate": 1.944011151030799e-06,
+      "loss": 0.5587,
+      "step": 17154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.060169428487998,
+      "learning_rate": 1.943612183535653e-06,
+      "loss": 0.6981,
+      "step": 17155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6659791982757646,
+      "learning_rate": 1.9432132904912783e-06,
+      "loss": 0.4339,
+      "step": 17156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6282606441342982,
+      "learning_rate": 1.9428144719060232e-06,
+      "loss": 0.4417,
+      "step": 17157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6825003032629673,
+      "learning_rate": 1.9424157277882376e-06,
+      "loss": 0.371,
+      "step": 17158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6768590060357942,
+      "learning_rate": 1.9420170581462687e-06,
+      "loss": 0.3931,
+      "step": 17159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.067489422109219,
+      "learning_rate": 1.9416184629884633e-06,
+      "loss": 0.6572,
+      "step": 17160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6560823423745605,
+      "learning_rate": 1.941219942323165e-06,
+      "loss": 0.5067,
+      "step": 17161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8086630404806054,
+      "learning_rate": 1.940821496158718e-06,
+      "loss": 0.3924,
+      "step": 17162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.60899562986024,
+      "learning_rate": 1.9404231245034613e-06,
+      "loss": 0.4223,
+      "step": 17163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4280632090156637,
+      "learning_rate": 1.9400248273657353e-06,
+      "loss": 0.4249,
+      "step": 17164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6691920396346278,
+      "learning_rate": 1.9396266047538796e-06,
+      "loss": 0.3762,
+      "step": 17165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5365293568163498,
+      "learning_rate": 1.9392284566762282e-06,
+      "loss": 0.4036,
+      "step": 17166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0515525045358523,
+      "learning_rate": 1.9388303831411176e-06,
+      "loss": 0.5562,
+      "step": 17167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.858121817159707,
+      "learning_rate": 1.9384323841568803e-06,
+      "loss": 0.548,
+      "step": 17168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9598053491469773,
+      "learning_rate": 1.9380344597318505e-06,
+      "loss": 0.5131,
+      "step": 17169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8235918932286848,
+      "learning_rate": 1.937636609874356e-06,
+      "loss": 0.5704,
+      "step": 17170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6078885559515408,
+      "learning_rate": 1.937238834592726e-06,
+      "loss": 0.4009,
+      "step": 17171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9568529889965254,
+      "learning_rate": 1.9368411338952895e-06,
+      "loss": 0.4601,
+      "step": 17172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6722557728794862,
+      "learning_rate": 1.93644350779037e-06,
+      "loss": 0.4341,
+      "step": 17173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.656233506319582,
+      "learning_rate": 1.9360459562862917e-06,
+      "loss": 0.5012,
+      "step": 17174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4196894244170926,
+      "learning_rate": 1.9356484793913783e-06,
+      "loss": 0.4057,
+      "step": 17175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9606673741522032,
+      "learning_rate": 1.93525107711395e-06,
+      "loss": 0.5301,
+      "step": 17176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6581478335702715,
+      "learning_rate": 1.934853749462325e-06,
+      "loss": 0.5241,
+      "step": 17177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9560791495627083,
+      "learning_rate": 1.9344564964448236e-06,
+      "loss": 0.4635,
+      "step": 17178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.53558151076316,
+      "learning_rate": 1.934059318069761e-06,
+      "loss": 0.3857,
+      "step": 17179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5747919433248274,
+      "learning_rate": 1.9336622143454515e-06,
+      "loss": 0.4477,
+      "step": 17180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8474061708364398,
+      "learning_rate": 1.9332651852802074e-06,
+      "loss": 0.5492,
+      "step": 17181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9002582076086099,
+      "learning_rate": 1.932868230882343e-06,
+      "loss": 0.5005,
+      "step": 17182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4969739383143361,
+      "learning_rate": 1.932471351160165e-06,
+      "loss": 0.4184,
+      "step": 17183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5558784002602823,
+      "learning_rate": 1.932074546121984e-06,
+      "loss": 0.3937,
+      "step": 17184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5736867445786655,
+      "learning_rate": 1.931677815776107e-06,
+      "loss": 0.4083,
+      "step": 17185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.142975617457727,
+      "learning_rate": 1.9312811601308373e-06,
+      "loss": 0.6452,
+      "step": 17186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.945568095555666,
+      "learning_rate": 1.930884579194481e-06,
+      "loss": 0.5252,
+      "step": 17187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6653485024149666,
+      "learning_rate": 1.930488072975339e-06,
+      "loss": 0.3354,
+      "step": 17188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7838540506600364,
+      "learning_rate": 1.9300916414817115e-06,
+      "loss": 0.414,
+      "step": 17189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.729823422602026,
+      "learning_rate": 1.9296952847218976e-06,
+      "loss": 0.4587,
+      "step": 17190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.724543037897941,
+      "learning_rate": 1.929299002704196e-06,
+      "loss": 0.4139,
+      "step": 17191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.939072843069859,
+      "learning_rate": 1.9289027954369025e-06,
+      "loss": 0.5304,
+      "step": 17192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5125556166835723,
+      "learning_rate": 1.92850666292831e-06,
+      "loss": 0.4151,
+      "step": 17193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.811987531621155,
+      "learning_rate": 1.928110605186713e-06,
+      "loss": 0.4585,
+      "step": 17194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7162845989536004,
+      "learning_rate": 1.927714622220402e-06,
+      "loss": 0.5001,
+      "step": 17195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1249153959498073,
+      "learning_rate": 1.9273187140376655e-06,
+      "loss": 0.6097,
+      "step": 17196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6628252361346936,
+      "learning_rate": 1.9269228806467928e-06,
+      "loss": 0.3624,
+      "step": 17197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7240111061774144,
+      "learning_rate": 1.9265271220560715e-06,
+      "loss": 0.3808,
+      "step": 17198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8696067253918895,
+      "learning_rate": 1.926131438273784e-06,
+      "loss": 0.5403,
+      "step": 17199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6575576494355952,
+      "learning_rate": 1.925735829308215e-06,
+      "loss": 0.5187,
+      "step": 17200
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.49055179953575134,
+      "eval_runtime": 67.934,
+      "eval_samples_per_second": 50.196,
+      "eval_steps_per_second": 1.575,
+      "step": 17200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7642986611261215,
+      "learning_rate": 1.925340295167648e-06,
+      "loss": 0.6455,
+      "step": 17201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.492817126875135,
+      "learning_rate": 1.9249448358603594e-06,
+      "loss": 0.445,
+      "step": 17202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4767428868912282,
+      "learning_rate": 1.9245494513946313e-06,
+      "loss": 0.4558,
+      "step": 17203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9783155548769409,
+      "learning_rate": 1.9241541417787393e-06,
+      "loss": 0.5589,
+      "step": 17204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8206249013608522,
+      "learning_rate": 1.92375890702096e-06,
+      "loss": 0.5724,
+      "step": 17205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5087821179670942,
+      "learning_rate": 1.923363747129566e-06,
+      "loss": 0.3864,
+      "step": 17206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8533130540948,
+      "learning_rate": 1.9229686621128304e-06,
+      "loss": 0.5074,
+      "step": 17207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.675896528851987,
+      "learning_rate": 1.922573651979024e-06,
+      "loss": 0.5888,
+      "step": 17208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6575249386940154,
+      "learning_rate": 1.922178716736416e-06,
+      "loss": 0.4042,
+      "step": 17209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5001686024658516,
+      "learning_rate": 1.9217838563932756e-06,
+      "loss": 0.4604,
+      "step": 17210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4173991198933031,
+      "learning_rate": 1.9213890709578662e-06,
+      "loss": 0.3546,
+      "step": 17211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.671125659423454,
+      "learning_rate": 1.9209943604384545e-06,
+      "loss": 0.5574,
+      "step": 17212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8613096253161079,
+      "learning_rate": 1.9205997248433028e-06,
+      "loss": 0.4256,
+      "step": 17213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8180181183909223,
+      "learning_rate": 1.920205164180673e-06,
+      "loss": 0.4804,
+      "step": 17214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.113751668521654,
+      "learning_rate": 1.919810678458824e-06,
+      "loss": 0.6426,
+      "step": 17215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.258495335673112,
+      "learning_rate": 1.9194162676860146e-06,
+      "loss": 0.5512,
+      "step": 17216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4182300636684229,
+      "learning_rate": 1.9190219318705028e-06,
+      "loss": 0.381,
+      "step": 17217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6628963417872074,
+      "learning_rate": 1.9186276710205417e-06,
+      "loss": 0.4455,
+      "step": 17218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.601730547014694,
+      "learning_rate": 1.9182334851443856e-06,
+      "loss": 0.3413,
+      "step": 17219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6456368503845478,
+      "learning_rate": 1.917839374250288e-06,
+      "loss": 0.3658,
+      "step": 17220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0025290996499163,
+      "learning_rate": 1.917445338346497e-06,
+      "loss": 0.5203,
+      "step": 17221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7471306289634265,
+      "learning_rate": 1.9170513774412626e-06,
+      "loss": 0.5434,
+      "step": 17222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5495148415518631,
+      "learning_rate": 1.916657491542833e-06,
+      "loss": 0.3025,
+      "step": 17223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2339829639923647,
+      "learning_rate": 1.9162636806594526e-06,
+      "loss": 0.5077,
+      "step": 17224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8531738376271023,
+      "learning_rate": 1.9158699447993657e-06,
+      "loss": 0.4233,
+      "step": 17225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6902117303751871,
+      "learning_rate": 1.915476283970815e-06,
+      "loss": 0.4727,
+      "step": 17226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4108623931050728,
+      "learning_rate": 1.9150826981820437e-06,
+      "loss": 0.3001,
+      "step": 17227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9618593122289756,
+      "learning_rate": 1.914689187441288e-06,
+      "loss": 0.5721,
+      "step": 17228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.723231866365178,
+      "learning_rate": 1.9142957517567875e-06,
+      "loss": 0.2936,
+      "step": 17229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5530994080244576,
+      "learning_rate": 1.9139023911367788e-06,
+      "loss": 0.4193,
+      "step": 17230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.151345956784808,
+      "learning_rate": 1.9135091055894954e-06,
+      "loss": 0.4399,
+      "step": 17231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0694806866354365,
+      "learning_rate": 1.913115895123171e-06,
+      "loss": 0.4993,
+      "step": 17232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.885167256840542,
+      "learning_rate": 1.912722759746039e-06,
+      "loss": 0.4167,
+      "step": 17233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.047722680575938,
+      "learning_rate": 1.9123296994663265e-06,
+      "loss": 0.4647,
+      "step": 17234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7562390396177574,
+      "learning_rate": 1.9119367142922635e-06,
+      "loss": 0.4581,
+      "step": 17235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.958393933004665,
+      "learning_rate": 1.9115438042320776e-06,
+      "loss": 0.5714,
+      "step": 17236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6002836878969973,
+      "learning_rate": 1.9111509692939924e-06,
+      "loss": 0.4697,
+      "step": 17237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6451618017957514,
+      "learning_rate": 1.9107582094862324e-06,
+      "loss": 0.558,
+      "step": 17238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7897189791337098,
+      "learning_rate": 1.91036552481702e-06,
+      "loss": 0.5439,
+      "step": 17239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9722834493135344,
+      "learning_rate": 1.909972915294577e-06,
+      "loss": 0.479,
+      "step": 17240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8142940736189224,
+      "learning_rate": 1.909580380927119e-06,
+      "loss": 0.3854,
+      "step": 17241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6906366975488918,
+      "learning_rate": 1.9091879217228672e-06,
+      "loss": 0.4809,
+      "step": 17242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4254547098852037,
+      "learning_rate": 1.9087955376900363e-06,
+      "loss": 0.3565,
+      "step": 17243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8510787465636371,
+      "learning_rate": 1.908403228836839e-06,
+      "loss": 0.4225,
+      "step": 17244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0319930377684297,
+      "learning_rate": 1.9080109951714894e-06,
+      "loss": 0.3614,
+      "step": 17245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7997447443773626,
+      "learning_rate": 1.9076188367021995e-06,
+      "loss": 0.5329,
+      "step": 17246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8108256654158974,
+      "learning_rate": 1.9072267534371775e-06,
+      "loss": 0.4273,
+      "step": 17247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.576229860246421,
+      "learning_rate": 1.906834745384631e-06,
+      "loss": 0.3865,
+      "step": 17248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7254213898099906,
+      "learning_rate": 1.906442812552769e-06,
+      "loss": 0.3834,
+      "step": 17249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6985497436305428,
+      "learning_rate": 1.9060509549497935e-06,
+      "loss": 0.4578,
+      "step": 17250
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.49049121141433716,
+      "eval_runtime": 67.9524,
+      "eval_samples_per_second": 50.182,
+      "eval_steps_per_second": 1.575,
+      "step": 17250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8007312514117506,
+      "learning_rate": 1.9056591725839085e-06,
+      "loss": 0.4267,
+      "step": 17251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6262978783161421,
+      "learning_rate": 1.9052674654633167e-06,
+      "loss": 0.4229,
+      "step": 17252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9288284218995542,
+      "learning_rate": 1.9048758335962187e-06,
+      "loss": 0.6236,
+      "step": 17253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6131415569960852,
+      "learning_rate": 1.904484276990811e-06,
+      "loss": 0.3836,
+      "step": 17254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.737672477386867,
+      "learning_rate": 1.9040927956552917e-06,
+      "loss": 0.4272,
+      "step": 17255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6392605834612355,
+      "learning_rate": 1.9037013895978576e-06,
+      "loss": 0.3821,
+      "step": 17256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3286216654794756,
+      "learning_rate": 1.9033100588266998e-06,
+      "loss": 0.5429,
+      "step": 17257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0828901902896417,
+      "learning_rate": 1.9029188033500107e-06,
+      "loss": 0.3821,
+      "step": 17258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9911928178159122,
+      "learning_rate": 1.9025276231759846e-06,
+      "loss": 0.4736,
+      "step": 17259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8239487039406481,
+      "learning_rate": 1.902136518312807e-06,
+      "loss": 0.4605,
+      "step": 17260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8236366089287646,
+      "learning_rate": 1.9017454887686665e-06,
+      "loss": 0.5458,
+      "step": 17261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7297752243086986,
+      "learning_rate": 1.9013545345517509e-06,
+      "loss": 0.4927,
+      "step": 17262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8630602532469935,
+      "learning_rate": 1.9009636556702416e-06,
+      "loss": 0.4577,
+      "step": 17263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.205924709845824,
+      "learning_rate": 1.9005728521323218e-06,
+      "loss": 0.5246,
+      "step": 17264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8837791223749942,
+      "learning_rate": 1.900182123946176e-06,
+      "loss": 0.495,
+      "step": 17265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6937779901541086,
+      "learning_rate": 1.89979147111998e-06,
+      "loss": 0.5003,
+      "step": 17266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.426790055493602,
+      "learning_rate": 1.8994008936619134e-06,
+      "loss": 0.4233,
+      "step": 17267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6691966626251042,
+      "learning_rate": 1.8990103915801539e-06,
+      "loss": 0.5824,
+      "step": 17268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9567016957554486,
+      "learning_rate": 1.8986199648828738e-06,
+      "loss": 0.5947,
+      "step": 17269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6149058245857166,
+      "learning_rate": 1.898229613578248e-06,
+      "loss": 0.4543,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.158408983807919,
+      "learning_rate": 1.897839337674448e-06,
+      "loss": 0.6041,
+      "step": 17271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.453824946088077,
+      "learning_rate": 1.8974491371796456e-06,
+      "loss": 0.3864,
+      "step": 17272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.186593394586213,
+      "learning_rate": 1.8970590121020063e-06,
+      "loss": 0.48,
+      "step": 17273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6936333319689683,
+      "learning_rate": 1.8966689624496992e-06,
+      "loss": 0.3989,
+      "step": 17274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.227643858167402,
+      "learning_rate": 1.8962789882308897e-06,
+      "loss": 0.3766,
+      "step": 17275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5253827541880047,
+      "learning_rate": 1.8958890894537401e-06,
+      "loss": 0.3653,
+      "step": 17276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.683076982385573,
+      "learning_rate": 1.8954992661264146e-06,
+      "loss": 0.4168,
+      "step": 17277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4782571664728819,
+      "learning_rate": 1.895109518257074e-06,
+      "loss": 0.4348,
+      "step": 17278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7752073280825,
+      "learning_rate": 1.8947198458538752e-06,
+      "loss": 0.5005,
+      "step": 17279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3500597280940532,
+      "learning_rate": 1.8943302489249772e-06,
+      "loss": 0.3695,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0735437133271213,
+      "learning_rate": 1.8939407274785366e-06,
+      "loss": 0.6653,
+      "step": 17281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.016430110726534,
+      "learning_rate": 1.8935512815227063e-06,
+      "loss": 0.5385,
+      "step": 17282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7165894818289367,
+      "learning_rate": 1.89316191106564e-06,
+      "loss": 0.4468,
+      "step": 17283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.912586313755925,
+      "learning_rate": 1.8927726161154886e-06,
+      "loss": 0.5744,
+      "step": 17284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.640091079269357,
+      "learning_rate": 1.892383396680403e-06,
+      "loss": 0.5051,
+      "step": 17285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8851930308323135,
+      "learning_rate": 1.8919942527685293e-06,
+      "loss": 0.6566,
+      "step": 17286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5784815122781357,
+      "learning_rate": 1.8916051843880148e-06,
+      "loss": 0.4019,
+      "step": 17287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1043141622333894,
+      "learning_rate": 1.8912161915470056e-06,
+      "loss": 0.4596,
+      "step": 17288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6225075917207181,
+      "learning_rate": 1.8908272742536428e-06,
+      "loss": 0.4077,
+      "step": 17289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0997422162285053,
+      "learning_rate": 1.890438432516069e-06,
+      "loss": 0.4921,
+      "step": 17290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5104372822833676,
+      "learning_rate": 1.890049666342426e-06,
+      "loss": 0.3497,
+      "step": 17291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5340166824336692,
+      "learning_rate": 1.8896609757408503e-06,
+      "loss": 0.5199,
+      "step": 17292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7334747206707792,
+      "learning_rate": 1.8892723607194797e-06,
+      "loss": 0.4608,
+      "step": 17293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6393172295355543,
+      "learning_rate": 1.88888382128645e-06,
+      "loss": 0.3755,
+      "step": 17294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8441268313584607,
+      "learning_rate": 1.8884953574498937e-06,
+      "loss": 0.584,
+      "step": 17295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7551712887822841,
+      "learning_rate": 1.8881069692179439e-06,
+      "loss": 0.627,
+      "step": 17296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5412504947047507,
+      "learning_rate": 1.8877186565987312e-06,
+      "loss": 0.4814,
+      "step": 17297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4741748495884588,
+      "learning_rate": 1.887330419600386e-06,
+      "loss": 0.4326,
+      "step": 17298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2868240245443454,
+      "learning_rate": 1.8869422582310332e-06,
+      "loss": 0.6008,
+      "step": 17299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0095537111182025,
+      "learning_rate": 1.8865541724988002e-06,
+      "loss": 0.6238,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.490773469209671,
+      "eval_runtime": 67.9319,
+      "eval_samples_per_second": 50.197,
+      "eval_steps_per_second": 1.575,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5460866184512159,
+      "learning_rate": 1.886166162411812e-06,
+      "loss": 0.419,
+      "step": 17301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.569214241362944,
+      "learning_rate": 1.88577822797819e-06,
+      "loss": 0.4252,
+      "step": 17302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9147909164991739,
+      "learning_rate": 1.885390369206056e-06,
+      "loss": 0.6497,
+      "step": 17303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6841485179160676,
+      "learning_rate": 1.8850025861035299e-06,
+      "loss": 0.5136,
+      "step": 17304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6844020387730883,
+      "learning_rate": 1.8846148786787289e-06,
+      "loss": 0.5153,
+      "step": 17305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5658238386958692,
+      "learning_rate": 1.8842272469397693e-06,
+      "loss": 0.5191,
+      "step": 17306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6308204186553392,
+      "learning_rate": 1.8838396908947665e-06,
+      "loss": 0.4158,
+      "step": 17307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9044186431383852,
+      "learning_rate": 1.883452210551834e-06,
+      "loss": 0.5381,
+      "step": 17308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9165029466211116,
+      "learning_rate": 1.8830648059190832e-06,
+      "loss": 0.4116,
+      "step": 17309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4274363512684831,
+      "learning_rate": 1.8826774770046247e-06,
+      "loss": 0.3312,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.337471545908776,
+      "learning_rate": 1.8822902238165658e-06,
+      "loss": 0.3392,
+      "step": 17311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8466432493909113,
+      "learning_rate": 1.8819030463630141e-06,
+      "loss": 0.5096,
+      "step": 17312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.336960918284143,
+      "learning_rate": 1.8815159446520753e-06,
+      "loss": 0.6506,
+      "step": 17313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5143862986682795,
+      "learning_rate": 1.8811289186918524e-06,
+      "loss": 0.6115,
+      "step": 17314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8877870033803277,
+      "learning_rate": 1.8807419684904474e-06,
+      "loss": 0.4805,
+      "step": 17315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5989715057826597,
+      "learning_rate": 1.8803550940559624e-06,
+      "loss": 0.4097,
+      "step": 17316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8843316499239093,
+      "learning_rate": 1.8799682953964937e-06,
+      "loss": 0.4006,
+      "step": 17317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7503910282874555,
+      "learning_rate": 1.8795815725201412e-06,
+      "loss": 0.4657,
+      "step": 17318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.700308251848182,
+      "learning_rate": 1.879194925434999e-06,
+      "loss": 0.4933,
+      "step": 17319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9845373114588156,
+      "learning_rate": 1.8788083541491637e-06,
+      "loss": 0.501,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6838480690227455,
+      "learning_rate": 1.8784218586707247e-06,
+      "loss": 0.542,
+      "step": 17321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.814584035181916,
+      "learning_rate": 1.878035439007775e-06,
+      "loss": 0.5031,
+      "step": 17322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8123591084767692,
+      "learning_rate": 1.877649095168405e-06,
+      "loss": 0.4021,
+      "step": 17323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.639045769243423,
+      "learning_rate": 1.8772628271606997e-06,
+      "loss": 0.4869,
+      "step": 17324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5602657019586728,
+      "learning_rate": 1.8768766349927469e-06,
+      "loss": 0.4749,
+      "step": 17325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.011279593039946,
+      "learning_rate": 1.876490518672633e-06,
+      "loss": 0.5489,
+      "step": 17326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6816207030253711,
+      "learning_rate": 1.8761044782084381e-06,
+      "loss": 0.3982,
+      "step": 17327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6400009722566704,
+      "learning_rate": 1.8757185136082444e-06,
+      "loss": 0.5383,
+      "step": 17328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7631652873070789,
+      "learning_rate": 1.8753326248801346e-06,
+      "loss": 0.5009,
+      "step": 17329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8421244790294053,
+      "learning_rate": 1.8749468120321828e-06,
+      "loss": 0.525,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0835711654241775,
+      "learning_rate": 1.874561075072469e-06,
+      "loss": 0.5272,
+      "step": 17331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3746362244811954,
+      "learning_rate": 1.8741754140090662e-06,
+      "loss": 0.3761,
+      "step": 17332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5796787562296675,
+      "learning_rate": 1.873789828850051e-06,
+      "loss": 0.387,
+      "step": 17333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8289419639552842,
+      "learning_rate": 1.8734043196034922e-06,
+      "loss": 0.5158,
+      "step": 17334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.581307906129545,
+      "learning_rate": 1.873018886277461e-06,
+      "loss": 0.4613,
+      "step": 17335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.008622306113101,
+      "learning_rate": 1.872633528880028e-06,
+      "loss": 0.6162,
+      "step": 17336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5027366985192623,
+      "learning_rate": 1.8722482474192576e-06,
+      "loss": 0.3329,
+      "step": 17337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4704672176803644,
+      "learning_rate": 1.871863041903217e-06,
+      "loss": 0.3771,
+      "step": 17338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4289974917336619,
+      "learning_rate": 1.8714779123399712e-06,
+      "loss": 0.3712,
+      "step": 17339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8489837471905692,
+      "learning_rate": 1.8710928587375807e-06,
+      "loss": 0.45,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.885080388776603,
+      "learning_rate": 1.8707078811041067e-06,
+      "loss": 0.4817,
+      "step": 17341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0046289908216677,
+      "learning_rate": 1.8703229794476107e-06,
+      "loss": 0.4262,
+      "step": 17342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9694903596653646,
+      "learning_rate": 1.8699381537761472e-06,
+      "loss": 0.5514,
+      "step": 17343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.788101010988369,
+      "learning_rate": 1.8695534040977737e-06,
+      "loss": 0.5641,
+      "step": 17344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0714991100394404,
+      "learning_rate": 1.8691687304205448e-06,
+      "loss": 0.5653,
+      "step": 17345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6849505382502037,
+      "learning_rate": 1.8687841327525147e-06,
+      "loss": 0.4236,
+      "step": 17346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.986584406770172,
+      "learning_rate": 1.8683996111017316e-06,
+      "loss": 0.4637,
+      "step": 17347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6180700167072508,
+      "learning_rate": 1.8680151654762475e-06,
+      "loss": 0.4598,
+      "step": 17348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0946982075757714,
+      "learning_rate": 1.867630795884111e-06,
+      "loss": 0.5891,
+      "step": 17349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9447378574083838,
+      "learning_rate": 1.8672465023333663e-06,
+      "loss": 0.5748,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.49020472168922424,
+      "eval_runtime": 67.9146,
+      "eval_samples_per_second": 50.21,
+      "eval_steps_per_second": 1.576,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8188952649658225,
+      "learning_rate": 1.86686228483206e-06,
+      "loss": 0.3759,
+      "step": 17351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.478608413241097,
+      "learning_rate": 1.866478143388236e-06,
+      "loss": 0.3083,
+      "step": 17352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7146820174742474,
+      "learning_rate": 1.8660940780099346e-06,
+      "loss": 0.2961,
+      "step": 17353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4149504353644282,
+      "learning_rate": 1.865710088705196e-06,
+      "loss": 0.419,
+      "step": 17354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8595070404827698,
+      "learning_rate": 1.8653261754820595e-06,
+      "loss": 0.5408,
+      "step": 17355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.600359885435122,
+      "learning_rate": 1.864942338348562e-06,
+      "loss": 0.348,
+      "step": 17356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8493334960554122,
+      "learning_rate": 1.8645585773127386e-06,
+      "loss": 0.5698,
+      "step": 17357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8237210167094493,
+      "learning_rate": 1.8641748923826247e-06,
+      "loss": 0.5642,
+      "step": 17358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5999060201612496,
+      "learning_rate": 1.8637912835662497e-06,
+      "loss": 0.5038,
+      "step": 17359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8753745046734986,
+      "learning_rate": 1.863407750871646e-06,
+      "loss": 0.4837,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.68719140843379,
+      "learning_rate": 1.8630242943068439e-06,
+      "loss": 0.5732,
+      "step": 17361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3679581448112685,
+      "learning_rate": 1.8626409138798667e-06,
+      "loss": 0.3351,
+      "step": 17362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7415349090628072,
+      "learning_rate": 1.8622576095987439e-06,
+      "loss": 0.5192,
+      "step": 17363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7504456008118041,
+      "learning_rate": 1.8618743814714982e-06,
+      "loss": 0.4244,
+      "step": 17364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4764837438243212,
+      "learning_rate": 1.8614912295061532e-06,
+      "loss": 0.3353,
+      "step": 17365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7818166590695697,
+      "learning_rate": 1.861108153710729e-06,
+      "loss": 0.5285,
+      "step": 17366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1069015836759513,
+      "learning_rate": 1.8607251540932452e-06,
+      "loss": 0.4976,
+      "step": 17367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8339391414157433,
+      "learning_rate": 1.8603422306617209e-06,
+      "loss": 0.4004,
+      "step": 17368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9429230329786165,
+      "learning_rate": 1.8599593834241703e-06,
+      "loss": 0.597,
+      "step": 17369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7137250046438124,
+      "learning_rate": 1.8595766123886092e-06,
+      "loss": 0.4471,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6384195622983726,
+      "learning_rate": 1.8591939175630514e-06,
+      "loss": 0.4914,
+      "step": 17371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7619416339038743,
+      "learning_rate": 1.8588112989555063e-06,
+      "loss": 0.3987,
+      "step": 17372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6514473012610134,
+      "learning_rate": 1.8584287565739853e-06,
+      "loss": 0.409,
+      "step": 17373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6214713691658638,
+      "learning_rate": 1.8580462904264976e-06,
+      "loss": 0.3326,
+      "step": 17374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9689560964045225,
+      "learning_rate": 1.8576639005210474e-06,
+      "loss": 0.6597,
+      "step": 17375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7742307871521141,
+      "learning_rate": 1.8572815868656407e-06,
+      "loss": 0.388,
+      "step": 17376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6728626111691876,
+      "learning_rate": 1.8568993494682818e-06,
+      "loss": 0.4795,
+      "step": 17377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5664114104335098,
+      "learning_rate": 1.8565171883369743e-06,
+      "loss": 0.3129,
+      "step": 17378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6652953607150012,
+      "learning_rate": 1.8561351034797138e-06,
+      "loss": 0.4444,
+      "step": 17379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7751985847906115,
+      "learning_rate": 1.8557530949045022e-06,
+      "loss": 0.5518,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9237976641737589,
+      "learning_rate": 1.8553711626193372e-06,
+      "loss": 0.4209,
+      "step": 17381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2454095219978873,
+      "learning_rate": 1.8549893066322122e-06,
+      "loss": 0.2586,
+      "step": 17382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6009317359575626,
+      "learning_rate": 1.8546075269511216e-06,
+      "loss": 0.5183,
+      "step": 17383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5290941928364858,
+      "learning_rate": 1.8542258235840599e-06,
+      "loss": 0.4235,
+      "step": 17384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6545263689582361,
+      "learning_rate": 1.8538441965390142e-06,
+      "loss": 0.5131,
+      "step": 17385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.818247167752203,
+      "learning_rate": 1.8534626458239763e-06,
+      "loss": 0.4907,
+      "step": 17386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.843525063825671,
+      "learning_rate": 1.8530811714469338e-06,
+      "loss": 0.4204,
+      "step": 17387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6339112221953134,
+      "learning_rate": 1.8526997734158703e-06,
+      "loss": 0.4461,
+      "step": 17388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.505764518748971,
+      "learning_rate": 1.8523184517387721e-06,
+      "loss": 0.4625,
+      "step": 17389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.625446303446375,
+      "learning_rate": 1.8519372064236214e-06,
+      "loss": 0.3545,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6197626068908355,
+      "learning_rate": 1.8515560374784005e-06,
+      "loss": 0.3862,
+      "step": 17391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.146943620255213,
+      "learning_rate": 1.8511749449110865e-06,
+      "loss": 0.6816,
+      "step": 17392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7747743087579244,
+      "learning_rate": 1.8507939287296589e-06,
+      "loss": 0.5261,
+      "step": 17393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.827385475244081,
+      "learning_rate": 1.8504129889420946e-06,
+      "loss": 0.4258,
+      "step": 17394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3991126203328332,
+      "learning_rate": 1.8500321255563669e-06,
+      "loss": 0.2993,
+      "step": 17395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7408338851288225,
+      "learning_rate": 1.8496513385804487e-06,
+      "loss": 0.3815,
+      "step": 17396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4994811199768996,
+      "learning_rate": 1.8492706280223144e-06,
+      "loss": 0.4836,
+      "step": 17397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2762602617508991,
+      "learning_rate": 1.8488899938899305e-06,
+      "loss": 0.3218,
+      "step": 17398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0375653734457697,
+      "learning_rate": 1.8485094361912667e-06,
+      "loss": 0.5795,
+      "step": 17399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.87596394702866,
+      "learning_rate": 1.8481289549342915e-06,
+      "loss": 0.4841,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.49048349261283875,
+      "eval_runtime": 67.9516,
+      "eval_samples_per_second": 50.183,
+      "eval_steps_per_second": 1.575,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8069089528307618,
+      "learning_rate": 1.847748550126967e-06,
+      "loss": 0.4333,
+      "step": 17401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6949644830485355,
+      "learning_rate": 1.8473682217772577e-06,
+      "loss": 0.5152,
+      "step": 17402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4689172220858222,
+      "learning_rate": 1.8469879698931259e-06,
+      "loss": 0.5364,
+      "step": 17403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.708417813107853,
+      "learning_rate": 1.8466077944825338e-06,
+      "loss": 0.4185,
+      "step": 17404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8768152631590258,
+      "learning_rate": 1.8462276955534352e-06,
+      "loss": 0.5033,
+      "step": 17405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6567472764942741,
+      "learning_rate": 1.8458476731137932e-06,
+      "loss": 0.4174,
+      "step": 17406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9458067654281328,
+      "learning_rate": 1.845467727171559e-06,
+      "loss": 0.713,
+      "step": 17407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5759118900596993,
+      "learning_rate": 1.8450878577346884e-06,
+      "loss": 0.4597,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2066042399993417,
+      "learning_rate": 1.8447080648111343e-06,
+      "loss": 0.6337,
+      "step": 17409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8787135384004232,
+      "learning_rate": 1.8443283484088459e-06,
+      "loss": 0.5258,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5527046780462612,
+      "learning_rate": 1.8439487085357722e-06,
+      "loss": 0.3778,
+      "step": 17411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3680869464313463,
+      "learning_rate": 1.8435691451998615e-06,
+      "loss": 0.3816,
+      "step": 17412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8122842562436787,
+      "learning_rate": 1.8431896584090605e-06,
+      "loss": 0.56,
+      "step": 17413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0156531572747824,
+      "learning_rate": 1.8428102481713123e-06,
+      "loss": 0.7005,
+      "step": 17414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9960538928864398,
+      "learning_rate": 1.8424309144945595e-06,
+      "loss": 0.4999,
+      "step": 17415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6301407227542688,
+      "learning_rate": 1.8420516573867446e-06,
+      "loss": 0.395,
+      "step": 17416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9365603812309784,
+      "learning_rate": 1.8416724768558056e-06,
+      "loss": 0.5477,
+      "step": 17417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5635423022354689,
+      "learning_rate": 1.841293372909681e-06,
+      "loss": 0.3828,
+      "step": 17418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6416130292070674,
+      "learning_rate": 1.8409143455563078e-06,
+      "loss": 0.4318,
+      "step": 17419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.019939565023098,
+      "learning_rate": 1.8405353948036194e-06,
+      "loss": 0.6346,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6050890391481696,
+      "learning_rate": 1.8401565206595496e-06,
+      "loss": 0.4115,
+      "step": 17421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6550799863414398,
+      "learning_rate": 1.8397777231320308e-06,
+      "loss": 0.4339,
+      "step": 17422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9113617636056504,
+      "learning_rate": 1.8393990022289904e-06,
+      "loss": 0.5665,
+      "step": 17423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9999198870842774,
+      "learning_rate": 1.8390203579583588e-06,
+      "loss": 0.5107,
+      "step": 17424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3585347092774018,
+      "learning_rate": 1.8386417903280614e-06,
+      "loss": 0.3055,
+      "step": 17425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7251107261955476,
+      "learning_rate": 1.8382632993460257e-06,
+      "loss": 0.38,
+      "step": 17426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6839211255662718,
+      "learning_rate": 1.837884885020172e-06,
+      "loss": 0.4854,
+      "step": 17427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8232563878492105,
+      "learning_rate": 1.8375065473584236e-06,
+      "loss": 0.5521,
+      "step": 17428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3519038801609202,
+      "learning_rate": 1.8371282863687019e-06,
+      "loss": 0.333,
+      "step": 17429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1181592614258005,
+      "learning_rate": 1.8367501020589234e-06,
+      "loss": 0.5014,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6910281946992447,
+      "learning_rate": 1.8363719944370063e-06,
+      "loss": 0.4506,
+      "step": 17431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8643695663195723,
+      "learning_rate": 1.835993963510867e-06,
+      "loss": 0.5621,
+      "step": 17432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7869683051069523,
+      "learning_rate": 1.8356160092884173e-06,
+      "loss": 0.4244,
+      "step": 17433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8355374174407773,
+      "learning_rate": 1.8352381317775703e-06,
+      "loss": 0.4519,
+      "step": 17434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5129586617734379,
+      "learning_rate": 1.8348603309862361e-06,
+      "loss": 0.3833,
+      "step": 17435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3617438779542204,
+      "learning_rate": 1.8344826069223265e-06,
+      "loss": 0.3517,
+      "step": 17436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8155482507766803,
+      "learning_rate": 1.8341049595937451e-06,
+      "loss": 0.4453,
+      "step": 17437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.775930362606321,
+      "learning_rate": 1.8337273890083994e-06,
+      "loss": 0.4606,
+      "step": 17438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8391544483318922,
+      "learning_rate": 1.8333498951741952e-06,
+      "loss": 0.4986,
+      "step": 17439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0457210469870515,
+      "learning_rate": 1.8329724780990315e-06,
+      "loss": 0.6568,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3958193618259738,
+      "learning_rate": 1.8325951377908119e-06,
+      "loss": 0.3471,
+      "step": 17441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.544258442749683,
+      "learning_rate": 1.8322178742574362e-06,
+      "loss": 0.4395,
+      "step": 17442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7913429339156157,
+      "learning_rate": 1.8318406875068e-06,
+      "loss": 0.4383,
+      "step": 17443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9281993509418645,
+      "learning_rate": 1.831463577546801e-06,
+      "loss": 0.3937,
+      "step": 17444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6175243993013915,
+      "learning_rate": 1.8310865443853344e-06,
+      "loss": 0.4507,
+      "step": 17445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.523470532710789,
+      "learning_rate": 1.8307095880302905e-06,
+      "loss": 0.4776,
+      "step": 17446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8913860310303146,
+      "learning_rate": 1.8303327084895623e-06,
+      "loss": 0.4995,
+      "step": 17447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.685780175283094,
+      "learning_rate": 1.8299559057710398e-06,
+      "loss": 0.4556,
+      "step": 17448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.086486079757472,
+      "learning_rate": 1.8295791798826123e-06,
+      "loss": 0.5012,
+      "step": 17449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6515440868468132,
+      "learning_rate": 1.8292025308321638e-06,
+      "loss": 0.3606,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.4903833568096161,
+      "eval_runtime": 67.9937,
+      "eval_samples_per_second": 50.152,
+      "eval_steps_per_second": 1.574,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9307645963629758,
+      "learning_rate": 1.8288259586275797e-06,
+      "loss": 0.4996,
+      "step": 17451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8915608030933562,
+      "learning_rate": 1.8284494632767457e-06,
+      "loss": 0.3939,
+      "step": 17452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.509620661745689,
+      "learning_rate": 1.82807304478754e-06,
+      "loss": 0.4146,
+      "step": 17453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.446887906417663,
+      "learning_rate": 1.8276967031678443e-06,
+      "loss": 0.3267,
+      "step": 17454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.867624700406405,
+      "learning_rate": 1.8273204384255372e-06,
+      "loss": 0.576,
+      "step": 17455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8069008034489817,
+      "learning_rate": 1.8269442505684962e-06,
+      "loss": 0.3509,
+      "step": 17456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0880637655731453,
+      "learning_rate": 1.8265681396045953e-06,
+      "loss": 0.6117,
+      "step": 17457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6948552129206491,
+      "learning_rate": 1.82619210554171e-06,
+      "loss": 0.4051,
+      "step": 17458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.058876235398031,
+      "learning_rate": 1.8258161483877101e-06,
+      "loss": 0.447,
+      "step": 17459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5329236268012536,
+      "learning_rate": 1.8254402681504668e-06,
+      "loss": 0.5832,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4381867898585583,
+      "learning_rate": 1.8250644648378506e-06,
+      "loss": 0.431,
+      "step": 17461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8491393023890368,
+      "learning_rate": 1.8246887384577261e-06,
+      "loss": 0.5486,
+      "step": 17462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5708695285207217,
+      "learning_rate": 1.8243130890179596e-06,
+      "loss": 0.3876,
+      "step": 17463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5944266301236945,
+      "learning_rate": 1.8239375165264172e-06,
+      "loss": 0.3844,
+      "step": 17464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.549102628580082,
+      "learning_rate": 1.823562020990958e-06,
+      "loss": 0.3977,
+      "step": 17465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5656555810069244,
+      "learning_rate": 1.8231866024194444e-06,
+      "loss": 0.3875,
+      "step": 17466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5976128383456822,
+      "learning_rate": 1.822811260819737e-06,
+      "loss": 0.5054,
+      "step": 17467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7562967834897134,
+      "learning_rate": 1.8224359961996906e-06,
+      "loss": 0.599,
+      "step": 17468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8237039354765439,
+      "learning_rate": 1.822060808567162e-06,
+      "loss": 0.5201,
+      "step": 17469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9920739890069465,
+      "learning_rate": 1.8216856979300063e-06,
+      "loss": 0.3742,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5495980263186182,
+      "learning_rate": 1.8213106642960772e-06,
+      "loss": 0.4301,
+      "step": 17471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8792222753268746,
+      "learning_rate": 1.820935707673223e-06,
+      "loss": 0.3867,
+      "step": 17472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1894287257469704,
+      "learning_rate": 1.820560828069295e-06,
+      "loss": 0.6294,
+      "step": 17473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9746202897765979,
+      "learning_rate": 1.8201860254921412e-06,
+      "loss": 0.5562,
+      "step": 17474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4332598477483494,
+      "learning_rate": 1.8198112999496062e-06,
+      "loss": 0.2784,
+      "step": 17475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1393405233590226,
+      "learning_rate": 1.819436651449536e-06,
+      "loss": 0.5577,
+      "step": 17476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8353931870352898,
+      "learning_rate": 1.8190620799997749e-06,
+      "loss": 0.4463,
+      "step": 17477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5995058264452924,
+      "learning_rate": 1.818687585608161e-06,
+      "loss": 0.5043,
+      "step": 17478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7911818204230807,
+      "learning_rate": 1.8183131682825367e-06,
+      "loss": 0.4792,
+      "step": 17479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6178782640925478,
+      "learning_rate": 1.8179388280307397e-06,
+      "loss": 0.3838,
+      "step": 17480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8418470976260484,
+      "learning_rate": 1.817564564860606e-06,
+      "loss": 0.5888,
+      "step": 17481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7880460829812281,
+      "learning_rate": 1.8171903787799706e-06,
+      "loss": 0.5225,
+      "step": 17482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.715049753938635,
+      "learning_rate": 1.8168162697966668e-06,
+      "loss": 0.4397,
+      "step": 17483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8298474580998236,
+      "learning_rate": 1.8164422379185288e-06,
+      "loss": 0.5296,
+      "step": 17484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7569210777304398,
+      "learning_rate": 1.8160682831533823e-06,
+      "loss": 0.4918,
+      "step": 17485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7358066051701502,
+      "learning_rate": 1.8156944055090587e-06,
+      "loss": 0.5441,
+      "step": 17486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5182216241337638,
+      "learning_rate": 1.8153206049933858e-06,
+      "loss": 0.4036,
+      "step": 17487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.041027748974636,
+      "learning_rate": 1.8149468816141858e-06,
+      "loss": 0.5511,
+      "step": 17488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0655765134432746,
+      "learning_rate": 1.814573235379284e-06,
+      "loss": 0.5025,
+      "step": 17489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8797400523973127,
+      "learning_rate": 1.8141996662965034e-06,
+      "loss": 0.4075,
+      "step": 17490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6667176079163637,
+      "learning_rate": 1.8138261743736624e-06,
+      "loss": 0.4363,
+      "step": 17491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.418987352666089,
+      "learning_rate": 1.8134527596185813e-06,
+      "loss": 0.3684,
+      "step": 17492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5512058453884459,
+      "learning_rate": 1.8130794220390774e-06,
+      "loss": 0.3518,
+      "step": 17493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6904470918096497,
+      "learning_rate": 1.812706161642965e-06,
+      "loss": 0.3944,
+      "step": 17494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3269321930630211,
+      "learning_rate": 1.8123329784380583e-06,
+      "loss": 0.3035,
+      "step": 17495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5009125661536968,
+      "learning_rate": 1.8119598724321711e-06,
+      "loss": 0.378,
+      "step": 17496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9338354726697935,
+      "learning_rate": 1.8115868436331135e-06,
+      "loss": 0.5616,
+      "step": 17497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1950447911352078,
+      "learning_rate": 1.8112138920486937e-06,
+      "loss": 0.3611,
+      "step": 17498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6587894753289778,
+      "learning_rate": 1.8108410176867198e-06,
+      "loss": 0.4924,
+      "step": 17499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.446084751259896,
+      "learning_rate": 1.8104682205549993e-06,
+      "loss": 0.3641,
+      "step": 17500
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.49008381366729736,
+      "eval_runtime": 67.9637,
+      "eval_samples_per_second": 50.174,
+      "eval_steps_per_second": 1.574,
+      "step": 17500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8264078200995757,
+      "learning_rate": 1.8100955006613335e-06,
+      "loss": 0.5641,
+      "step": 17501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5286308299056734,
+      "learning_rate": 1.8097228580135264e-06,
+      "loss": 0.3802,
+      "step": 17502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7429793638155595,
+      "learning_rate": 1.8093502926193804e-06,
+      "loss": 0.5582,
+      "step": 17503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6845089595235463,
+      "learning_rate": 1.8089778044866913e-06,
+      "loss": 0.4259,
+      "step": 17504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8225058253751687,
+      "learning_rate": 1.8086053936232603e-06,
+      "loss": 0.4223,
+      "step": 17505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.609059544190087,
+      "learning_rate": 1.8082330600368836e-06,
+      "loss": 0.5046,
+      "step": 17506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2609959948549534,
+      "learning_rate": 1.807860803735354e-06,
+      "loss": 0.4365,
+      "step": 17507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4274363640367984,
+      "learning_rate": 1.8074886247264655e-06,
+      "loss": 0.3969,
+      "step": 17508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8116544571463813,
+      "learning_rate": 1.80711652301801e-06,
+      "loss": 0.5957,
+      "step": 17509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.277344413903061,
+      "learning_rate": 1.8067444986177749e-06,
+      "loss": 0.5811,
+      "step": 17510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.806924525940317,
+      "learning_rate": 1.8063725515335495e-06,
+      "loss": 0.5009,
+      "step": 17511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.388348666212769,
+      "learning_rate": 1.8060006817731227e-06,
+      "loss": 0.5691,
+      "step": 17512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9880511517027835,
+      "learning_rate": 1.8056288893442753e-06,
+      "loss": 0.5523,
+      "step": 17513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0032634269396765,
+      "learning_rate": 1.8052571742547926e-06,
+      "loss": 0.4992,
+      "step": 17514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5038659327837685,
+      "learning_rate": 1.8048855365124557e-06,
+      "loss": 0.4522,
+      "step": 17515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1139933645898332,
+      "learning_rate": 1.8045139761250468e-06,
+      "loss": 0.4947,
+      "step": 17516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.154942751099358,
+      "learning_rate": 1.8041424931003413e-06,
+      "loss": 0.5845,
+      "step": 17517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5700077508252868,
+      "learning_rate": 1.8037710874461168e-06,
+      "loss": 0.4774,
+      "step": 17518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7010896028823634,
+      "learning_rate": 1.8033997591701497e-06,
+      "loss": 0.4532,
+      "step": 17519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.885591033078797,
+      "learning_rate": 1.8030285082802116e-06,
+      "loss": 0.4212,
+      "step": 17520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.362863015669182,
+      "learning_rate": 1.8026573347840759e-06,
+      "loss": 0.3304,
+      "step": 17521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0147687921834536,
+      "learning_rate": 1.802286238689513e-06,
+      "loss": 0.4594,
+      "step": 17522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.803639668123008,
+      "learning_rate": 1.8019152200042898e-06,
+      "loss": 0.5124,
+      "step": 17523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8145411970904877,
+      "learning_rate": 1.8015442787361747e-06,
+      "loss": 0.5537,
+      "step": 17524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4074407889043352,
+      "learning_rate": 1.8011734148929338e-06,
+      "loss": 0.4135,
+      "step": 17525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6201285393347655,
+      "learning_rate": 1.8008026284823293e-06,
+      "loss": 0.41,
+      "step": 17526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9644674223716718,
+      "learning_rate": 1.800431919512124e-06,
+      "loss": 0.5367,
+      "step": 17527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6848950713179156,
+      "learning_rate": 1.8000612879900788e-06,
+      "loss": 0.5395,
+      "step": 17528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.661818241074193,
+      "learning_rate": 1.7996907339239527e-06,
+      "loss": 0.3957,
+      "step": 17529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8483400182878338,
+      "learning_rate": 1.7993202573215023e-06,
+      "loss": 0.4624,
+      "step": 17530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1385071494970602,
+      "learning_rate": 1.798949858190484e-06,
+      "loss": 0.485,
+      "step": 17531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3442663928942344,
+      "learning_rate": 1.798579536538652e-06,
+      "loss": 0.2813,
+      "step": 17532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9089264023884196,
+      "learning_rate": 1.7982092923737582e-06,
+      "loss": 0.5031,
+      "step": 17533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2981993808032677,
+      "learning_rate": 1.7978391257035529e-06,
+      "loss": 0.2532,
+      "step": 17534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6381781883498379,
+      "learning_rate": 1.7974690365357872e-06,
+      "loss": 0.3666,
+      "step": 17535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6331366766650874,
+      "learning_rate": 1.7970990248782067e-06,
+      "loss": 0.3733,
+      "step": 17536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8644368615278573,
+      "learning_rate": 1.7967290907385574e-06,
+      "loss": 0.4849,
+      "step": 17537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9866560673839795,
+      "learning_rate": 1.796359234124586e-06,
+      "loss": 0.3958,
+      "step": 17538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6513841033017163,
+      "learning_rate": 1.7959894550440326e-06,
+      "loss": 0.3883,
+      "step": 17539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5284605718233406,
+      "learning_rate": 1.7956197535046391e-06,
+      "loss": 0.5711,
+      "step": 17540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1262950359555197,
+      "learning_rate": 1.795250129514145e-06,
+      "loss": 0.7372,
+      "step": 17541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8676243562091832,
+      "learning_rate": 1.7948805830802898e-06,
+      "loss": 0.5292,
+      "step": 17542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6418986671854114,
+      "learning_rate": 1.7945111142108075e-06,
+      "loss": 0.5383,
+      "step": 17543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5345174497582792,
+      "learning_rate": 1.7941417229134328e-06,
+      "loss": 0.4213,
+      "step": 17544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9758020660258626,
+      "learning_rate": 1.7937724091959008e-06,
+      "loss": 0.5739,
+      "step": 17545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.788054439778776,
+      "learning_rate": 1.79340317306594e-06,
+      "loss": 0.552,
+      "step": 17546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5323614269989907,
+      "learning_rate": 1.7930340145312817e-06,
+      "loss": 0.5376,
+      "step": 17547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.839475134779638,
+      "learning_rate": 1.7926649335996548e-06,
+      "loss": 0.5119,
+      "step": 17548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5101270957561632,
+      "learning_rate": 1.792295930278784e-06,
+      "loss": 0.3524,
+      "step": 17549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.198553998666507,
+      "learning_rate": 1.7919270045763948e-06,
+      "loss": 0.4834,
+      "step": 17550
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.49013519287109375,
+      "eval_runtime": 67.9659,
+      "eval_samples_per_second": 50.172,
+      "eval_steps_per_second": 1.574,
+      "step": 17550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7067448227195774,
+      "learning_rate": 1.7915581565002117e-06,
+      "loss": 0.4566,
+      "step": 17551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5469155560456538,
+      "learning_rate": 1.7911893860579537e-06,
+      "loss": 0.3533,
+      "step": 17552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6336110744917192,
+      "learning_rate": 1.7908206932573418e-06,
+      "loss": 0.3828,
+      "step": 17553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5363358837486534,
+      "learning_rate": 1.7904520781060969e-06,
+      "loss": 0.4144,
+      "step": 17554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7123056493073485,
+      "learning_rate": 1.7900835406119326e-06,
+      "loss": 0.473,
+      "step": 17555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8053463731291224,
+      "learning_rate": 1.7897150807825647e-06,
+      "loss": 0.4345,
+      "step": 17556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.454586733295852,
+      "learning_rate": 1.789346698625708e-06,
+      "loss": 0.4283,
+      "step": 17557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1597352793660365,
+      "learning_rate": 1.7889783941490728e-06,
+      "loss": 0.4936,
+      "step": 17558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8998451939183918,
+      "learning_rate": 1.7886101673603698e-06,
+      "loss": 0.5147,
+      "step": 17559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5190257967280927,
+      "learning_rate": 1.7882420182673086e-06,
+      "loss": 0.3621,
+      "step": 17560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.481600134630902,
+      "learning_rate": 1.7878739468775945e-06,
+      "loss": 0.3623,
+      "step": 17561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.69493543158906,
+      "learning_rate": 1.787505953198933e-06,
+      "loss": 0.5752,
+      "step": 17562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7252163732807908,
+      "learning_rate": 1.7871380372390293e-06,
+      "loss": 0.4692,
+      "step": 17563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4849533874502507,
+      "learning_rate": 1.7867701990055853e-06,
+      "loss": 0.4321,
+      "step": 17564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6712699797167825,
+      "learning_rate": 1.7864024385062995e-06,
+      "loss": 0.4153,
+      "step": 17565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3337522090520426,
+      "learning_rate": 1.7860347557488725e-06,
+      "loss": 0.4597,
+      "step": 17566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8116902290290169,
+      "learning_rate": 1.7856671507410022e-06,
+      "loss": 0.5281,
+      "step": 17567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8700063787017904,
+      "learning_rate": 1.7852996234903813e-06,
+      "loss": 0.4538,
+      "step": 17568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7770892605395365,
+      "learning_rate": 1.7849321740047066e-06,
+      "loss": 0.4984,
+      "step": 17569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4818432533356283,
+      "learning_rate": 1.7845648022916695e-06,
+      "loss": 0.4098,
+      "step": 17570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7344245207979299,
+      "learning_rate": 1.7841975083589597e-06,
+      "loss": 0.5281,
+      "step": 17571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6695599050464471,
+      "learning_rate": 1.7838302922142669e-06,
+      "loss": 0.4054,
+      "step": 17572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7945677773637039,
+      "learning_rate": 1.7834631538652802e-06,
+      "loss": 0.4759,
+      "step": 17573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7783429136660605,
+      "learning_rate": 1.7830960933196823e-06,
+      "loss": 0.4199,
+      "step": 17574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7921408741907452,
+      "learning_rate": 1.7827291105851598e-06,
+      "loss": 0.383,
+      "step": 17575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.910379669046273,
+      "learning_rate": 1.782362205669394e-06,
+      "loss": 0.404,
+      "step": 17576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9358170034733408,
+      "learning_rate": 1.7819953785800674e-06,
+      "loss": 0.5943,
+      "step": 17577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4780307708071903,
+      "learning_rate": 1.7816286293248568e-06,
+      "loss": 0.3842,
+      "step": 17578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7981742393519202,
+      "learning_rate": 1.7812619579114421e-06,
+      "loss": 0.4577,
+      "step": 17579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5466628331633125,
+      "learning_rate": 1.7808953643474992e-06,
+      "loss": 0.5321,
+      "step": 17580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6802224734544915,
+      "learning_rate": 1.7805288486407008e-06,
+      "loss": 0.4734,
+      "step": 17581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7931477229319348,
+      "learning_rate": 1.7801624107987202e-06,
+      "loss": 0.487,
+      "step": 17582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.028226542957065,
+      "learning_rate": 1.779796050829231e-06,
+      "loss": 0.5451,
+      "step": 17583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8033963083224038,
+      "learning_rate": 1.779429768739899e-06,
+      "loss": 0.4911,
+      "step": 17584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7484711114893252,
+      "learning_rate": 1.779063564538394e-06,
+      "loss": 0.5003,
+      "step": 17585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.688548315425474,
+      "learning_rate": 1.7786974382323835e-06,
+      "loss": 0.4442,
+      "step": 17586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7409935891140655,
+      "learning_rate": 1.7783313898295297e-06,
+      "loss": 0.4842,
+      "step": 17587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5754374916709482,
+      "learning_rate": 1.7779654193374962e-06,
+      "loss": 0.6777,
+      "step": 17588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1834262416576316,
+      "learning_rate": 1.777599526763945e-06,
+      "loss": 0.6837,
+      "step": 17589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9928081188888416,
+      "learning_rate": 1.7772337121165367e-06,
+      "loss": 0.5228,
+      "step": 17590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.516927223091929,
+      "learning_rate": 1.7768679754029273e-06,
+      "loss": 0.4466,
+      "step": 17591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8611190125445747,
+      "learning_rate": 1.7765023166307744e-06,
+      "loss": 0.511,
+      "step": 17592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.78888372054183,
+      "learning_rate": 1.7761367358077342e-06,
+      "loss": 0.5442,
+      "step": 17593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1094479505373966,
+      "learning_rate": 1.7757712329414564e-06,
+      "loss": 0.5149,
+      "step": 17594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9518649601207365,
+      "learning_rate": 1.7754058080395953e-06,
+      "loss": 0.5262,
+      "step": 17595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.786874853147806,
+      "learning_rate": 1.775040461109801e-06,
+      "loss": 0.7386,
+      "step": 17596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.229041422027444,
+      "learning_rate": 1.7746751921597199e-06,
+      "loss": 0.5497,
+      "step": 17597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9592028537997177,
+      "learning_rate": 1.774310001197e-06,
+      "loss": 0.5125,
+      "step": 17598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5358933938295745,
+      "learning_rate": 1.7739448882292876e-06,
+      "loss": 0.4132,
+      "step": 17599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6935434040912123,
+      "learning_rate": 1.7735798532642225e-06,
+      "loss": 0.4165,
+      "step": 17600
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.4896329939365387,
+      "eval_runtime": 68.0258,
+      "eval_samples_per_second": 50.128,
+      "eval_steps_per_second": 1.573,
+      "step": 17600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.272608898399909,
+      "learning_rate": 1.7732148963094491e-06,
+      "loss": 0.5412,
+      "step": 17601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7600405876915384,
+      "learning_rate": 1.772850017372607e-06,
+      "loss": 0.6027,
+      "step": 17602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7934832868706285,
+      "learning_rate": 1.772485216461335e-06,
+      "loss": 0.5387,
+      "step": 17603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6209598211417728,
+      "learning_rate": 1.7721204935832697e-06,
+      "loss": 0.5296,
+      "step": 17604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8949004104352214,
+      "learning_rate": 1.7717558487460474e-06,
+      "loss": 0.4413,
+      "step": 17605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4209158205405368,
+      "learning_rate": 1.7713912819572998e-06,
+      "loss": 0.3112,
+      "step": 17606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9195827668350927,
+      "learning_rate": 1.7710267932246591e-06,
+      "loss": 0.4955,
+      "step": 17607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.600951488414876,
+      "learning_rate": 1.7706623825557573e-06,
+      "loss": 0.3159,
+      "step": 17608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.282056401058003,
+      "learning_rate": 1.7702980499582226e-06,
+      "loss": 0.7592,
+      "step": 17609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6136047029788974,
+      "learning_rate": 1.7699337954396805e-06,
+      "loss": 0.4573,
+      "step": 17610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9682195629365948,
+      "learning_rate": 1.7695696190077577e-06,
+      "loss": 0.4796,
+      "step": 17611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0084319420550134,
+      "learning_rate": 1.7692055206700791e-06,
+      "loss": 0.6454,
+      "step": 17612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.031172927749069,
+      "learning_rate": 1.7688415004342646e-06,
+      "loss": 0.4694,
+      "step": 17613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8280230558263053,
+      "learning_rate": 1.7684775583079357e-06,
+      "loss": 0.5288,
+      "step": 17614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1487119964114685,
+      "learning_rate": 1.768113694298712e-06,
+      "loss": 0.5591,
+      "step": 17615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.018463387992053,
+      "learning_rate": 1.76774990841421e-06,
+      "loss": 0.5046,
+      "step": 17616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8691625688043245,
+      "learning_rate": 1.767386200662045e-06,
+      "loss": 0.4409,
+      "step": 17617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.600733171974171,
+      "learning_rate": 1.7670225710498322e-06,
+      "loss": 0.4371,
+      "step": 17618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5117089979074352,
+      "learning_rate": 1.7666590195851823e-06,
+      "loss": 0.3901,
+      "step": 17619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0780579612157046,
+      "learning_rate": 1.7662955462757076e-06,
+      "loss": 0.535,
+      "step": 17620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.269341175308383,
+      "learning_rate": 1.765932151129016e-06,
+      "loss": 0.4337,
+      "step": 17621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8681973444178275,
+      "learning_rate": 1.7655688341527166e-06,
+      "loss": 0.4926,
+      "step": 17622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8389658200090566,
+      "learning_rate": 1.765205595354413e-06,
+      "loss": 0.4456,
+      "step": 17623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5576224527222975,
+      "learning_rate": 1.7648424347417105e-06,
+      "loss": 0.4123,
+      "step": 17624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5404868867870134,
+      "learning_rate": 1.764479352322213e-06,
+      "loss": 0.4931,
+      "step": 17625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8984745099686902,
+      "learning_rate": 1.7641163481035188e-06,
+      "loss": 0.4663,
+      "step": 17626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6290820378266118,
+      "learning_rate": 1.7637534220932284e-06,
+      "loss": 0.46,
+      "step": 17627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.38537367964091,
+      "learning_rate": 1.7633905742989404e-06,
+      "loss": 0.5604,
+      "step": 17628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9308445157610334,
+      "learning_rate": 1.7630278047282484e-06,
+      "loss": 0.5248,
+      "step": 17629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.986407572747295,
+      "learning_rate": 1.762665113388749e-06,
+      "loss": 0.4181,
+      "step": 17630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8619891292657016,
+      "learning_rate": 1.7623025002880348e-06,
+      "loss": 0.4462,
+      "step": 17631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8488939863968634,
+      "learning_rate": 1.7619399654336949e-06,
+      "loss": 0.4998,
+      "step": 17632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.061691944442164,
+      "learning_rate": 1.7615775088333203e-06,
+      "loss": 0.5745,
+      "step": 17633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9641059215565595,
+      "learning_rate": 1.761215130494499e-06,
+      "loss": 0.5276,
+      "step": 17634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9317482718326981,
+      "learning_rate": 1.7608528304248175e-06,
+      "loss": 0.4416,
+      "step": 17635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7102989136420577,
+      "learning_rate": 1.760490608631858e-06,
+      "loss": 0.4542,
+      "step": 17636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5874068724993098,
+      "learning_rate": 1.7601284651232049e-06,
+      "loss": 0.438,
+      "step": 17637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7487116965628038,
+      "learning_rate": 1.7597663999064407e-06,
+      "loss": 0.4672,
+      "step": 17638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6561010936359186,
+      "learning_rate": 1.759404412989143e-06,
+      "loss": 0.5206,
+      "step": 17639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8470617926198727,
+      "learning_rate": 1.75904250437889e-06,
+      "loss": 0.4781,
+      "step": 17640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0319806937679568,
+      "learning_rate": 1.7586806740832601e-06,
+      "loss": 0.4724,
+      "step": 17641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7761777547875457,
+      "learning_rate": 1.7583189221098256e-06,
+      "loss": 0.4369,
+      "step": 17642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8892646228955559,
+      "learning_rate": 1.75795724846616e-06,
+      "loss": 0.5829,
+      "step": 17643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8970971322637056,
+      "learning_rate": 1.7575956531598359e-06,
+      "loss": 0.4202,
+      "step": 17644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.768375475907685,
+      "learning_rate": 1.757234136198422e-06,
+      "loss": 0.4467,
+      "step": 17645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.653898889093113,
+      "learning_rate": 1.7568726975894863e-06,
+      "loss": 0.329,
+      "step": 17646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7273031647543706,
+      "learning_rate": 1.7565113373405957e-06,
+      "loss": 0.5388,
+      "step": 17647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0956797032252297,
+      "learning_rate": 1.7561500554593163e-06,
+      "loss": 0.5467,
+      "step": 17648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9640954906002097,
+      "learning_rate": 1.755788851953209e-06,
+      "loss": 0.5746,
+      "step": 17649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2865142575579,
+      "learning_rate": 1.7554277268298364e-06,
+      "loss": 0.4367,
+      "step": 17650
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.48940569162368774,
+      "eval_runtime": 67.9436,
+      "eval_samples_per_second": 50.189,
+      "eval_steps_per_second": 1.575,
+      "step": 17650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.908960712744336,
+      "learning_rate": 1.7550666800967595e-06,
+      "loss": 0.527,
+      "step": 17651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6833313459061485,
+      "learning_rate": 1.7547057117615334e-06,
+      "loss": 0.3138,
+      "step": 17652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.091213267909794,
+      "learning_rate": 1.7543448218317194e-06,
+      "loss": 0.4701,
+      "step": 17653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8082781283642317,
+      "learning_rate": 1.7539840103148687e-06,
+      "loss": 0.3864,
+      "step": 17654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7250960494487133,
+      "learning_rate": 1.7536232772185363e-06,
+      "loss": 0.4985,
+      "step": 17655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7650046950884122,
+      "learning_rate": 1.7532626225502736e-06,
+      "loss": 0.5264,
+      "step": 17656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.554718727491108,
+      "learning_rate": 1.7529020463176316e-06,
+      "loss": 0.4395,
+      "step": 17657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.942415765705716,
+      "learning_rate": 1.752541548528157e-06,
+      "loss": 0.4604,
+      "step": 17658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5094224972926684,
+      "learning_rate": 1.752181129189397e-06,
+      "loss": 0.5807,
+      "step": 17659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7260326887904018,
+      "learning_rate": 1.7518207883088986e-06,
+      "loss": 0.5809,
+      "step": 17660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6379146926414074,
+      "learning_rate": 1.751460525894203e-06,
+      "loss": 0.4938,
+      "step": 17661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0442361408425302,
+      "learning_rate": 1.7511003419528527e-06,
+      "loss": 0.4936,
+      "step": 17662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0713682530222353,
+      "learning_rate": 1.7507402364923892e-06,
+      "loss": 0.5271,
+      "step": 17663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.716785302351284,
+      "learning_rate": 1.7503802095203492e-06,
+      "loss": 0.4744,
+      "step": 17664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.837566143937413,
+      "learning_rate": 1.7500202610442705e-06,
+      "loss": 0.6696,
+      "step": 17665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6197354041619438,
+      "learning_rate": 1.7496603910716896e-06,
+      "loss": 0.3218,
+      "step": 17666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5878508431862446,
+      "learning_rate": 1.749300599610138e-06,
+      "loss": 0.4643,
+      "step": 17667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6390627170772465,
+      "learning_rate": 1.7489408866671481e-06,
+      "loss": 0.4708,
+      "step": 17668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.559343035801497,
+      "learning_rate": 1.7485812522502515e-06,
+      "loss": 0.415,
+      "step": 17669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5268710710675093,
+      "learning_rate": 1.7482216963669774e-06,
+      "loss": 0.3946,
+      "step": 17670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8309477363465183,
+      "learning_rate": 1.7478622190248503e-06,
+      "loss": 0.4828,
+      "step": 17671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8818911176349058,
+      "learning_rate": 1.7475028202313976e-06,
+      "loss": 0.4652,
+      "step": 17672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7284817618153447,
+      "learning_rate": 1.7471434999941434e-06,
+      "loss": 0.3834,
+      "step": 17673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8637077237792068,
+      "learning_rate": 1.7467842583206073e-06,
+      "loss": 0.3747,
+      "step": 17674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8783397270102737,
+      "learning_rate": 1.7464250952183124e-06,
+      "loss": 0.4574,
+      "step": 17675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8933155366398147,
+      "learning_rate": 1.7460660106947773e-06,
+      "loss": 0.4951,
+      "step": 17676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6022677979354207,
+      "learning_rate": 1.7457070047575184e-06,
+      "loss": 0.4176,
+      "step": 17677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.235374837164074,
+      "learning_rate": 1.7453480774140507e-06,
+      "loss": 0.6437,
+      "step": 17678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7515027161713692,
+      "learning_rate": 1.7449892286718904e-06,
+      "loss": 0.4016,
+      "step": 17679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6088873890464677,
+      "learning_rate": 1.7446304585385465e-06,
+      "loss": 0.436,
+      "step": 17680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.008257285770542,
+      "learning_rate": 1.744271767021532e-06,
+      "loss": 0.4438,
+      "step": 17681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5523394876369376,
+      "learning_rate": 1.743913154128355e-06,
+      "loss": 0.4214,
+      "step": 17682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.922918837440199,
+      "learning_rate": 1.7435546198665246e-06,
+      "loss": 0.5381,
+      "step": 17683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.315259208081055,
+      "learning_rate": 1.7431961642435435e-06,
+      "loss": 0.4202,
+      "step": 17684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4613349484506504,
+      "learning_rate": 1.7428377872669173e-06,
+      "loss": 0.3603,
+      "step": 17685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6752076119107728,
+      "learning_rate": 1.7424794889441497e-06,
+      "loss": 0.5404,
+      "step": 17686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.729513350051952,
+      "learning_rate": 1.7421212692827387e-06,
+      "loss": 0.571,
+      "step": 17687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6651003231140924,
+      "learning_rate": 1.7417631282901853e-06,
+      "loss": 0.4195,
+      "step": 17688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6156791663290775,
+      "learning_rate": 1.7414050659739873e-06,
+      "loss": 0.4946,
+      "step": 17689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7229800168728737,
+      "learning_rate": 1.7410470823416387e-06,
+      "loss": 0.511,
+      "step": 17690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.762909474562556,
+      "learning_rate": 1.7406891774006347e-06,
+      "loss": 0.444,
+      "step": 17691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3452669885960455,
+      "learning_rate": 1.7403313511584674e-06,
+      "loss": 0.3548,
+      "step": 17692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0767168968565017,
+      "learning_rate": 1.7399736036226295e-06,
+      "loss": 0.5653,
+      "step": 17693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0219741629552064,
+      "learning_rate": 1.7396159348006075e-06,
+      "loss": 0.5509,
+      "step": 17694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6105768268560712,
+      "learning_rate": 1.73925834469989e-06,
+      "loss": 0.3259,
+      "step": 17695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8441132523231292,
+      "learning_rate": 1.738900833327965e-06,
+      "loss": 0.3799,
+      "step": 17696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.553453622888933,
+      "learning_rate": 1.738543400692313e-06,
+      "loss": 0.4531,
+      "step": 17697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8355751651884036,
+      "learning_rate": 1.7381860468004194e-06,
+      "loss": 0.5679,
+      "step": 17698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8323638990498015,
+      "learning_rate": 1.7378287716597647e-06,
+      "loss": 0.4324,
+      "step": 17699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8291580764886137,
+      "learning_rate": 1.7374715752778277e-06,
+      "loss": 0.5959,
+      "step": 17700
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.4890903830528259,
+      "eval_runtime": 68.0032,
+      "eval_samples_per_second": 50.145,
+      "eval_steps_per_second": 1.573,
+      "step": 17700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9093455761568188,
+      "learning_rate": 1.7371144576620847e-06,
+      "loss": 0.503,
+      "step": 17701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1248602941472035,
+      "learning_rate": 1.7367574188200149e-06,
+      "loss": 0.5907,
+      "step": 17702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.016516256462269,
+      "learning_rate": 1.7364004587590904e-06,
+      "loss": 0.463,
+      "step": 17703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8089681834350182,
+      "learning_rate": 1.7360435774867854e-06,
+      "loss": 0.4318,
+      "step": 17704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8066736669767496,
+      "learning_rate": 1.7356867750105707e-06,
+      "loss": 0.5794,
+      "step": 17705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3930448233351358,
+      "learning_rate": 1.735330051337914e-06,
+      "loss": 0.3452,
+      "step": 17706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5547721170207733,
+      "learning_rate": 1.7349734064762847e-06,
+      "loss": 0.4504,
+      "step": 17707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.066499679345603,
+      "learning_rate": 1.7346168404331493e-06,
+      "loss": 0.5113,
+      "step": 17708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.173653669514861,
+      "learning_rate": 1.7342603532159707e-06,
+      "loss": 0.441,
+      "step": 17709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9715728545280482,
+      "learning_rate": 1.7339039448322128e-06,
+      "loss": 0.3828,
+      "step": 17710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.724481857732103,
+      "learning_rate": 1.7335476152893372e-06,
+      "loss": 0.4837,
+      "step": 17711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.995136020426555,
+      "learning_rate": 1.733191364594802e-06,
+      "loss": 0.5862,
+      "step": 17712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4435226652119548,
+      "learning_rate": 1.732835192756066e-06,
+      "loss": 0.3138,
+      "step": 17713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0737015688464635,
+      "learning_rate": 1.7324790997805858e-06,
+      "loss": 0.4761,
+      "step": 17714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8590855845688115,
+      "learning_rate": 1.7321230856758161e-06,
+      "loss": 0.5284,
+      "step": 17715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7854171857454735,
+      "learning_rate": 1.7317671504492084e-06,
+      "loss": 0.4982,
+      "step": 17716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6211244671713592,
+      "learning_rate": 1.7314112941082148e-06,
+      "loss": 0.4996,
+      "step": 17717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0384022678666107,
+      "learning_rate": 1.7310555166602861e-06,
+      "loss": 0.4393,
+      "step": 17718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4882002151826852,
+      "learning_rate": 1.7306998181128675e-06,
+      "loss": 0.448,
+      "step": 17719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.431680227435781,
+      "learning_rate": 1.7303441984734082e-06,
+      "loss": 0.461,
+      "step": 17720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1768794341062705,
+      "learning_rate": 1.7299886577493518e-06,
+      "loss": 0.7022,
+      "step": 17721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5242565326864743,
+      "learning_rate": 1.7296331959481405e-06,
+      "loss": 0.6451,
+      "step": 17722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8999587730959355,
+      "learning_rate": 1.7292778130772164e-06,
+      "loss": 0.4115,
+      "step": 17723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8517829323350086,
+      "learning_rate": 1.7289225091440205e-06,
+      "loss": 0.5576,
+      "step": 17724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6667815432684514,
+      "learning_rate": 1.728567284155988e-06,
+      "loss": 0.4805,
+      "step": 17725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7305184480109113,
+      "learning_rate": 1.728212138120557e-06,
+      "loss": 0.5332,
+      "step": 17726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.506585376406294,
+      "learning_rate": 1.7278570710451623e-06,
+      "loss": 0.4652,
+      "step": 17727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8650026726536375,
+      "learning_rate": 1.7275020829372376e-06,
+      "loss": 0.4634,
+      "step": 17728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.804004142174372,
+      "learning_rate": 1.7271471738042128e-06,
+      "loss": 0.567,
+      "step": 17729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2131254259220974,
+      "learning_rate": 1.7267923436535185e-06,
+      "loss": 0.4456,
+      "step": 17730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8924349033926682,
+      "learning_rate": 1.726437592492584e-06,
+      "loss": 0.4147,
+      "step": 17731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8311744691688592,
+      "learning_rate": 1.726082920328834e-06,
+      "loss": 0.4574,
+      "step": 17732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5980419776969428,
+      "learning_rate": 1.725728327169693e-06,
+      "loss": 0.4025,
+      "step": 17733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5632670695444248,
+      "learning_rate": 1.7253738130225872e-06,
+      "loss": 0.402,
+      "step": 17734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6035969840395161,
+      "learning_rate": 1.7250193778949342e-06,
+      "loss": 0.4794,
+      "step": 17735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1295881574487607,
+      "learning_rate": 1.7246650217941563e-06,
+      "loss": 0.4667,
+      "step": 17736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7512056562241585,
+      "learning_rate": 1.7243107447276724e-06,
+      "loss": 0.537,
+      "step": 17737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.084393135313723,
+      "learning_rate": 1.7239565467028966e-06,
+      "loss": 0.4805,
+      "step": 17738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.543543667608268,
+      "learning_rate": 1.723602427727245e-06,
+      "loss": 0.3971,
+      "step": 17739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5446305258704958,
+      "learning_rate": 1.7232483878081312e-06,
+      "loss": 0.4203,
+      "step": 17740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.846541650311322,
+      "learning_rate": 1.7228944269529673e-06,
+      "loss": 0.4539,
+      "step": 17741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7323009064847212,
+      "learning_rate": 1.7225405451691623e-06,
+      "loss": 0.4797,
+      "step": 17742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.501723023464921,
+      "learning_rate": 1.7221867424641243e-06,
+      "loss": 0.3308,
+      "step": 17743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7920191787593374,
+      "learning_rate": 1.721833018845261e-06,
+      "loss": 0.4332,
+      "step": 17744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7444049685554823,
+      "learning_rate": 1.7214793743199759e-06,
+      "loss": 0.4619,
+      "step": 17745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9013595087065807,
+      "learning_rate": 1.7211258088956733e-06,
+      "loss": 0.5584,
+      "step": 17746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6666258421013496,
+      "learning_rate": 1.7207723225797554e-06,
+      "loss": 0.3855,
+      "step": 17747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8539963532568338,
+      "learning_rate": 1.720418915379621e-06,
+      "loss": 0.6118,
+      "step": 17748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7497029019759651,
+      "learning_rate": 1.7200655873026689e-06,
+      "loss": 0.4268,
+      "step": 17749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.667578464368842,
+      "learning_rate": 1.719712338356296e-06,
+      "loss": 0.4313,
+      "step": 17750
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.4896278381347656,
+      "eval_runtime": 67.9755,
+      "eval_samples_per_second": 50.165,
+      "eval_steps_per_second": 1.574,
+      "step": 17750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8858665307127243,
+      "learning_rate": 1.7193591685478972e-06,
+      "loss": 0.4743,
+      "step": 17751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3366406780184366,
+      "learning_rate": 1.7190060778848656e-06,
+      "loss": 0.3326,
+      "step": 17752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.737032668976193,
+      "learning_rate": 1.7186530663745948e-06,
+      "loss": 0.5058,
+      "step": 17753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0153399610037726,
+      "learning_rate": 1.718300134024472e-06,
+      "loss": 0.5564,
+      "step": 17754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0208096650954293,
+      "learning_rate": 1.7179472808418873e-06,
+      "loss": 0.8584,
+      "step": 17755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8389785702504988,
+      "learning_rate": 1.7175945068342278e-06,
+      "loss": 0.6186,
+      "step": 17756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6247957584783659,
+      "learning_rate": 1.7172418120088776e-06,
+      "loss": 0.4701,
+      "step": 17757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.802939089719498,
+      "learning_rate": 1.7168891963732199e-06,
+      "loss": 0.4201,
+      "step": 17758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8238674693213468,
+      "learning_rate": 1.7165366599346383e-06,
+      "loss": 0.5116,
+      "step": 17759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6816929856970313,
+      "learning_rate": 1.716184202700511e-06,
+      "loss": 0.4287,
+      "step": 17760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7499905079006086,
+      "learning_rate": 1.7158318246782165e-06,
+      "loss": 0.5466,
+      "step": 17761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.455000524356565,
+      "learning_rate": 1.7154795258751327e-06,
+      "loss": 0.3587,
+      "step": 17762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.119824260321046,
+      "learning_rate": 1.715127306298636e-06,
+      "loss": 0.2595,
+      "step": 17763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6038290885343953,
+      "learning_rate": 1.7147751659560964e-06,
+      "loss": 0.4242,
+      "step": 17764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.051184178443803,
+      "learning_rate": 1.7144231048548881e-06,
+      "loss": 0.5069,
+      "step": 17765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.524322309279894,
+      "learning_rate": 1.7140711230023816e-06,
+      "loss": 0.352,
+      "step": 17766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8923922730515503,
+      "learning_rate": 1.7137192204059433e-06,
+      "loss": 0.5225,
+      "step": 17767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6538432003564625,
+      "learning_rate": 1.713367397072942e-06,
+      "loss": 0.4352,
+      "step": 17768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5330686988296032,
+      "learning_rate": 1.713015653010743e-06,
+      "loss": 0.3645,
+      "step": 17769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7445998522092696,
+      "learning_rate": 1.712663988226708e-06,
+      "loss": 0.3878,
+      "step": 17770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.65458729175764,
+      "learning_rate": 1.7123124027281997e-06,
+      "loss": 0.4581,
+      "step": 17771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.815738839501325,
+      "learning_rate": 1.7119608965225792e-06,
+      "loss": 0.4852,
+      "step": 17772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.253621461150411,
+      "learning_rate": 1.7116094696172058e-06,
+      "loss": 0.5777,
+      "step": 17773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0669781530840896,
+      "learning_rate": 1.7112581220194334e-06,
+      "loss": 0.5844,
+      "step": 17774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6179429934944207,
+      "learning_rate": 1.7109068537366192e-06,
+      "loss": 0.4857,
+      "step": 17775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7805651175760528,
+      "learning_rate": 1.710555664776117e-06,
+      "loss": 0.4031,
+      "step": 17776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.61127688702832,
+      "learning_rate": 1.710204555145278e-06,
+      "loss": 0.4282,
+      "step": 17777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.807433497898556,
+      "learning_rate": 1.7098535248514524e-06,
+      "loss": 0.5058,
+      "step": 17778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8303612848296271,
+      "learning_rate": 1.70950257390199e-06,
+      "loss": 0.4581,
+      "step": 17779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8846899491501685,
+      "learning_rate": 1.7091517023042365e-06,
+      "loss": 0.44,
+      "step": 17780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.792589870133611,
+      "learning_rate": 1.7088009100655368e-06,
+      "loss": 0.6645,
+      "step": 17781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.321117729040542,
+      "learning_rate": 1.7084501971932365e-06,
+      "loss": 0.3983,
+      "step": 17782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1559349942476422,
+      "learning_rate": 1.708099563694675e-06,
+      "loss": 0.5224,
+      "step": 17783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.919849870253366,
+      "learning_rate": 1.7077490095771943e-06,
+      "loss": 0.5065,
+      "step": 17784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.162633480470845,
+      "learning_rate": 1.7073985348481325e-06,
+      "loss": 0.4322,
+      "step": 17785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.49065133034464,
+      "learning_rate": 1.7070481395148274e-06,
+      "loss": 0.2897,
+      "step": 17786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8841257004530239,
+      "learning_rate": 1.7066978235846132e-06,
+      "loss": 0.4064,
+      "step": 17787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1087725299476165,
+      "learning_rate": 1.7063475870648239e-06,
+      "loss": 0.4662,
+      "step": 17788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9985723058331,
+      "learning_rate": 1.7059974299627918e-06,
+      "loss": 0.5197,
+      "step": 17789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9614201527985766,
+      "learning_rate": 1.7056473522858468e-06,
+      "loss": 0.4069,
+      "step": 17790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7034483864066592,
+      "learning_rate": 1.7052973540413174e-06,
+      "loss": 0.596,
+      "step": 17791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7090232322377477,
+      "learning_rate": 1.704947435236532e-06,
+      "loss": 0.42,
+      "step": 17792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4853032678636058,
+      "learning_rate": 1.7045975958788136e-06,
+      "loss": 0.4524,
+      "step": 17793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6601240060867195,
+      "learning_rate": 1.7042478359754877e-06,
+      "loss": 0.4936,
+      "step": 17794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8963980693672127,
+      "learning_rate": 1.703898155533876e-06,
+      "loss": 0.4354,
+      "step": 17795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5044704040749686,
+      "learning_rate": 1.7035485545612978e-06,
+      "loss": 0.4208,
+      "step": 17796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7036646585884698,
+      "learning_rate": 1.7031990330650727e-06,
+      "loss": 0.3904,
+      "step": 17797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.210647120002637,
+      "learning_rate": 1.7028495910525176e-06,
+      "loss": 0.5163,
+      "step": 17798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5298037899722274,
+      "learning_rate": 1.7025002285309471e-06,
+      "loss": 0.3595,
+      "step": 17799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0729754995134093,
+      "learning_rate": 1.7021509455076764e-06,
+      "loss": 0.547,
+      "step": 17800
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.48920074105262756,
+      "eval_runtime": 67.984,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.574,
+      "step": 17800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7416939295210978,
+      "learning_rate": 1.7018017419900178e-06,
+      "loss": 0.5147,
+      "step": 17801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.109277639994772,
+      "learning_rate": 1.7014526179852792e-06,
+      "loss": 0.4839,
+      "step": 17802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3533673123010483,
+      "learning_rate": 1.7011035735007708e-06,
+      "loss": 0.3124,
+      "step": 17803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5896563282688922,
+      "learning_rate": 1.7007546085438004e-06,
+      "loss": 0.5252,
+      "step": 17804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0794620966715462,
+      "learning_rate": 1.7004057231216714e-06,
+      "loss": 0.4902,
+      "step": 17805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8829992529831068,
+      "learning_rate": 1.7000569172416891e-06,
+      "loss": 0.4864,
+      "step": 17806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.103193139739063,
+      "learning_rate": 1.6997081909111547e-06,
+      "loss": 0.5051,
+      "step": 17807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.749883840218196,
+      "learning_rate": 1.6993595441373695e-06,
+      "loss": 0.5399,
+      "step": 17808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8257101288031032,
+      "learning_rate": 1.6990109769276308e-06,
+      "loss": 0.4516,
+      "step": 17809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5279752339129062,
+      "learning_rate": 1.6986624892892367e-06,
+      "loss": 0.3877,
+      "step": 17810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8570858420865968,
+      "learning_rate": 1.6983140812294829e-06,
+      "loss": 0.5521,
+      "step": 17811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.100326056059012,
+      "learning_rate": 1.6979657527556618e-06,
+      "loss": 0.5693,
+      "step": 17812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9559624244965395,
+      "learning_rate": 1.6976175038750658e-06,
+      "loss": 0.6454,
+      "step": 17813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.407292514834881,
+      "learning_rate": 1.6972693345949873e-06,
+      "loss": 0.3674,
+      "step": 17814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9556148445051187,
+      "learning_rate": 1.6969212449227113e-06,
+      "loss": 0.5854,
+      "step": 17815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6271758252261634,
+      "learning_rate": 1.6965732348655278e-06,
+      "loss": 0.4021,
+      "step": 17816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4263003045068503,
+      "learning_rate": 1.6962253044307214e-06,
+      "loss": 0.2544,
+      "step": 17817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7893798558496228,
+      "learning_rate": 1.6958774536255753e-06,
+      "loss": 0.5061,
+      "step": 17818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4244695636185478,
+      "learning_rate": 1.695529682457372e-06,
+      "loss": 0.3289,
+      "step": 17819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9513764494328367,
+      "learning_rate": 1.6951819909333911e-06,
+      "loss": 0.6039,
+      "step": 17820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5489789498300368,
+      "learning_rate": 1.6948343790609135e-06,
+      "loss": 0.3976,
+      "step": 17821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9587959976764309,
+      "learning_rate": 1.6944868468472131e-06,
+      "loss": 0.4643,
+      "step": 17822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.37784162835324,
+      "learning_rate": 1.6941393942995676e-06,
+      "loss": 0.3722,
+      "step": 17823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9260766099629552,
+      "learning_rate": 1.6937920214252507e-06,
+      "loss": 0.5815,
+      "step": 17824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7155719165192196,
+      "learning_rate": 1.6934447282315324e-06,
+      "loss": 0.4967,
+      "step": 17825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.631796037122757,
+      "learning_rate": 1.6930975147256845e-06,
+      "loss": 0.4561,
+      "step": 17826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8487705199036726,
+      "learning_rate": 1.6927503809149768e-06,
+      "loss": 0.4832,
+      "step": 17827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8759009880412485,
+      "learning_rate": 1.6924033268066742e-06,
+      "loss": 0.3839,
+      "step": 17828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.963519669844743,
+      "learning_rate": 1.6920563524080422e-06,
+      "loss": 0.3917,
+      "step": 17829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.154963220743573,
+      "learning_rate": 1.691709457726347e-06,
+      "loss": 0.5666,
+      "step": 17830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9624293379347595,
+      "learning_rate": 1.6913626427688478e-06,
+      "loss": 0.5813,
+      "step": 17831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5958662383975244,
+      "learning_rate": 1.6910159075428056e-06,
+      "loss": 0.5357,
+      "step": 17832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8246536145076264,
+      "learning_rate": 1.6906692520554803e-06,
+      "loss": 0.5928,
+      "step": 17833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7172554086871181,
+      "learning_rate": 1.6903226763141283e-06,
+      "loss": 0.5265,
+      "step": 17834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7725126037530128,
+      "learning_rate": 1.6899761803260045e-06,
+      "loss": 0.5207,
+      "step": 17835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9000091777568149,
+      "learning_rate": 1.6896297640983624e-06,
+      "loss": 0.4485,
+      "step": 17836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4826536299339252,
+      "learning_rate": 1.6892834276384557e-06,
+      "loss": 0.3561,
+      "step": 17837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7401422249081762,
+      "learning_rate": 1.6889371709535327e-06,
+      "loss": 0.4524,
+      "step": 17838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3023471501696955,
+      "learning_rate": 1.6885909940508427e-06,
+      "loss": 0.8078,
+      "step": 17839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.712101790022007,
+      "learning_rate": 1.6882448969376335e-06,
+      "loss": 0.5361,
+      "step": 17840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6686221342287262,
+      "learning_rate": 1.6878988796211492e-06,
+      "loss": 0.4357,
+      "step": 17841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5729675255097502,
+      "learning_rate": 1.687552942108634e-06,
+      "loss": 0.4797,
+      "step": 17842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.636711088757852,
+      "learning_rate": 1.6872070844073314e-06,
+      "loss": 0.487,
+      "step": 17843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6209156829141433,
+      "learning_rate": 1.6868613065244787e-06,
+      "loss": 0.4848,
+      "step": 17844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.213728257026626,
+      "learning_rate": 1.6865156084673156e-06,
+      "loss": 0.4515,
+      "step": 17845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8361070308936078,
+      "learning_rate": 1.6861699902430805e-06,
+      "loss": 0.4648,
+      "step": 17846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6078593892617206,
+      "learning_rate": 1.6858244518590088e-06,
+      "loss": 0.3484,
+      "step": 17847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4910786911085647,
+      "learning_rate": 1.685478993322331e-06,
+      "loss": 0.345,
+      "step": 17848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.752177586508777,
+      "learning_rate": 1.685133614640283e-06,
+      "loss": 0.7565,
+      "step": 17849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7137295893656184,
+      "learning_rate": 1.6847883158200922e-06,
+      "loss": 0.4995,
+      "step": 17850
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.4889281988143921,
+      "eval_runtime": 67.9199,
+      "eval_samples_per_second": 50.206,
+      "eval_steps_per_second": 1.575,
+      "step": 17850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7340697390740507,
+      "learning_rate": 1.6844430968689884e-06,
+      "loss": 0.4722,
+      "step": 17851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.517048954218517,
+      "learning_rate": 1.6840979577941984e-06,
+      "loss": 0.3654,
+      "step": 17852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8443853313407603,
+      "learning_rate": 1.6837528986029489e-06,
+      "loss": 0.5141,
+      "step": 17853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.079961938024283,
+      "learning_rate": 1.6834079193024605e-06,
+      "loss": 0.5596,
+      "step": 17854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8641427169760019,
+      "learning_rate": 1.6830630198999573e-06,
+      "loss": 0.4191,
+      "step": 17855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.837947227174918,
+      "learning_rate": 1.68271820040266e-06,
+      "loss": 0.3706,
+      "step": 17856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.631684552894669,
+      "learning_rate": 1.6823734608177843e-06,
+      "loss": 0.3899,
+      "step": 17857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7213478967505036,
+      "learning_rate": 1.6820288011525501e-06,
+      "loss": 0.5435,
+      "step": 17858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8656005761452825,
+      "learning_rate": 1.6816842214141718e-06,
+      "loss": 0.4389,
+      "step": 17859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2769357733311764,
+      "learning_rate": 1.6813397216098621e-06,
+      "loss": 0.3008,
+      "step": 17860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.670337566048488,
+      "learning_rate": 1.6809953017468333e-06,
+      "loss": 0.4709,
+      "step": 17861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8375335973509817,
+      "learning_rate": 1.6806509618322966e-06,
+      "loss": 0.5457,
+      "step": 17862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.708566225780765,
+      "learning_rate": 1.6803067018734588e-06,
+      "loss": 0.309,
+      "step": 17863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5614149871636707,
+      "learning_rate": 1.679962521877528e-06,
+      "loss": 0.3604,
+      "step": 17864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7858493894823801,
+      "learning_rate": 1.6796184218517086e-06,
+      "loss": 0.5918,
+      "step": 17865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4155013394092444,
+      "learning_rate": 1.679274401803206e-06,
+      "loss": 0.2849,
+      "step": 17866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9661121374192319,
+      "learning_rate": 1.6789304617392194e-06,
+      "loss": 0.4055,
+      "step": 17867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7873772955446674,
+      "learning_rate": 1.6785866016669505e-06,
+      "loss": 0.4531,
+      "step": 17868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.99177868844314,
+      "learning_rate": 1.6782428215935985e-06,
+      "loss": 0.3959,
+      "step": 17869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8453798226038418,
+      "learning_rate": 1.6778991215263579e-06,
+      "loss": 0.4869,
+      "step": 17870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8974770889690382,
+      "learning_rate": 1.6775555014724257e-06,
+      "loss": 0.4846,
+      "step": 17871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6200707359701247,
+      "learning_rate": 1.6772119614389952e-06,
+      "loss": 0.3329,
+      "step": 17872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4173513375303437,
+      "learning_rate": 1.6768685014332575e-06,
+      "loss": 0.489,
+      "step": 17873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0935013130813114,
+      "learning_rate": 1.6765251214624028e-06,
+      "loss": 0.4821,
+      "step": 17874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7988801576749525,
+      "learning_rate": 1.6761818215336212e-06,
+      "loss": 0.4877,
+      "step": 17875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9657300904252522,
+      "learning_rate": 1.6758386016540967e-06,
+      "loss": 0.6904,
+      "step": 17876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.566215900869329,
+      "learning_rate": 1.675495461831016e-06,
+      "loss": 0.4944,
+      "step": 17877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0226812498857263,
+      "learning_rate": 1.675152402071562e-06,
+      "loss": 0.6263,
+      "step": 17878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3231225605495516,
+      "learning_rate": 1.674809422382918e-06,
+      "loss": 0.3106,
+      "step": 17879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.933506820911824,
+      "learning_rate": 1.6744665227722617e-06,
+      "loss": 0.4665,
+      "step": 17880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5165442638561075,
+      "learning_rate": 1.6741237032467728e-06,
+      "loss": 0.4681,
+      "step": 17881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7191146191568094,
+      "learning_rate": 1.6737809638136288e-06,
+      "loss": 0.4877,
+      "step": 17882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6067982335455637,
+      "learning_rate": 1.673438304480003e-06,
+      "loss": 0.4439,
+      "step": 17883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1434221762948784,
+      "learning_rate": 1.6730957252530689e-06,
+      "loss": 0.3914,
+      "step": 17884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5722891115594282,
+      "learning_rate": 1.6727532261400004e-06,
+      "loss": 0.4858,
+      "step": 17885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.785531334104964,
+      "learning_rate": 1.6724108071479644e-06,
+      "loss": 0.5393,
+      "step": 17886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2709483232330876,
+      "learning_rate": 1.6720684682841309e-06,
+      "loss": 0.5913,
+      "step": 17887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8082927303260232,
+      "learning_rate": 1.671726209555668e-06,
+      "loss": 0.4864,
+      "step": 17888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.081598378816443,
+      "learning_rate": 1.6713840309697376e-06,
+      "loss": 0.4534,
+      "step": 17889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8573132784188826,
+      "learning_rate": 1.6710419325335042e-06,
+      "loss": 0.5506,
+      "step": 17890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4662335118715804,
+      "learning_rate": 1.6706999142541303e-06,
+      "loss": 0.376,
+      "step": 17891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8165456497758203,
+      "learning_rate": 1.6703579761387755e-06,
+      "loss": 0.5209,
+      "step": 17892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5982483753005567,
+      "learning_rate": 1.6700161181945975e-06,
+      "loss": 0.4766,
+      "step": 17893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.139848461621255,
+      "learning_rate": 1.669674340428753e-06,
+      "loss": 0.4787,
+      "step": 17894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0867042482732727,
+      "learning_rate": 1.6693326428483978e-06,
+      "loss": 0.5295,
+      "step": 17895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7516909217264705,
+      "learning_rate": 1.6689910254606834e-06,
+      "loss": 0.4578,
+      "step": 17896
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.101227598976595,
+      "learning_rate": 1.6686494882727628e-06,
+      "loss": 0.5934,
+      "step": 17897
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6809453791907767,
+      "learning_rate": 1.6683080312917848e-06,
+      "loss": 0.3735,
+      "step": 17898
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0007083527775635,
+      "learning_rate": 1.6679666545248988e-06,
+      "loss": 0.4393,
+      "step": 17899
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6943200091677453,
+      "learning_rate": 1.66762535797925e-06,
+      "loss": 0.5048,
+      "step": 17900
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.4888206422328949,
+      "eval_runtime": 67.9731,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 17900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6647732557919213,
+      "learning_rate": 1.6672841416619855e-06,
+      "loss": 0.369,
+      "step": 17901
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8512935359899865,
+      "learning_rate": 1.666943005580246e-06,
+      "loss": 0.4066,
+      "step": 17902
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6437458713738935,
+      "learning_rate": 1.666601949741174e-06,
+      "loss": 0.3804,
+      "step": 17903
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8764556358581337,
+      "learning_rate": 1.6662609741519098e-06,
+      "loss": 0.6666,
+      "step": 17904
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9353449807421623,
+      "learning_rate": 1.6659200788195895e-06,
+      "loss": 0.4964,
+      "step": 17905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0878339797927774,
+      "learning_rate": 1.6655792637513515e-06,
+      "loss": 0.5134,
+      "step": 17906
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2711473291054178,
+      "learning_rate": 1.665238528954331e-06,
+      "loss": 0.2756,
+      "step": 17907
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.018925752184008,
+      "learning_rate": 1.6648978744356592e-06,
+      "loss": 0.6294,
+      "step": 17908
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8004373777235247,
+      "learning_rate": 1.6645573002024678e-06,
+      "loss": 0.5502,
+      "step": 17909
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2078768345942903,
+      "learning_rate": 1.664216806261888e-06,
+      "loss": 0.5503,
+      "step": 17910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.48779546155432,
+      "learning_rate": 1.6638763926210462e-06,
+      "loss": 0.4181,
+      "step": 17911
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5214045081683054,
+      "learning_rate": 1.663536059287069e-06,
+      "loss": 0.4357,
+      "step": 17912
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6054218914740956,
+      "learning_rate": 1.6631958062670822e-06,
+      "loss": 0.3148,
+      "step": 17913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8242975970062805,
+      "learning_rate": 1.6628556335682086e-06,
+      "loss": 0.4144,
+      "step": 17914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3592402772003491,
+      "learning_rate": 1.6625155411975682e-06,
+      "loss": 0.3786,
+      "step": 17915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8244809716808168,
+      "learning_rate": 1.6621755291622814e-06,
+      "loss": 0.5402,
+      "step": 17916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9818730953659807,
+      "learning_rate": 1.661835597469467e-06,
+      "loss": 0.4794,
+      "step": 17917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7859600160634905,
+      "learning_rate": 1.661495746126239e-06,
+      "loss": 0.4634,
+      "step": 17918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.605271016939042,
+      "learning_rate": 1.6611559751397138e-06,
+      "loss": 0.3919,
+      "step": 17919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6838397454385714,
+      "learning_rate": 1.660816284517005e-06,
+      "loss": 0.3804,
+      "step": 17920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.410447340086089,
+      "learning_rate": 1.6604766742652218e-06,
+      "loss": 0.3955,
+      "step": 17921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1440734556645085,
+      "learning_rate": 1.6601371443914737e-06,
+      "loss": 0.4948,
+      "step": 17922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9542652562163392,
+      "learning_rate": 1.6597976949028716e-06,
+      "loss": 0.7531,
+      "step": 17923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.868285213522245,
+      "learning_rate": 1.6594583258065177e-06,
+      "loss": 0.5121,
+      "step": 17924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.115110592931213,
+      "learning_rate": 1.6591190371095184e-06,
+      "loss": 0.7405,
+      "step": 17925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7344825672165645,
+      "learning_rate": 1.6587798288189767e-06,
+      "loss": 0.5286,
+      "step": 17926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.433939893413226,
+      "learning_rate": 1.6584407009419941e-06,
+      "loss": 0.7002,
+      "step": 17927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3272639966564739,
+      "learning_rate": 1.658101653485669e-06,
+      "loss": 0.2911,
+      "step": 17928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5968791850406547,
+      "learning_rate": 1.6577626864570984e-06,
+      "loss": 0.4821,
+      "step": 17929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7780425581379018,
+      "learning_rate": 1.6574237998633814e-06,
+      "loss": 0.4399,
+      "step": 17930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4546876883926114,
+      "learning_rate": 1.657084993711609e-06,
+      "loss": 0.5477,
+      "step": 17931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.827096889039146,
+      "learning_rate": 1.6567462680088755e-06,
+      "loss": 0.5317,
+      "step": 17932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7237461102183433,
+      "learning_rate": 1.6564076227622727e-06,
+      "loss": 0.4764,
+      "step": 17933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6562770724037137,
+      "learning_rate": 1.656069057978888e-06,
+      "loss": 0.4946,
+      "step": 17934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9078462834777605,
+      "learning_rate": 1.6557305736658097e-06,
+      "loss": 0.4789,
+      "step": 17935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4463107358003604,
+      "learning_rate": 1.6553921698301257e-06,
+      "loss": 0.3609,
+      "step": 17936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5925011451127777,
+      "learning_rate": 1.6550538464789174e-06,
+      "loss": 0.4092,
+      "step": 17937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.731644913487992,
+      "learning_rate": 1.6547156036192685e-06,
+      "loss": 0.5516,
+      "step": 17938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8390974017935668,
+      "learning_rate": 1.6543774412582603e-06,
+      "loss": 0.6009,
+      "step": 17939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4120604233766991,
+      "learning_rate": 1.6540393594029724e-06,
+      "loss": 0.3085,
+      "step": 17940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7212125112719465,
+      "learning_rate": 1.6537013580604807e-06,
+      "loss": 0.4196,
+      "step": 17941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7962885662413095,
+      "learning_rate": 1.653363437237862e-06,
+      "loss": 0.4234,
+      "step": 17942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7590408539614453,
+      "learning_rate": 1.6530255969421922e-06,
+      "loss": 0.3786,
+      "step": 17943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3515836990112557,
+      "learning_rate": 1.6526878371805405e-06,
+      "loss": 0.4147,
+      "step": 17944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8576562147889302,
+      "learning_rate": 1.6523501579599793e-06,
+      "loss": 0.6468,
+      "step": 17945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.144191079695376,
+      "learning_rate": 1.652012559287579e-06,
+      "loss": 0.6478,
+      "step": 17946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.871597395968997,
+      "learning_rate": 1.6516750411704035e-06,
+      "loss": 0.5005,
+      "step": 17947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7597386112042899,
+      "learning_rate": 1.6513376036155216e-06,
+      "loss": 0.5949,
+      "step": 17948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.025815082140771,
+      "learning_rate": 1.6510002466299973e-06,
+      "loss": 0.6121,
+      "step": 17949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.76631137337077,
+      "learning_rate": 1.6506629702208914e-06,
+      "loss": 0.4366,
+      "step": 17950
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.48902013897895813,
+      "eval_runtime": 68.009,
+      "eval_samples_per_second": 50.14,
+      "eval_steps_per_second": 1.573,
+      "step": 17950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8582606375488213,
+      "learning_rate": 1.6503257743952656e-06,
+      "loss": 0.5627,
+      "step": 17951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.061676991852867,
+      "learning_rate": 1.6499886591601794e-06,
+      "loss": 0.48,
+      "step": 17952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8051059590878435,
+      "learning_rate": 1.6496516245226879e-06,
+      "loss": 0.4756,
+      "step": 17953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2131095926877893,
+      "learning_rate": 1.6493146704898484e-06,
+      "loss": 0.513,
+      "step": 17954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8370220850700127,
+      "learning_rate": 1.6489777970687154e-06,
+      "loss": 0.4569,
+      "step": 17955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9763063531195635,
+      "learning_rate": 1.6486410042663398e-06,
+      "loss": 0.4761,
+      "step": 17956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1835580208527756,
+      "learning_rate": 1.6483042920897718e-06,
+      "loss": 0.6186,
+      "step": 17957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7504295263289429,
+      "learning_rate": 1.6479676605460615e-06,
+      "loss": 0.5217,
+      "step": 17958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5622136202224717,
+      "learning_rate": 1.6476311096422566e-06,
+      "loss": 0.4615,
+      "step": 17959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.308003323979965,
+      "learning_rate": 1.6472946393854006e-06,
+      "loss": 0.3349,
+      "step": 17960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6009220962332706,
+      "learning_rate": 1.646958249782538e-06,
+      "loss": 0.4299,
+      "step": 17961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.924236152522732,
+      "learning_rate": 1.646621940840713e-06,
+      "loss": 0.4422,
+      "step": 17962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8266505236745327,
+      "learning_rate": 1.6462857125669624e-06,
+      "loss": 0.5382,
+      "step": 17963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.880886624951815,
+      "learning_rate": 1.6459495649683272e-06,
+      "loss": 0.5139,
+      "step": 17964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4406563252136513,
+      "learning_rate": 1.6456134980518452e-06,
+      "loss": 0.5277,
+      "step": 17965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8236915277294872,
+      "learning_rate": 1.6452775118245493e-06,
+      "loss": 0.5351,
+      "step": 17966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.427653032998671,
+      "learning_rate": 1.6449416062934745e-06,
+      "loss": 0.4443,
+      "step": 17967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6980162885000119,
+      "learning_rate": 1.6446057814656536e-06,
+      "loss": 0.4781,
+      "step": 17968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0171690125083055,
+      "learning_rate": 1.6442700373481154e-06,
+      "loss": 0.5417,
+      "step": 17969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6945406311100162,
+      "learning_rate": 1.6439343739478886e-06,
+      "loss": 0.4771,
+      "step": 17970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9580348559941492,
+      "learning_rate": 1.643598791272001e-06,
+      "loss": 0.5782,
+      "step": 17971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2431846391839594,
+      "learning_rate": 1.6432632893274781e-06,
+      "loss": 0.7176,
+      "step": 17972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9413255342539348,
+      "learning_rate": 1.6429278681213418e-06,
+      "loss": 0.4487,
+      "step": 17973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6800558727233854,
+      "learning_rate": 1.6425925276606148e-06,
+      "loss": 0.4334,
+      "step": 17974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4926493380617178,
+      "learning_rate": 1.6422572679523185e-06,
+      "loss": 0.3371,
+      "step": 17975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6694897825483892,
+      "learning_rate": 1.6419220890034693e-06,
+      "loss": 0.5141,
+      "step": 17976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3452501675262685,
+      "learning_rate": 1.641586990821084e-06,
+      "loss": 0.485,
+      "step": 17977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6948704527291478,
+      "learning_rate": 1.6412519734121802e-06,
+      "loss": 0.5341,
+      "step": 17978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0164131911696384,
+      "learning_rate": 1.6409170367837685e-06,
+      "loss": 0.5286,
+      "step": 17979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7277739045612532,
+      "learning_rate": 1.6405821809428613e-06,
+      "loss": 0.448,
+      "step": 17980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6237286872946086,
+      "learning_rate": 1.6402474058964708e-06,
+      "loss": 0.4865,
+      "step": 17981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.064931461632856,
+      "learning_rate": 1.639912711651602e-06,
+      "loss": 0.4463,
+      "step": 17982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1425160052083214,
+      "learning_rate": 1.639578098215263e-06,
+      "loss": 0.5876,
+      "step": 17983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1528832844715104,
+      "learning_rate": 1.6392435655944586e-06,
+      "loss": 0.5565,
+      "step": 17984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.913184444890831,
+      "learning_rate": 1.6389091137961926e-06,
+      "loss": 0.4895,
+      "step": 17985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6598511089472887,
+      "learning_rate": 1.6385747428274662e-06,
+      "loss": 0.467,
+      "step": 17986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9204212913678853,
+      "learning_rate": 1.6382404526952786e-06,
+      "loss": 0.4943,
+      "step": 17987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1229242874007364,
+      "learning_rate": 1.6379062434066297e-06,
+      "loss": 0.5526,
+      "step": 17988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7720378219981554,
+      "learning_rate": 1.637572114968513e-06,
+      "loss": 0.4386,
+      "step": 17989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1306252625577677,
+      "learning_rate": 1.6372380673879257e-06,
+      "loss": 0.579,
+      "step": 17990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0343921140752186,
+      "learning_rate": 1.6369041006718617e-06,
+      "loss": 0.697,
+      "step": 17991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.724491316427975,
+      "learning_rate": 1.6365702148273094e-06,
+      "loss": 0.3995,
+      "step": 17992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0056472302258923,
+      "learning_rate": 1.6362364098612599e-06,
+      "loss": 0.4606,
+      "step": 17993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6694268187515977,
+      "learning_rate": 1.6359026857807022e-06,
+      "loss": 0.4735,
+      "step": 17994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.908180496352647,
+      "learning_rate": 1.6355690425926207e-06,
+      "loss": 0.4204,
+      "step": 17995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.592828683540559,
+      "learning_rate": 1.6352354803040001e-06,
+      "loss": 0.379,
+      "step": 17996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.558219676421323,
+      "learning_rate": 1.6349019989218264e-06,
+      "loss": 0.6465,
+      "step": 17997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5845746716941946,
+      "learning_rate": 1.6345685984530773e-06,
+      "loss": 0.4018,
+      "step": 17998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5828110210087956,
+      "learning_rate": 1.634235278904734e-06,
+      "loss": 0.5084,
+      "step": 17999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5543679705433973,
+      "learning_rate": 1.6339020402837746e-06,
+      "loss": 0.4227,
+      "step": 18000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.4891813397407532,
+      "eval_runtime": 67.9753,
+      "eval_samples_per_second": 50.165,
+      "eval_steps_per_second": 1.574,
+      "step": 18000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6446210925107871,
+      "learning_rate": 1.6335688825971746e-06,
+      "loss": 0.5649,
+      "step": 18001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7832285800230216,
+      "learning_rate": 1.6332358058519077e-06,
+      "loss": 0.4842,
+      "step": 18002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6699829337371548,
+      "learning_rate": 1.6329028100549484e-06,
+      "loss": 0.4433,
+      "step": 18003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8759019666116281,
+      "learning_rate": 1.6325698952132662e-06,
+      "loss": 0.5015,
+      "step": 18004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.227775773018721,
+      "learning_rate": 1.6322370613338312e-06,
+      "loss": 0.5495,
+      "step": 18005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.946481765830022,
+      "learning_rate": 1.6319043084236109e-06,
+      "loss": 0.4312,
+      "step": 18006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1902849984825754,
+      "learning_rate": 1.6315716364895717e-06,
+      "loss": 0.5145,
+      "step": 18007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8407181494207954,
+      "learning_rate": 1.6312390455386768e-06,
+      "loss": 0.4557,
+      "step": 18008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5137003800885138,
+      "learning_rate": 1.6309065355778901e-06,
+      "loss": 0.3747,
+      "step": 18009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1035540500966583,
+      "learning_rate": 1.6305741066141724e-06,
+      "loss": 0.4598,
+      "step": 18010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7005484153164963,
+      "learning_rate": 1.6302417586544813e-06,
+      "loss": 0.5511,
+      "step": 18011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.707610927361301,
+      "learning_rate": 1.6299094917057757e-06,
+      "loss": 0.5331,
+      "step": 18012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.548513029464305,
+      "learning_rate": 1.629577305775012e-06,
+      "loss": 0.4603,
+      "step": 18013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6876928337893269,
+      "learning_rate": 1.6292452008691423e-06,
+      "loss": 0.725,
+      "step": 18014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5808165143730906,
+      "learning_rate": 1.62891317699512e-06,
+      "loss": 0.3898,
+      "step": 18015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.675852679601955,
+      "learning_rate": 1.6285812341598972e-06,
+      "loss": 0.4923,
+      "step": 18016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0138305246723682,
+      "learning_rate": 1.6282493723704207e-06,
+      "loss": 0.5729,
+      "step": 18017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8800058982392525,
+      "learning_rate": 1.6279175916336381e-06,
+      "loss": 0.4492,
+      "step": 18018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6486066891562967,
+      "learning_rate": 1.6275858919564966e-06,
+      "loss": 0.3884,
+      "step": 18019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9059329083965295,
+      "learning_rate": 1.6272542733459396e-06,
+      "loss": 0.5324,
+      "step": 18020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3923381946316282,
+      "learning_rate": 1.6269227358089078e-06,
+      "loss": 0.3096,
+      "step": 18021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8259422542325239,
+      "learning_rate": 1.6265912793523437e-06,
+      "loss": 0.4982,
+      "step": 18022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6263999669076568,
+      "learning_rate": 1.6262599039831854e-06,
+      "loss": 0.4443,
+      "step": 18023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7902413172411105,
+      "learning_rate": 1.6259286097083693e-06,
+      "loss": 0.4902,
+      "step": 18024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6620155293794001,
+      "learning_rate": 1.6255973965348315e-06,
+      "loss": 0.4733,
+      "step": 18025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5420536445160324,
+      "learning_rate": 1.6252662644695067e-06,
+      "loss": 0.3775,
+      "step": 18026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3709878792152477,
+      "learning_rate": 1.6249352135193253e-06,
+      "loss": 0.4687,
+      "step": 18027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9083603971143954,
+      "learning_rate": 1.624604243691218e-06,
+      "loss": 0.3529,
+      "step": 18028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5924277327548675,
+      "learning_rate": 1.6242733549921133e-06,
+      "loss": 0.5479,
+      "step": 18029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5699986949839855,
+      "learning_rate": 1.6239425474289403e-06,
+      "loss": 0.3963,
+      "step": 18030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9141726483009553,
+      "learning_rate": 1.6236118210086213e-06,
+      "loss": 0.5042,
+      "step": 18031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8091139808971266,
+      "learning_rate": 1.6232811757380812e-06,
+      "loss": 0.4001,
+      "step": 18032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9361366306760495,
+      "learning_rate": 1.6229506116242428e-06,
+      "loss": 0.5309,
+      "step": 18033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5821718637877422,
+      "learning_rate": 1.622620128674024e-06,
+      "loss": 0.3917,
+      "step": 18034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0538561976303678,
+      "learning_rate": 1.6222897268943447e-06,
+      "loss": 0.5608,
+      "step": 18035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6706170433768153,
+      "learning_rate": 1.6219594062921224e-06,
+      "loss": 0.5719,
+      "step": 18036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4445181154607074,
+      "learning_rate": 1.6216291668742701e-06,
+      "loss": 0.3255,
+      "step": 18037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7246771259089981,
+      "learning_rate": 1.621299008647702e-06,
+      "loss": 0.6031,
+      "step": 18038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9111480217868277,
+      "learning_rate": 1.620968931619331e-06,
+      "loss": 0.5705,
+      "step": 18039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2520332619390728,
+      "learning_rate": 1.6206389357960656e-06,
+      "loss": 0.3201,
+      "step": 18040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9482263741251535,
+      "learning_rate": 1.6203090211848143e-06,
+      "loss": 0.464,
+      "step": 18041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7842092019154157,
+      "learning_rate": 1.619979187792484e-06,
+      "loss": 0.5056,
+      "step": 18042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7257449314850144,
+      "learning_rate": 1.6196494356259798e-06,
+      "loss": 0.5757,
+      "step": 18043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8387233060364634,
+      "learning_rate": 1.619319764692204e-06,
+      "loss": 0.692,
+      "step": 18044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7648129785057538,
+      "learning_rate": 1.6189901749980585e-06,
+      "loss": 0.4895,
+      "step": 18045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9287645819951738,
+      "learning_rate": 1.6186606665504434e-06,
+      "loss": 0.6055,
+      "step": 18046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6319678898902357,
+      "learning_rate": 1.6183312393562559e-06,
+      "loss": 0.4672,
+      "step": 18047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4678911039172382,
+      "learning_rate": 1.6180018934223946e-06,
+      "loss": 0.4057,
+      "step": 18048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7166701509438829,
+      "learning_rate": 1.617672628755751e-06,
+      "loss": 0.4382,
+      "step": 18049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8262792800864938,
+      "learning_rate": 1.6173434453632195e-06,
+      "loss": 0.4494,
+      "step": 18050
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.48932769894599915,
+      "eval_runtime": 68.0037,
+      "eval_samples_per_second": 50.144,
+      "eval_steps_per_second": 1.573,
+      "step": 18050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1674847542452198,
+      "learning_rate": 1.6170143432516914e-06,
+      "loss": 0.2661,
+      "step": 18051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7827091100076904,
+      "learning_rate": 1.616685322428057e-06,
+      "loss": 0.5484,
+      "step": 18052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6867729356064063,
+      "learning_rate": 1.6163563828992024e-06,
+      "loss": 0.3911,
+      "step": 18053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5507172183800826,
+      "learning_rate": 1.616027524672015e-06,
+      "loss": 0.4222,
+      "step": 18054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.564127917485188,
+      "learning_rate": 1.61569874775338e-06,
+      "loss": 0.5006,
+      "step": 18055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6206382989384358,
+      "learning_rate": 1.6153700521501781e-06,
+      "loss": 0.4343,
+      "step": 18056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.303677018597286,
+      "learning_rate": 1.6150414378692907e-06,
+      "loss": 0.5095,
+      "step": 18057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6983575077110415,
+      "learning_rate": 1.6147129049175995e-06,
+      "loss": 0.3389,
+      "step": 18058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5106174576568827,
+      "learning_rate": 1.6143844533019789e-06,
+      "loss": 0.5306,
+      "step": 18059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6753901257523853,
+      "learning_rate": 1.6140560830293065e-06,
+      "loss": 0.3577,
+      "step": 18060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4571563146283335,
+      "learning_rate": 1.6137277941064571e-06,
+      "loss": 0.4864,
+      "step": 18061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.708206271511418,
+      "learning_rate": 1.6133995865403018e-06,
+      "loss": 0.3958,
+      "step": 18062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4121175754259632,
+      "learning_rate": 1.6130714603377125e-06,
+      "loss": 0.3866,
+      "step": 18063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7677263536811236,
+      "learning_rate": 1.6127434155055574e-06,
+      "loss": 0.6844,
+      "step": 18064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9249547391022686,
+      "learning_rate": 1.6124154520507055e-06,
+      "loss": 0.6533,
+      "step": 18065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.837138643661641,
+      "learning_rate": 1.6120875699800206e-06,
+      "loss": 0.5361,
+      "step": 18066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.033450949968566,
+      "learning_rate": 1.6117597693003672e-06,
+      "loss": 0.4647,
+      "step": 18067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.931235434431349,
+      "learning_rate": 1.611432050018609e-06,
+      "loss": 0.5981,
+      "step": 18068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6940378988756228,
+      "learning_rate": 1.611104412141605e-06,
+      "loss": 0.5067,
+      "step": 18069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7223365727175024,
+      "learning_rate": 1.6107768556762142e-06,
+      "loss": 0.3495,
+      "step": 18070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.629386764715139,
+      "learning_rate": 1.6104493806292953e-06,
+      "loss": 0.4646,
+      "step": 18071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9393174284877308,
+      "learning_rate": 1.6101219870077017e-06,
+      "loss": 0.5896,
+      "step": 18072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1022186279666926,
+      "learning_rate": 1.6097946748182889e-06,
+      "loss": 0.4921,
+      "step": 18073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.998373794075749,
+      "learning_rate": 1.6094674440679082e-06,
+      "loss": 0.4785,
+      "step": 18074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6069731038984854,
+      "learning_rate": 1.6091402947634096e-06,
+      "loss": 0.399,
+      "step": 18075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5599481934002204,
+      "learning_rate": 1.6088132269116422e-06,
+      "loss": 0.489,
+      "step": 18076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.540781682238891,
+      "learning_rate": 1.608486240519453e-06,
+      "loss": 0.3588,
+      "step": 18077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9928399250399744,
+      "learning_rate": 1.608159335593688e-06,
+      "loss": 0.3983,
+      "step": 18078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.442271955332506,
+      "learning_rate": 1.6078325121411895e-06,
+      "loss": 0.5703,
+      "step": 18079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.251766296976343,
+      "learning_rate": 1.6075057701687995e-06,
+      "loss": 0.4508,
+      "step": 18080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2455278812628923,
+      "learning_rate": 1.6071791096833597e-06,
+      "loss": 0.2666,
+      "step": 18081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6659789869782526,
+      "learning_rate": 1.6068525306917064e-06,
+      "loss": 0.4265,
+      "step": 18082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6365283285274297,
+      "learning_rate": 1.6065260332006768e-06,
+      "loss": 0.4419,
+      "step": 18083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7918389376582808,
+      "learning_rate": 1.6061996172171078e-06,
+      "loss": 0.3977,
+      "step": 18084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7348281977561373,
+      "learning_rate": 1.6058732827478295e-06,
+      "loss": 0.4422,
+      "step": 18085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.605758738895417,
+      "learning_rate": 1.6055470297996761e-06,
+      "loss": 0.3607,
+      "step": 18086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4914750006904733,
+      "learning_rate": 1.605220858379478e-06,
+      "loss": 0.4013,
+      "step": 18087
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6695145579666673,
+      "learning_rate": 1.6048947684940603e-06,
+      "loss": 0.32,
+      "step": 18088
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6020331578301545,
+      "learning_rate": 1.6045687601502516e-06,
+      "loss": 0.3899,
+      "step": 18089
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7638223645439153,
+      "learning_rate": 1.604242833354876e-06,
+      "loss": 0.5162,
+      "step": 18090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0628648561638596,
+      "learning_rate": 1.6039169881147588e-06,
+      "loss": 0.4933,
+      "step": 18091
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6385249815651635,
+      "learning_rate": 1.6035912244367178e-06,
+      "loss": 0.3855,
+      "step": 18092
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5826772144576926,
+      "learning_rate": 1.6032655423275744e-06,
+      "loss": 0.4331,
+      "step": 18093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.335055845195989,
+      "learning_rate": 1.6029399417941477e-06,
+      "loss": 0.5148,
+      "step": 18094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8080970264890808,
+      "learning_rate": 1.6026144228432502e-06,
+      "loss": 0.6241,
+      "step": 18095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4516104862731276,
+      "learning_rate": 1.6022889854817018e-06,
+      "loss": 0.3406,
+      "step": 18096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5626241235245368,
+      "learning_rate": 1.6019636297163116e-06,
+      "loss": 0.4707,
+      "step": 18097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4264975819034436,
+      "learning_rate": 1.601638355553891e-06,
+      "loss": 0.3735,
+      "step": 18098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.759959694262101,
+      "learning_rate": 1.6013131630012502e-06,
+      "loss": 0.4089,
+      "step": 18099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2960824262565016,
+      "learning_rate": 1.6009880520651983e-06,
+      "loss": 0.3565,
+      "step": 18100
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.489234983921051,
+      "eval_runtime": 67.9309,
+      "eval_samples_per_second": 50.198,
+      "eval_steps_per_second": 1.575,
+      "step": 18100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8334448723976506,
+      "learning_rate": 1.6006630227525383e-06,
+      "loss": 0.4206,
+      "step": 18101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.935037459683274,
+      "learning_rate": 1.6003380750700758e-06,
+      "loss": 0.4935,
+      "step": 18102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5536779612377054,
+      "learning_rate": 1.6000132090246144e-06,
+      "loss": 0.554,
+      "step": 18103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4052523742289016,
+      "learning_rate": 1.5996884246229538e-06,
+      "loss": 0.3115,
+      "step": 18104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6396432808157264,
+      "learning_rate": 1.599363721871893e-06,
+      "loss": 0.2954,
+      "step": 18105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8214948884474915,
+      "learning_rate": 1.5990391007782312e-06,
+      "loss": 0.4516,
+      "step": 18106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.697813985343434,
+      "learning_rate": 1.5987145613487618e-06,
+      "loss": 0.5109,
+      "step": 18107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1476710281090896,
+      "learning_rate": 1.59839010359028e-06,
+      "loss": 0.4402,
+      "step": 18108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7528238727201257,
+      "learning_rate": 1.5980657275095774e-06,
+      "loss": 0.4354,
+      "step": 18109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7707392156071968,
+      "learning_rate": 1.5977414331134467e-06,
+      "loss": 0.5257,
+      "step": 18110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3293838806783986,
+      "learning_rate": 1.597417220408674e-06,
+      "loss": 0.3099,
+      "step": 18111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6996481847104343,
+      "learning_rate": 1.5970930894020482e-06,
+      "loss": 0.441,
+      "step": 18112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.110631308822521,
+      "learning_rate": 1.5967690401003552e-06,
+      "loss": 0.5425,
+      "step": 18113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.649602066527166,
+      "learning_rate": 1.596445072510377e-06,
+      "loss": 0.484,
+      "step": 18114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0163642604356036,
+      "learning_rate": 1.5961211866388962e-06,
+      "loss": 0.5475,
+      "step": 18115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.676883772014978,
+      "learning_rate": 1.595797382492695e-06,
+      "loss": 0.403,
+      "step": 18116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6884418677944715,
+      "learning_rate": 1.5954736600785498e-06,
+      "loss": 0.3587,
+      "step": 18117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6641915850003144,
+      "learning_rate": 1.5951500194032377e-06,
+      "loss": 0.3752,
+      "step": 18118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.170744101174386,
+      "learning_rate": 1.5948264604735363e-06,
+      "loss": 0.5495,
+      "step": 18119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1170778521255396,
+      "learning_rate": 1.5945029832962162e-06,
+      "loss": 0.5977,
+      "step": 18120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.686229392454998,
+      "learning_rate": 1.5941795878780497e-06,
+      "loss": 0.4438,
+      "step": 18121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6865859232542744,
+      "learning_rate": 1.5938562742258083e-06,
+      "loss": 0.4738,
+      "step": 18122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9146433087271313,
+      "learning_rate": 1.59353304234626e-06,
+      "loss": 0.354,
+      "step": 18123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8591328795626698,
+      "learning_rate": 1.5932098922461708e-06,
+      "loss": 0.4381,
+      "step": 18124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.542958683575325,
+      "learning_rate": 1.592886823932305e-06,
+      "loss": 0.3592,
+      "step": 18125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7203640465254677,
+      "learning_rate": 1.5925638374114283e-06,
+      "loss": 0.4325,
+      "step": 18126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7023291469918211,
+      "learning_rate": 1.5922409326902997e-06,
+      "loss": 0.399,
+      "step": 18127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.959436811854366,
+      "learning_rate": 1.5919181097756794e-06,
+      "loss": 0.4611,
+      "step": 18128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5456665856930814,
+      "learning_rate": 1.5915953686743274e-06,
+      "loss": 0.4581,
+      "step": 18129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1213911174179305,
+      "learning_rate": 1.5912727093929984e-06,
+      "loss": 0.5109,
+      "step": 18130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.762132386114437,
+      "learning_rate": 1.5909501319384468e-06,
+      "loss": 0.6164,
+      "step": 18131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0131768884073855,
+      "learning_rate": 1.590627636317427e-06,
+      "loss": 0.3412,
+      "step": 18132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.995235809900696,
+      "learning_rate": 1.5903052225366891e-06,
+      "loss": 0.5041,
+      "step": 18133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1482811352663527,
+      "learning_rate": 1.5899828906029824e-06,
+      "loss": 0.4602,
+      "step": 18134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.20224635229336,
+      "learning_rate": 1.5896606405230553e-06,
+      "loss": 0.4861,
+      "step": 18135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8980328372860586,
+      "learning_rate": 1.5893384723036554e-06,
+      "loss": 0.5447,
+      "step": 18136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8426709577464402,
+      "learning_rate": 1.589016385951524e-06,
+      "loss": 0.4746,
+      "step": 18137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8958960896954422,
+      "learning_rate": 1.5886943814734054e-06,
+      "loss": 0.54,
+      "step": 18138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8199890026127938,
+      "learning_rate": 1.588372458876042e-06,
+      "loss": 0.4496,
+      "step": 18139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.339169427570952,
+      "learning_rate": 1.58805061816617e-06,
+      "loss": 0.3552,
+      "step": 18140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7668478456327419,
+      "learning_rate": 1.5877288593505287e-06,
+      "loss": 0.5765,
+      "step": 18141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8407081130071188,
+      "learning_rate": 1.587407182435855e-06,
+      "loss": 0.4381,
+      "step": 18142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0383825195665426,
+      "learning_rate": 1.5870855874288805e-06,
+      "loss": 0.568,
+      "step": 18143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5386340886204293,
+      "learning_rate": 1.586764074336338e-06,
+      "loss": 0.4497,
+      "step": 18144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5725930275842945,
+      "learning_rate": 1.586442643164961e-06,
+      "loss": 0.3771,
+      "step": 18145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9137302914635255,
+      "learning_rate": 1.586121293921475e-06,
+      "loss": 0.4895,
+      "step": 18146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.931965578459393,
+      "learning_rate": 1.5858000266126092e-06,
+      "loss": 0.5024,
+      "step": 18147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9726535827493723,
+      "learning_rate": 1.5854788412450895e-06,
+      "loss": 0.487,
+      "step": 18148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7292160124434741,
+      "learning_rate": 1.585157737825638e-06,
+      "loss": 0.511,
+      "step": 18149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.809754210042552,
+      "learning_rate": 1.5848367163609773e-06,
+      "loss": 0.4879,
+      "step": 18150
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.4887146055698395,
+      "eval_runtime": 67.914,
+      "eval_samples_per_second": 50.211,
+      "eval_steps_per_second": 1.576,
+      "step": 18150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7244957669085446,
+      "learning_rate": 1.5845157768578295e-06,
+      "loss": 0.4802,
+      "step": 18151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6761759867165535,
+      "learning_rate": 1.5841949193229109e-06,
+      "loss": 0.4455,
+      "step": 18152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7502572463052848,
+      "learning_rate": 1.583874143762939e-06,
+      "loss": 0.5143,
+      "step": 18153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9922501779413322,
+      "learning_rate": 1.583553450184631e-06,
+      "loss": 0.5881,
+      "step": 18154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7406679155928992,
+      "learning_rate": 1.5832328385946979e-06,
+      "loss": 0.5683,
+      "step": 18155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6351535636698575,
+      "learning_rate": 1.5829123089998524e-06,
+      "loss": 0.3612,
+      "step": 18156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.812130595253507,
+      "learning_rate": 1.5825918614068048e-06,
+      "loss": 0.4821,
+      "step": 18157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8315020504698165,
+      "learning_rate": 1.582271495822264e-06,
+      "loss": 0.4531,
+      "step": 18158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.188142781585546,
+      "learning_rate": 1.5819512122529352e-06,
+      "loss": 0.7711,
+      "step": 18159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9929311747994105,
+      "learning_rate": 1.5816310107055243e-06,
+      "loss": 0.5753,
+      "step": 18160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.553610649845305,
+      "learning_rate": 1.5813108911867358e-06,
+      "loss": 0.4376,
+      "step": 18161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.673332088353959,
+      "learning_rate": 1.5809908537032687e-06,
+      "loss": 0.3942,
+      "step": 18162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9523802379539668,
+      "learning_rate": 1.5806708982618232e-06,
+      "loss": 0.62,
+      "step": 18163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6562358262161208,
+      "learning_rate": 1.5803510248690992e-06,
+      "loss": 0.5341,
+      "step": 18164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.615861062370222,
+      "learning_rate": 1.5800312335317912e-06,
+      "loss": 0.5108,
+      "step": 18165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5818629305470846,
+      "learning_rate": 1.579711524256595e-06,
+      "loss": 0.4064,
+      "step": 18166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6679342031822801,
+      "learning_rate": 1.5793918970502034e-06,
+      "loss": 0.4506,
+      "step": 18167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.280052574446427,
+      "learning_rate": 1.5790723519193065e-06,
+      "loss": 0.6044,
+      "step": 18168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6731546617467135,
+      "learning_rate": 1.5787528888705945e-06,
+      "loss": 0.3727,
+      "step": 18169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8570320022536402,
+      "learning_rate": 1.578433507910755e-06,
+      "loss": 0.4239,
+      "step": 18170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.740204940094752,
+      "learning_rate": 1.578114209046476e-06,
+      "loss": 0.4069,
+      "step": 18171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8087046762733128,
+      "learning_rate": 1.5777949922844383e-06,
+      "loss": 0.457,
+      "step": 18172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8217622155026623,
+      "learning_rate": 1.5774758576313264e-06,
+      "loss": 0.3743,
+      "step": 18173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8709224041182169,
+      "learning_rate": 1.577156805093822e-06,
+      "loss": 0.542,
+      "step": 18174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0159490979676105,
+      "learning_rate": 1.5768378346786027e-06,
+      "loss": 0.4261,
+      "step": 18175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9308576141043967,
+      "learning_rate": 1.576518946392346e-06,
+      "loss": 0.5691,
+      "step": 18176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7554409708998198,
+      "learning_rate": 1.5762001402417294e-06,
+      "loss": 0.4576,
+      "step": 18177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6495511535079463,
+      "learning_rate": 1.5758814162334249e-06,
+      "loss": 0.4321,
+      "step": 18178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9851218648060034,
+      "learning_rate": 1.5755627743741053e-06,
+      "loss": 0.5046,
+      "step": 18179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5525996896408396,
+      "learning_rate": 1.5752442146704427e-06,
+      "loss": 0.3902,
+      "step": 18180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3665833796934916,
+      "learning_rate": 1.5749257371291037e-06,
+      "loss": 0.3344,
+      "step": 18181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7811719538650292,
+      "learning_rate": 1.5746073417567564e-06,
+      "loss": 0.5258,
+      "step": 18182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.393789491820379,
+      "learning_rate": 1.574289028560066e-06,
+      "loss": 0.4654,
+      "step": 18183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5082331104489282,
+      "learning_rate": 1.573970797545698e-06,
+      "loss": 0.3723,
+      "step": 18184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.581170873982437,
+      "learning_rate": 1.5736526487203113e-06,
+      "loss": 0.4292,
+      "step": 18185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8943386700355707,
+      "learning_rate": 1.5733345820905676e-06,
+      "loss": 0.5069,
+      "step": 18186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.540294631515973,
+      "learning_rate": 1.573016597663127e-06,
+      "loss": 0.4646,
+      "step": 18187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9245852125161311,
+      "learning_rate": 1.5726986954446432e-06,
+      "loss": 0.5263,
+      "step": 18188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7137758492810558,
+      "learning_rate": 1.5723808754417736e-06,
+      "loss": 0.4531,
+      "step": 18189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7141895928477777,
+      "learning_rate": 1.5720631376611712e-06,
+      "loss": 0.4604,
+      "step": 18190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.764456884415055,
+      "learning_rate": 1.5717454821094866e-06,
+      "loss": 0.4702,
+      "step": 18191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0792608468660303,
+      "learning_rate": 1.5714279087933704e-06,
+      "loss": 0.5638,
+      "step": 18192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8995879071359485,
+      "learning_rate": 1.571110417719471e-06,
+      "loss": 0.6066,
+      "step": 18193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.576537244259019,
+      "learning_rate": 1.5707930088944337e-06,
+      "loss": 0.5272,
+      "step": 18194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.655017375607884,
+      "learning_rate": 1.570475682324905e-06,
+      "loss": 0.4071,
+      "step": 18195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7382513509471407,
+      "learning_rate": 1.570158438017528e-06,
+      "loss": 0.4414,
+      "step": 18196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3073911626703765,
+      "learning_rate": 1.5698412759789421e-06,
+      "loss": 0.2478,
+      "step": 18197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5352196814126258,
+      "learning_rate": 1.5695241962157887e-06,
+      "loss": 0.3439,
+      "step": 18198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.041889940675486,
+      "learning_rate": 1.569207198734705e-06,
+      "loss": 0.4848,
+      "step": 18199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6562877714512567,
+      "learning_rate": 1.5688902835423268e-06,
+      "loss": 0.4069,
+      "step": 18200
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.4887249767780304,
+      "eval_runtime": 67.9445,
+      "eval_samples_per_second": 50.188,
+      "eval_steps_per_second": 1.575,
+      "step": 18200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8394681366625962,
+      "learning_rate": 1.5685734506452885e-06,
+      "loss": 0.5476,
+      "step": 18201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9154260162727383,
+      "learning_rate": 1.568256700050223e-06,
+      "loss": 0.4825,
+      "step": 18202
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7458709446483285,
+      "learning_rate": 1.5679400317637625e-06,
+      "loss": 0.4993,
+      "step": 18203
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6757149474128181,
+      "learning_rate": 1.567623445792534e-06,
+      "loss": 0.5519,
+      "step": 18204
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0183975568474675,
+      "learning_rate": 1.5673069421431664e-06,
+      "loss": 0.5067,
+      "step": 18205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.995160102196782,
+      "learning_rate": 1.566990520822286e-06,
+      "loss": 0.5441,
+      "step": 18206
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.65834074669724,
+      "learning_rate": 1.5666741818365156e-06,
+      "loss": 0.4744,
+      "step": 18207
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6748203592133486,
+      "learning_rate": 1.5663579251924778e-06,
+      "loss": 0.5682,
+      "step": 18208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5235628822831697,
+      "learning_rate": 1.5660417508967946e-06,
+      "loss": 0.3878,
+      "step": 18209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9328966936087169,
+      "learning_rate": 1.5657256589560831e-06,
+      "loss": 0.5247,
+      "step": 18210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.112554012312721,
+      "learning_rate": 1.5654096493769608e-06,
+      "loss": 0.5451,
+      "step": 18211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7568545865953304,
+      "learning_rate": 1.565093722166045e-06,
+      "loss": 0.5078,
+      "step": 18212
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.680480655855263,
+      "learning_rate": 1.5647778773299468e-06,
+      "loss": 0.3726,
+      "step": 18213
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7788687014087492,
+      "learning_rate": 1.56446211487528e-06,
+      "loss": 0.474,
+      "step": 18214
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9407113681429584,
+      "learning_rate": 1.5641464348086537e-06,
+      "loss": 0.5322,
+      "step": 18215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9321739485693756,
+      "learning_rate": 1.5638308371366784e-06,
+      "loss": 0.4536,
+      "step": 18216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1472687647469693,
+      "learning_rate": 1.5635153218659588e-06,
+      "loss": 0.4896,
+      "step": 18217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0397917446886304,
+      "learning_rate": 1.563199889003101e-06,
+      "loss": 0.549,
+      "step": 18218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.665093333995485,
+      "learning_rate": 1.5628845385547093e-06,
+      "loss": 0.4453,
+      "step": 18219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6369309727787453,
+      "learning_rate": 1.5625692705273828e-06,
+      "loss": 0.3169,
+      "step": 18220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9299609979169783,
+      "learning_rate": 1.5622540849277238e-06,
+      "loss": 0.5214,
+      "step": 18221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.46681140341588,
+      "learning_rate": 1.5619389817623297e-06,
+      "loss": 0.4155,
+      "step": 18222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8461267619466437,
+      "learning_rate": 1.5616239610377967e-06,
+      "loss": 0.482,
+      "step": 18223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9682491829520108,
+      "learning_rate": 1.56130902276072e-06,
+      "loss": 0.4408,
+      "step": 18224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5432650188643116,
+      "learning_rate": 1.5609941669376932e-06,
+      "loss": 0.4226,
+      "step": 18225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.015394119817658,
+      "learning_rate": 1.5606793935753059e-06,
+      "loss": 0.4743,
+      "step": 18226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.422078090487435,
+      "learning_rate": 1.5603647026801486e-06,
+      "loss": 0.4437,
+      "step": 18227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.741383185063499,
+      "learning_rate": 1.5600500942588093e-06,
+      "loss": 0.3811,
+      "step": 18228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6881545446687969,
+      "learning_rate": 1.5597355683178753e-06,
+      "loss": 0.3726,
+      "step": 18229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6396416678030543,
+      "learning_rate": 1.5594211248639285e-06,
+      "loss": 0.332,
+      "step": 18230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.687229368954768,
+      "learning_rate": 1.5591067639035532e-06,
+      "loss": 0.3696,
+      "step": 18231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8075505851269607,
+      "learning_rate": 1.5587924854433307e-06,
+      "loss": 0.4048,
+      "step": 18232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4012841959949163,
+      "learning_rate": 1.5584782894898386e-06,
+      "loss": 0.3949,
+      "step": 18233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8397220198226052,
+      "learning_rate": 1.5581641760496552e-06,
+      "loss": 0.4348,
+      "step": 18234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1713604879604538,
+      "learning_rate": 1.5578501451293576e-06,
+      "loss": 0.5062,
+      "step": 18235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2183951960390964,
+      "learning_rate": 1.557536196735518e-06,
+      "loss": 0.2441,
+      "step": 18236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5222949297764523,
+      "learning_rate": 1.557222330874709e-06,
+      "loss": 0.5886,
+      "step": 18237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.140653957186249,
+      "learning_rate": 1.5569085475535026e-06,
+      "loss": 0.6031,
+      "step": 18238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.855790842571412,
+      "learning_rate": 1.5565948467784656e-06,
+      "loss": 0.5157,
+      "step": 18239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9966101358200392,
+      "learning_rate": 1.5562812285561662e-06,
+      "loss": 0.4881,
+      "step": 18240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5270436904790319,
+      "learning_rate": 1.55596769289317e-06,
+      "loss": 0.4154,
+      "step": 18241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.06806237446527,
+      "learning_rate": 1.5556542397960396e-06,
+      "loss": 0.5324,
+      "step": 18242
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9472001587673724,
+      "learning_rate": 1.5553408692713386e-06,
+      "loss": 0.5591,
+      "step": 18243
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7605887944777452,
+      "learning_rate": 1.555027581325627e-06,
+      "loss": 0.426,
+      "step": 18244
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6317476289704915,
+      "learning_rate": 1.5547143759654616e-06,
+      "loss": 0.4283,
+      "step": 18245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7519542429645099,
+      "learning_rate": 1.5544012531974003e-06,
+      "loss": 0.4178,
+      "step": 18246
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8835787897172833,
+      "learning_rate": 1.554088213027999e-06,
+      "loss": 0.4417,
+      "step": 18247
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5381549480523389,
+      "learning_rate": 1.5537752554638096e-06,
+      "loss": 0.339,
+      "step": 18248
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9293368377194886,
+      "learning_rate": 1.5534623805113836e-06,
+      "loss": 0.6311,
+      "step": 18249
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6387128476799488,
+      "learning_rate": 1.5531495881772716e-06,
+      "loss": 0.4399,
+      "step": 18250
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.4883028566837311,
+      "eval_runtime": 67.9658,
+      "eval_samples_per_second": 50.172,
+      "eval_steps_per_second": 1.574,
+      "step": 18250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.627805316112722,
+      "learning_rate": 1.5528368784680225e-06,
+      "loss": 0.4011,
+      "step": 18251
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.720032291173241,
+      "learning_rate": 1.55252425139018e-06,
+      "loss": 0.5092,
+      "step": 18252
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.883461955200488,
+      "learning_rate": 1.5522117069502912e-06,
+      "loss": 0.4382,
+      "step": 18253
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.248162539221375,
+      "learning_rate": 1.5518992451548993e-06,
+      "loss": 0.4966,
+      "step": 18254
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8429131692680742,
+      "learning_rate": 1.5515868660105431e-06,
+      "loss": 0.4645,
+      "step": 18255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8145701175841682,
+      "learning_rate": 1.5512745695237635e-06,
+      "loss": 0.4668,
+      "step": 18256
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5731489516694959,
+      "learning_rate": 1.550962355701099e-06,
+      "loss": 0.4484,
+      "step": 18257
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.618334662627029,
+      "learning_rate": 1.550650224549084e-06,
+      "loss": 0.4832,
+      "step": 18258
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7572510323931476,
+      "learning_rate": 1.5503381760742537e-06,
+      "loss": 0.5822,
+      "step": 18259
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.877002621264424,
+      "learning_rate": 1.5500262102831413e-06,
+      "loss": 0.4164,
+      "step": 18260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3520771133911906,
+      "learning_rate": 1.5497143271822753e-06,
+      "loss": 0.5039,
+      "step": 18261
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7392216281095971,
+      "learning_rate": 1.5494025267781865e-06,
+      "loss": 0.5143,
+      "step": 18262
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8817520880740317,
+      "learning_rate": 1.5490908090774024e-06,
+      "loss": 0.5793,
+      "step": 18263
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.127050812117998,
+      "learning_rate": 1.5487791740864488e-06,
+      "loss": 0.7346,
+      "step": 18264
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4437342427348816,
+      "learning_rate": 1.5484676218118477e-06,
+      "loss": 0.3135,
+      "step": 18265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6917738691319402,
+      "learning_rate": 1.5481561522601234e-06,
+      "loss": 0.4777,
+      "step": 18266
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5486210236923577,
+      "learning_rate": 1.5478447654377959e-06,
+      "loss": 0.4423,
+      "step": 18267
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6901445072210353,
+      "learning_rate": 1.5475334613513827e-06,
+      "loss": 0.3847,
+      "step": 18268
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7909264729527985,
+      "learning_rate": 1.5472222400074011e-06,
+      "loss": 0.482,
+      "step": 18269
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6735687845938976,
+      "learning_rate": 1.5469111014123684e-06,
+      "loss": 0.4724,
+      "step": 18270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1218486422904017,
+      "learning_rate": 1.546600045572795e-06,
+      "loss": 0.5964,
+      "step": 18271
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.93241312968579,
+      "learning_rate": 1.546289072495194e-06,
+      "loss": 0.54,
+      "step": 18272
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7848890092699234,
+      "learning_rate": 1.5459781821860764e-06,
+      "loss": 0.4316,
+      "step": 18273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.653903146394871,
+      "learning_rate": 1.5456673746519493e-06,
+      "loss": 0.4733,
+      "step": 18274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1604850903299564,
+      "learning_rate": 1.5453566498993188e-06,
+      "loss": 0.4918,
+      "step": 18275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6773699263533246,
+      "learning_rate": 1.5450460079346915e-06,
+      "loss": 0.4453,
+      "step": 18276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.650947780941871,
+      "learning_rate": 1.5447354487645697e-06,
+      "loss": 0.4228,
+      "step": 18277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6420595043183974,
+      "learning_rate": 1.544424972395454e-06,
+      "loss": 0.4263,
+      "step": 18278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5682768502171158,
+      "learning_rate": 1.5441145788338441e-06,
+      "loss": 0.4097,
+      "step": 18279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.852060975206656,
+      "learning_rate": 1.54380426808624e-06,
+      "loss": 0.4319,
+      "step": 18280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.624934521328875,
+      "learning_rate": 1.5434940401591347e-06,
+      "loss": 0.3302,
+      "step": 18281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4668563126956276,
+      "learning_rate": 1.543183895059025e-06,
+      "loss": 0.323,
+      "step": 18282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5187804853351035,
+      "learning_rate": 1.5428738327924032e-06,
+      "loss": 0.4979,
+      "step": 18283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.022255086123948,
+      "learning_rate": 1.5425638533657583e-06,
+      "loss": 0.5465,
+      "step": 18284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7788689436504703,
+      "learning_rate": 1.542253956785582e-06,
+      "loss": 0.4643,
+      "step": 18285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8466050714130857,
+      "learning_rate": 1.5419441430583603e-06,
+      "loss": 0.5299,
+      "step": 18286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7402544705226743,
+      "learning_rate": 1.5416344121905808e-06,
+      "loss": 0.627,
+      "step": 18287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8333507284648303,
+      "learning_rate": 1.5413247641887253e-06,
+      "loss": 0.3656,
+      "step": 18288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3951308254732946,
+      "learning_rate": 1.541015199059277e-06,
+      "loss": 0.4045,
+      "step": 18289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.048607763839476,
+      "learning_rate": 1.5407057168087177e-06,
+      "loss": 0.5369,
+      "step": 18290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.73257457388413,
+      "learning_rate": 1.5403963174435236e-06,
+      "loss": 0.3965,
+      "step": 18291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.060018929897149,
+      "learning_rate": 1.5400870009701733e-06,
+      "loss": 0.5955,
+      "step": 18292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6349432884170831,
+      "learning_rate": 1.5397777673951421e-06,
+      "loss": 0.3641,
+      "step": 18293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3950466865023738,
+      "learning_rate": 1.5394686167249036e-06,
+      "loss": 0.2836,
+      "step": 18294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8658563700703585,
+      "learning_rate": 1.5391595489659295e-06,
+      "loss": 0.6735,
+      "step": 18295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9354923077782236,
+      "learning_rate": 1.5388505641246911e-06,
+      "loss": 0.5434,
+      "step": 18296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1630461498339795,
+      "learning_rate": 1.5385416622076551e-06,
+      "loss": 0.6134,
+      "step": 18297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9033854593076125,
+      "learning_rate": 1.5382328432212884e-06,
+      "loss": 0.5214,
+      "step": 18298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8709508662988363,
+      "learning_rate": 1.5379241071720574e-06,
+      "loss": 0.4432,
+      "step": 18299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9424291200748676,
+      "learning_rate": 1.5376154540664235e-06,
+      "loss": 0.5083,
+      "step": 18300
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.48745080828666687,
+      "eval_runtime": 67.965,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 18300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.805024752266619,
+      "learning_rate": 1.5373068839108492e-06,
+      "loss": 0.4589,
+      "step": 18301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7570851133622634,
+      "learning_rate": 1.5369983967117944e-06,
+      "loss": 0.4273,
+      "step": 18302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7142633757725136,
+      "learning_rate": 1.536689992475716e-06,
+      "loss": 0.3952,
+      "step": 18303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6366683458798257,
+      "learning_rate": 1.5363816712090712e-06,
+      "loss": 0.4199,
+      "step": 18304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.755364259479191,
+      "learning_rate": 1.5360734329183148e-06,
+      "loss": 0.4985,
+      "step": 18305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8194753664606638,
+      "learning_rate": 1.5357652776098975e-06,
+      "loss": 0.4939,
+      "step": 18306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6155413717584588,
+      "learning_rate": 1.5354572052902727e-06,
+      "loss": 0.4288,
+      "step": 18307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5184962793498062,
+      "learning_rate": 1.5351492159658882e-06,
+      "loss": 0.5531,
+      "step": 18308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.867933004203818,
+      "learning_rate": 1.5348413096431938e-06,
+      "loss": 0.4548,
+      "step": 18309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5568892985028502,
+      "learning_rate": 1.534533486328632e-06,
+      "loss": 0.4485,
+      "step": 18310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.653091228250222,
+      "learning_rate": 1.534225746028649e-06,
+      "loss": 0.3804,
+      "step": 18311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8312627412594062,
+      "learning_rate": 1.5339180887496881e-06,
+      "loss": 0.4958,
+      "step": 18312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8838009124320425,
+      "learning_rate": 1.5336105144981871e-06,
+      "loss": 0.6034,
+      "step": 18313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8210942722138568,
+      "learning_rate": 1.5333030232805871e-06,
+      "loss": 0.5058,
+      "step": 18314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7485842601053851,
+      "learning_rate": 1.5329956151033252e-06,
+      "loss": 0.5655,
+      "step": 18315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8463813239392997,
+      "learning_rate": 1.5326882899728342e-06,
+      "loss": 0.409,
+      "step": 18316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.982476272326636,
+      "learning_rate": 1.5323810478955507e-06,
+      "loss": 0.4596,
+      "step": 18317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.749069669901905,
+      "learning_rate": 1.5320738888779063e-06,
+      "loss": 0.3442,
+      "step": 18318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1180126137181428,
+      "learning_rate": 1.5317668129263292e-06,
+      "loss": 0.5964,
+      "step": 18319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7683019104523474,
+      "learning_rate": 1.5314598200472494e-06,
+      "loss": 0.3453,
+      "step": 18320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.875186118758257,
+      "learning_rate": 1.531152910247093e-06,
+      "loss": 0.4395,
+      "step": 18321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8511953096583216,
+      "learning_rate": 1.5308460835322867e-06,
+      "loss": 0.4797,
+      "step": 18322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.424513138953014,
+      "learning_rate": 1.530539339909251e-06,
+      "loss": 0.3316,
+      "step": 18323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4506305013620606,
+      "learning_rate": 1.5302326793844092e-06,
+      "loss": 0.3928,
+      "step": 18324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9717642227152736,
+      "learning_rate": 1.5299261019641809e-06,
+      "loss": 0.4955,
+      "step": 18325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7457711244807392,
+      "learning_rate": 1.5296196076549833e-06,
+      "loss": 0.5661,
+      "step": 18326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8258372915101881,
+      "learning_rate": 1.5293131964632327e-06,
+      "loss": 0.4565,
+      "step": 18327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.647964287542992,
+      "learning_rate": 1.5290068683953452e-06,
+      "loss": 0.4516,
+      "step": 18328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6843387858028631,
+      "learning_rate": 1.5287006234577312e-06,
+      "loss": 0.4961,
+      "step": 18329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7941620749502452,
+      "learning_rate": 1.528394461656803e-06,
+      "loss": 0.5295,
+      "step": 18330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7250197508413627,
+      "learning_rate": 1.528088382998971e-06,
+      "loss": 0.4984,
+      "step": 18331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7034894813746893,
+      "learning_rate": 1.5277823874906408e-06,
+      "loss": 0.5396,
+      "step": 18332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1807506579717097,
+      "learning_rate": 1.5274764751382194e-06,
+      "loss": 0.6058,
+      "step": 18333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0696794037650643,
+      "learning_rate": 1.5271706459481101e-06,
+      "loss": 0.5218,
+      "step": 18334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8351570811943059,
+      "learning_rate": 1.5268648999267163e-06,
+      "loss": 0.5074,
+      "step": 18335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.485846222912594,
+      "learning_rate": 1.526559237080438e-06,
+      "loss": 0.3842,
+      "step": 18336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3447529090575763,
+      "learning_rate": 1.5262536574156734e-06,
+      "loss": 0.4734,
+      "step": 18337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7895465659400331,
+      "learning_rate": 1.5259481609388215e-06,
+      "loss": 0.4194,
+      "step": 18338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7772001537200823,
+      "learning_rate": 1.5256427476562754e-06,
+      "loss": 0.4636,
+      "step": 18339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.833807935728796,
+      "learning_rate": 1.5253374175744301e-06,
+      "loss": 0.4897,
+      "step": 18340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6521759767869006,
+      "learning_rate": 1.525032170699677e-06,
+      "loss": 0.534,
+      "step": 18341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.619537196927596,
+      "learning_rate": 1.5247270070384066e-06,
+      "loss": 0.5199,
+      "step": 18342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.045884606991698,
+      "learning_rate": 1.524421926597007e-06,
+      "loss": 0.508,
+      "step": 18343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.105354907681282,
+      "learning_rate": 1.5241169293818658e-06,
+      "loss": 0.5275,
+      "step": 18344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9695868906085447,
+      "learning_rate": 1.5238120153993663e-06,
+      "loss": 0.3879,
+      "step": 18345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9725346040890503,
+      "learning_rate": 1.5235071846558927e-06,
+      "loss": 0.3902,
+      "step": 18346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.842221905300136,
+      "learning_rate": 1.5232024371578274e-06,
+      "loss": 0.4728,
+      "step": 18347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7413015707088182,
+      "learning_rate": 1.5228977729115479e-06,
+      "loss": 0.4133,
+      "step": 18348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3799947494345686,
+      "learning_rate": 1.5225931919234332e-06,
+      "loss": 0.6258,
+      "step": 18349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7471009068783245,
+      "learning_rate": 1.5222886941998605e-06,
+      "loss": 0.4254,
+      "step": 18350
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.48753178119659424,
+      "eval_runtime": 68.0175,
+      "eval_samples_per_second": 50.134,
+      "eval_steps_per_second": 1.573,
+      "step": 18350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6753413500810577,
+      "learning_rate": 1.5219842797472025e-06,
+      "loss": 0.4605,
+      "step": 18351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7160962159499331,
+      "learning_rate": 1.5216799485718329e-06,
+      "loss": 0.5092,
+      "step": 18352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6037266056113695,
+      "learning_rate": 1.521375700680123e-06,
+      "loss": 0.3995,
+      "step": 18353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.880347702385222,
+      "learning_rate": 1.5210715360784412e-06,
+      "loss": 0.5737,
+      "step": 18354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6541520405271566,
+      "learning_rate": 1.5207674547731554e-06,
+      "loss": 0.4883,
+      "step": 18355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.939811264598828,
+      "learning_rate": 1.5204634567706309e-06,
+      "loss": 0.5203,
+      "step": 18356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9119792393516626,
+      "learning_rate": 1.5201595420772334e-06,
+      "loss": 0.4991,
+      "step": 18357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9897718379812648,
+      "learning_rate": 1.5198557106993226e-06,
+      "loss": 0.3876,
+      "step": 18358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5799270671761259,
+      "learning_rate": 1.5195519626432606e-06,
+      "loss": 0.5012,
+      "step": 18359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6815707733016751,
+      "learning_rate": 1.519248297915407e-06,
+      "loss": 0.4479,
+      "step": 18360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.320520915821955,
+      "learning_rate": 1.5189447165221166e-06,
+      "loss": 0.3449,
+      "step": 18361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.909741628149111,
+      "learning_rate": 1.5186412184697458e-06,
+      "loss": 0.5929,
+      "step": 18362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8884849963137582,
+      "learning_rate": 1.518337803764649e-06,
+      "loss": 0.5641,
+      "step": 18363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.50079527448848,
+      "learning_rate": 1.5180344724131768e-06,
+      "loss": 0.368,
+      "step": 18364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9539167727963336,
+      "learning_rate": 1.5177312244216791e-06,
+      "loss": 0.6593,
+      "step": 18365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2253252222946704,
+      "learning_rate": 1.517428059796505e-06,
+      "loss": 0.5478,
+      "step": 18366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7784295420147895,
+      "learning_rate": 1.5171249785440019e-06,
+      "loss": 0.4668,
+      "step": 18367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.137905363230241,
+      "learning_rate": 1.5168219806705123e-06,
+      "loss": 0.4845,
+      "step": 18368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5300999436562726,
+      "learning_rate": 1.5165190661823805e-06,
+      "loss": 0.315,
+      "step": 18369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.021627993256453,
+      "learning_rate": 1.5162162350859488e-06,
+      "loss": 0.5015,
+      "step": 18370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9250994319723629,
+      "learning_rate": 1.515913487387555e-06,
+      "loss": 0.4154,
+      "step": 18371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3760159972877368,
+      "learning_rate": 1.5156108230935371e-06,
+      "loss": 0.3055,
+      "step": 18372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.132629316428909,
+      "learning_rate": 1.5153082422102335e-06,
+      "loss": 0.4469,
+      "step": 18373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5846485406775763,
+      "learning_rate": 1.515005744743976e-06,
+      "loss": 0.4381,
+      "step": 18374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7746049488153894,
+      "learning_rate": 1.5147033307010973e-06,
+      "loss": 0.3868,
+      "step": 18375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5649286835084544,
+      "learning_rate": 1.5144010000879304e-06,
+      "loss": 0.4463,
+      "step": 18376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.38767625159916,
+      "learning_rate": 1.514098752910802e-06,
+      "loss": 0.3132,
+      "step": 18377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8529431844001931,
+      "learning_rate": 1.5137965891760406e-06,
+      "loss": 0.4485,
+      "step": 18378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7188370202414793,
+      "learning_rate": 1.5134945088899716e-06,
+      "loss": 0.4978,
+      "step": 18379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6652267296717065,
+      "learning_rate": 1.5131925120589197e-06,
+      "loss": 0.4665,
+      "step": 18380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9700517608701869,
+      "learning_rate": 1.5128905986892057e-06,
+      "loss": 0.4844,
+      "step": 18381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5370720792104517,
+      "learning_rate": 1.5125887687871505e-06,
+      "loss": 0.4099,
+      "step": 18382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5956169001537945,
+      "learning_rate": 1.5122870223590736e-06,
+      "loss": 0.5071,
+      "step": 18383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.158061862295293,
+      "learning_rate": 1.5119853594112898e-06,
+      "loss": 0.553,
+      "step": 18384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6346834897750773,
+      "learning_rate": 1.511683779950116e-06,
+      "loss": 0.4764,
+      "step": 18385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9566319339959402,
+      "learning_rate": 1.5113822839818656e-06,
+      "loss": 0.4644,
+      "step": 18386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7142327994289377,
+      "learning_rate": 1.5110808715128484e-06,
+      "loss": 0.4849,
+      "step": 18387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3783537431443003,
+      "learning_rate": 1.5107795425493762e-06,
+      "loss": 0.388,
+      "step": 18388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6646400738686964,
+      "learning_rate": 1.5104782970977575e-06,
+      "loss": 0.4171,
+      "step": 18389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7191693143900928,
+      "learning_rate": 1.5101771351642949e-06,
+      "loss": 0.3751,
+      "step": 18390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.660485592985246,
+      "learning_rate": 1.5098760567552975e-06,
+      "loss": 0.5761,
+      "step": 18391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4912675320142825,
+      "learning_rate": 1.5095750618770672e-06,
+      "loss": 0.4035,
+      "step": 18392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4691516431657836,
+      "learning_rate": 1.5092741505359037e-06,
+      "loss": 0.4358,
+      "step": 18393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4106978500158236,
+      "learning_rate": 1.5089733227381071e-06,
+      "loss": 0.4139,
+      "step": 18394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.529418968456016,
+      "learning_rate": 1.5086725784899752e-06,
+      "loss": 0.6379,
+      "step": 18395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8987926238941262,
+      "learning_rate": 1.508371917797804e-06,
+      "loss": 0.422,
+      "step": 18396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.937308841958788,
+      "learning_rate": 1.5080713406678866e-06,
+      "loss": 0.6042,
+      "step": 18397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4693587297725261,
+      "learning_rate": 1.5077708471065172e-06,
+      "loss": 0.4238,
+      "step": 18398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6980072601501868,
+      "learning_rate": 1.507470437119985e-06,
+      "loss": 0.3899,
+      "step": 18399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6648434033803619,
+      "learning_rate": 1.507170110714579e-06,
+      "loss": 0.5318,
+      "step": 18400
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.4873120188713074,
+      "eval_runtime": 67.9899,
+      "eval_samples_per_second": 50.155,
+      "eval_steps_per_second": 1.574,
+      "step": 18400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.487629855651557,
+      "learning_rate": 1.5068698678965867e-06,
+      "loss": 0.4071,
+      "step": 18401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5885950455100808,
+      "learning_rate": 1.5065697086722946e-06,
+      "loss": 0.4615,
+      "step": 18402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4486707848283003,
+      "learning_rate": 1.5062696330479841e-06,
+      "loss": 0.3693,
+      "step": 18403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8762848203511464,
+      "learning_rate": 1.5059696410299385e-06,
+      "loss": 0.4775,
+      "step": 18404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8433266246135522,
+      "learning_rate": 1.5056697326244386e-06,
+      "loss": 0.5119,
+      "step": 18405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8349095825988038,
+      "learning_rate": 1.505369907837761e-06,
+      "loss": 0.3523,
+      "step": 18406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0723765848652507,
+      "learning_rate": 1.5050701666761824e-06,
+      "loss": 0.4982,
+      "step": 18407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8729119585710283,
+      "learning_rate": 1.5047705091459805e-06,
+      "loss": 0.5538,
+      "step": 18408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8302421321136295,
+      "learning_rate": 1.5044709352534247e-06,
+      "loss": 0.5219,
+      "step": 18409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7694724345145316,
+      "learning_rate": 1.504171445004789e-06,
+      "loss": 0.479,
+      "step": 18410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6849177116456249,
+      "learning_rate": 1.5038720384063426e-06,
+      "loss": 0.5033,
+      "step": 18411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9033307908718244,
+      "learning_rate": 1.5035727154643523e-06,
+      "loss": 0.4923,
+      "step": 18412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6330012390800406,
+      "learning_rate": 1.5032734761850847e-06,
+      "loss": 0.3479,
+      "step": 18413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.379486061338672,
+      "learning_rate": 1.5029743205748043e-06,
+      "loss": 0.3149,
+      "step": 18414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.926014539530863,
+      "learning_rate": 1.5026752486397752e-06,
+      "loss": 0.4732,
+      "step": 18415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.050468187526245,
+      "learning_rate": 1.5023762603862558e-06,
+      "loss": 0.4881,
+      "step": 18416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8048670050967608,
+      "learning_rate": 1.5020773558205063e-06,
+      "loss": 0.3814,
+      "step": 18417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7969015436234164,
+      "learning_rate": 1.5017785349487851e-06,
+      "loss": 0.5801,
+      "step": 18418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6287371617163828,
+      "learning_rate": 1.5014797977773459e-06,
+      "loss": 0.4532,
+      "step": 18419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7676662972446413,
+      "learning_rate": 1.5011811443124441e-06,
+      "loss": 0.4899,
+      "step": 18420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6513516726762105,
+      "learning_rate": 1.5008825745603318e-06,
+      "loss": 0.6116,
+      "step": 18421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7132237881029075,
+      "learning_rate": 1.5005840885272574e-06,
+      "loss": 0.5124,
+      "step": 18422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8348759448345477,
+      "learning_rate": 1.5002856862194715e-06,
+      "loss": 0.4446,
+      "step": 18423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7028481470284393,
+      "learning_rate": 1.4999873676432214e-06,
+      "loss": 0.4418,
+      "step": 18424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7210754368032455,
+      "learning_rate": 1.4996891328047503e-06,
+      "loss": 0.483,
+      "step": 18425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.980366253124744,
+      "learning_rate": 1.4993909817103026e-06,
+      "loss": 0.6406,
+      "step": 18426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7945012772809659,
+      "learning_rate": 1.4990929143661197e-06,
+      "loss": 0.4283,
+      "step": 18427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.20988486273548,
+      "learning_rate": 1.4987949307784425e-06,
+      "loss": 0.5707,
+      "step": 18428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9162018117773951,
+      "learning_rate": 1.4984970309535075e-06,
+      "loss": 0.6095,
+      "step": 18429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8362102194840642,
+      "learning_rate": 1.4981992148975517e-06,
+      "loss": 0.5244,
+      "step": 18430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5374441713126465,
+      "learning_rate": 1.49790148261681e-06,
+      "loss": 0.4653,
+      "step": 18431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.417509677551532,
+      "learning_rate": 1.4976038341175148e-06,
+      "loss": 0.6533,
+      "step": 18432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7213944704466977,
+      "learning_rate": 1.4973062694058972e-06,
+      "loss": 0.4116,
+      "step": 18433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.488813601356728,
+      "learning_rate": 1.497008788488187e-06,
+      "loss": 0.3575,
+      "step": 18434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8143810658470834,
+      "learning_rate": 1.4967113913706111e-06,
+      "loss": 0.35,
+      "step": 18435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.616640803515855,
+      "learning_rate": 1.4964140780593954e-06,
+      "loss": 0.4727,
+      "step": 18436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5008084987940815,
+      "learning_rate": 1.4961168485607658e-06,
+      "loss": 0.5478,
+      "step": 18437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.380883468362571,
+      "learning_rate": 1.4958197028809418e-06,
+      "loss": 0.724,
+      "step": 18438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7167129772608059,
+      "learning_rate": 1.4955226410261443e-06,
+      "loss": 0.4888,
+      "step": 18439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.831836881110179,
+      "learning_rate": 1.4952256630025944e-06,
+      "loss": 0.4045,
+      "step": 18440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.487162767068395,
+      "learning_rate": 1.4949287688165076e-06,
+      "loss": 0.3933,
+      "step": 18441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.684980505861756,
+      "learning_rate": 1.4946319584740993e-06,
+      "loss": 0.4839,
+      "step": 18442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7858069206955516,
+      "learning_rate": 1.4943352319815837e-06,
+      "loss": 0.5538,
+      "step": 18443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6989135304215357,
+      "learning_rate": 1.494038589345171e-06,
+      "loss": 0.4343,
+      "step": 18444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.459104180565695,
+      "learning_rate": 1.4937420305710726e-06,
+      "loss": 0.3562,
+      "step": 18445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8721902297677093,
+      "learning_rate": 1.4934455556654961e-06,
+      "loss": 0.6135,
+      "step": 18446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.009702943916069,
+      "learning_rate": 1.4931491646346496e-06,
+      "loss": 0.6303,
+      "step": 18447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8181750290912067,
+      "learning_rate": 1.4928528574847353e-06,
+      "loss": 0.4678,
+      "step": 18448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9194559583217434,
+      "learning_rate": 1.4925566342219577e-06,
+      "loss": 0.4533,
+      "step": 18449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5021954930820431,
+      "learning_rate": 1.4922604948525187e-06,
+      "loss": 0.4396,
+      "step": 18450
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.48696228861808777,
+      "eval_runtime": 67.9714,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 18450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9728181332202073,
+      "learning_rate": 1.4919644393826164e-06,
+      "loss": 0.4764,
+      "step": 18451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9785061781888897,
+      "learning_rate": 1.4916684678184487e-06,
+      "loss": 0.666,
+      "step": 18452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7555679738781207,
+      "learning_rate": 1.4913725801662127e-06,
+      "loss": 0.5364,
+      "step": 18453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8617662206039822,
+      "learning_rate": 1.4910767764321014e-06,
+      "loss": 0.5956,
+      "step": 18454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.986508509427842,
+      "learning_rate": 1.4907810566223071e-06,
+      "loss": 0.4263,
+      "step": 18455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2161262995937734,
+      "learning_rate": 1.490485420743023e-06,
+      "loss": 0.264,
+      "step": 18456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.071069973222304,
+      "learning_rate": 1.4901898688004343e-06,
+      "loss": 0.4098,
+      "step": 18457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6878198852285378,
+      "learning_rate": 1.489894400800731e-06,
+      "loss": 0.4742,
+      "step": 18458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.234314084291364,
+      "learning_rate": 1.4895990167500974e-06,
+      "loss": 0.5162,
+      "step": 18459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8124906938281726,
+      "learning_rate": 1.4893037166547177e-06,
+      "loss": 0.4147,
+      "step": 18460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7001163110699233,
+      "learning_rate": 1.4890085005207736e-06,
+      "loss": 0.4501,
+      "step": 18461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3946705268225223,
+      "learning_rate": 1.4887133683544448e-06,
+      "loss": 0.4341,
+      "step": 18462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.883042404095535,
+      "learning_rate": 1.488418320161911e-06,
+      "loss": 0.4217,
+      "step": 18463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.540664758132168,
+      "learning_rate": 1.4881233559493468e-06,
+      "loss": 0.3595,
+      "step": 18464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3966870566151193,
+      "learning_rate": 1.4878284757229287e-06,
+      "loss": 0.3058,
+      "step": 18465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.63898860173338,
+      "learning_rate": 1.4875336794888304e-06,
+      "loss": 0.4049,
+      "step": 18466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.666537510861145,
+      "learning_rate": 1.487238967253221e-06,
+      "loss": 0.4461,
+      "step": 18467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.84393420543702,
+      "learning_rate": 1.4869443390222718e-06,
+      "loss": 0.3809,
+      "step": 18468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6930413853732393,
+      "learning_rate": 1.4866497948021507e-06,
+      "loss": 0.5005,
+      "step": 18469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7619863992985796,
+      "learning_rate": 1.4863553345990228e-06,
+      "loss": 0.5658,
+      "step": 18470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6729958873009199,
+      "learning_rate": 1.4860609584190527e-06,
+      "loss": 0.5749,
+      "step": 18471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7308790088055086,
+      "learning_rate": 1.4857666662684034e-06,
+      "loss": 0.4094,
+      "step": 18472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7435713723898514,
+      "learning_rate": 1.4854724581532366e-06,
+      "loss": 0.5262,
+      "step": 18473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8773998102246579,
+      "learning_rate": 1.4851783340797096e-06,
+      "loss": 0.416,
+      "step": 18474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7557020017114418,
+      "learning_rate": 1.48488429405398e-06,
+      "loss": 0.4791,
+      "step": 18475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.726185541457342,
+      "learning_rate": 1.4845903380822059e-06,
+      "loss": 0.5048,
+      "step": 18476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5796671764433674,
+      "learning_rate": 1.484296466170537e-06,
+      "loss": 0.4281,
+      "step": 18477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7503785577652688,
+      "learning_rate": 1.4840026783251277e-06,
+      "loss": 0.4285,
+      "step": 18478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5666956394778495,
+      "learning_rate": 1.4837089745521285e-06,
+      "loss": 0.4157,
+      "step": 18479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6006491128579337,
+      "learning_rate": 1.4834153548576864e-06,
+      "loss": 0.3641,
+      "step": 18480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8628167313797896,
+      "learning_rate": 1.4831218192479497e-06,
+      "loss": 0.4463,
+      "step": 18481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8102671751000505,
+      "learning_rate": 1.4828283677290628e-06,
+      "loss": 0.4873,
+      "step": 18482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6795234228755584,
+      "learning_rate": 1.4825350003071688e-06,
+      "loss": 0.4895,
+      "step": 18483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9195672162788469,
+      "learning_rate": 1.4822417169884086e-06,
+      "loss": 0.5022,
+      "step": 18484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.477733661477896,
+      "learning_rate": 1.481948517778923e-06,
+      "loss": 0.4747,
+      "step": 18485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0631583642430287,
+      "learning_rate": 1.4816554026848496e-06,
+      "loss": 0.4504,
+      "step": 18486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.655353623734069,
+      "learning_rate": 1.4813623717123243e-06,
+      "loss": 0.5476,
+      "step": 18487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.816213218778969,
+      "learning_rate": 1.4810694248674817e-06,
+      "loss": 0.4943,
+      "step": 18488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7640940199199244,
+      "learning_rate": 1.480776562156454e-06,
+      "loss": 0.4965,
+      "step": 18489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9644508714848739,
+      "learning_rate": 1.480483783585373e-06,
+      "loss": 0.5958,
+      "step": 18490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.830780339384569,
+      "learning_rate": 1.480191089160368e-06,
+      "loss": 0.6163,
+      "step": 18491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6509733048965027,
+      "learning_rate": 1.4798984788875647e-06,
+      "loss": 0.5864,
+      "step": 18492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0147907816281667,
+      "learning_rate": 1.4796059527730897e-06,
+      "loss": 0.5252,
+      "step": 18493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.151888503988618,
+      "learning_rate": 1.4793135108230673e-06,
+      "loss": 0.7225,
+      "step": 18494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7217805255181182,
+      "learning_rate": 1.47902115304362e-06,
+      "loss": 0.4791,
+      "step": 18495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7585098069009133,
+      "learning_rate": 1.4787288794408666e-06,
+      "loss": 0.4478,
+      "step": 18496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6425660448185198,
+      "learning_rate": 1.4784366900209257e-06,
+      "loss": 0.4344,
+      "step": 18497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7561202545230374,
+      "learning_rate": 1.4781445847899163e-06,
+      "loss": 0.4518,
+      "step": 18498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.606626365290943,
+      "learning_rate": 1.4778525637539515e-06,
+      "loss": 0.4795,
+      "step": 18499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7562082171323805,
+      "learning_rate": 1.477560626919144e-06,
+      "loss": 0.3999,
+      "step": 18500
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.48708176612854004,
+      "eval_runtime": 67.8739,
+      "eval_samples_per_second": 50.24,
+      "eval_steps_per_second": 1.576,
+      "step": 18500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7678332846147569,
+      "learning_rate": 1.4772687742916074e-06,
+      "loss": 0.5325,
+      "step": 18501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9684209168190958,
+      "learning_rate": 1.4769770058774502e-06,
+      "loss": 0.524,
+      "step": 18502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.026513224405927,
+      "learning_rate": 1.4766853216827803e-06,
+      "loss": 0.5651,
+      "step": 18503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5301287239215218,
+      "learning_rate": 1.4763937217137048e-06,
+      "loss": 0.337,
+      "step": 18504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0267127431939307,
+      "learning_rate": 1.4761022059763274e-06,
+      "loss": 0.4628,
+      "step": 18505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.431592186121744,
+      "learning_rate": 1.4758107744767506e-06,
+      "loss": 0.3444,
+      "step": 18506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.713287913509138,
+      "learning_rate": 1.4755194272210751e-06,
+      "loss": 0.3757,
+      "step": 18507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5595611402070586,
+      "learning_rate": 1.4752281642154021e-06,
+      "loss": 0.3858,
+      "step": 18508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9341834933871338,
+      "learning_rate": 1.474936985465827e-06,
+      "loss": 0.5253,
+      "step": 18509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.056903165485001,
+      "learning_rate": 1.4746458909784455e-06,
+      "loss": 0.9181,
+      "step": 18510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7967389372483609,
+      "learning_rate": 1.4743548807593532e-06,
+      "loss": 0.528,
+      "step": 18511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6670289945853063,
+      "learning_rate": 1.4740639548146406e-06,
+      "loss": 0.4856,
+      "step": 18512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.048317058554802,
+      "learning_rate": 1.4737731131503979e-06,
+      "loss": 0.5487,
+      "step": 18513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.186047149965525,
+      "learning_rate": 1.4734823557727153e-06,
+      "loss": 0.5901,
+      "step": 18514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8151572427700737,
+      "learning_rate": 1.473191682687678e-06,
+      "loss": 0.4988,
+      "step": 18515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7502854079247332,
+      "learning_rate": 1.4729010939013713e-06,
+      "loss": 0.4595,
+      "step": 18516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4222155936686123,
+      "learning_rate": 1.4726105894198798e-06,
+      "loss": 0.3006,
+      "step": 18517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5685126719375326,
+      "learning_rate": 1.4723201692492836e-06,
+      "loss": 0.4169,
+      "step": 18518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9649842697075905,
+      "learning_rate": 1.4720298333956628e-06,
+      "loss": 0.5279,
+      "step": 18519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4210598080489296,
+      "learning_rate": 1.4717395818650954e-06,
+      "loss": 0.3308,
+      "step": 18520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8887781895686067,
+      "learning_rate": 1.4714494146636587e-06,
+      "loss": 0.4622,
+      "step": 18521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.706054342920461,
+      "learning_rate": 1.4711593317974254e-06,
+      "loss": 0.4297,
+      "step": 18522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7965479023444426,
+      "learning_rate": 1.4708693332724693e-06,
+      "loss": 0.509,
+      "step": 18523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6271999200226899,
+      "learning_rate": 1.470579419094862e-06,
+      "loss": 0.4644,
+      "step": 18524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.243995938515282,
+      "learning_rate": 1.4702895892706706e-06,
+      "loss": 0.2509,
+      "step": 18525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5333524387493929,
+      "learning_rate": 1.469999843805964e-06,
+      "loss": 0.4416,
+      "step": 18526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.769520173398147,
+      "learning_rate": 1.469710182706808e-06,
+      "loss": 0.4384,
+      "step": 18527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9874234660014662,
+      "learning_rate": 1.4694206059792658e-06,
+      "loss": 0.5415,
+      "step": 18528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6363550457810103,
+      "learning_rate": 1.4691311136293992e-06,
+      "loss": 0.4344,
+      "step": 18529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9522525248642593,
+      "learning_rate": 1.4688417056632704e-06,
+      "loss": 0.4873,
+      "step": 18530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9068311875698623,
+      "learning_rate": 1.4685523820869355e-06,
+      "loss": 0.4894,
+      "step": 18531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1428329985467953,
+      "learning_rate": 1.468263142906452e-06,
+      "loss": 0.5577,
+      "step": 18532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.699500190524005,
+      "learning_rate": 1.467973988127876e-06,
+      "loss": 0.4868,
+      "step": 18533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7837736991653295,
+      "learning_rate": 1.467684917757261e-06,
+      "loss": 0.5238,
+      "step": 18534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4637959914229695,
+      "learning_rate": 1.4673959318006567e-06,
+      "loss": 0.4488,
+      "step": 18535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.569744534138187,
+      "learning_rate": 1.4671070302641135e-06,
+      "loss": 0.506,
+      "step": 18536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8256928200576987,
+      "learning_rate": 1.466818213153681e-06,
+      "loss": 0.4942,
+      "step": 18537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6080774106787221,
+      "learning_rate": 1.4665294804754026e-06,
+      "loss": 0.3962,
+      "step": 18538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9351320029323953,
+      "learning_rate": 1.4662408322353247e-06,
+      "loss": 0.4706,
+      "step": 18539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6813472631263984,
+      "learning_rate": 1.4659522684394906e-06,
+      "loss": 0.4897,
+      "step": 18540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8398610612967246,
+      "learning_rate": 1.465663789093939e-06,
+      "loss": 0.6712,
+      "step": 18541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.956721838497998,
+      "learning_rate": 1.4653753942047102e-06,
+      "loss": 0.4162,
+      "step": 18542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5115955872150233,
+      "learning_rate": 1.4650870837778422e-06,
+      "loss": 0.3822,
+      "step": 18543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.002324007226471,
+      "learning_rate": 1.4647988578193694e-06,
+      "loss": 0.6639,
+      "step": 18544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9384067386570951,
+      "learning_rate": 1.4645107163353262e-06,
+      "loss": 0.7193,
+      "step": 18545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6739120775399614,
+      "learning_rate": 1.4642226593317458e-06,
+      "loss": 0.426,
+      "step": 18546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.683472214654018,
+      "learning_rate": 1.4639346868146559e-06,
+      "loss": 0.4891,
+      "step": 18547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5137539002125995,
+      "learning_rate": 1.4636467987900865e-06,
+      "loss": 0.3643,
+      "step": 18548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.722295770790093,
+      "learning_rate": 1.4633589952640647e-06,
+      "loss": 0.6534,
+      "step": 18549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0010290017094188,
+      "learning_rate": 1.463071276242615e-06,
+      "loss": 0.5443,
+      "step": 18550
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.486773282289505,
+      "eval_runtime": 67.9203,
+      "eval_samples_per_second": 50.206,
+      "eval_steps_per_second": 1.575,
+      "step": 18550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6927398163432075,
+      "learning_rate": 1.4627836417317607e-06,
+      "loss": 0.4554,
+      "step": 18551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9948983918062808,
+      "learning_rate": 1.4624960917375231e-06,
+      "loss": 0.4864,
+      "step": 18552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5322933924112818,
+      "learning_rate": 1.4622086262659227e-06,
+      "loss": 0.4004,
+      "step": 18553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5656543776033287,
+      "learning_rate": 1.4619212453229756e-06,
+      "loss": 0.366,
+      "step": 18554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8840712460999496,
+      "learning_rate": 1.4616339489146997e-06,
+      "loss": 0.5676,
+      "step": 18555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6459576328003507,
+      "learning_rate": 1.4613467370471091e-06,
+      "loss": 0.521,
+      "step": 18556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6869356411456669,
+      "learning_rate": 1.4610596097262153e-06,
+      "loss": 0.4227,
+      "step": 18557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7645853048245217,
+      "learning_rate": 1.4607725669580303e-06,
+      "loss": 0.4293,
+      "step": 18558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7363957934180667,
+      "learning_rate": 1.4604856087485634e-06,
+      "loss": 0.5984,
+      "step": 18559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.732786876831471,
+      "learning_rate": 1.4601987351038202e-06,
+      "loss": 0.5478,
+      "step": 18560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0220323251772903,
+      "learning_rate": 1.4599119460298072e-06,
+      "loss": 0.6146,
+      "step": 18561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7006393251522824,
+      "learning_rate": 1.4596252415325296e-06,
+      "loss": 0.5412,
+      "step": 18562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9133872948636192,
+      "learning_rate": 1.4593386216179867e-06,
+      "loss": 0.3839,
+      "step": 18563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7299190446389754,
+      "learning_rate": 1.4590520862921804e-06,
+      "loss": 0.4811,
+      "step": 18564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.205019932902222,
+      "learning_rate": 1.4587656355611082e-06,
+      "loss": 0.5176,
+      "step": 18565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5079793203020069,
+      "learning_rate": 1.4584792694307687e-06,
+      "loss": 0.4335,
+      "step": 18566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9934528952112947,
+      "learning_rate": 1.4581929879071544e-06,
+      "loss": 0.7288,
+      "step": 18567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5668265415814338,
+      "learning_rate": 1.457906790996259e-06,
+      "loss": 0.3988,
+      "step": 18568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0190222587745237,
+      "learning_rate": 1.4576206787040759e-06,
+      "loss": 0.459,
+      "step": 18569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0909666307887305,
+      "learning_rate": 1.457334651036592e-06,
+      "loss": 0.4936,
+      "step": 18570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2940494976769785,
+      "learning_rate": 1.4570487079997961e-06,
+      "loss": 0.3937,
+      "step": 18571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1203834628657807,
+      "learning_rate": 1.4567628495996748e-06,
+      "loss": 0.4445,
+      "step": 18572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8140498219316858,
+      "learning_rate": 1.4564770758422111e-06,
+      "loss": 0.4894,
+      "step": 18573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6351540033242509,
+      "learning_rate": 1.4561913867333889e-06,
+      "loss": 0.3721,
+      "step": 18574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6076394615347918,
+      "learning_rate": 1.455905782279189e-06,
+      "loss": 0.4253,
+      "step": 18575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5976704945791942,
+      "learning_rate": 1.455620262485588e-06,
+      "loss": 0.5273,
+      "step": 18576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.705249464226741,
+      "learning_rate": 1.455334827358566e-06,
+      "loss": 0.454,
+      "step": 18577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6311267721606855,
+      "learning_rate": 1.4550494769040962e-06,
+      "loss": 0.3551,
+      "step": 18578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7679789034406053,
+      "learning_rate": 1.4547642111281545e-06,
+      "loss": 0.3998,
+      "step": 18579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9045710940105698,
+      "learning_rate": 1.4544790300367104e-06,
+      "loss": 0.5628,
+      "step": 18580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7818699670446498,
+      "learning_rate": 1.4541939336357352e-06,
+      "loss": 0.5568,
+      "step": 18581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2982271566333243,
+      "learning_rate": 1.453908921931198e-06,
+      "loss": 0.6012,
+      "step": 18582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8179797855976023,
+      "learning_rate": 1.4536239949290634e-06,
+      "loss": 0.4091,
+      "step": 18583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0300869815198843,
+      "learning_rate": 1.4533391526352972e-06,
+      "loss": 0.5369,
+      "step": 18584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6850702388183383,
+      "learning_rate": 1.4530543950558628e-06,
+      "loss": 0.4233,
+      "step": 18585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.389034280777242,
+      "learning_rate": 1.4527697221967204e-06,
+      "loss": 0.4058,
+      "step": 18586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8854348670999608,
+      "learning_rate": 1.4524851340638294e-06,
+      "loss": 0.5011,
+      "step": 18587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.874775581523451,
+      "learning_rate": 1.4522006306631497e-06,
+      "loss": 0.4836,
+      "step": 18588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7092323922230808,
+      "learning_rate": 1.4519162120006344e-06,
+      "loss": 0.4616,
+      "step": 18589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.642804321565294,
+      "learning_rate": 1.4516318780822395e-06,
+      "loss": 0.3784,
+      "step": 18590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.707885638049138,
+      "learning_rate": 1.4513476289139176e-06,
+      "loss": 0.5247,
+      "step": 18591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6134089549268007,
+      "learning_rate": 1.4510634645016167e-06,
+      "loss": 0.4336,
+      "step": 18592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6403712009163227,
+      "learning_rate": 1.4507793848512877e-06,
+      "loss": 0.5594,
+      "step": 18593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9318024893795456,
+      "learning_rate": 1.4504953899688785e-06,
+      "loss": 0.4543,
+      "step": 18594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6503801909299378,
+      "learning_rate": 1.4502114798603317e-06,
+      "loss": 0.3918,
+      "step": 18595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2662020843212507,
+      "learning_rate": 1.4499276545315922e-06,
+      "loss": 0.7086,
+      "step": 18596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0150976878330518,
+      "learning_rate": 1.4496439139886029e-06,
+      "loss": 0.5335,
+      "step": 18597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5026972537404315,
+      "learning_rate": 1.4493602582373008e-06,
+      "loss": 0.3834,
+      "step": 18598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8595819063390695,
+      "learning_rate": 1.4490766872836267e-06,
+      "loss": 0.4678,
+      "step": 18599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.62116365697443,
+      "learning_rate": 1.4487932011335155e-06,
+      "loss": 0.4478,
+      "step": 18600
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.48608872294425964,
+      "eval_runtime": 67.9286,
+      "eval_samples_per_second": 50.2,
+      "eval_steps_per_second": 1.575,
+      "step": 18600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.628766181436704,
+      "learning_rate": 1.4485097997929034e-06,
+      "loss": 0.548,
+      "step": 18601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7904585022649826,
+      "learning_rate": 1.4482264832677213e-06,
+      "loss": 0.438,
+      "step": 18602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8216696660006109,
+      "learning_rate": 1.4479432515639014e-06,
+      "loss": 0.4151,
+      "step": 18603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.562551636296128,
+      "learning_rate": 1.4476601046873732e-06,
+      "loss": 0.317,
+      "step": 18604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9275869227499465,
+      "learning_rate": 1.4473770426440623e-06,
+      "loss": 0.3212,
+      "step": 18605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6178765906171377,
+      "learning_rate": 1.4470940654398969e-06,
+      "loss": 0.373,
+      "step": 18606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.639686115557846,
+      "learning_rate": 1.4468111730807996e-06,
+      "loss": 0.3961,
+      "step": 18607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0370933438710215,
+      "learning_rate": 1.4465283655726929e-06,
+      "loss": 0.4291,
+      "step": 18608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.89112789023854,
+      "learning_rate": 1.446245642921497e-06,
+      "loss": 0.4159,
+      "step": 18609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1468875313283338,
+      "learning_rate": 1.4459630051331313e-06,
+      "loss": 0.6123,
+      "step": 18610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.257324311141186,
+      "learning_rate": 1.4456804522135114e-06,
+      "loss": 0.6782,
+      "step": 18611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.573025832163235,
+      "learning_rate": 1.445397984168553e-06,
+      "loss": 0.4749,
+      "step": 18612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.042982277769414,
+      "learning_rate": 1.4451156010041694e-06,
+      "loss": 0.4989,
+      "step": 18613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7304413938281773,
+      "learning_rate": 1.4448333027262725e-06,
+      "loss": 0.506,
+      "step": 18614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.475404979777099,
+      "learning_rate": 1.4445510893407716e-06,
+      "loss": 0.3857,
+      "step": 18615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.691436832576687,
+      "learning_rate": 1.4442689608535743e-06,
+      "loss": 0.4904,
+      "step": 18616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7639964788994533,
+      "learning_rate": 1.4439869172705878e-06,
+      "loss": 0.4316,
+      "step": 18617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9334041439944873,
+      "learning_rate": 1.4437049585977154e-06,
+      "loss": 0.5279,
+      "step": 18618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9848497738191089,
+      "learning_rate": 1.4434230848408603e-06,
+      "loss": 0.5179,
+      "step": 18619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2778350565588152,
+      "learning_rate": 1.4431412960059238e-06,
+      "loss": 0.3015,
+      "step": 18620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6710980074096642,
+      "learning_rate": 1.442859592098804e-06,
+      "loss": 0.3779,
+      "step": 18621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2055185129190957,
+      "learning_rate": 1.4425779731253984e-06,
+      "loss": 0.539,
+      "step": 18622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.526883299745944,
+      "learning_rate": 1.4422964390916024e-06,
+      "loss": 0.4164,
+      "step": 18623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3467822608649027,
+      "learning_rate": 1.4420149900033118e-06,
+      "loss": 0.528,
+      "step": 18624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0217313703987774,
+      "learning_rate": 1.4417336258664156e-06,
+      "loss": 0.5182,
+      "step": 18625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8355503615093498,
+      "learning_rate": 1.441452346686805e-06,
+      "loss": 0.5885,
+      "step": 18626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6984561134278693,
+      "learning_rate": 1.4411711524703697e-06,
+      "loss": 0.4502,
+      "step": 18627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5994165971041316,
+      "learning_rate": 1.4408900432229944e-06,
+      "loss": 0.4339,
+      "step": 18628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1844283549829937,
+      "learning_rate": 1.4406090189505653e-06,
+      "loss": 0.6427,
+      "step": 18629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5382016631442406,
+      "learning_rate": 1.4403280796589655e-06,
+      "loss": 0.3167,
+      "step": 18630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.023575269159068,
+      "learning_rate": 1.4400472253540749e-06,
+      "loss": 0.5591,
+      "step": 18631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3889275429168206,
+      "learning_rate": 1.439766456041774e-06,
+      "loss": 0.3284,
+      "step": 18632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4162401976040517,
+      "learning_rate": 1.4394857717279413e-06,
+      "loss": 0.396,
+      "step": 18633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9934001943348394,
+      "learning_rate": 1.4392051724184508e-06,
+      "loss": 0.598,
+      "step": 18634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.113908157896898,
+      "learning_rate": 1.4389246581191786e-06,
+      "loss": 0.6041,
+      "step": 18635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7626832466319968,
+      "learning_rate": 1.438644228835995e-06,
+      "loss": 0.6483,
+      "step": 18636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.535065648204609,
+      "learning_rate": 1.4383638845747728e-06,
+      "loss": 0.4376,
+      "step": 18637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5865355066027125,
+      "learning_rate": 1.4380836253413797e-06,
+      "loss": 0.393,
+      "step": 18638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5850997374029827,
+      "learning_rate": 1.4378034511416836e-06,
+      "loss": 0.3882,
+      "step": 18639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6458618004681835,
+      "learning_rate": 1.4375233619815479e-06,
+      "loss": 0.4284,
+      "step": 18640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0084829662903285,
+      "learning_rate": 1.4372433578668379e-06,
+      "loss": 0.4854,
+      "step": 18641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6908399045877667,
+      "learning_rate": 1.4369634388034157e-06,
+      "loss": 0.4003,
+      "step": 18642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6611445137143168,
+      "learning_rate": 1.436683604797139e-06,
+      "loss": 0.4118,
+      "step": 18643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9740043659597384,
+      "learning_rate": 1.4364038558538677e-06,
+      "loss": 0.4155,
+      "step": 18644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7050163412040884,
+      "learning_rate": 1.4361241919794571e-06,
+      "loss": 0.3791,
+      "step": 18645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7314751747858392,
+      "learning_rate": 1.4358446131797635e-06,
+      "loss": 0.3685,
+      "step": 18646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6894879985536455,
+      "learning_rate": 1.435565119460638e-06,
+      "loss": 0.4968,
+      "step": 18647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.447762662603801,
+      "learning_rate": 1.4352857108279316e-06,
+      "loss": 0.3033,
+      "step": 18648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7141047572381887,
+      "learning_rate": 1.4350063872874955e-06,
+      "loss": 0.4428,
+      "step": 18649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9146814541991737,
+      "learning_rate": 1.4347271488451748e-06,
+      "loss": 0.4101,
+      "step": 18650
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.48602399230003357,
+      "eval_runtime": 67.9724,
+      "eval_samples_per_second": 50.167,
+      "eval_steps_per_second": 1.574,
+      "step": 18650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.755952869614432,
+      "learning_rate": 1.4344479955068164e-06,
+      "loss": 0.4667,
+      "step": 18651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7900961850232218,
+      "learning_rate": 1.4341689272782643e-06,
+      "loss": 0.4632,
+      "step": 18652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4543538572971704,
+      "learning_rate": 1.43388994416536e-06,
+      "loss": 0.3349,
+      "step": 18653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9632250474657726,
+      "learning_rate": 1.4336110461739441e-06,
+      "loss": 0.481,
+      "step": 18654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.558421178470156,
+      "learning_rate": 1.433332233309856e-06,
+      "loss": 0.4292,
+      "step": 18655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7181822086900915,
+      "learning_rate": 1.4330535055789308e-06,
+      "loss": 0.3618,
+      "step": 18656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5257275186382586,
+      "learning_rate": 1.4327748629870042e-06,
+      "loss": 0.4484,
+      "step": 18657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5060747492402877,
+      "learning_rate": 1.43249630553991e-06,
+      "loss": 0.4915,
+      "step": 18658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4476355524310935,
+      "learning_rate": 1.4322178332434799e-06,
+      "loss": 0.3671,
+      "step": 18659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.924907556047062,
+      "learning_rate": 1.4319394461035419e-06,
+      "loss": 0.4948,
+      "step": 18660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6361002122370742,
+      "learning_rate": 1.4316611441259253e-06,
+      "loss": 0.3304,
+      "step": 18661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.748902916119055,
+      "learning_rate": 1.4313829273164561e-06,
+      "loss": 0.4347,
+      "step": 18662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6249947716497368,
+      "learning_rate": 1.4311047956809576e-06,
+      "loss": 0.5001,
+      "step": 18663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5229384778377135,
+      "learning_rate": 1.4308267492252526e-06,
+      "loss": 0.3642,
+      "step": 18664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5969598807839551,
+      "learning_rate": 1.430548787955163e-06,
+      "loss": 0.5551,
+      "step": 18665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7810330514336983,
+      "learning_rate": 1.4302709118765064e-06,
+      "loss": 0.407,
+      "step": 18666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0006230435438304,
+      "learning_rate": 1.4299931209951001e-06,
+      "loss": 0.4249,
+      "step": 18667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.077812856696819,
+      "learning_rate": 1.429715415316761e-06,
+      "loss": 0.632,
+      "step": 18668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.73830262412774,
+      "learning_rate": 1.4294377948473003e-06,
+      "loss": 0.6152,
+      "step": 18669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7794556496227603,
+      "learning_rate": 1.4291602595925313e-06,
+      "loss": 0.4645,
+      "step": 18670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5836397031560117,
+      "learning_rate": 1.4288828095582642e-06,
+      "loss": 0.4298,
+      "step": 18671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3971053755405782,
+      "learning_rate": 1.4286054447503069e-06,
+      "loss": 0.3112,
+      "step": 18672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.145483320197565,
+      "learning_rate": 1.4283281651744654e-06,
+      "loss": 0.4089,
+      "step": 18673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6522924148090352,
+      "learning_rate": 1.4280509708365443e-06,
+      "loss": 0.4735,
+      "step": 18674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6868556295652464,
+      "learning_rate": 1.4277738617423482e-06,
+      "loss": 0.4693,
+      "step": 18675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.46747540698905,
+      "learning_rate": 1.4274968378976762e-06,
+      "loss": 0.4689,
+      "step": 18676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1011281009751497,
+      "learning_rate": 1.4272198993083284e-06,
+      "loss": 0.539,
+      "step": 18677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3094072180704304,
+      "learning_rate": 1.4269430459801025e-06,
+      "loss": 0.3863,
+      "step": 18678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.983635526364402,
+      "learning_rate": 1.426666277918794e-06,
+      "loss": 0.5767,
+      "step": 18679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5757329017387973,
+      "learning_rate": 1.4263895951301964e-06,
+      "loss": 0.366,
+      "step": 18680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.805224152883743,
+      "learning_rate": 1.4261129976201035e-06,
+      "loss": 0.5468,
+      "step": 18681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5012407133042198,
+      "learning_rate": 1.4258364853943039e-06,
+      "loss": 0.3833,
+      "step": 18682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9466851965361105,
+      "learning_rate": 1.4255600584585863e-06,
+      "loss": 0.4479,
+      "step": 18683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9127410140645675,
+      "learning_rate": 1.4252837168187388e-06,
+      "loss": 0.5211,
+      "step": 18684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5841073296410078,
+      "learning_rate": 1.4250074604805456e-06,
+      "loss": 0.4175,
+      "step": 18685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7765363683550386,
+      "learning_rate": 1.42473128944979e-06,
+      "loss": 0.4333,
+      "step": 18686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6442394524991084,
+      "learning_rate": 1.4244552037322546e-06,
+      "loss": 0.3685,
+      "step": 18687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2072451249747367,
+      "learning_rate": 1.4241792033337173e-06,
+      "loss": 0.4245,
+      "step": 18688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.902641992686544,
+      "learning_rate": 1.4239032882599568e-06,
+      "loss": 0.5391,
+      "step": 18689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.713921628687052,
+      "learning_rate": 1.4236274585167497e-06,
+      "loss": 0.473,
+      "step": 18690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.800345670877937,
+      "learning_rate": 1.4233517141098692e-06,
+      "loss": 0.4521,
+      "step": 18691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4396511879743694,
+      "learning_rate": 1.4230760550450889e-06,
+      "loss": 0.3486,
+      "step": 18692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.520226872292256,
+      "learning_rate": 1.4228004813281788e-06,
+      "loss": 0.3161,
+      "step": 18693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7494532923637576,
+      "learning_rate": 1.4225249929649086e-06,
+      "loss": 0.5429,
+      "step": 18694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7297135928865568,
+      "learning_rate": 1.422249589961045e-06,
+      "loss": 0.6213,
+      "step": 18695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6911403942067802,
+      "learning_rate": 1.421974272322353e-06,
+      "loss": 0.3905,
+      "step": 18696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8300406786611227,
+      "learning_rate": 1.4216990400545974e-06,
+      "loss": 0.5623,
+      "step": 18697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7364941488140464,
+      "learning_rate": 1.4214238931635382e-06,
+      "loss": 0.4912,
+      "step": 18698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0149027083593665,
+      "learning_rate": 1.421148831654937e-06,
+      "loss": 0.5298,
+      "step": 18699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5512805493592152,
+      "learning_rate": 1.4208738555345523e-06,
+      "loss": 0.4762,
+      "step": 18700
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.4855850040912628,
+      "eval_runtime": 67.9581,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.574,
+      "step": 18700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8314103254993717,
+      "learning_rate": 1.4205989648081387e-06,
+      "loss": 0.6538,
+      "step": 18701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7492370096240735,
+      "learning_rate": 1.4203241594814523e-06,
+      "loss": 0.4754,
+      "step": 18702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7727949554676738,
+      "learning_rate": 1.4200494395602452e-06,
+      "loss": 0.4377,
+      "step": 18703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9714948862578314,
+      "learning_rate": 1.4197748050502703e-06,
+      "loss": 0.4449,
+      "step": 18704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7857680055812466,
+      "learning_rate": 1.4195002559572746e-06,
+      "loss": 0.5579,
+      "step": 18705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5956977913803125,
+      "learning_rate": 1.4192257922870062e-06,
+      "loss": 0.5461,
+      "step": 18706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6474898417454382,
+      "learning_rate": 1.4189514140452121e-06,
+      "loss": 0.4027,
+      "step": 18707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.786930562294219,
+      "learning_rate": 1.4186771212376349e-06,
+      "loss": 0.4055,
+      "step": 18708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.867337197300523,
+      "learning_rate": 1.4184029138700167e-06,
+      "loss": 0.4599,
+      "step": 18709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.707112565336049,
+      "learning_rate": 1.418128791948099e-06,
+      "loss": 0.4231,
+      "step": 18710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5806189823876622,
+      "learning_rate": 1.4178547554776195e-06,
+      "loss": 0.4204,
+      "step": 18711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8695712104880755,
+      "learning_rate": 1.4175808044643147e-06,
+      "loss": 0.5246,
+      "step": 18712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9901842325918866,
+      "learning_rate": 1.417306938913921e-06,
+      "loss": 0.5408,
+      "step": 18713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8351970657095238,
+      "learning_rate": 1.4170331588321698e-06,
+      "loss": 0.5086,
+      "step": 18714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8255349582799123,
+      "learning_rate": 1.4167594642247933e-06,
+      "loss": 0.573,
+      "step": 18715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9655224128818778,
+      "learning_rate": 1.4164858550975216e-06,
+      "loss": 0.5783,
+      "step": 18716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.273224915637351,
+      "learning_rate": 1.4162123314560828e-06,
+      "loss": 0.4268,
+      "step": 18717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9178081254628556,
+      "learning_rate": 1.4159388933062018e-06,
+      "loss": 0.4248,
+      "step": 18718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7076967549117261,
+      "learning_rate": 1.415665540653603e-06,
+      "loss": 0.3437,
+      "step": 18719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8357513355168145,
+      "learning_rate": 1.41539227350401e-06,
+      "loss": 0.5393,
+      "step": 18720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.799462536185395,
+      "learning_rate": 1.4151190918631422e-06,
+      "loss": 0.4774,
+      "step": 18721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6648408671352486,
+      "learning_rate": 1.4148459957367186e-06,
+      "loss": 0.4794,
+      "step": 18722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.404746645539775,
+      "learning_rate": 1.4145729851304576e-06,
+      "loss": 0.4261,
+      "step": 18723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0659486539694893,
+      "learning_rate": 1.4143000600500727e-06,
+      "loss": 0.5415,
+      "step": 18724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.145664896179307,
+      "learning_rate": 1.4140272205012787e-06,
+      "loss": 0.5595,
+      "step": 18725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2152501455338185,
+      "learning_rate": 1.4137544664897874e-06,
+      "loss": 0.594,
+      "step": 18726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7400843324915696,
+      "learning_rate": 1.413481798021308e-06,
+      "loss": 0.4975,
+      "step": 18727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2468661367200498,
+      "learning_rate": 1.4132092151015486e-06,
+      "loss": 0.5573,
+      "step": 18728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.562798033063625,
+      "learning_rate": 1.4129367177362158e-06,
+      "loss": 0.379,
+      "step": 18729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4435613970309087,
+      "learning_rate": 1.412664305931015e-06,
+      "loss": 0.5233,
+      "step": 18730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3258298559412274,
+      "learning_rate": 1.4123919796916478e-06,
+      "loss": 0.3314,
+      "step": 18731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5625544878051965,
+      "learning_rate": 1.412119739023815e-06,
+      "loss": 0.3932,
+      "step": 18732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7826792730041938,
+      "learning_rate": 1.411847583933218e-06,
+      "loss": 0.4373,
+      "step": 18733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8076460788751163,
+      "learning_rate": 1.4115755144255507e-06,
+      "loss": 0.3205,
+      "step": 18734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9254726936827735,
+      "learning_rate": 1.4113035305065116e-06,
+      "loss": 0.5563,
+      "step": 18735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7666527896819568,
+      "learning_rate": 1.4110316321817929e-06,
+      "loss": 0.5751,
+      "step": 18736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.209845311653088,
+      "learning_rate": 1.410759819457087e-06,
+      "loss": 0.5859,
+      "step": 18737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.597565667306428,
+      "learning_rate": 1.410488092338085e-06,
+      "loss": 0.4557,
+      "step": 18738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0733209826018615,
+      "learning_rate": 1.4102164508304756e-06,
+      "loss": 0.3955,
+      "step": 18739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7546868583335482,
+      "learning_rate": 1.4099448949399432e-06,
+      "loss": 0.5723,
+      "step": 18740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9076677503319186,
+      "learning_rate": 1.4096734246721742e-06,
+      "loss": 0.5871,
+      "step": 18741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5421461130964969,
+      "learning_rate": 1.4094020400328526e-06,
+      "loss": 0.5194,
+      "step": 18742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6268569646986244,
+      "learning_rate": 1.4091307410276573e-06,
+      "loss": 0.4642,
+      "step": 18743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7826541734859294,
+      "learning_rate": 1.408859527662269e-06,
+      "loss": 0.5388,
+      "step": 18744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8015273749064935,
+      "learning_rate": 1.408588399942366e-06,
+      "loss": 0.508,
+      "step": 18745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.51114553746014,
+      "learning_rate": 1.4083173578736231e-06,
+      "loss": 0.4671,
+      "step": 18746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9032732317679881,
+      "learning_rate": 1.408046401461715e-06,
+      "loss": 0.5588,
+      "step": 18747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9176703073331791,
+      "learning_rate": 1.4077755307123141e-06,
+      "loss": 0.4732,
+      "step": 18748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8573066291881402,
+      "learning_rate": 1.4075047456310906e-06,
+      "loss": 0.4973,
+      "step": 18749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.874190405381722,
+      "learning_rate": 1.4072340462237134e-06,
+      "loss": 0.4933,
+      "step": 18750
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.486182302236557,
+      "eval_runtime": 68.0039,
+      "eval_samples_per_second": 50.144,
+      "eval_steps_per_second": 1.573,
+      "step": 18750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.441789351422939,
+      "learning_rate": 1.406963432495849e-06,
+      "loss": 0.3861,
+      "step": 18751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3024685605338409,
+      "learning_rate": 1.4066929044531636e-06,
+      "loss": 0.3353,
+      "step": 18752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0325976146737763,
+      "learning_rate": 1.4064224621013186e-06,
+      "loss": 0.4523,
+      "step": 18753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8169994514376686,
+      "learning_rate": 1.406152105445977e-06,
+      "loss": 0.4809,
+      "step": 18754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7405388409807392,
+      "learning_rate": 1.4058818344927994e-06,
+      "loss": 0.3745,
+      "step": 18755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7020634212678776,
+      "learning_rate": 1.4056116492474418e-06,
+      "loss": 0.4304,
+      "step": 18756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5962635597775527,
+      "learning_rate": 1.4053415497155611e-06,
+      "loss": 0.6135,
+      "step": 18757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6806452418241828,
+      "learning_rate": 1.4050715359028125e-06,
+      "loss": 0.4234,
+      "step": 18758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6275706467161803,
+      "learning_rate": 1.4048016078148468e-06,
+      "loss": 0.3759,
+      "step": 18759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.639619587093258,
+      "learning_rate": 1.4045317654573163e-06,
+      "loss": 0.4813,
+      "step": 18760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.463189201072205,
+      "learning_rate": 1.4042620088358704e-06,
+      "loss": 0.3814,
+      "step": 18761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8788300556749367,
+      "learning_rate": 1.403992337956154e-06,
+      "loss": 0.4738,
+      "step": 18762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6769667643675665,
+      "learning_rate": 1.4037227528238145e-06,
+      "loss": 0.4557,
+      "step": 18763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8146755241515966,
+      "learning_rate": 1.4034532534444947e-06,
+      "loss": 0.365,
+      "step": 18764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8106805930872343,
+      "learning_rate": 1.403183839823837e-06,
+      "loss": 0.475,
+      "step": 18765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8010839883595018,
+      "learning_rate": 1.4029145119674806e-06,
+      "loss": 0.4536,
+      "step": 18766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7831409135867196,
+      "learning_rate": 1.4026452698810637e-06,
+      "loss": 0.6527,
+      "step": 18767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8453838007002,
+      "learning_rate": 1.4023761135702245e-06,
+      "loss": 0.4056,
+      "step": 18768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.048615392578888,
+      "learning_rate": 1.4021070430405948e-06,
+      "loss": 0.5133,
+      "step": 18769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4652236315292537,
+      "learning_rate": 1.4018380582978092e-06,
+      "loss": 0.3632,
+      "step": 18770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8874678817739368,
+      "learning_rate": 1.401569159347499e-06,
+      "loss": 0.5021,
+      "step": 18771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.061564897914615,
+      "learning_rate": 1.4013003461952924e-06,
+      "loss": 0.4929,
+      "step": 18772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8189099331146716,
+      "learning_rate": 1.4010316188468165e-06,
+      "loss": 0.4951,
+      "step": 18773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7797788278981417,
+      "learning_rate": 1.4007629773076992e-06,
+      "loss": 0.5335,
+      "step": 18774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.030724696002239,
+      "learning_rate": 1.4004944215835616e-06,
+      "loss": 0.5303,
+      "step": 18775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9799072236010375,
+      "learning_rate": 1.4002259516800277e-06,
+      "loss": 0.4826,
+      "step": 18776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.976952095115144,
+      "learning_rate": 1.3999575676027163e-06,
+      "loss": 0.6458,
+      "step": 18777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7617579824671805,
+      "learning_rate": 1.3996892693572478e-06,
+      "loss": 0.4282,
+      "step": 18778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7172141705426287,
+      "learning_rate": 1.3994210569492366e-06,
+      "loss": 0.3767,
+      "step": 18779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7324656974444963,
+      "learning_rate": 1.3991529303842989e-06,
+      "loss": 0.5293,
+      "step": 18780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4439556757300083,
+      "learning_rate": 1.3988848896680479e-06,
+      "loss": 0.4227,
+      "step": 18781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2087797074125834,
+      "learning_rate": 1.398616934806094e-06,
+      "loss": 0.5062,
+      "step": 18782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8883375910180225,
+      "learning_rate": 1.3983490658040469e-06,
+      "loss": 0.5207,
+      "step": 18783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9425706178663371,
+      "learning_rate": 1.3980812826675143e-06,
+      "loss": 0.3573,
+      "step": 18784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6011942434352036,
+      "learning_rate": 1.3978135854021027e-06,
+      "loss": 0.3601,
+      "step": 18785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1926349398072515,
+      "learning_rate": 1.3975459740134156e-06,
+      "loss": 0.4602,
+      "step": 18786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5489150797200504,
+      "learning_rate": 1.3972784485070558e-06,
+      "loss": 0.3833,
+      "step": 18787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4907900248628676,
+      "learning_rate": 1.397011008888623e-06,
+      "loss": 0.4507,
+      "step": 18788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6044542503984738,
+      "learning_rate": 1.3967436551637162e-06,
+      "loss": 0.4832,
+      "step": 18789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5883270492268402,
+      "learning_rate": 1.3964763873379334e-06,
+      "loss": 0.3462,
+      "step": 18790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.737913080930985,
+      "learning_rate": 1.3962092054168672e-06,
+      "loss": 0.472,
+      "step": 18791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7935301147455653,
+      "learning_rate": 1.3959421094061123e-06,
+      "loss": 0.5852,
+      "step": 18792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0747236614903684,
+      "learning_rate": 1.3956750993112618e-06,
+      "loss": 0.4646,
+      "step": 18793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5643614880521834,
+      "learning_rate": 1.3954081751379022e-06,
+      "loss": 0.3676,
+      "step": 18794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5533363213626414,
+      "learning_rate": 1.395141336891624e-06,
+      "loss": 0.5253,
+      "step": 18795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5123486908120676,
+      "learning_rate": 1.3948745845780116e-06,
+      "loss": 0.3498,
+      "step": 18796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.185403639401238,
+      "learning_rate": 1.3946079182026504e-06,
+      "loss": 0.3917,
+      "step": 18797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7011489644276616,
+      "learning_rate": 1.3943413377711221e-06,
+      "loss": 0.4652,
+      "step": 18798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7083077905333024,
+      "learning_rate": 1.3940748432890077e-06,
+      "loss": 0.4928,
+      "step": 18799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7384901147669722,
+      "learning_rate": 1.393808434761887e-06,
+      "loss": 0.4681,
+      "step": 18800
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.48615458607673645,
+      "eval_runtime": 67.9643,
+      "eval_samples_per_second": 50.173,
+      "eval_steps_per_second": 1.574,
+      "step": 18800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8651516174986509,
+      "learning_rate": 1.3935421121953361e-06,
+      "loss": 0.4554,
+      "step": 18801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.637794379825399,
+      "learning_rate": 1.3932758755949296e-06,
+      "loss": 0.3815,
+      "step": 18802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0544839940286184,
+      "learning_rate": 1.3930097249662433e-06,
+      "loss": 0.7522,
+      "step": 18803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9713800187789698,
+      "learning_rate": 1.392743660314846e-06,
+      "loss": 0.5768,
+      "step": 18804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9260216857952743,
+      "learning_rate": 1.3924776816463095e-06,
+      "loss": 0.6386,
+      "step": 18805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7419420058014425,
+      "learning_rate": 1.3922117889662018e-06,
+      "loss": 0.4151,
+      "step": 18806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.665330019586647,
+      "learning_rate": 1.3919459822800887e-06,
+      "loss": 0.5103,
+      "step": 18807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0545800634593507,
+      "learning_rate": 1.3916802615935346e-06,
+      "loss": 0.5737,
+      "step": 18808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3699846695758104,
+      "learning_rate": 1.3914146269121023e-06,
+      "loss": 0.3389,
+      "step": 18809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7884334550867016,
+      "learning_rate": 1.3911490782413538e-06,
+      "loss": 0.6149,
+      "step": 18810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8945250553656763,
+      "learning_rate": 1.3908836155868463e-06,
+      "loss": 0.506,
+      "step": 18811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7797403959376952,
+      "learning_rate": 1.390618238954138e-06,
+      "loss": 0.3522,
+      "step": 18812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2710565887411636,
+      "learning_rate": 1.3903529483487858e-06,
+      "loss": 0.6444,
+      "step": 18813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7355774210522652,
+      "learning_rate": 1.390087743776341e-06,
+      "loss": 0.4817,
+      "step": 18814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0663709811010773,
+      "learning_rate": 1.389822625242356e-06,
+      "loss": 0.4818,
+      "step": 18815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6660729988768284,
+      "learning_rate": 1.3895575927523824e-06,
+      "loss": 0.3459,
+      "step": 18816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3045144631152006,
+      "learning_rate": 1.389292646311967e-06,
+      "loss": 0.3213,
+      "step": 18817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6500933361724546,
+      "learning_rate": 1.3890277859266566e-06,
+      "loss": 0.4911,
+      "step": 18818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.691562033832541,
+      "learning_rate": 1.3887630116019968e-06,
+      "loss": 0.4145,
+      "step": 18819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9979133667154372,
+      "learning_rate": 1.3884983233435293e-06,
+      "loss": 0.4546,
+      "step": 18820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.102406362882506,
+      "learning_rate": 1.3882337211567953e-06,
+      "loss": 0.5782,
+      "step": 18821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6499992656002385,
+      "learning_rate": 1.3879692050473348e-06,
+      "loss": 0.4357,
+      "step": 18822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2655433106780856,
+      "learning_rate": 1.3877047750206853e-06,
+      "loss": 0.4705,
+      "step": 18823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.904022158824098,
+      "learning_rate": 1.3874404310823816e-06,
+      "loss": 0.4168,
+      "step": 18824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5611176486035918,
+      "learning_rate": 1.387176173237958e-06,
+      "loss": 0.4822,
+      "step": 18825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6303394920034069,
+      "learning_rate": 1.3869120014929474e-06,
+      "loss": 0.3447,
+      "step": 18826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.885645477390403,
+      "learning_rate": 1.386647915852879e-06,
+      "loss": 0.5067,
+      "step": 18827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5257073916988935,
+      "learning_rate": 1.3863839163232806e-06,
+      "loss": 0.3763,
+      "step": 18828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5859610372563369,
+      "learning_rate": 1.3861200029096813e-06,
+      "loss": 0.579,
+      "step": 18829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9027269031557352,
+      "learning_rate": 1.3858561756176036e-06,
+      "loss": 0.3549,
+      "step": 18830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4914105756722051,
+      "learning_rate": 1.3855924344525712e-06,
+      "loss": 0.3968,
+      "step": 18831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8069938846965827,
+      "learning_rate": 1.385328779420107e-06,
+      "loss": 0.5394,
+      "step": 18832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6222227415412314,
+      "learning_rate": 1.3850652105257272e-06,
+      "loss": 0.4602,
+      "step": 18833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4110658247886751,
+      "learning_rate": 1.3848017277749522e-06,
+      "loss": 0.4579,
+      "step": 18834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.76683882584578,
+      "learning_rate": 1.3845383311732977e-06,
+      "loss": 0.4607,
+      "step": 18835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.545851063336041,
+      "learning_rate": 1.384275020726276e-06,
+      "loss": 0.5261,
+      "step": 18836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.552513161006532,
+      "learning_rate": 1.3840117964394007e-06,
+      "loss": 0.4279,
+      "step": 18837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.963176091879523,
+      "learning_rate": 1.3837486583181825e-06,
+      "loss": 0.4618,
+      "step": 18838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7396919528501222,
+      "learning_rate": 1.3834856063681288e-06,
+      "loss": 0.4397,
+      "step": 18839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7603778238797947,
+      "learning_rate": 1.383222640594747e-06,
+      "loss": 0.4891,
+      "step": 18840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.675320502594105,
+      "learning_rate": 1.3829597610035427e-06,
+      "loss": 0.3925,
+      "step": 18841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6449084573581707,
+      "learning_rate": 1.3826969676000183e-06,
+      "loss": 0.3874,
+      "step": 18842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6086836535399571,
+      "learning_rate": 1.3824342603896757e-06,
+      "loss": 0.3952,
+      "step": 18843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3694531921123756,
+      "learning_rate": 1.3821716393780139e-06,
+      "loss": 0.3306,
+      "step": 18844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9031747757268656,
+      "learning_rate": 1.3819091045705322e-06,
+      "loss": 0.5185,
+      "step": 18845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1247791129875355,
+      "learning_rate": 1.3816466559727244e-06,
+      "loss": 0.5983,
+      "step": 18846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.665553272285004,
+      "learning_rate": 1.3813842935900863e-06,
+      "loss": 0.3254,
+      "step": 18847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7625660759216175,
+      "learning_rate": 1.38112201742811e-06,
+      "loss": 0.5977,
+      "step": 18848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0796389483694457,
+      "learning_rate": 1.3808598274922856e-06,
+      "loss": 0.5545,
+      "step": 18849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1795821896055223,
+      "learning_rate": 1.3805977237881022e-06,
+      "loss": 0.4889,
+      "step": 18850
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.48594823479652405,
+      "eval_runtime": 68.0113,
+      "eval_samples_per_second": 50.139,
+      "eval_steps_per_second": 1.573,
+      "step": 18850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6061534032232998,
+      "learning_rate": 1.3803357063210476e-06,
+      "loss": 0.4349,
+      "step": 18851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8383801071632238,
+      "learning_rate": 1.3800737750966054e-06,
+      "loss": 0.4859,
+      "step": 18852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9931597308983646,
+      "learning_rate": 1.3798119301202594e-06,
+      "loss": 0.4389,
+      "step": 18853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7176088466094779,
+      "learning_rate": 1.3795501713974923e-06,
+      "loss": 0.4888,
+      "step": 18854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6588387011018695,
+      "learning_rate": 1.3792884989337827e-06,
+      "loss": 0.4221,
+      "step": 18855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6242820401879905,
+      "learning_rate": 1.379026912734609e-06,
+      "loss": 0.4209,
+      "step": 18856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4732817921484163,
+      "learning_rate": 1.3787654128054467e-06,
+      "loss": 0.44,
+      "step": 18857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.278952455436421,
+      "learning_rate": 1.378503999151772e-06,
+      "loss": 0.6835,
+      "step": 18858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.183363263655533,
+      "learning_rate": 1.378242671779055e-06,
+      "loss": 0.4265,
+      "step": 18859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5897586023950891,
+      "learning_rate": 1.3779814306927677e-06,
+      "loss": 0.3735,
+      "step": 18860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8073944254449545,
+      "learning_rate": 1.3777202758983794e-06,
+      "loss": 0.4426,
+      "step": 18861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8753318668688481,
+      "learning_rate": 1.3774592074013563e-06,
+      "loss": 0.5349,
+      "step": 18862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6280076010682873,
+      "learning_rate": 1.3771982252071637e-06,
+      "loss": 0.4816,
+      "step": 18863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5298524669600095,
+      "learning_rate": 1.3769373293212666e-06,
+      "loss": 0.3948,
+      "step": 18864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9635150374335228,
+      "learning_rate": 1.3766765197491243e-06,
+      "loss": 0.4604,
+      "step": 18865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9082855186533982,
+      "learning_rate": 1.3764157964961984e-06,
+      "loss": 0.4982,
+      "step": 18866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4520545331219237,
+      "learning_rate": 1.3761551595679474e-06,
+      "loss": 0.399,
+      "step": 18867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.917780556329695,
+      "learning_rate": 1.375894608969826e-06,
+      "loss": 0.4555,
+      "step": 18868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.049340760286531,
+      "learning_rate": 1.3756341447072888e-06,
+      "loss": 0.5116,
+      "step": 18869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4097345243176782,
+      "learning_rate": 1.3753737667857896e-06,
+      "loss": 0.3687,
+      "step": 18870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9919931230317045,
+      "learning_rate": 1.3751134752107791e-06,
+      "loss": 0.4967,
+      "step": 18871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4727471989992902,
+      "learning_rate": 1.3748532699877054e-06,
+      "loss": 0.3314,
+      "step": 18872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.093470923589418,
+      "learning_rate": 1.374593151122016e-06,
+      "loss": 0.5249,
+      "step": 18873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8028633979494502,
+      "learning_rate": 1.3743331186191578e-06,
+      "loss": 0.5389,
+      "step": 18874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9627809037291344,
+      "learning_rate": 1.3740731724845723e-06,
+      "loss": 0.5534,
+      "step": 18875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1627727178961833,
+      "learning_rate": 1.3738133127237019e-06,
+      "loss": 0.5137,
+      "step": 18876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7889355790392016,
+      "learning_rate": 1.3735535393419883e-06,
+      "loss": 0.5469,
+      "step": 18877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.157698578910381,
+      "learning_rate": 1.373293852344867e-06,
+      "loss": 0.7512,
+      "step": 18878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5112117110790027,
+      "learning_rate": 1.373034251737776e-06,
+      "loss": 0.4183,
+      "step": 18879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7126788605879801,
+      "learning_rate": 1.372774737526149e-06,
+      "loss": 0.3493,
+      "step": 18880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7481981649393088,
+      "learning_rate": 1.372515309715421e-06,
+      "loss": 0.479,
+      "step": 18881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5851344080714096,
+      "learning_rate": 1.3722559683110192e-06,
+      "loss": 0.5132,
+      "step": 18882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5953266954355838,
+      "learning_rate": 1.3719967133183768e-06,
+      "loss": 0.4873,
+      "step": 18883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7388813929544573,
+      "learning_rate": 1.3717375447429182e-06,
+      "loss": 0.444,
+      "step": 18884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.098677422174755,
+      "learning_rate": 1.3714784625900698e-06,
+      "loss": 0.5582,
+      "step": 18885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1579234686489888,
+      "learning_rate": 1.3712194668652568e-06,
+      "loss": 0.516,
+      "step": 18886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5022325338045077,
+      "learning_rate": 1.3709605575738985e-06,
+      "loss": 0.3429,
+      "step": 18887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7432605065745246,
+      "learning_rate": 1.3707017347214163e-06,
+      "loss": 0.4443,
+      "step": 18888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7127555642212842,
+      "learning_rate": 1.3704429983132287e-06,
+      "loss": 0.4949,
+      "step": 18889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7212232789813369,
+      "learning_rate": 1.3701843483547524e-06,
+      "loss": 0.4291,
+      "step": 18890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5635213690072474,
+      "learning_rate": 1.3699257848514015e-06,
+      "loss": 0.3784,
+      "step": 18891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7528283725398344,
+      "learning_rate": 1.3696673078085881e-06,
+      "loss": 0.4678,
+      "step": 18892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.813038112627379,
+      "learning_rate": 1.3694089172317258e-06,
+      "loss": 0.5007,
+      "step": 18893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8091061842325447,
+      "learning_rate": 1.3691506131262209e-06,
+      "loss": 0.4621,
+      "step": 18894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6654639753183367,
+      "learning_rate": 1.3688923954974825e-06,
+      "loss": 0.3568,
+      "step": 18895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7544122022082786,
+      "learning_rate": 1.3686342643509165e-06,
+      "loss": 0.4566,
+      "step": 18896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0320035300701784,
+      "learning_rate": 1.3683762196919253e-06,
+      "loss": 0.5749,
+      "step": 18897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8177444574762132,
+      "learning_rate": 1.3681182615259115e-06,
+      "loss": 0.5605,
+      "step": 18898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8916924891045632,
+      "learning_rate": 1.3678603898582767e-06,
+      "loss": 0.5412,
+      "step": 18899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.516975121927065,
+      "learning_rate": 1.3676026046944168e-06,
+      "loss": 0.4105,
+      "step": 18900
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.4858224391937256,
+      "eval_runtime": 67.9361,
+      "eval_samples_per_second": 50.194,
+      "eval_steps_per_second": 1.575,
+      "step": 18900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3272569773249585,
+      "learning_rate": 1.3673449060397302e-06,
+      "loss": 0.5426,
+      "step": 18901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7259451742995944,
+      "learning_rate": 1.3670872938996108e-06,
+      "loss": 0.5836,
+      "step": 18902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5939625369482564,
+      "learning_rate": 1.3668297682794527e-06,
+      "loss": 0.3359,
+      "step": 18903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8802540410625883,
+      "learning_rate": 1.3665723291846452e-06,
+      "loss": 0.5079,
+      "step": 18904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.96666905898679,
+      "learning_rate": 1.3663149766205785e-06,
+      "loss": 0.4575,
+      "step": 18905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4972975497939394,
+      "learning_rate": 1.3660577105926415e-06,
+      "loss": 0.461,
+      "step": 18906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4600802062775269,
+      "learning_rate": 1.3658005311062175e-06,
+      "loss": 0.356,
+      "step": 18907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7800427417147586,
+      "learning_rate": 1.3655434381666915e-06,
+      "loss": 0.6066,
+      "step": 18908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9575338962573658,
+      "learning_rate": 1.3652864317794472e-06,
+      "loss": 0.4541,
+      "step": 18909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8800410813826156,
+      "learning_rate": 1.3650295119498615e-06,
+      "loss": 0.5019,
+      "step": 18910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.633835567873395,
+      "learning_rate": 1.3647726786833149e-06,
+      "loss": 0.3924,
+      "step": 18911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9444464297071842,
+      "learning_rate": 1.3645159319851852e-06,
+      "loss": 0.505,
+      "step": 18912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7475757318767915,
+      "learning_rate": 1.3642592718608444e-06,
+      "loss": 0.458,
+      "step": 18913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3595093786998484,
+      "learning_rate": 1.364002698315667e-06,
+      "loss": 0.3182,
+      "step": 18914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4946425749210734,
+      "learning_rate": 1.3637462113550243e-06,
+      "loss": 0.3747,
+      "step": 18915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8566386703067992,
+      "learning_rate": 1.3634898109842864e-06,
+      "loss": 0.4313,
+      "step": 18916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7478108625863498,
+      "learning_rate": 1.3632334972088188e-06,
+      "loss": 0.4938,
+      "step": 18917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5157159208562954,
+      "learning_rate": 1.362977270033989e-06,
+      "loss": 0.4267,
+      "step": 18918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8721708892290447,
+      "learning_rate": 1.3627211294651606e-06,
+      "loss": 0.707,
+      "step": 18919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.697572323200994,
+      "learning_rate": 1.3624650755076956e-06,
+      "loss": 0.5592,
+      "step": 18920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8163155711581946,
+      "learning_rate": 1.362209108166954e-06,
+      "loss": 0.4636,
+      "step": 18921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5384294653177315,
+      "learning_rate": 1.3619532274482955e-06,
+      "loss": 0.3788,
+      "step": 18922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1191071460923765,
+      "learning_rate": 1.361697433357075e-06,
+      "loss": 0.6296,
+      "step": 18923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.747892814022656,
+      "learning_rate": 1.3614417258986484e-06,
+      "loss": 0.5673,
+      "step": 18924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1638996337662397,
+      "learning_rate": 1.3611861050783693e-06,
+      "loss": 0.6413,
+      "step": 18925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.302624179673139,
+      "learning_rate": 1.3609305709015877e-06,
+      "loss": 0.5838,
+      "step": 18926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5751328744886774,
+      "learning_rate": 1.3606751233736538e-06,
+      "loss": 0.3736,
+      "step": 18927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.631673496542604,
+      "learning_rate": 1.3604197624999153e-06,
+      "loss": 0.4752,
+      "step": 18928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5286012678643897,
+      "learning_rate": 1.3601644882857187e-06,
+      "loss": 0.4191,
+      "step": 18929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.554963681815799,
+      "learning_rate": 1.359909300736406e-06,
+      "loss": 0.4921,
+      "step": 18930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.934253776626442,
+      "learning_rate": 1.3596541998573213e-06,
+      "loss": 0.4435,
+      "step": 18931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0719575801122505,
+      "learning_rate": 1.3593991856538035e-06,
+      "loss": 0.4564,
+      "step": 18932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6966197785255634,
+      "learning_rate": 1.359144258131192e-06,
+      "loss": 0.4688,
+      "step": 18933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.49384498330078,
+      "learning_rate": 1.3588894172948247e-06,
+      "loss": 0.4487,
+      "step": 18934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5180352260949717,
+      "learning_rate": 1.3586346631500342e-06,
+      "loss": 0.3132,
+      "step": 18935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6346687661669475,
+      "learning_rate": 1.358379995702155e-06,
+      "loss": 0.4096,
+      "step": 18936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.036592636641319,
+      "learning_rate": 1.358125414956518e-06,
+      "loss": 0.629,
+      "step": 18937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5297315121116468,
+      "learning_rate": 1.3578709209184534e-06,
+      "loss": 0.3498,
+      "step": 18938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8882018576522988,
+      "learning_rate": 1.3576165135932875e-06,
+      "loss": 0.505,
+      "step": 18939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8889282356822557,
+      "learning_rate": 1.3573621929863474e-06,
+      "loss": 0.4921,
+      "step": 18940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8252806349940596,
+      "learning_rate": 1.3571079591029573e-06,
+      "loss": 0.5141,
+      "step": 18941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1003657434343155,
+      "learning_rate": 1.3568538119484378e-06,
+      "loss": 0.4507,
+      "step": 18942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.03099568526013,
+      "learning_rate": 1.3565997515281108e-06,
+      "loss": 0.4925,
+      "step": 18943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5345530193616677,
+      "learning_rate": 1.3563457778472947e-06,
+      "loss": 0.3378,
+      "step": 18944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.515832609924512,
+      "learning_rate": 1.3560918909113054e-06,
+      "loss": 0.4903,
+      "step": 18945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8491586845985308,
+      "learning_rate": 1.3558380907254586e-06,
+      "loss": 0.4698,
+      "step": 18946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6291701001125753,
+      "learning_rate": 1.3555843772950683e-06,
+      "loss": 0.452,
+      "step": 18947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5920526810884104,
+      "learning_rate": 1.3553307506254437e-06,
+      "loss": 0.4195,
+      "step": 18948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1275812235935176,
+      "learning_rate": 1.3550772107218958e-06,
+      "loss": 0.5296,
+      "step": 18949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9089302542680242,
+      "learning_rate": 1.3548237575897319e-06,
+      "loss": 0.4682,
+      "step": 18950
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.48605790734291077,
+      "eval_runtime": 67.9434,
+      "eval_samples_per_second": 50.189,
+      "eval_steps_per_second": 1.575,
+      "step": 18950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.131592118237047,
+      "learning_rate": 1.354570391234259e-06,
+      "loss": 0.5513,
+      "step": 18951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.68977356756904,
+      "learning_rate": 1.354317111660779e-06,
+      "loss": 0.4768,
+      "step": 18952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9431324709102882,
+      "learning_rate": 1.3540639188745948e-06,
+      "loss": 0.3354,
+      "step": 18953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9131825476551882,
+      "learning_rate": 1.353810812881009e-06,
+      "loss": 0.5871,
+      "step": 18954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9192416476824987,
+      "learning_rate": 1.3535577936853174e-06,
+      "loss": 0.5847,
+      "step": 18955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8331921469368175,
+      "learning_rate": 1.353304861292818e-06,
+      "loss": 0.5085,
+      "step": 18956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7764996314896278,
+      "learning_rate": 1.3530520157088062e-06,
+      "loss": 0.4723,
+      "step": 18957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.107181331348901,
+      "learning_rate": 1.3527992569385746e-06,
+      "loss": 0.669,
+      "step": 18958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0006704341475703,
+      "learning_rate": 1.352546584987414e-06,
+      "loss": 0.4369,
+      "step": 18959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5906593777074227,
+      "learning_rate": 1.3522939998606146e-06,
+      "loss": 0.4801,
+      "step": 18960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6214749458450326,
+      "learning_rate": 1.352041501563465e-06,
+      "loss": 0.4845,
+      "step": 18961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7735270896201176,
+      "learning_rate": 1.3517890901012492e-06,
+      "loss": 0.4557,
+      "step": 18962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.004583358477269,
+      "learning_rate": 1.3515367654792527e-06,
+      "loss": 0.5627,
+      "step": 18963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9148513153124203,
+      "learning_rate": 1.351284527702758e-06,
+      "loss": 0.3532,
+      "step": 18964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.542663497499276,
+      "learning_rate": 1.3510323767770439e-06,
+      "loss": 0.376,
+      "step": 18965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.069186450275586,
+      "learning_rate": 1.35078031270739e-06,
+      "loss": 0.5301,
+      "step": 18966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0175581713726367,
+      "learning_rate": 1.350528335499074e-06,
+      "loss": 0.5116,
+      "step": 18967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4768467193541785,
+      "learning_rate": 1.3502764451573692e-06,
+      "loss": 0.3114,
+      "step": 18968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0909320922219696,
+      "learning_rate": 1.350024641687549e-06,
+      "loss": 0.5603,
+      "step": 18969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4998031361347073,
+      "learning_rate": 1.3497729250948862e-06,
+      "loss": 0.4224,
+      "step": 18970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.891429624996851,
+      "learning_rate": 1.3495212953846493e-06,
+      "loss": 0.5615,
+      "step": 18971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2202015287053674,
+      "learning_rate": 1.3492697525621054e-06,
+      "loss": 0.5888,
+      "step": 18972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8842065791413296,
+      "learning_rate": 1.3490182966325216e-06,
+      "loss": 0.4813,
+      "step": 18973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8104195765224051,
+      "learning_rate": 1.348766927601162e-06,
+      "loss": 0.5135,
+      "step": 18974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6833714495297474,
+      "learning_rate": 1.3485156454732875e-06,
+      "loss": 0.4768,
+      "step": 18975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8481446616756205,
+      "learning_rate": 1.3482644502541597e-06,
+      "loss": 0.367,
+      "step": 18976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.594738471838104,
+      "learning_rate": 1.3480133419490374e-06,
+      "loss": 0.5633,
+      "step": 18977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.738434694227496,
+      "learning_rate": 1.3477623205631763e-06,
+      "loss": 0.3842,
+      "step": 18978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8815877599474569,
+      "learning_rate": 1.3475113861018318e-06,
+      "loss": 0.4961,
+      "step": 18979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4499876654898982,
+      "learning_rate": 1.3472605385702577e-06,
+      "loss": 0.408,
+      "step": 18980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.827930881278402,
+      "learning_rate": 1.347009777973704e-06,
+      "loss": 0.4816,
+      "step": 18981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8518903800721582,
+      "learning_rate": 1.3467591043174216e-06,
+      "loss": 0.4156,
+      "step": 18982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8010789175972175,
+      "learning_rate": 1.3465085176066584e-06,
+      "loss": 0.6626,
+      "step": 18983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8955598630681483,
+      "learning_rate": 1.346258017846659e-06,
+      "loss": 0.5338,
+      "step": 18984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.768852558349546,
+      "learning_rate": 1.3460076050426674e-06,
+      "loss": 0.4727,
+      "step": 18985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8602811178555623,
+      "learning_rate": 1.345757279199928e-06,
+      "loss": 0.5661,
+      "step": 18986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.912780150743744,
+      "learning_rate": 1.3455070403236786e-06,
+      "loss": 0.5106,
+      "step": 18987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8268906271107606,
+      "learning_rate": 1.3452568884191582e-06,
+      "loss": 0.5267,
+      "step": 18988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8011784165294162,
+      "learning_rate": 1.3450068234916058e-06,
+      "loss": 0.4764,
+      "step": 18989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5949464442715329,
+      "learning_rate": 1.344756845546254e-06,
+      "loss": 0.395,
+      "step": 18990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9222448059134802,
+      "learning_rate": 1.344506954588336e-06,
+      "loss": 0.5369,
+      "step": 18991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.88532207858785,
+      "learning_rate": 1.3442571506230853e-06,
+      "loss": 0.3828,
+      "step": 18992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7168669924575142,
+      "learning_rate": 1.344007433655729e-06,
+      "loss": 0.3982,
+      "step": 18993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3248828614313783,
+      "learning_rate": 1.3437578036914956e-06,
+      "loss": 0.5752,
+      "step": 18994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5445110038982501,
+      "learning_rate": 1.3435082607356115e-06,
+      "loss": 0.3764,
+      "step": 18995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9546931007446477,
+      "learning_rate": 1.3432588047933001e-06,
+      "loss": 0.5033,
+      "step": 18996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7777592583436763,
+      "learning_rate": 1.3430094358697837e-06,
+      "loss": 0.4755,
+      "step": 18997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1549771175840142,
+      "learning_rate": 1.3427601539702824e-06,
+      "loss": 0.6513,
+      "step": 18998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.035385203819495,
+      "learning_rate": 1.3425109591000163e-06,
+      "loss": 0.554,
+      "step": 18999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.76047739289975,
+      "learning_rate": 1.3422618512642002e-06,
+      "loss": 0.4093,
+      "step": 19000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.4860801100730896,
+      "eval_runtime": 67.9761,
+      "eval_samples_per_second": 50.165,
+      "eval_steps_per_second": 1.574,
+      "step": 19000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5238023932041245,
+      "learning_rate": 1.3420128304680496e-06,
+      "loss": 0.3694,
+      "step": 19001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.475630589549328,
+      "learning_rate": 1.3417638967167782e-06,
+      "loss": 0.3806,
+      "step": 19002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.000299584312442,
+      "learning_rate": 1.3415150500155966e-06,
+      "loss": 0.4961,
+      "step": 19003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8797431216398839,
+      "learning_rate": 1.3412662903697143e-06,
+      "loss": 0.4536,
+      "step": 19004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7097040005832174,
+      "learning_rate": 1.3410176177843398e-06,
+      "loss": 0.3386,
+      "step": 19005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8337758894481249,
+      "learning_rate": 1.3407690322646777e-06,
+      "loss": 0.5333,
+      "step": 19006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.901976779341683,
+      "learning_rate": 1.3405205338159322e-06,
+      "loss": 0.4442,
+      "step": 19007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8065062085665753,
+      "learning_rate": 1.3402721224433057e-06,
+      "loss": 0.4923,
+      "step": 19008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3919054485424576,
+      "learning_rate": 1.3400237981519999e-06,
+      "loss": 0.3526,
+      "step": 19009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9630569966797038,
+      "learning_rate": 1.3397755609472108e-06,
+      "loss": 0.6056,
+      "step": 19010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.625468328516527,
+      "learning_rate": 1.3395274108341366e-06,
+      "loss": 0.4513,
+      "step": 19011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5627082378112982,
+      "learning_rate": 1.3392793478179723e-06,
+      "loss": 0.4512,
+      "step": 19012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5928381864166554,
+      "learning_rate": 1.33903137190391e-06,
+      "loss": 0.5563,
+      "step": 19013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4745005457444635,
+      "learning_rate": 1.3387834830971417e-06,
+      "loss": 0.4878,
+      "step": 19014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4963082214240297,
+      "learning_rate": 1.3385356814028568e-06,
+      "loss": 0.3817,
+      "step": 19015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7150735979118998,
+      "learning_rate": 1.3382879668262422e-06,
+      "loss": 0.4057,
+      "step": 19016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7569733903879026,
+      "learning_rate": 1.3380403393724843e-06,
+      "loss": 0.6514,
+      "step": 19017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7781506952273542,
+      "learning_rate": 1.3377927990467674e-06,
+      "loss": 0.4979,
+      "step": 19018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.113138830238662,
+      "learning_rate": 1.337545345854272e-06,
+      "loss": 0.6567,
+      "step": 19019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5544352386135956,
+      "learning_rate": 1.3372979798001792e-06,
+      "loss": 0.3545,
+      "step": 19020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0524886343938826,
+      "learning_rate": 1.3370507008896683e-06,
+      "loss": 0.5613,
+      "step": 19021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4252174867621,
+      "learning_rate": 1.3368035091279157e-06,
+      "loss": 0.3198,
+      "step": 19022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.664613492318731,
+      "learning_rate": 1.3365564045200953e-06,
+      "loss": 0.6498,
+      "step": 19023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6948746591623336,
+      "learning_rate": 1.3363093870713798e-06,
+      "loss": 0.4103,
+      "step": 19024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.668193542385167,
+      "learning_rate": 1.3360624567869423e-06,
+      "loss": 0.4668,
+      "step": 19025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7844370378975989,
+      "learning_rate": 1.3358156136719502e-06,
+      "loss": 0.4085,
+      "step": 19026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5804191428506347,
+      "learning_rate": 1.335568857731572e-06,
+      "loss": 0.4166,
+      "step": 19027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8961471157598864,
+      "learning_rate": 1.3353221889709735e-06,
+      "loss": 0.4934,
+      "step": 19028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3911804836497668,
+      "learning_rate": 1.3350756073953175e-06,
+      "loss": 0.3771,
+      "step": 19029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4484320860524493,
+      "learning_rate": 1.3348291130097669e-06,
+      "loss": 0.4061,
+      "step": 19030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.940456987445275,
+      "learning_rate": 1.3345827058194825e-06,
+      "loss": 0.5243,
+      "step": 19031
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.793806378627424,
+      "learning_rate": 1.3343363858296207e-06,
+      "loss": 0.4072,
+      "step": 19032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6315002185320642,
+      "learning_rate": 1.3340901530453396e-06,
+      "loss": 0.6701,
+      "step": 19033
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.634088614377671,
+      "learning_rate": 1.3338440074717945e-06,
+      "loss": 0.4748,
+      "step": 19034
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.816200321143502,
+      "learning_rate": 1.3335979491141367e-06,
+      "loss": 0.5408,
+      "step": 19035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4768279641223863,
+      "learning_rate": 1.3333519779775181e-06,
+      "loss": 0.3763,
+      "step": 19036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7087373789319418,
+      "learning_rate": 1.333106094067088e-06,
+      "loss": 0.4829,
+      "step": 19037
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9815931137887157,
+      "learning_rate": 1.3328602973879935e-06,
+      "loss": 0.423,
+      "step": 19038
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1259626377958423,
+      "learning_rate": 1.33261458794538e-06,
+      "loss": 0.7611,
+      "step": 19039
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4770005529511525,
+      "learning_rate": 1.3323689657443922e-06,
+      "loss": 0.4634,
+      "step": 19040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.616014810539146,
+      "learning_rate": 1.3321234307901718e-06,
+      "loss": 0.3941,
+      "step": 19041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.692204678813279,
+      "learning_rate": 1.3318779830878578e-06,
+      "loss": 0.4106,
+      "step": 19042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6986576059356866,
+      "learning_rate": 1.3316326226425895e-06,
+      "loss": 0.507,
+      "step": 19043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6217393473983672,
+      "learning_rate": 1.331387349459504e-06,
+      "loss": 0.4859,
+      "step": 19044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.816581983114763,
+      "learning_rate": 1.3311421635437337e-06,
+      "loss": 0.5287,
+      "step": 19045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8590232523088492,
+      "learning_rate": 1.3308970649004138e-06,
+      "loss": 0.5122,
+      "step": 19046
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.972458967223412,
+      "learning_rate": 1.3306520535346746e-06,
+      "loss": 0.4253,
+      "step": 19047
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9747177014195119,
+      "learning_rate": 1.330407129451644e-06,
+      "loss": 0.5031,
+      "step": 19048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5486537951006505,
+      "learning_rate": 1.3301622926564512e-06,
+      "loss": 0.4631,
+      "step": 19049
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.357811150966073,
+      "learning_rate": 1.3299175431542205e-06,
+      "loss": 0.5893,
+      "step": 19050
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.48597410321235657,
+      "eval_runtime": 68.0284,
+      "eval_samples_per_second": 50.126,
+      "eval_steps_per_second": 1.573,
+      "step": 19050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8760619172985058,
+      "learning_rate": 1.3296728809500756e-06,
+      "loss": 0.6681,
+      "step": 19051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9233206699311494,
+      "learning_rate": 1.3294283060491386e-06,
+      "loss": 0.535,
+      "step": 19052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7700872593332404,
+      "learning_rate": 1.3291838184565299e-06,
+      "loss": 0.4682,
+      "step": 19053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.04244396489846,
+      "learning_rate": 1.328939418177368e-06,
+      "loss": 0.5232,
+      "step": 19054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7709702575323116,
+      "learning_rate": 1.3286951052167676e-06,
+      "loss": 0.6144,
+      "step": 19055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.448134491560425,
+      "learning_rate": 1.328450879579844e-06,
+      "loss": 0.3054,
+      "step": 19056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0110232621006023,
+      "learning_rate": 1.3282067412717115e-06,
+      "loss": 0.5305,
+      "step": 19057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.716587798655478,
+      "learning_rate": 1.3279626902974788e-06,
+      "loss": 0.5089,
+      "step": 19058
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.780116265924029,
+      "learning_rate": 1.327718726662256e-06,
+      "loss": 0.4165,
+      "step": 19059
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.694905720684201,
+      "learning_rate": 1.3274748503711504e-06,
+      "loss": 0.5799,
+      "step": 19060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5597133723430818,
+      "learning_rate": 1.327231061429267e-06,
+      "loss": 0.4412,
+      "step": 19061
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6560862144530533,
+      "learning_rate": 1.326987359841709e-06,
+      "loss": 0.4283,
+      "step": 19062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4880081009557065,
+      "learning_rate": 1.3267437456135798e-06,
+      "loss": 0.3575,
+      "step": 19063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.846203359666702,
+      "learning_rate": 1.3265002187499777e-06,
+      "loss": 0.5178,
+      "step": 19064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.629646308050344,
+      "learning_rate": 1.3262567792560002e-06,
+      "loss": 0.4953,
+      "step": 19065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.158009888060299,
+      "learning_rate": 1.3260134271367456e-06,
+      "loss": 0.438,
+      "step": 19066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4833463249265888,
+      "learning_rate": 1.3257701623973076e-06,
+      "loss": 0.2935,
+      "step": 19067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7857923507822715,
+      "learning_rate": 1.3255269850427781e-06,
+      "loss": 0.5182,
+      "step": 19068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3100747229417398,
+      "learning_rate": 1.3252838950782482e-06,
+      "loss": 0.3246,
+      "step": 19069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4509180468601794,
+      "learning_rate": 1.3250408925088077e-06,
+      "loss": 0.7098,
+      "step": 19070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5808842727831283,
+      "learning_rate": 1.324797977339542e-06,
+      "loss": 0.4401,
+      "step": 19071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0584163572877006,
+      "learning_rate": 1.3245551495755377e-06,
+      "loss": 0.5271,
+      "step": 19072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0853667043341306,
+      "learning_rate": 1.3243124092218784e-06,
+      "loss": 0.4518,
+      "step": 19073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7246374494944567,
+      "learning_rate": 1.3240697562836444e-06,
+      "loss": 0.5629,
+      "step": 19074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8018098240973845,
+      "learning_rate": 1.3238271907659165e-06,
+      "loss": 0.4786,
+      "step": 19075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.361122098915296,
+      "learning_rate": 1.323584712673773e-06,
+      "loss": 0.3337,
+      "step": 19076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5654652201619814,
+      "learning_rate": 1.3233423220122886e-06,
+      "loss": 0.4039,
+      "step": 19077
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.059869604958678,
+      "learning_rate": 1.3231000187865389e-06,
+      "loss": 0.5685,
+      "step": 19078
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7213776759043664,
+      "learning_rate": 1.322857803001596e-06,
+      "loss": 0.44,
+      "step": 19079
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.724960011003039,
+      "learning_rate": 1.3226156746625296e-06,
+      "loss": 0.4308,
+      "step": 19080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5302519262811112,
+      "learning_rate": 1.3223736337744102e-06,
+      "loss": 0.4456,
+      "step": 19081
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9735493042659566,
+      "learning_rate": 1.3221316803423046e-06,
+      "loss": 0.513,
+      "step": 19082
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.108532844377859,
+      "learning_rate": 1.3218898143712762e-06,
+      "loss": 0.5311,
+      "step": 19083
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5656550362623125,
+      "learning_rate": 1.3216480358663896e-06,
+      "loss": 0.5047,
+      "step": 19084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.109560727447157,
+      "learning_rate": 1.3214063448327065e-06,
+      "loss": 0.5603,
+      "step": 19085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6873907902872485,
+      "learning_rate": 1.3211647412752862e-06,
+      "loss": 0.4873,
+      "step": 19086
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9137094421675513,
+      "learning_rate": 1.320923225199186e-06,
+      "loss": 0.4826,
+      "step": 19087
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.814270918451527,
+      "learning_rate": 1.3206817966094624e-06,
+      "loss": 0.5599,
+      "step": 19088
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6215807612324609,
+      "learning_rate": 1.3204404555111704e-06,
+      "loss": 0.4892,
+      "step": 19089
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6733897273378109,
+      "learning_rate": 1.3201992019093604e-06,
+      "loss": 0.4609,
+      "step": 19090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6636057471493717,
+      "learning_rate": 1.3199580358090842e-06,
+      "loss": 0.5359,
+      "step": 19091
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7853132803261416,
+      "learning_rate": 1.3197169572153906e-06,
+      "loss": 0.5284,
+      "step": 19092
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7587091821439083,
+      "learning_rate": 1.3194759661333256e-06,
+      "loss": 0.4135,
+      "step": 19093
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6065390197978833,
+      "learning_rate": 1.3192350625679345e-06,
+      "loss": 0.4221,
+      "step": 19094
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5095299429000277,
+      "learning_rate": 1.3189942465242614e-06,
+      "loss": 0.3449,
+      "step": 19095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8076864713285363,
+      "learning_rate": 1.3187535180073467e-06,
+      "loss": 0.4978,
+      "step": 19096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6110701768522449,
+      "learning_rate": 1.3185128770222289e-06,
+      "loss": 0.3935,
+      "step": 19097
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6999143253369902,
+      "learning_rate": 1.318272323573948e-06,
+      "loss": 0.4733,
+      "step": 19098
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5646116915099593,
+      "learning_rate": 1.3180318576675384e-06,
+      "loss": 0.3698,
+      "step": 19099
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8852129139329616,
+      "learning_rate": 1.3177914793080337e-06,
+      "loss": 0.4263,
+      "step": 19100
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.4851611852645874,
+      "eval_runtime": 67.9846,
+      "eval_samples_per_second": 50.158,
+      "eval_steps_per_second": 1.574,
+      "step": 19100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.404956366908714,
+      "learning_rate": 1.3175511885004665e-06,
+      "loss": 0.4822,
+      "step": 19101
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6580335561465094,
+      "learning_rate": 1.3173109852498688e-06,
+      "loss": 0.5047,
+      "step": 19102
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5232465607808214,
+      "learning_rate": 1.3170708695612666e-06,
+      "loss": 0.4395,
+      "step": 19103
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5591354876536772,
+      "learning_rate": 1.316830841439688e-06,
+      "loss": 0.383,
+      "step": 19104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7547403367463243,
+      "learning_rate": 1.3165909008901578e-06,
+      "loss": 0.4345,
+      "step": 19105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7268337670037026,
+      "learning_rate": 1.316351047917698e-06,
+      "loss": 0.4256,
+      "step": 19106
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3335790581253637,
+      "learning_rate": 1.3161112825273301e-06,
+      "loss": 0.3915,
+      "step": 19107
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.015159674579594,
+      "learning_rate": 1.3158716047240752e-06,
+      "loss": 0.4358,
+      "step": 19108
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4467283958019084,
+      "learning_rate": 1.3156320145129482e-06,
+      "loss": 0.7222,
+      "step": 19109
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.457500816498477,
+      "learning_rate": 1.3153925118989664e-06,
+      "loss": 0.6407,
+      "step": 19110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.870385568343053,
+      "learning_rate": 1.3151530968871438e-06,
+      "loss": 0.3152,
+      "step": 19111
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.552101945489463,
+      "learning_rate": 1.3149137694824905e-06,
+      "loss": 0.4134,
+      "step": 19112
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8566901084261815,
+      "learning_rate": 1.3146745296900184e-06,
+      "loss": 0.3725,
+      "step": 19113
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6681694582367732,
+      "learning_rate": 1.3144353775147355e-06,
+      "loss": 0.4834,
+      "step": 19114
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8628368579215575,
+      "learning_rate": 1.3141963129616492e-06,
+      "loss": 0.5093,
+      "step": 19115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.537432902117688,
+      "learning_rate": 1.3139573360357618e-06,
+      "loss": 0.3801,
+      "step": 19116
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5466012099067235,
+      "learning_rate": 1.3137184467420777e-06,
+      "loss": 0.4086,
+      "step": 19117
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7615386634350256,
+      "learning_rate": 1.3134796450855992e-06,
+      "loss": 0.5118,
+      "step": 19118
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6897723125856614,
+      "learning_rate": 1.3132409310713224e-06,
+      "loss": 0.3857,
+      "step": 19119
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.93643060756895,
+      "learning_rate": 1.3130023047042461e-06,
+      "loss": 0.4641,
+      "step": 19120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8083065746461906,
+      "learning_rate": 1.3127637659893676e-06,
+      "loss": 0.4727,
+      "step": 19121
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5300712082172754,
+      "learning_rate": 1.3125253149316774e-06,
+      "loss": 0.4768,
+      "step": 19122
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8187979314702833,
+      "learning_rate": 1.3122869515361688e-06,
+      "loss": 0.4249,
+      "step": 19123
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5871300817830984,
+      "learning_rate": 1.3120486758078326e-06,
+      "loss": 0.4838,
+      "step": 19124
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3966176729954265,
+      "learning_rate": 1.3118104877516555e-06,
+      "loss": 0.3571,
+      "step": 19125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1032595442434623,
+      "learning_rate": 1.3115723873726246e-06,
+      "loss": 0.5232,
+      "step": 19126
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.129712160445214,
+      "learning_rate": 1.311334374675724e-06,
+      "loss": 0.4929,
+      "step": 19127
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8183926377706536,
+      "learning_rate": 1.3110964496659375e-06,
+      "loss": 0.4322,
+      "step": 19128
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.793562159933742,
+      "learning_rate": 1.3108586123482439e-06,
+      "loss": 0.5255,
+      "step": 19129
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8852079440483538,
+      "learning_rate": 1.310620862727624e-06,
+      "loss": 0.4413,
+      "step": 19130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6737051340806297,
+      "learning_rate": 1.310383200809054e-06,
+      "loss": 0.4601,
+      "step": 19131
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9040634680429285,
+      "learning_rate": 1.310145626597509e-06,
+      "loss": 0.5527,
+      "step": 19132
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6448870355048046,
+      "learning_rate": 1.3099081400979632e-06,
+      "loss": 0.4704,
+      "step": 19133
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4407053721577916,
+      "learning_rate": 1.3096707413153887e-06,
+      "loss": 0.5487,
+      "step": 19134
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0957386843352444,
+      "learning_rate": 1.3094334302547537e-06,
+      "loss": 0.5648,
+      "step": 19135
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.618102463686538,
+      "learning_rate": 1.309196206921027e-06,
+      "loss": 0.4615,
+      "step": 19136
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.590761329490795,
+      "learning_rate": 1.3089590713191753e-06,
+      "loss": 0.5712,
+      "step": 19137
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.596464766308397,
+      "learning_rate": 1.3087220234541615e-06,
+      "loss": 0.3739,
+      "step": 19138
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6526494031915946,
+      "learning_rate": 1.3084850633309487e-06,
+      "loss": 0.4056,
+      "step": 19139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5152521964878929,
+      "learning_rate": 1.308248190954499e-06,
+      "loss": 0.4396,
+      "step": 19140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9645294783714653,
+      "learning_rate": 1.3080114063297689e-06,
+      "loss": 0.5553,
+      "step": 19141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7700835089354,
+      "learning_rate": 1.3077747094617158e-06,
+      "loss": 0.4524,
+      "step": 19142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2765949811413124,
+      "learning_rate": 1.3075381003552963e-06,
+      "loss": 0.521,
+      "step": 19143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5664100245690737,
+      "learning_rate": 1.307301579015462e-06,
+      "loss": 0.4232,
+      "step": 19144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5198738173931943,
+      "learning_rate": 1.3070651454471648e-06,
+      "loss": 0.3222,
+      "step": 19145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2091103478214698,
+      "learning_rate": 1.3068287996553548e-06,
+      "loss": 0.6268,
+      "step": 19146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7673888493252108,
+      "learning_rate": 1.3065925416449799e-06,
+      "loss": 0.5355,
+      "step": 19147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0978963429364037,
+      "learning_rate": 1.3063563714209846e-06,
+      "loss": 0.5176,
+      "step": 19148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9607225423487349,
+      "learning_rate": 1.306120288988314e-06,
+      "loss": 0.3688,
+      "step": 19149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.094367002015707,
+      "learning_rate": 1.3058842943519111e-06,
+      "loss": 0.5438,
+      "step": 19150
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.4855349659919739,
+      "eval_runtime": 67.967,
+      "eval_samples_per_second": 50.171,
+      "eval_steps_per_second": 1.574,
+      "step": 19150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3216293669619232,
+      "learning_rate": 1.3056483875167148e-06,
+      "loss": 0.6482,
+      "step": 19151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7224516718431584,
+      "learning_rate": 1.3054125684876643e-06,
+      "loss": 0.4279,
+      "step": 19152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7242613424659283,
+      "learning_rate": 1.3051768372696963e-06,
+      "loss": 0.528,
+      "step": 19153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.715397757226119,
+      "learning_rate": 1.3049411938677457e-06,
+      "loss": 0.561,
+      "step": 19154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1513877906382195,
+      "learning_rate": 1.3047056382867453e-06,
+      "loss": 0.5594,
+      "step": 19155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9230502607827333,
+      "learning_rate": 1.3044701705316277e-06,
+      "loss": 0.5921,
+      "step": 19156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6171854100641725,
+      "learning_rate": 1.3042347906073202e-06,
+      "loss": 0.4382,
+      "step": 19157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.688156728678473,
+      "learning_rate": 1.3039994985187507e-06,
+      "loss": 0.4039,
+      "step": 19158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5184816338673308,
+      "learning_rate": 1.3037642942708462e-06,
+      "loss": 0.3688,
+      "step": 19159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8665872193888369,
+      "learning_rate": 1.3035291778685302e-06,
+      "loss": 0.514,
+      "step": 19160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7304536811202809,
+      "learning_rate": 1.303294149316724e-06,
+      "loss": 0.4836,
+      "step": 19161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5724159874360992,
+      "learning_rate": 1.3030592086203478e-06,
+      "loss": 0.4759,
+      "step": 19162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8310645785334854,
+      "learning_rate": 1.302824355784321e-06,
+      "loss": 0.717,
+      "step": 19163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7466227312729645,
+      "learning_rate": 1.3025895908135588e-06,
+      "loss": 0.4463,
+      "step": 19164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7707115750181468,
+      "learning_rate": 1.302354913712976e-06,
+      "loss": 0.4385,
+      "step": 19165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.667538948099911,
+      "learning_rate": 1.302120324487487e-06,
+      "loss": 0.3376,
+      "step": 19166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6824990328662157,
+      "learning_rate": 1.3018858231420005e-06,
+      "loss": 0.4876,
+      "step": 19167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6622289001992205,
+      "learning_rate": 1.301651409681427e-06,
+      "loss": 0.416,
+      "step": 19168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.585498324357495,
+      "learning_rate": 1.3014170841106743e-06,
+      "loss": 0.4833,
+      "step": 19169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.971217189054375,
+      "learning_rate": 1.3011828464346462e-06,
+      "loss": 0.5745,
+      "step": 19170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6445446615437205,
+      "learning_rate": 1.3009486966582475e-06,
+      "loss": 0.5408,
+      "step": 19171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2109644935909154,
+      "learning_rate": 1.3007146347863795e-06,
+      "loss": 0.5549,
+      "step": 19172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.853936087141491,
+      "learning_rate": 1.3004806608239428e-06,
+      "loss": 0.5876,
+      "step": 19173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5762373667020848,
+      "learning_rate": 1.3002467747758347e-06,
+      "loss": 0.551,
+      "step": 19174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8085608647021507,
+      "learning_rate": 1.3000129766469516e-06,
+      "loss": 0.4687,
+      "step": 19175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5489899912839167,
+      "learning_rate": 1.299779266442189e-06,
+      "loss": 0.4068,
+      "step": 19176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.500612169652159,
+      "learning_rate": 1.2995456441664377e-06,
+      "loss": 0.3203,
+      "step": 19177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5721234847425292,
+      "learning_rate": 1.2993121098245895e-06,
+      "loss": 0.3283,
+      "step": 19178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.633080025389689,
+      "learning_rate": 1.2990786634215335e-06,
+      "loss": 0.3841,
+      "step": 19179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9717678923325417,
+      "learning_rate": 1.2988453049621561e-06,
+      "loss": 0.5494,
+      "step": 19180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.02494006035859,
+      "learning_rate": 1.2986120344513423e-06,
+      "loss": 0.5469,
+      "step": 19181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8481533224950846,
+      "learning_rate": 1.2983788518939772e-06,
+      "loss": 0.4927,
+      "step": 19182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.311757670281534,
+      "learning_rate": 1.2981457572949407e-06,
+      "loss": 0.4825,
+      "step": 19183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.559791346999783,
+      "learning_rate": 1.2979127506591125e-06,
+      "loss": 0.3419,
+      "step": 19184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.739268785114654,
+      "learning_rate": 1.2976798319913715e-06,
+      "loss": 0.48,
+      "step": 19185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5366908743975887,
+      "learning_rate": 1.2974470012965924e-06,
+      "loss": 0.3586,
+      "step": 19186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7197621805359204,
+      "learning_rate": 1.2972142585796502e-06,
+      "loss": 0.4133,
+      "step": 19187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8240009097058967,
+      "learning_rate": 1.2969816038454171e-06,
+      "loss": 0.5465,
+      "step": 19188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8921861308550791,
+      "learning_rate": 1.2967490370987635e-06,
+      "loss": 0.5287,
+      "step": 19189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2063152962211183,
+      "learning_rate": 1.2965165583445583e-06,
+      "loss": 0.6564,
+      "step": 19190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1220050400942867,
+      "learning_rate": 1.2962841675876681e-06,
+      "loss": 0.316,
+      "step": 19191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8863844652490078,
+      "learning_rate": 1.2960518648329579e-06,
+      "loss": 0.4144,
+      "step": 19192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.206184812503941,
+      "learning_rate": 1.29581965008529e-06,
+      "loss": 0.5625,
+      "step": 19193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7510547586466483,
+      "learning_rate": 1.2955875233495265e-06,
+      "loss": 0.3602,
+      "step": 19194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.976634904480027,
+      "learning_rate": 1.2953554846305277e-06,
+      "loss": 0.4681,
+      "step": 19195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8553706967334875,
+      "learning_rate": 1.2951235339331498e-06,
+      "loss": 0.468,
+      "step": 19196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8556742048535506,
+      "learning_rate": 1.294891671262248e-06,
+      "loss": 0.5156,
+      "step": 19197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8648689567466017,
+      "learning_rate": 1.294659896622679e-06,
+      "loss": 0.4809,
+      "step": 19198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8076840548845428,
+      "learning_rate": 1.2944282100192915e-06,
+      "loss": 0.3979,
+      "step": 19199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2504993262622945,
+      "learning_rate": 1.294196611456938e-06,
+      "loss": 0.6577,
+      "step": 19200
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.4850960969924927,
+      "eval_runtime": 67.9768,
+      "eval_samples_per_second": 50.164,
+      "eval_steps_per_second": 1.574,
+      "step": 19200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6771103615290408,
+      "learning_rate": 1.293965100940466e-06,
+      "loss": 0.5549,
+      "step": 19201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8560387139452639,
+      "learning_rate": 1.2937336784747217e-06,
+      "loss": 0.5487,
+      "step": 19202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.812443593681043,
+      "learning_rate": 1.2935023440645505e-06,
+      "loss": 0.4475,
+      "step": 19203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6825464000306476,
+      "learning_rate": 1.2932710977147955e-06,
+      "loss": 0.43,
+      "step": 19204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5919620933380798,
+      "learning_rate": 1.2930399394302966e-06,
+      "loss": 0.3704,
+      "step": 19205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8773448643055382,
+      "learning_rate": 1.2928088692158935e-06,
+      "loss": 0.5775,
+      "step": 19206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.731270621056594,
+      "learning_rate": 1.2925778870764232e-06,
+      "loss": 0.5359,
+      "step": 19207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8638951348097883,
+      "learning_rate": 1.2923469930167225e-06,
+      "loss": 0.5324,
+      "step": 19208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5645850965891683,
+      "learning_rate": 1.2921161870416235e-06,
+      "loss": 0.354,
+      "step": 19209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0128730086677513,
+      "learning_rate": 1.2918854691559584e-06,
+      "loss": 0.4108,
+      "step": 19210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5139337644081012,
+      "learning_rate": 1.2916548393645577e-06,
+      "loss": 0.2822,
+      "step": 19211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4886418807953503,
+      "learning_rate": 1.2914242976722487e-06,
+      "loss": 0.3258,
+      "step": 19212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7268465568234939,
+      "learning_rate": 1.291193844083858e-06,
+      "loss": 0.5454,
+      "step": 19213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2498519714466012,
+      "learning_rate": 1.2909634786042105e-06,
+      "loss": 0.2656,
+      "step": 19214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.637425587064244,
+      "learning_rate": 1.2907332012381275e-06,
+      "loss": 0.4356,
+      "step": 19215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.280868069052146,
+      "learning_rate": 1.2905030119904305e-06,
+      "loss": 0.5918,
+      "step": 19216
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8161903927228018,
+      "learning_rate": 1.2902729108659382e-06,
+      "loss": 0.5657,
+      "step": 19217
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7094295012328071,
+      "learning_rate": 1.2900428978694687e-06,
+      "loss": 0.4621,
+      "step": 19218
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1058011372083048,
+      "learning_rate": 1.2898129730058353e-06,
+      "loss": 0.4891,
+      "step": 19219
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9738914628809683,
+      "learning_rate": 1.2895831362798522e-06,
+      "loss": 0.5359,
+      "step": 19220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8896815222995782,
+      "learning_rate": 1.289353387696332e-06,
+      "loss": 0.5422,
+      "step": 19221
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8235766491222554,
+      "learning_rate": 1.2891237272600824e-06,
+      "loss": 0.5778,
+      "step": 19222
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9420210894494883,
+      "learning_rate": 1.288894154975912e-06,
+      "loss": 0.5435,
+      "step": 19223
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5127657802684833,
+      "learning_rate": 1.2886646708486275e-06,
+      "loss": 0.3736,
+      "step": 19224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1615001904459894,
+      "learning_rate": 1.288435274883032e-06,
+      "loss": 0.5694,
+      "step": 19225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7922363305594105,
+      "learning_rate": 1.288205967083928e-06,
+      "loss": 0.465,
+      "step": 19226
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6700883296823033,
+      "learning_rate": 1.287976747456116e-06,
+      "loss": 0.4414,
+      "step": 19227
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8284707048375366,
+      "learning_rate": 1.2877476160043947e-06,
+      "loss": 0.5006,
+      "step": 19228
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5168765517383926,
+      "learning_rate": 1.2875185727335606e-06,
+      "loss": 0.4234,
+      "step": 19229
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5867089872626565,
+      "learning_rate": 1.2872896176484093e-06,
+      "loss": 0.4107,
+      "step": 19230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4586558282626425,
+      "learning_rate": 1.287060750753733e-06,
+      "loss": 0.4254,
+      "step": 19231
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0161363414020177,
+      "learning_rate": 1.2868319720543226e-06,
+      "loss": 0.591,
+      "step": 19232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1072785389353252,
+      "learning_rate": 1.286603281554969e-06,
+      "loss": 0.6025,
+      "step": 19233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7314192442424277,
+      "learning_rate": 1.2863746792604578e-06,
+      "loss": 0.4988,
+      "step": 19234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8134821823596,
+      "learning_rate": 1.2861461651755759e-06,
+      "loss": 0.7069,
+      "step": 19235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.870385637272269,
+      "learning_rate": 1.2859177393051065e-06,
+      "loss": 0.4459,
+      "step": 19236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7064186247114173,
+      "learning_rate": 1.2856894016538315e-06,
+      "loss": 0.4403,
+      "step": 19237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5652537884782196,
+      "learning_rate": 1.285461152226532e-06,
+      "loss": 0.4388,
+      "step": 19238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8018516059790655,
+      "learning_rate": 1.2852329910279846e-06,
+      "loss": 0.4932,
+      "step": 19239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5399957727002542,
+      "learning_rate": 1.2850049180629676e-06,
+      "loss": 0.3468,
+      "step": 19240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.731704843822414,
+      "learning_rate": 1.2847769333362545e-06,
+      "loss": 0.4793,
+      "step": 19241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6994011199839487,
+      "learning_rate": 1.2845490368526176e-06,
+      "loss": 0.3911,
+      "step": 19242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.895560465738505,
+      "learning_rate": 1.2843212286168285e-06,
+      "loss": 0.5142,
+      "step": 19243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2719349332847747,
+      "learning_rate": 1.284093508633656e-06,
+      "loss": 0.5641,
+      "step": 19244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.794619059269743,
+      "learning_rate": 1.283865876907867e-06,
+      "loss": 0.5398,
+      "step": 19245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7925433934221733,
+      "learning_rate": 1.2836383334442278e-06,
+      "loss": 0.4402,
+      "step": 19246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5517965418982695,
+      "learning_rate": 1.283410878247501e-06,
+      "loss": 0.543,
+      "step": 19247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6833966098873403,
+      "learning_rate": 1.2831835113224481e-06,
+      "loss": 0.4241,
+      "step": 19248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.883785639281733,
+      "learning_rate": 1.2829562326738298e-06,
+      "loss": 0.4338,
+      "step": 19249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4578479747321835,
+      "learning_rate": 1.2827290423064029e-06,
+      "loss": 0.4338,
+      "step": 19250
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.4848049581050873,
+      "eval_runtime": 67.9381,
+      "eval_samples_per_second": 50.193,
+      "eval_steps_per_second": 1.575,
+      "step": 19250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8381968268984237,
+      "learning_rate": 1.2825019402249246e-06,
+      "loss": 0.4795,
+      "step": 19251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9456893411096325,
+      "learning_rate": 1.282274926434148e-06,
+      "loss": 0.6257,
+      "step": 19252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0269475358785747,
+      "learning_rate": 1.2820480009388268e-06,
+      "loss": 0.4391,
+      "step": 19253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7856597426691314,
+      "learning_rate": 1.2818211637437103e-06,
+      "loss": 0.335,
+      "step": 19254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.825208227609709,
+      "learning_rate": 1.2815944148535475e-06,
+      "loss": 0.5299,
+      "step": 19255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3434626579208122,
+      "learning_rate": 1.281367754273086e-06,
+      "loss": 0.3498,
+      "step": 19256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.426957729218285,
+      "learning_rate": 1.2811411820070702e-06,
+      "loss": 0.3453,
+      "step": 19257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6897042160639164,
+      "learning_rate": 1.2809146980602427e-06,
+      "loss": 0.3837,
+      "step": 19258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7623372512107498,
+      "learning_rate": 1.2806883024373464e-06,
+      "loss": 0.5003,
+      "step": 19259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5967451775137018,
+      "learning_rate": 1.280461995143119e-06,
+      "loss": 0.4333,
+      "step": 19260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5920498023635534,
+      "learning_rate": 1.2802357761822995e-06,
+      "loss": 0.5272,
+      "step": 19261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.761790383526699,
+      "learning_rate": 1.280009645559623e-06,
+      "loss": 0.5681,
+      "step": 19262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.663246492235968,
+      "learning_rate": 1.279783603279823e-06,
+      "loss": 0.3925,
+      "step": 19263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6920865799107456,
+      "learning_rate": 1.2795576493476319e-06,
+      "loss": 0.4172,
+      "step": 19264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8773238916219026,
+      "learning_rate": 1.27933178376778e-06,
+      "loss": 0.3956,
+      "step": 19265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.098389425247737,
+      "learning_rate": 1.2791060065449964e-06,
+      "loss": 0.5126,
+      "step": 19266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8976464359400793,
+      "learning_rate": 1.2788803176840062e-06,
+      "loss": 0.61,
+      "step": 19267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5482390614605692,
+      "learning_rate": 1.2786547171895347e-06,
+      "loss": 0.4141,
+      "step": 19268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.673121307630942,
+      "learning_rate": 1.2784292050663052e-06,
+      "loss": 0.3909,
+      "step": 19269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2073233759977797,
+      "learning_rate": 1.278203781319038e-06,
+      "loss": 0.4496,
+      "step": 19270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.799544244120596,
+      "learning_rate": 1.2779784459524523e-06,
+      "loss": 0.4104,
+      "step": 19271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5416650255513735,
+      "learning_rate": 1.2777531989712655e-06,
+      "loss": 0.4272,
+      "step": 19272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2724012972330887,
+      "learning_rate": 1.277528040380193e-06,
+      "loss": 0.5333,
+      "step": 19273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8310123771931268,
+      "learning_rate": 1.2773029701839476e-06,
+      "loss": 0.4018,
+      "step": 19274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8259032017771093,
+      "learning_rate": 1.2770779883872425e-06,
+      "loss": 0.5104,
+      "step": 19275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0772632851830144,
+      "learning_rate": 1.2768530949947857e-06,
+      "loss": 0.5609,
+      "step": 19276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.805841756851233,
+      "learning_rate": 1.2766282900112866e-06,
+      "loss": 0.3912,
+      "step": 19277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6276013705839267,
+      "learning_rate": 1.276403573441452e-06,
+      "loss": 0.5026,
+      "step": 19278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.716672470050274,
+      "learning_rate": 1.276178945289985e-06,
+      "loss": 0.5543,
+      "step": 19279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6677107321527047,
+      "learning_rate": 1.2759544055615875e-06,
+      "loss": 0.3744,
+      "step": 19280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7490790199780442,
+      "learning_rate": 1.275729954260962e-06,
+      "loss": 0.5027,
+      "step": 19281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6853229801869154,
+      "learning_rate": 1.2755055913928051e-06,
+      "loss": 0.5571,
+      "step": 19282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9587469557442008,
+      "learning_rate": 1.275281316961815e-06,
+      "loss": 0.4441,
+      "step": 19283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6247679329714435,
+      "learning_rate": 1.275057130972687e-06,
+      "loss": 0.4387,
+      "step": 19284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.542662136817828,
+      "learning_rate": 1.274833033430113e-06,
+      "loss": 0.4278,
+      "step": 19285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.994281965078843,
+      "learning_rate": 1.2746090243387854e-06,
+      "loss": 0.5566,
+      "step": 19286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4995464797968687,
+      "learning_rate": 1.274385103703393e-06,
+      "loss": 0.418,
+      "step": 19287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6071390221261592,
+      "learning_rate": 1.2741612715286244e-06,
+      "loss": 0.4432,
+      "step": 19288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.776357791577831,
+      "learning_rate": 1.2739375278191646e-06,
+      "loss": 0.5667,
+      "step": 19289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7974398245345629,
+      "learning_rate": 1.2737138725796976e-06,
+      "loss": 0.524,
+      "step": 19290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8682635204569356,
+      "learning_rate": 1.273490305814906e-06,
+      "loss": 0.5529,
+      "step": 19291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5515418137883994,
+      "learning_rate": 1.2732668275294697e-06,
+      "loss": 0.3947,
+      "step": 19292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9416170650821547,
+      "learning_rate": 1.2730434377280666e-06,
+      "loss": 0.4711,
+      "step": 19293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7088177349881553,
+      "learning_rate": 1.2728201364153745e-06,
+      "loss": 0.4425,
+      "step": 19294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2220464694712607,
+      "learning_rate": 1.2725969235960665e-06,
+      "loss": 0.4331,
+      "step": 19295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1443270491787016,
+      "learning_rate": 1.2723737992748156e-06,
+      "loss": 0.6128,
+      "step": 19296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.329692616306861,
+      "learning_rate": 1.272150763456294e-06,
+      "loss": 0.6256,
+      "step": 19297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9125115041910163,
+      "learning_rate": 1.2719278161451706e-06,
+      "loss": 0.5682,
+      "step": 19298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6608275683114142,
+      "learning_rate": 1.2717049573461116e-06,
+      "loss": 0.5174,
+      "step": 19299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.447127330392087,
+      "learning_rate": 1.2714821870637832e-06,
+      "loss": 0.4873,
+      "step": 19300
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.4848194718360901,
+      "eval_runtime": 67.9854,
+      "eval_samples_per_second": 50.158,
+      "eval_steps_per_second": 1.574,
+      "step": 19300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6831280467422898,
+      "learning_rate": 1.271259505302849e-06,
+      "loss": 0.4957,
+      "step": 19301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7616123661955945,
+      "learning_rate": 1.2710369120679702e-06,
+      "loss": 0.3996,
+      "step": 19302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2988684998515165,
+      "learning_rate": 1.2708144073638066e-06,
+      "loss": 0.3298,
+      "step": 19303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8034355925592245,
+      "learning_rate": 1.2705919911950177e-06,
+      "loss": 0.4122,
+      "step": 19304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6249024498490774,
+      "learning_rate": 1.2703696635662573e-06,
+      "loss": 0.4968,
+      "step": 19305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.813054511276687,
+      "learning_rate": 1.2701474244821813e-06,
+      "loss": 0.3984,
+      "step": 19306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5776558915690775,
+      "learning_rate": 1.2699252739474421e-06,
+      "loss": 0.4423,
+      "step": 19307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6954185872883814,
+      "learning_rate": 1.2697032119666897e-06,
+      "loss": 0.5885,
+      "step": 19308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.037546171633407,
+      "learning_rate": 1.2694812385445723e-06,
+      "loss": 0.5633,
+      "step": 19309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8595274927706813,
+      "learning_rate": 1.2692593536857379e-06,
+      "loss": 0.6313,
+      "step": 19310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7693921767577068,
+      "learning_rate": 1.269037557394832e-06,
+      "loss": 0.426,
+      "step": 19311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5691403563820276,
+      "learning_rate": 1.2688158496764958e-06,
+      "loss": 0.4195,
+      "step": 19312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8202380373119587,
+      "learning_rate": 1.2685942305353715e-06,
+      "loss": 0.3971,
+      "step": 19313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1228716091092403,
+      "learning_rate": 1.2683726999760998e-06,
+      "loss": 0.6382,
+      "step": 19314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6805969551434872,
+      "learning_rate": 1.2681512580033168e-06,
+      "loss": 0.4255,
+      "step": 19315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5629688886541704,
+      "learning_rate": 1.2679299046216587e-06,
+      "loss": 0.3823,
+      "step": 19316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6310758135585526,
+      "learning_rate": 1.2677086398357596e-06,
+      "loss": 0.4306,
+      "step": 19317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4931180917678206,
+      "learning_rate": 1.2674874636502512e-06,
+      "loss": 0.3692,
+      "step": 19318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8841758787018588,
+      "learning_rate": 1.2672663760697637e-06,
+      "loss": 0.4753,
+      "step": 19319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.334577201695796,
+      "learning_rate": 1.2670453770989261e-06,
+      "loss": 0.3788,
+      "step": 19320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.39109464056266,
+      "learning_rate": 1.2668244667423635e-06,
+      "loss": 0.4899,
+      "step": 19321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7176787542762615,
+      "learning_rate": 1.2666036450047017e-06,
+      "loss": 0.3791,
+      "step": 19322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.738749461290889,
+      "learning_rate": 1.266382911890563e-06,
+      "loss": 0.4411,
+      "step": 19323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7415067742328632,
+      "learning_rate": 1.2661622674045693e-06,
+      "loss": 0.3247,
+      "step": 19324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8549427839779231,
+      "learning_rate": 1.2659417115513378e-06,
+      "loss": 0.4744,
+      "step": 19325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.189355211923817,
+      "learning_rate": 1.2657212443354869e-06,
+      "loss": 0.521,
+      "step": 19326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.042438157676952,
+      "learning_rate": 1.265500865761632e-06,
+      "loss": 0.4443,
+      "step": 19327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7393895393760608,
+      "learning_rate": 1.2652805758343856e-06,
+      "loss": 0.3482,
+      "step": 19328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9225394996108862,
+      "learning_rate": 1.2650603745583614e-06,
+      "loss": 0.5684,
+      "step": 19329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7689678332890806,
+      "learning_rate": 1.2648402619381667e-06,
+      "loss": 0.4219,
+      "step": 19330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9885849672700504,
+      "learning_rate": 1.2646202379784107e-06,
+      "loss": 0.4583,
+      "step": 19331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7317498156903723,
+      "learning_rate": 1.2644003026836997e-06,
+      "loss": 0.428,
+      "step": 19332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.999311814273342,
+      "learning_rate": 1.2641804560586377e-06,
+      "loss": 0.5276,
+      "step": 19333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4726630376864795,
+      "learning_rate": 1.2639606981078265e-06,
+      "loss": 0.3804,
+      "step": 19334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8621384705234156,
+      "learning_rate": 1.2637410288358668e-06,
+      "loss": 0.5114,
+      "step": 19335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5614530239663678,
+      "learning_rate": 1.2635214482473585e-06,
+      "loss": 0.5977,
+      "step": 19336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4200025034241175,
+      "learning_rate": 1.2633019563468966e-06,
+      "loss": 0.4537,
+      "step": 19337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5278256035705076,
+      "learning_rate": 1.2630825531390762e-06,
+      "loss": 0.4709,
+      "step": 19338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.730413103506827,
+      "learning_rate": 1.262863238628492e-06,
+      "loss": 0.4401,
+      "step": 19339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.409805592992743,
+      "learning_rate": 1.262644012819733e-06,
+      "loss": 0.4889,
+      "step": 19340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.704469806930063,
+      "learning_rate": 1.2624248757173902e-06,
+      "loss": 0.42,
+      "step": 19341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6955534285834228,
+      "learning_rate": 1.262205827326051e-06,
+      "loss": 0.3755,
+      "step": 19342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8891408261113238,
+      "learning_rate": 1.2619868676502996e-06,
+      "loss": 0.5202,
+      "step": 19343
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3266073054588547,
+      "learning_rate": 1.2617679966947213e-06,
+      "loss": 0.369,
+      "step": 19344
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8492022066751113,
+      "learning_rate": 1.2615492144638974e-06,
+      "loss": 0.5142,
+      "step": 19345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5603140554613195,
+      "learning_rate": 1.2613305209624089e-06,
+      "loss": 0.3815,
+      "step": 19346
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8755283973047305,
+      "learning_rate": 1.2611119161948323e-06,
+      "loss": 0.4895,
+      "step": 19347
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4867509128106315,
+      "learning_rate": 1.2608934001657451e-06,
+      "loss": 0.6308,
+      "step": 19348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4152616897302668,
+      "learning_rate": 1.2606749728797218e-06,
+      "loss": 0.3648,
+      "step": 19349
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5116180721378187,
+      "learning_rate": 1.2604566343413347e-06,
+      "loss": 0.4692,
+      "step": 19350
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.48429396748542786,
+      "eval_runtime": 67.9601,
+      "eval_samples_per_second": 50.176,
+      "eval_steps_per_second": 1.574,
+      "step": 19350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.446117719503219,
+      "learning_rate": 1.2602383845551542e-06,
+      "loss": 0.2859,
+      "step": 19351
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7627367766596538,
+      "learning_rate": 1.2600202235257508e-06,
+      "loss": 0.5351,
+      "step": 19352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8396365651179374,
+      "learning_rate": 1.2598021512576893e-06,
+      "loss": 0.4859,
+      "step": 19353
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8089799804512638,
+      "learning_rate": 1.2595841677555363e-06,
+      "loss": 0.3881,
+      "step": 19354
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8747470141398432,
+      "learning_rate": 1.259366273023856e-06,
+      "loss": 0.5441,
+      "step": 19355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.724031154310817,
+      "learning_rate": 1.2591484670672075e-06,
+      "loss": 0.4769,
+      "step": 19356
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.643742961997322,
+      "learning_rate": 1.2589307498901524e-06,
+      "loss": 0.4666,
+      "step": 19357
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8649476647388623,
+      "learning_rate": 1.258713121497247e-06,
+      "loss": 0.4948,
+      "step": 19358
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9332233551212112,
+      "learning_rate": 1.2584955818930491e-06,
+      "loss": 0.4452,
+      "step": 19359
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8001794773436315,
+      "learning_rate": 1.2582781310821117e-06,
+      "loss": 0.495,
+      "step": 19360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3949567522711477,
+      "learning_rate": 1.2580607690689861e-06,
+      "loss": 0.6172,
+      "step": 19361
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0859080128124625,
+      "learning_rate": 1.2578434958582245e-06,
+      "loss": 0.5273,
+      "step": 19362
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.782441664627956,
+      "learning_rate": 1.2576263114543738e-06,
+      "loss": 0.4904,
+      "step": 19363
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8890907378828254,
+      "learning_rate": 1.2574092158619813e-06,
+      "loss": 0.4667,
+      "step": 19364
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7562017914601709,
+      "learning_rate": 1.2571922090855923e-06,
+      "loss": 0.3571,
+      "step": 19365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7585313349797185,
+      "learning_rate": 1.2569752911297483e-06,
+      "loss": 0.4442,
+      "step": 19366
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5695115020679709,
+      "learning_rate": 1.2567584619989913e-06,
+      "loss": 0.4425,
+      "step": 19367
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.028934430945229,
+      "learning_rate": 1.256541721697861e-06,
+      "loss": 0.5436,
+      "step": 19368
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6296224473904939,
+      "learning_rate": 1.2563250702308932e-06,
+      "loss": 0.4951,
+      "step": 19369
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7681640373295278,
+      "learning_rate": 1.256108507602624e-06,
+      "loss": 0.4445,
+      "step": 19370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.524286767874523,
+      "learning_rate": 1.2558920338175879e-06,
+      "loss": 0.4337,
+      "step": 19371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6572315602052574,
+      "learning_rate": 1.255675648880316e-06,
+      "loss": 0.3699,
+      "step": 19372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7787432512065535,
+      "learning_rate": 1.2554593527953377e-06,
+      "loss": 0.4705,
+      "step": 19373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5569095113440665,
+      "learning_rate": 1.2552431455671817e-06,
+      "loss": 0.4301,
+      "step": 19374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.587935860594093,
+      "learning_rate": 1.2550270272003736e-06,
+      "loss": 0.4534,
+      "step": 19375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8586400918639567,
+      "learning_rate": 1.2548109976994383e-06,
+      "loss": 0.3944,
+      "step": 19376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6993673101755968,
+      "learning_rate": 1.2545950570688974e-06,
+      "loss": 0.655,
+      "step": 19377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6289648846519946,
+      "learning_rate": 1.2543792053132737e-06,
+      "loss": 0.4054,
+      "step": 19378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.610243310097218,
+      "learning_rate": 1.254163442437083e-06,
+      "loss": 0.528,
+      "step": 19379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5530416320533673,
+      "learning_rate": 1.2539477684448433e-06,
+      "loss": 0.5045,
+      "step": 19380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8484025757216225,
+      "learning_rate": 1.2537321833410706e-06,
+      "loss": 0.3834,
+      "step": 19381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.786664090529838,
+      "learning_rate": 1.2535166871302766e-06,
+      "loss": 0.5783,
+      "step": 19382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7831584801661033,
+      "learning_rate": 1.2533012798169732e-06,
+      "loss": 0.4971,
+      "step": 19383
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7207498936874008,
+      "learning_rate": 1.2530859614056707e-06,
+      "loss": 0.4927,
+      "step": 19384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1330091952186345,
+      "learning_rate": 1.2528707319008745e-06,
+      "loss": 0.5793,
+      "step": 19385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.995757885202971,
+      "learning_rate": 1.252655591307092e-06,
+      "loss": 0.6523,
+      "step": 19386
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7998510397332224,
+      "learning_rate": 1.252440539628827e-06,
+      "loss": 0.426,
+      "step": 19387
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.791236747033463,
+      "learning_rate": 1.2522255768705801e-06,
+      "loss": 0.4702,
+      "step": 19388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6859591564486591,
+      "learning_rate": 1.2520107030368527e-06,
+      "loss": 0.4819,
+      "step": 19389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9678784561553189,
+      "learning_rate": 1.251795918132143e-06,
+      "loss": 0.5734,
+      "step": 19390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.914859539542749,
+      "learning_rate": 1.2515812221609471e-06,
+      "loss": 0.4434,
+      "step": 19391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9867950894034787,
+      "learning_rate": 1.2513666151277592e-06,
+      "loss": 0.5677,
+      "step": 19392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9586372505338279,
+      "learning_rate": 1.2511520970370723e-06,
+      "loss": 0.4472,
+      "step": 19393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9437810949660415,
+      "learning_rate": 1.2509376678933783e-06,
+      "loss": 0.578,
+      "step": 19394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.034762484646971,
+      "learning_rate": 1.2507233277011635e-06,
+      "loss": 0.472,
+      "step": 19395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5622970326611392,
+      "learning_rate": 1.2505090764649171e-06,
+      "loss": 0.5862,
+      "step": 19396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.726765440500031,
+      "learning_rate": 1.2502949141891243e-06,
+      "loss": 0.5313,
+      "step": 19397
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.577795599872673,
+      "learning_rate": 1.250080840878267e-06,
+      "loss": 0.4324,
+      "step": 19398
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4021011554540979,
+      "learning_rate": 1.2498668565368278e-06,
+      "loss": 0.3283,
+      "step": 19399
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7812039286189174,
+      "learning_rate": 1.2496529611692868e-06,
+      "loss": 0.4382,
+      "step": 19400
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.48435500264167786,
+      "eval_runtime": 67.9798,
+      "eval_samples_per_second": 50.162,
+      "eval_steps_per_second": 1.574,
+      "step": 19400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2121074307176256,
+      "learning_rate": 1.24943915478012e-06,
+      "loss": 0.5419,
+      "step": 19401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7485176776936113,
+      "learning_rate": 1.2492254373738048e-06,
+      "loss": 0.4906,
+      "step": 19402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.826232523100813,
+      "learning_rate": 1.2490118089548148e-06,
+      "loss": 0.4804,
+      "step": 19403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.681309949553318,
+      "learning_rate": 1.2487982695276225e-06,
+      "loss": 0.4764,
+      "step": 19404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.726952517989721,
+      "learning_rate": 1.2485848190966976e-06,
+      "loss": 0.461,
+      "step": 19405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.566963547171659,
+      "learning_rate": 1.2483714576665085e-06,
+      "loss": 0.3354,
+      "step": 19406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.922477380240632,
+      "learning_rate": 1.2481581852415232e-06,
+      "loss": 0.3946,
+      "step": 19407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6427560410843747,
+      "learning_rate": 1.2479450018262044e-06,
+      "loss": 0.4065,
+      "step": 19408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0655559596932465,
+      "learning_rate": 1.2477319074250163e-06,
+      "loss": 0.5813,
+      "step": 19409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.717041074912101,
+      "learning_rate": 1.2475189020424197e-06,
+      "loss": 0.3935,
+      "step": 19410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.691740672253212,
+      "learning_rate": 1.2473059856828734e-06,
+      "loss": 0.4874,
+      "step": 19411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5514128345146372,
+      "learning_rate": 1.2470931583508345e-06,
+      "loss": 0.3061,
+      "step": 19412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0725131274869755,
+      "learning_rate": 1.2468804200507595e-06,
+      "loss": 0.4911,
+      "step": 19413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.209568942705372,
+      "learning_rate": 1.2466677707871008e-06,
+      "loss": 0.3691,
+      "step": 19414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8327547602616896,
+      "learning_rate": 1.24645521056431e-06,
+      "loss": 0.447,
+      "step": 19415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3544359740731047,
+      "learning_rate": 1.246242739386838e-06,
+      "loss": 0.3718,
+      "step": 19416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9142562613226228,
+      "learning_rate": 1.2460303572591325e-06,
+      "loss": 0.4237,
+      "step": 19417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.683766243749453,
+      "learning_rate": 1.2458180641856386e-06,
+      "loss": 0.4843,
+      "step": 19418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5665341083290618,
+      "learning_rate": 1.245605860170801e-06,
+      "loss": 0.333,
+      "step": 19419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7439130981776652,
+      "learning_rate": 1.2453937452190629e-06,
+      "loss": 0.549,
+      "step": 19420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7321806986619157,
+      "learning_rate": 1.2451817193348637e-06,
+      "loss": 0.5039,
+      "step": 19421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5073432641873385,
+      "learning_rate": 1.2449697825226422e-06,
+      "loss": 0.5052,
+      "step": 19422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8456654090667364,
+      "learning_rate": 1.2447579347868358e-06,
+      "loss": 0.4911,
+      "step": 19423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.566615573296997,
+      "learning_rate": 1.2445461761318783e-06,
+      "loss": 0.5198,
+      "step": 19424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9741269133888673,
+      "learning_rate": 1.244334506562204e-06,
+      "loss": 0.3746,
+      "step": 19425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4260723952432393,
+      "learning_rate": 1.2441229260822437e-06,
+      "loss": 0.3248,
+      "step": 19426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7831804881778914,
+      "learning_rate": 1.2439114346964259e-06,
+      "loss": 0.4513,
+      "step": 19427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.075742719364963,
+      "learning_rate": 1.2437000324091793e-06,
+      "loss": 0.481,
+      "step": 19428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6103381763185791,
+      "learning_rate": 1.243488719224929e-06,
+      "loss": 0.4175,
+      "step": 19429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6899996542444742,
+      "learning_rate": 1.2432774951480978e-06,
+      "loss": 0.5278,
+      "step": 19430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4687683400269937,
+      "learning_rate": 1.2430663601831086e-06,
+      "loss": 0.3824,
+      "step": 19431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2685138646368836,
+      "learning_rate": 1.2428553143343814e-06,
+      "loss": 0.3085,
+      "step": 19432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.382940900333848,
+      "learning_rate": 1.2426443576063338e-06,
+      "loss": 0.587,
+      "step": 19433
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7988962382942446,
+      "learning_rate": 1.2424334900033819e-06,
+      "loss": 0.4713,
+      "step": 19434
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8938500537903882,
+      "learning_rate": 1.2422227115299413e-06,
+      "loss": 0.4138,
+      "step": 19435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7075191194461004,
+      "learning_rate": 1.242012022190423e-06,
+      "loss": 0.426,
+      "step": 19436
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.07584720228961,
+      "learning_rate": 1.2418014219892378e-06,
+      "loss": 0.5537,
+      "step": 19437
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.723078733053972,
+      "learning_rate": 1.2415909109307954e-06,
+      "loss": 0.4052,
+      "step": 19438
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3718056129982337,
+      "learning_rate": 1.2413804890195027e-06,
+      "loss": 0.3941,
+      "step": 19439
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8382476085429498,
+      "learning_rate": 1.241170156259764e-06,
+      "loss": 0.5095,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8533368382128397,
+      "learning_rate": 1.2409599126559825e-06,
+      "loss": 0.3663,
+      "step": 19441
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6501944502031856,
+      "learning_rate": 1.2407497582125605e-06,
+      "loss": 0.4621,
+      "step": 19442
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6932318218187,
+      "learning_rate": 1.2405396929338962e-06,
+      "loss": 0.3345,
+      "step": 19443
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4999330311372012,
+      "learning_rate": 1.240329716824388e-06,
+      "loss": 0.3262,
+      "step": 19444
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7159982796263564,
+      "learning_rate": 1.2401198298884317e-06,
+      "loss": 0.4958,
+      "step": 19445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8462870760637018,
+      "learning_rate": 1.2399100321304204e-06,
+      "loss": 0.488,
+      "step": 19446
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9136532866508316,
+      "learning_rate": 1.2397003235547466e-06,
+      "loss": 0.5291,
+      "step": 19447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.813051401909076,
+      "learning_rate": 1.239490704165801e-06,
+      "loss": 0.4397,
+      "step": 19448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.168107107312165,
+      "learning_rate": 1.2392811739679703e-06,
+      "loss": 0.5235,
+      "step": 19449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7382811989412001,
+      "learning_rate": 1.239071732965642e-06,
+      "loss": 0.5297,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.4842434525489807,
+      "eval_runtime": 68.0062,
+      "eval_samples_per_second": 50.142,
+      "eval_steps_per_second": 1.573,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7210506569053197,
+      "learning_rate": 1.2388623811632003e-06,
+      "loss": 0.5236,
+      "step": 19451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.573907954029878,
+      "learning_rate": 1.2386531185650287e-06,
+      "loss": 0.4518,
+      "step": 19452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7196652763783122,
+      "learning_rate": 1.2384439451755068e-06,
+      "loss": 0.434,
+      "step": 19453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9642563836496085,
+      "learning_rate": 1.2382348609990134e-06,
+      "loss": 0.5877,
+      "step": 19454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7038462848424474,
+      "learning_rate": 1.238025866039927e-06,
+      "loss": 0.5699,
+      "step": 19455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.905674745912362,
+      "learning_rate": 1.2378169603026215e-06,
+      "loss": 0.6105,
+      "step": 19456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4954494247297936,
+      "learning_rate": 1.2376081437914708e-06,
+      "loss": 0.4164,
+      "step": 19457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6732953207443457,
+      "learning_rate": 1.2373994165108462e-06,
+      "loss": 0.546,
+      "step": 19458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8128559517789629,
+      "learning_rate": 1.2371907784651166e-06,
+      "loss": 0.5724,
+      "step": 19459
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7573076967929515,
+      "learning_rate": 1.2369822296586507e-06,
+      "loss": 0.3965,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.44674268657923,
+      "learning_rate": 1.2367737700958141e-06,
+      "loss": 0.3443,
+      "step": 19461
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8781402817348702,
+      "learning_rate": 1.2365653997809706e-06,
+      "loss": 0.4691,
+      "step": 19462
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6788460811876438,
+      "learning_rate": 1.236357118718482e-06,
+      "loss": 0.3733,
+      "step": 19463
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9601382765096036,
+      "learning_rate": 1.2361489269127088e-06,
+      "loss": 0.537,
+      "step": 19464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4821208041727134,
+      "learning_rate": 1.2359408243680104e-06,
+      "loss": 0.3604,
+      "step": 19465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.038039856681082,
+      "learning_rate": 1.2357328110887415e-06,
+      "loss": 0.5613,
+      "step": 19466
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8015319179157285,
+      "learning_rate": 1.2355248870792575e-06,
+      "loss": 0.4931,
+      "step": 19467
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6785523927920725,
+      "learning_rate": 1.2353170523439118e-06,
+      "loss": 0.4231,
+      "step": 19468
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7506783994406858,
+      "learning_rate": 1.2351093068870541e-06,
+      "loss": 0.5082,
+      "step": 19469
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6393949042235645,
+      "learning_rate": 1.2349016507130341e-06,
+      "loss": 0.3923,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9489533105290708,
+      "learning_rate": 1.2346940838261996e-06,
+      "loss": 0.5942,
+      "step": 19471
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5059710741648071,
+      "learning_rate": 1.2344866062308947e-06,
+      "loss": 0.3554,
+      "step": 19472
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6234039156304898,
+      "learning_rate": 1.2342792179314623e-06,
+      "loss": 0.5074,
+      "step": 19473
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.228203281969742,
+      "learning_rate": 1.2340719189322463e-06,
+      "loss": 0.5831,
+      "step": 19474
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8997314162904828,
+      "learning_rate": 1.2338647092375844e-06,
+      "loss": 0.443,
+      "step": 19475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4631734981802818,
+      "learning_rate": 1.233657588851815e-06,
+      "loss": 0.3866,
+      "step": 19476
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7880457695217107,
+      "learning_rate": 1.233450557779275e-06,
+      "loss": 0.5882,
+      "step": 19477
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5935180278028422,
+      "learning_rate": 1.2332436160242964e-06,
+      "loss": 0.3666,
+      "step": 19478
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8832968954574807,
+      "learning_rate": 1.233036763591213e-06,
+      "loss": 0.4863,
+      "step": 19479
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.723331307522722,
+      "learning_rate": 1.232830000484355e-06,
+      "loss": 0.4818,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7396905133987945,
+      "learning_rate": 1.2326233267080499e-06,
+      "loss": 0.5061,
+      "step": 19481
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6016154040073003,
+      "learning_rate": 1.232416742266625e-06,
+      "loss": 0.3799,
+      "step": 19482
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.786559405336753,
+      "learning_rate": 1.2322102471644045e-06,
+      "loss": 0.4871,
+      "step": 19483
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.615366884864307,
+      "learning_rate": 1.2320038414057128e-06,
+      "loss": 0.375,
+      "step": 19484
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8982032059777483,
+      "learning_rate": 1.2317975249948691e-06,
+      "loss": 0.6342,
+      "step": 19485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7626448991828518,
+      "learning_rate": 1.2315912979361926e-06,
+      "loss": 0.3392,
+      "step": 19486
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7912943787924858,
+      "learning_rate": 1.2313851602340023e-06,
+      "loss": 0.6871,
+      "step": 19487
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.551124547456214,
+      "learning_rate": 1.2311791118926113e-06,
+      "loss": 0.3312,
+      "step": 19488
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5695690165892884,
+      "learning_rate": 1.230973152916334e-06,
+      "loss": 0.436,
+      "step": 19489
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8883220634073454,
+      "learning_rate": 1.2307672833094831e-06,
+      "loss": 0.4693,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.589381288437062,
+      "learning_rate": 1.2305615030763667e-06,
+      "loss": 0.4186,
+      "step": 19491
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.034252787417209,
+      "learning_rate": 1.2303558122212936e-06,
+      "loss": 0.4094,
+      "step": 19492
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.826839494892519,
+      "learning_rate": 1.2301502107485697e-06,
+      "loss": 0.5208,
+      "step": 19493
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5168898837353262,
+      "learning_rate": 1.2299446986624988e-06,
+      "loss": 0.3399,
+      "step": 19494
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8274278122755412,
+      "learning_rate": 1.2297392759673835e-06,
+      "loss": 0.6335,
+      "step": 19495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8555161865194272,
+      "learning_rate": 1.229533942667524e-06,
+      "loss": 0.5147,
+      "step": 19496
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6440584091179835,
+      "learning_rate": 1.2293286987672193e-06,
+      "loss": 0.3775,
+      "step": 19497
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8330864261501811,
+      "learning_rate": 1.2291235442707652e-06,
+      "loss": 0.644,
+      "step": 19498
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1756352251087665,
+      "learning_rate": 1.228918479182457e-06,
+      "loss": 0.5189,
+      "step": 19499
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.042325718185931,
+      "learning_rate": 1.2287135035065885e-06,
+      "loss": 0.7344,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.4842815697193146,
+      "eval_runtime": 67.9802,
+      "eval_samples_per_second": 50.162,
+      "eval_steps_per_second": 1.574,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8394166350443892,
+      "learning_rate": 1.2285086172474488e-06,
+      "loss": 0.5061,
+      "step": 19501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1244762582861774,
+      "learning_rate": 1.2283038204093279e-06,
+      "loss": 0.5663,
+      "step": 19502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7752799031444628,
+      "learning_rate": 1.2280991129965145e-06,
+      "loss": 0.4317,
+      "step": 19503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7309942173042583,
+      "learning_rate": 1.2278944950132918e-06,
+      "loss": 0.4771,
+      "step": 19504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8345392803804976,
+      "learning_rate": 1.227689966463944e-06,
+      "loss": 0.4843,
+      "step": 19505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6764038029520323,
+      "learning_rate": 1.2274855273527545e-06,
+      "loss": 0.4523,
+      "step": 19506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8910730743043778,
+      "learning_rate": 1.2272811776840006e-06,
+      "loss": 0.5056,
+      "step": 19507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8782253001796623,
+      "learning_rate": 1.2270769174619617e-06,
+      "loss": 0.3516,
+      "step": 19508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9186080048594296,
+      "learning_rate": 1.2268727466909134e-06,
+      "loss": 0.5945,
+      "step": 19509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.632371291618098,
+      "learning_rate": 1.2266686653751301e-06,
+      "loss": 0.3692,
+      "step": 19510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1516484004707688,
+      "learning_rate": 1.2264646735188838e-06,
+      "loss": 0.6305,
+      "step": 19511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6937795298041214,
+      "learning_rate": 1.2262607711264452e-06,
+      "loss": 0.5212,
+      "step": 19512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0012945299335123,
+      "learning_rate": 1.2260569582020838e-06,
+      "loss": 0.5089,
+      "step": 19513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.636489956965482,
+      "learning_rate": 1.2258532347500644e-06,
+      "loss": 0.5037,
+      "step": 19514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.067406001742835,
+      "learning_rate": 1.2256496007746528e-06,
+      "loss": 0.5112,
+      "step": 19515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.802738926740523,
+      "learning_rate": 1.225446056280112e-06,
+      "loss": 0.4005,
+      "step": 19516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5044706450005323,
+      "learning_rate": 1.2252426012707032e-06,
+      "loss": 0.4181,
+      "step": 19517
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5422714792962906,
+      "learning_rate": 1.2250392357506847e-06,
+      "loss": 0.4113,
+      "step": 19518
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.593479046905531,
+      "learning_rate": 1.2248359597243157e-06,
+      "loss": 0.4301,
+      "step": 19519
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8906186254955435,
+      "learning_rate": 1.22463277319585e-06,
+      "loss": 0.5099,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9014688159350992,
+      "learning_rate": 1.2244296761695412e-06,
+      "loss": 0.6785,
+      "step": 19521
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9797951303653147,
+      "learning_rate": 1.2242266686496413e-06,
+      "loss": 0.5632,
+      "step": 19522
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9369668372390938,
+      "learning_rate": 1.2240237506404008e-06,
+      "loss": 0.4517,
+      "step": 19523
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.50940662888181,
+      "learning_rate": 1.2238209221460665e-06,
+      "loss": 0.4027,
+      "step": 19524
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4601536127289658,
+      "learning_rate": 1.2236181831708866e-06,
+      "loss": 0.3599,
+      "step": 19525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.147581812558778,
+      "learning_rate": 1.2234155337191027e-06,
+      "loss": 0.5138,
+      "step": 19526
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9443996244503259,
+      "learning_rate": 1.2232129737949588e-06,
+      "loss": 0.4265,
+      "step": 19527
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6665381588528108,
+      "learning_rate": 1.2230105034026948e-06,
+      "loss": 0.4126,
+      "step": 19528
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9229920180884768,
+      "learning_rate": 1.2228081225465493e-06,
+      "loss": 0.6318,
+      "step": 19529
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.963253631757162,
+      "learning_rate": 1.222605831230759e-06,
+      "loss": 0.5741,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.564150563995819,
+      "learning_rate": 1.2224036294595586e-06,
+      "loss": 0.4022,
+      "step": 19531
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7495937515025886,
+      "learning_rate": 1.2222015172371824e-06,
+      "loss": 0.5214,
+      "step": 19532
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7191859033631827,
+      "learning_rate": 1.2219994945678595e-06,
+      "loss": 0.4351,
+      "step": 19533
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.16735226175505,
+      "learning_rate": 1.2217975614558202e-06,
+      "loss": 0.4988,
+      "step": 19534
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7799362336453557,
+      "learning_rate": 1.2215957179052918e-06,
+      "loss": 0.4634,
+      "step": 19535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4822264977993962,
+      "learning_rate": 1.2213939639204996e-06,
+      "loss": 0.3062,
+      "step": 19536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.624057362167606,
+      "learning_rate": 1.2211922995056668e-06,
+      "loss": 0.3941,
+      "step": 19537
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5505621679378416,
+      "learning_rate": 1.2209907246650166e-06,
+      "loss": 0.3741,
+      "step": 19538
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3426441764416346,
+      "learning_rate": 1.2207892394027666e-06,
+      "loss": 0.3822,
+      "step": 19539
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7343251909009663,
+      "learning_rate": 1.2205878437231364e-06,
+      "loss": 0.5344,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6888666402397856,
+      "learning_rate": 1.2203865376303422e-06,
+      "loss": 0.4705,
+      "step": 19541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8739949774209128,
+      "learning_rate": 1.2201853211285965e-06,
+      "loss": 0.4324,
+      "step": 19542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5146136256859444,
+      "learning_rate": 1.2199841942221137e-06,
+      "loss": 0.3881,
+      "step": 19543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7016745739795003,
+      "learning_rate": 1.2197831569151025e-06,
+      "loss": 0.3825,
+      "step": 19544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6915531634054572,
+      "learning_rate": 1.2195822092117734e-06,
+      "loss": 0.4107,
+      "step": 19545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1348925519869906,
+      "learning_rate": 1.2193813511163319e-06,
+      "loss": 0.4879,
+      "step": 19546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5902701100612486,
+      "learning_rate": 1.2191805826329824e-06,
+      "loss": 0.4593,
+      "step": 19547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.722693441938288,
+      "learning_rate": 1.2189799037659293e-06,
+      "loss": 0.4009,
+      "step": 19548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6204943402153722,
+      "learning_rate": 1.2187793145193722e-06,
+      "loss": 0.356,
+      "step": 19549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8482355393041667,
+      "learning_rate": 1.2185788148975111e-06,
+      "loss": 0.4185,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.4840215742588043,
+      "eval_runtime": 67.9487,
+      "eval_samples_per_second": 50.185,
+      "eval_steps_per_second": 1.575,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.022214898804781,
+      "learning_rate": 1.2183784049045438e-06,
+      "loss": 0.4912,
+      "step": 19551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4541820918512522,
+      "learning_rate": 1.2181780845446648e-06,
+      "loss": 0.3632,
+      "step": 19552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9636567459941878,
+      "learning_rate": 1.2179778538220673e-06,
+      "loss": 0.4821,
+      "step": 19553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8015673536425365,
+      "learning_rate": 1.2177777127409442e-06,
+      "loss": 0.564,
+      "step": 19554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6202174296561676,
+      "learning_rate": 1.2175776613054862e-06,
+      "loss": 0.4005,
+      "step": 19555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7794179766751754,
+      "learning_rate": 1.2173776995198786e-06,
+      "loss": 0.4472,
+      "step": 19556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6182574119066795,
+      "learning_rate": 1.217177827388309e-06,
+      "loss": 0.3128,
+      "step": 19557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8946340802213821,
+      "learning_rate": 1.216978044914962e-06,
+      "loss": 0.4674,
+      "step": 19558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.397144097115403,
+      "learning_rate": 1.2167783521040188e-06,
+      "loss": 0.412,
+      "step": 19559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8888109128874235,
+      "learning_rate": 1.2165787489596603e-06,
+      "loss": 0.4413,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6355290792765882,
+      "learning_rate": 1.2163792354860657e-06,
+      "loss": 0.4961,
+      "step": 19561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.540216603899418,
+      "learning_rate": 1.2161798116874105e-06,
+      "loss": 0.5716,
+      "step": 19562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6395025293386762,
+      "learning_rate": 1.2159804775678705e-06,
+      "loss": 0.4477,
+      "step": 19563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9201585297916954,
+      "learning_rate": 1.2157812331316188e-06,
+      "loss": 0.4777,
+      "step": 19564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8005804245011416,
+      "learning_rate": 1.215582078382825e-06,
+      "loss": 0.3604,
+      "step": 19565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.000755085118462,
+      "learning_rate": 1.2153830133256598e-06,
+      "loss": 0.5567,
+      "step": 19566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.571115393555732,
+      "learning_rate": 1.2151840379642895e-06,
+      "loss": 0.47,
+      "step": 19567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6975804126603813,
+      "learning_rate": 1.2149851523028805e-06,
+      "loss": 0.4253,
+      "step": 19568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.640455432671996,
+      "learning_rate": 1.2147863563455956e-06,
+      "loss": 0.3434,
+      "step": 19569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8714522792665509,
+      "learning_rate": 1.2145876500965963e-06,
+      "loss": 0.515,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.208418974455163,
+      "learning_rate": 1.2143890335600437e-06,
+      "loss": 0.5686,
+      "step": 19571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8156297194033597,
+      "learning_rate": 1.2141905067400934e-06,
+      "loss": 0.5999,
+      "step": 19572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.070253478047603,
+      "learning_rate": 1.2139920696409044e-06,
+      "loss": 0.5176,
+      "step": 19573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8817682601325272,
+      "learning_rate": 1.2137937222666282e-06,
+      "loss": 0.4843,
+      "step": 19574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7090953208290696,
+      "learning_rate": 1.2135954646214185e-06,
+      "loss": 0.4214,
+      "step": 19575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0830603521010604,
+      "learning_rate": 1.2133972967094249e-06,
+      "loss": 0.4927,
+      "step": 19576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.939065354020831,
+      "learning_rate": 1.2131992185347973e-06,
+      "loss": 0.4541,
+      "step": 19577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6454064732991598,
+      "learning_rate": 1.2130012301016813e-06,
+      "loss": 0.4539,
+      "step": 19578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.961469594481315,
+      "learning_rate": 1.2128033314142209e-06,
+      "loss": 0.5312,
+      "step": 19579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.768391983464125,
+      "learning_rate": 1.212605522476561e-06,
+      "loss": 0.5363,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8823963734459672,
+      "learning_rate": 1.2124078032928405e-06,
+      "loss": 0.4139,
+      "step": 19581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8288711423990704,
+      "learning_rate": 1.2122101738672001e-06,
+      "loss": 0.5031,
+      "step": 19582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7777149057809023,
+      "learning_rate": 1.2120126342037767e-06,
+      "loss": 0.5561,
+      "step": 19583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6471017197193918,
+      "learning_rate": 1.211815184306705e-06,
+      "loss": 0.4388,
+      "step": 19584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5970781673681418,
+      "learning_rate": 1.2116178241801188e-06,
+      "loss": 0.4376,
+      "step": 19585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6849016575825855,
+      "learning_rate": 1.2114205538281508e-06,
+      "loss": 0.4037,
+      "step": 19586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8945931695107716,
+      "learning_rate": 1.2112233732549282e-06,
+      "loss": 0.6125,
+      "step": 19587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6688867178775542,
+      "learning_rate": 1.2110262824645812e-06,
+      "loss": 0.42,
+      "step": 19588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.367362839229205,
+      "learning_rate": 1.210829281461235e-06,
+      "loss": 0.6096,
+      "step": 19589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2934589060934267,
+      "learning_rate": 1.2106323702490141e-06,
+      "loss": 0.4019,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6011892209608234,
+      "learning_rate": 1.2104355488320397e-06,
+      "loss": 0.5201,
+      "step": 19591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1945200444041078,
+      "learning_rate": 1.2102388172144331e-06,
+      "loss": 0.6523,
+      "step": 19592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9185112109413525,
+      "learning_rate": 1.2100421754003127e-06,
+      "loss": 0.4313,
+      "step": 19593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6821142291594386,
+      "learning_rate": 1.2098456233937941e-06,
+      "loss": 0.3737,
+      "step": 19594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.797268416200228,
+      "learning_rate": 1.2096491611989926e-06,
+      "loss": 0.4519,
+      "step": 19595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.884892151973323,
+      "learning_rate": 1.2094527888200218e-06,
+      "loss": 0.616,
+      "step": 19596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2036738572302057,
+      "learning_rate": 1.2092565062609912e-06,
+      "loss": 0.4692,
+      "step": 19597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.669533433692798,
+      "learning_rate": 1.2090603135260106e-06,
+      "loss": 0.485,
+      "step": 19598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0464535303025873,
+      "learning_rate": 1.2088642106191878e-06,
+      "loss": 0.5364,
+      "step": 19599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.698359840338805,
+      "learning_rate": 1.2086681975446268e-06,
+      "loss": 0.4322,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.48379337787628174,
+      "eval_runtime": 67.952,
+      "eval_samples_per_second": 50.182,
+      "eval_steps_per_second": 1.575,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4292185669866064,
+      "learning_rate": 1.2084722743064317e-06,
+      "loss": 0.3803,
+      "step": 19601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7991979639018385,
+      "learning_rate": 1.208276440908704e-06,
+      "loss": 0.5222,
+      "step": 19602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8263029279702387,
+      "learning_rate": 1.2080806973555437e-06,
+      "loss": 0.561,
+      "step": 19603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7679071598267262,
+      "learning_rate": 1.207885043651048e-06,
+      "loss": 0.3884,
+      "step": 19604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5592319173192466,
+      "learning_rate": 1.207689479799313e-06,
+      "loss": 0.4502,
+      "step": 19605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.85272795312922,
+      "learning_rate": 1.2074940058044327e-06,
+      "loss": 0.7446,
+      "step": 19606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4609315369440568,
+      "learning_rate": 1.2072986216704988e-06,
+      "loss": 0.4487,
+      "step": 19607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5560585756297904,
+      "learning_rate": 1.2071033274016022e-06,
+      "loss": 0.4829,
+      "step": 19608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3892070222752662,
+      "learning_rate": 1.2069081230018313e-06,
+      "loss": 0.3722,
+      "step": 19609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5792949382418715,
+      "learning_rate": 1.2067130084752723e-06,
+      "loss": 0.2642,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5366846455352947,
+      "learning_rate": 1.2065179838260093e-06,
+      "loss": 0.3913,
+      "step": 19611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5681263204520743,
+      "learning_rate": 1.2063230490581264e-06,
+      "loss": 0.4171,
+      "step": 19612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.613624711963599,
+      "learning_rate": 1.2061282041757024e-06,
+      "loss": 0.5106,
+      "step": 19613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.092535241607814,
+      "learning_rate": 1.2059334491828182e-06,
+      "loss": 0.4658,
+      "step": 19614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8400416063975056,
+      "learning_rate": 1.2057387840835495e-06,
+      "loss": 0.4917,
+      "step": 19615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1795009671162204,
+      "learning_rate": 1.2055442088819733e-06,
+      "loss": 0.5087,
+      "step": 19616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0778384657908475,
+      "learning_rate": 1.2053497235821604e-06,
+      "loss": 0.5347,
+      "step": 19617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.557038702883696,
+      "learning_rate": 1.205155328188184e-06,
+      "loss": 0.3717,
+      "step": 19618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4491272355080702,
+      "learning_rate": 1.2049610227041135e-06,
+      "loss": 0.3545,
+      "step": 19619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7366699181954377,
+      "learning_rate": 1.2047668071340155e-06,
+      "loss": 0.4487,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6824409916713075,
+      "learning_rate": 1.2045726814819573e-06,
+      "loss": 0.4218,
+      "step": 19621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6754054313343245,
+      "learning_rate": 1.2043786457520014e-06,
+      "loss": 0.3852,
+      "step": 19622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7933100855336757,
+      "learning_rate": 1.2041846999482103e-06,
+      "loss": 0.5975,
+      "step": 19623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6293377666818822,
+      "learning_rate": 1.2039908440746448e-06,
+      "loss": 0.4379,
+      "step": 19624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5891034446378982,
+      "learning_rate": 1.203797078135363e-06,
+      "loss": 0.4891,
+      "step": 19625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6092315048622414,
+      "learning_rate": 1.2036034021344202e-06,
+      "loss": 0.3916,
+      "step": 19626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3677637872592534,
+      "learning_rate": 1.203409816075872e-06,
+      "loss": 0.316,
+      "step": 19627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9982514293084448,
+      "learning_rate": 1.2032163199637708e-06,
+      "loss": 0.651,
+      "step": 19628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.733671287696918,
+      "learning_rate": 1.2030229138021662e-06,
+      "loss": 0.3699,
+      "step": 19629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7664138570504868,
+      "learning_rate": 1.2028295975951087e-06,
+      "loss": 0.3941,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.59487618030114,
+      "learning_rate": 1.2026363713466445e-06,
+      "loss": 0.5132,
+      "step": 19631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.363582636221904,
+      "learning_rate": 1.2024432350608182e-06,
+      "loss": 0.3886,
+      "step": 19632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7839685793926572,
+      "learning_rate": 1.2022501887416735e-06,
+      "loss": 0.4303,
+      "step": 19633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.145696599528129,
+      "learning_rate": 1.2020572323932517e-06,
+      "loss": 0.6247,
+      "step": 19634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.733307728668349,
+      "learning_rate": 1.2018643660195931e-06,
+      "loss": 0.4054,
+      "step": 19635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.906879803070804,
+      "learning_rate": 1.2016715896247332e-06,
+      "loss": 0.4184,
+      "step": 19636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.605921166108611,
+      "learning_rate": 1.2014789032127091e-06,
+      "loss": 0.3698,
+      "step": 19637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.04321988233492,
+      "learning_rate": 1.201286306787555e-06,
+      "loss": 0.5305,
+      "step": 19638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6008585448035466,
+      "learning_rate": 1.201093800353301e-06,
+      "loss": 0.3177,
+      "step": 19639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7871154305159527,
+      "learning_rate": 1.2009013839139783e-06,
+      "loss": 0.4919,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0118905668447224,
+      "learning_rate": 1.2007090574736156e-06,
+      "loss": 0.7026,
+      "step": 19641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5195592259096813,
+      "learning_rate": 1.2005168210362377e-06,
+      "loss": 0.3337,
+      "step": 19642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6740920517190994,
+      "learning_rate": 1.2003246746058694e-06,
+      "loss": 0.413,
+      "step": 19643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0647183563422202,
+      "learning_rate": 1.2001326181865343e-06,
+      "loss": 0.4108,
+      "step": 19644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7602609125283637,
+      "learning_rate": 1.1999406517822518e-06,
+      "loss": 0.3633,
+      "step": 19645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4612927282868393,
+      "learning_rate": 1.1997487753970403e-06,
+      "loss": 0.4285,
+      "step": 19646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8073852008147875,
+      "learning_rate": 1.1995569890349173e-06,
+      "loss": 0.6332,
+      "step": 19647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.092814720758521,
+      "learning_rate": 1.1993652926998987e-06,
+      "loss": 0.4434,
+      "step": 19648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.157601826219726,
+      "learning_rate": 1.1991736863959955e-06,
+      "loss": 0.5182,
+      "step": 19649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6660914279300665,
+      "learning_rate": 1.1989821701272197e-06,
+      "loss": 0.2909,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.4831776022911072,
+      "eval_runtime": 67.9013,
+      "eval_samples_per_second": 50.22,
+      "eval_steps_per_second": 1.576,
+      "step": 19650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8615742526072325,
+      "learning_rate": 1.1987907438975816e-06,
+      "loss": 0.4893,
+      "step": 19651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.142590846037697,
+      "learning_rate": 1.1985994077110868e-06,
+      "loss": 0.5412,
+      "step": 19652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8226298142467483,
+      "learning_rate": 1.1984081615717418e-06,
+      "loss": 0.4298,
+      "step": 19653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0707110954364296,
+      "learning_rate": 1.198217005483551e-06,
+      "loss": 0.5117,
+      "step": 19654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7752706956921516,
+      "learning_rate": 1.1980259394505144e-06,
+      "loss": 0.4732,
+      "step": 19655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.545235699515149,
+      "learning_rate": 1.197834963476633e-06,
+      "loss": 0.518,
+      "step": 19656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.811281044059486,
+      "learning_rate": 1.197644077565905e-06,
+      "loss": 0.5072,
+      "step": 19657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1355056374466432,
+      "learning_rate": 1.1974532817223245e-06,
+      "loss": 0.728,
+      "step": 19658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8223805859201834,
+      "learning_rate": 1.1972625759498884e-06,
+      "loss": 0.4396,
+      "step": 19659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.678078458467315,
+      "learning_rate": 1.197071960252587e-06,
+      "loss": 0.3335,
+      "step": 19660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9820818182857391,
+      "learning_rate": 1.1968814346344121e-06,
+      "loss": 0.5968,
+      "step": 19661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5688359864854302,
+      "learning_rate": 1.196690999099351e-06,
+      "loss": 0.3252,
+      "step": 19662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.013550491776473,
+      "learning_rate": 1.196500653651391e-06,
+      "loss": 0.6971,
+      "step": 19663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0110771917990053,
+      "learning_rate": 1.1963103982945177e-06,
+      "loss": 0.441,
+      "step": 19664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5122528349119237,
+      "learning_rate": 1.196120233032712e-06,
+      "loss": 0.3869,
+      "step": 19665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8528203360184936,
+      "learning_rate": 1.1959301578699565e-06,
+      "loss": 0.4688,
+      "step": 19666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.625129319797278,
+      "learning_rate": 1.19574017281023e-06,
+      "loss": 0.4699,
+      "step": 19667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7377945048286525,
+      "learning_rate": 1.1955502778575092e-06,
+      "loss": 0.432,
+      "step": 19668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.004418248606421,
+      "learning_rate": 1.1953604730157695e-06,
+      "loss": 0.5086,
+      "step": 19669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8573782770151326,
+      "learning_rate": 1.1951707582889848e-06,
+      "loss": 0.5632,
+      "step": 19670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7859873009544938,
+      "learning_rate": 1.194981133681126e-06,
+      "loss": 0.4984,
+      "step": 19671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6798262052291129,
+      "learning_rate": 1.1947915991961638e-06,
+      "loss": 0.409,
+      "step": 19672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2394806405925016,
+      "learning_rate": 1.1946021548380656e-06,
+      "loss": 0.3909,
+      "step": 19673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1412447480252066,
+      "learning_rate": 1.1944128006107965e-06,
+      "loss": 0.4674,
+      "step": 19674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2339924208118622,
+      "learning_rate": 1.1942235365183217e-06,
+      "loss": 0.3478,
+      "step": 19675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8857282795850099,
+      "learning_rate": 1.1940343625646028e-06,
+      "loss": 0.547,
+      "step": 19676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.035484486855795,
+      "learning_rate": 1.1938452787535993e-06,
+      "loss": 0.5625,
+      "step": 19677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.623546074746792,
+      "learning_rate": 1.1936562850892707e-06,
+      "loss": 0.4258,
+      "step": 19678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.309167947600049,
+      "learning_rate": 1.1934673815755733e-06,
+      "loss": 0.3632,
+      "step": 19679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6732887853568517,
+      "learning_rate": 1.1932785682164606e-06,
+      "loss": 0.4408,
+      "step": 19680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.801242876265641,
+      "learning_rate": 1.1930898450158867e-06,
+      "loss": 0.4173,
+      "step": 19681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6420521315428336,
+      "learning_rate": 1.1929012119778012e-06,
+      "loss": 0.3914,
+      "step": 19682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8675312013709924,
+      "learning_rate": 1.1927126691061543e-06,
+      "loss": 0.5523,
+      "step": 19683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.770533917457641,
+      "learning_rate": 1.1925242164048917e-06,
+      "loss": 0.5214,
+      "step": 19684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0385575836443115,
+      "learning_rate": 1.1923358538779595e-06,
+      "loss": 0.4989,
+      "step": 19685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6848040646432827,
+      "learning_rate": 1.1921475815293006e-06,
+      "loss": 0.3736,
+      "step": 19686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6292427246212646,
+      "learning_rate": 1.1919593993628562e-06,
+      "loss": 0.3899,
+      "step": 19687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8224660355813778,
+      "learning_rate": 1.1917713073825662e-06,
+      "loss": 0.4948,
+      "step": 19688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7466076718807042,
+      "learning_rate": 1.191583305592368e-06,
+      "loss": 0.4131,
+      "step": 19689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.72116736415107,
+      "learning_rate": 1.1913953939961968e-06,
+      "loss": 0.3888,
+      "step": 19690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.767719112863756,
+      "learning_rate": 1.191207572597987e-06,
+      "loss": 0.5272,
+      "step": 19691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7574209622057009,
+      "learning_rate": 1.1910198414016706e-06,
+      "loss": 0.3698,
+      "step": 19692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.495489206764438,
+      "learning_rate": 1.1908322004111771e-06,
+      "loss": 0.6122,
+      "step": 19693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7429403367050476,
+      "learning_rate": 1.1906446496304353e-06,
+      "loss": 0.389,
+      "step": 19694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5638640964358452,
+      "learning_rate": 1.1904571890633705e-06,
+      "loss": 0.4069,
+      "step": 19695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0252169843984977,
+      "learning_rate": 1.1902698187139084e-06,
+      "loss": 0.4733,
+      "step": 19696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8939985075207204,
+      "learning_rate": 1.1900825385859708e-06,
+      "loss": 0.4996,
+      "step": 19697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.100731249592974,
+      "learning_rate": 1.1898953486834778e-06,
+      "loss": 0.5812,
+      "step": 19698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.991589804359461,
+      "learning_rate": 1.1897082490103492e-06,
+      "loss": 0.4504,
+      "step": 19699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7372143513031852,
+      "learning_rate": 1.1895212395705005e-06,
+      "loss": 0.3755,
+      "step": 19700
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.48350057005882263,
+      "eval_runtime": 67.8678,
+      "eval_samples_per_second": 50.245,
+      "eval_steps_per_second": 1.577,
+      "step": 19700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9619764859905657,
+      "learning_rate": 1.1893343203678475e-06,
+      "loss": 0.5339,
+      "step": 19701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.88752954244107,
+      "learning_rate": 1.189147491406303e-06,
+      "loss": 0.4834,
+      "step": 19702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7522109492732416,
+      "learning_rate": 1.1889607526897786e-06,
+      "loss": 0.5169,
+      "step": 19703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3502362510818677,
+      "learning_rate": 1.1887741042221825e-06,
+      "loss": 0.3774,
+      "step": 19704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9557421442137861,
+      "learning_rate": 1.1885875460074235e-06,
+      "loss": 0.5086,
+      "step": 19705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.900676886651564,
+      "learning_rate": 1.1884010780494058e-06,
+      "loss": 0.527,
+      "step": 19706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7875386977929852,
+      "learning_rate": 1.1882147003520331e-06,
+      "loss": 0.5224,
+      "step": 19707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0553360123661397,
+      "learning_rate": 1.1880284129192082e-06,
+      "loss": 0.5779,
+      "step": 19708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4041423288113688,
+      "learning_rate": 1.1878422157548303e-06,
+      "loss": 0.2807,
+      "step": 19709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.351065117415156,
+      "learning_rate": 1.1876561088627965e-06,
+      "loss": 0.3208,
+      "step": 19710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.975420789865122,
+      "learning_rate": 1.1874700922470037e-06,
+      "loss": 0.4136,
+      "step": 19711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5850010948672955,
+      "learning_rate": 1.187284165911347e-06,
+      "loss": 0.3363,
+      "step": 19712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6344570890637313,
+      "learning_rate": 1.1870983298597164e-06,
+      "loss": 0.3803,
+      "step": 19713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.932996948696917,
+      "learning_rate": 1.186912584096004e-06,
+      "loss": 0.6149,
+      "step": 19714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6634605445676942,
+      "learning_rate": 1.1867269286240973e-06,
+      "loss": 0.514,
+      "step": 19715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3748686686440745,
+      "learning_rate": 1.1865413634478837e-06,
+      "loss": 0.4725,
+      "step": 19716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3665634425789606,
+      "learning_rate": 1.186355888571247e-06,
+      "loss": 0.3434,
+      "step": 19717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6547863442801534,
+      "learning_rate": 1.186170503998071e-06,
+      "loss": 0.3725,
+      "step": 19718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6183687429099183,
+      "learning_rate": 1.1859852097322352e-06,
+      "loss": 0.4036,
+      "step": 19719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.06090669316911,
+      "learning_rate": 1.1858000057776197e-06,
+      "loss": 0.594,
+      "step": 19720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8229700008603522,
+      "learning_rate": 1.1856148921381023e-06,
+      "loss": 0.4599,
+      "step": 19721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6684952762846144,
+      "learning_rate": 1.1854298688175565e-06,
+      "loss": 0.4983,
+      "step": 19722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0012180294918136,
+      "learning_rate": 1.185244935819857e-06,
+      "loss": 0.5046,
+      "step": 19723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6281879223434694,
+      "learning_rate": 1.185060093148875e-06,
+      "loss": 0.492,
+      "step": 19724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7748261664855578,
+      "learning_rate": 1.1848753408084792e-06,
+      "loss": 0.531,
+      "step": 19725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.661708520794373,
+      "learning_rate": 1.1846906788025382e-06,
+      "loss": 0.5904,
+      "step": 19726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.142203114789165,
+      "learning_rate": 1.184506107134917e-06,
+      "loss": 0.6439,
+      "step": 19727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8264676397039143,
+      "learning_rate": 1.184321625809481e-06,
+      "loss": 0.4279,
+      "step": 19728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.813837119556389,
+      "learning_rate": 1.1841372348300908e-06,
+      "loss": 0.4567,
+      "step": 19729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5266157828543963,
+      "learning_rate": 1.1839529342006068e-06,
+      "loss": 0.4472,
+      "step": 19730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6326451481426196,
+      "learning_rate": 1.1837687239248875e-06,
+      "loss": 0.4383,
+      "step": 19731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.641675067607523,
+      "learning_rate": 1.1835846040067886e-06,
+      "loss": 0.6003,
+      "step": 19732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6654209819507384,
+      "learning_rate": 1.1834005744501655e-06,
+      "loss": 0.3413,
+      "step": 19733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2637714404245393,
+      "learning_rate": 1.1832166352588703e-06,
+      "loss": 0.637,
+      "step": 19734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8173068950327462,
+      "learning_rate": 1.183032786436753e-06,
+      "loss": 0.4221,
+      "step": 19735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9631416695179016,
+      "learning_rate": 1.1828490279876633e-06,
+      "loss": 0.561,
+      "step": 19736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5197893910932219,
+      "learning_rate": 1.1826653599154477e-06,
+      "loss": 0.4278,
+      "step": 19737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4957864442707045,
+      "learning_rate": 1.1824817822239516e-06,
+      "loss": 0.3448,
+      "step": 19738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1564836825987155,
+      "learning_rate": 1.1822982949170166e-06,
+      "loss": 0.5192,
+      "step": 19739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8925571836579984,
+      "learning_rate": 1.1821148979984854e-06,
+      "loss": 0.5614,
+      "step": 19740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3739891779021427,
+      "learning_rate": 1.1819315914721976e-06,
+      "loss": 0.3712,
+      "step": 19741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7435202318263767,
+      "learning_rate": 1.1817483753419892e-06,
+      "loss": 0.4399,
+      "step": 19742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5747845687175923,
+      "learning_rate": 1.1815652496116964e-06,
+      "loss": 0.3765,
+      "step": 19743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6392504688011063,
+      "learning_rate": 1.1813822142851535e-06,
+      "loss": 0.3828,
+      "step": 19744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8961394756845182,
+      "learning_rate": 1.1811992693661908e-06,
+      "loss": 0.3943,
+      "step": 19745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8777632420976214,
+      "learning_rate": 1.1810164148586392e-06,
+      "loss": 0.367,
+      "step": 19746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8481975479948256,
+      "learning_rate": 1.1808336507663265e-06,
+      "loss": 0.4009,
+      "step": 19747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9494824848961079,
+      "learning_rate": 1.1806509770930785e-06,
+      "loss": 0.6027,
+      "step": 19748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0852491230347265,
+      "learning_rate": 1.1804683938427188e-06,
+      "loss": 0.7364,
+      "step": 19749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.679145145566303,
+      "learning_rate": 1.1802859010190715e-06,
+      "loss": 0.4141,
+      "step": 19750
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.4837722182273865,
+      "eval_runtime": 67.8974,
+      "eval_samples_per_second": 50.223,
+      "eval_steps_per_second": 1.576,
+      "step": 19750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4972814444440208,
+      "learning_rate": 1.180103498625955e-06,
+      "loss": 0.3748,
+      "step": 19751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6267725423010455,
+      "learning_rate": 1.1799211866671883e-06,
+      "loss": 0.4027,
+      "step": 19752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7357724052161985,
+      "learning_rate": 1.1797389651465887e-06,
+      "loss": 0.444,
+      "step": 19753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6186339729867878,
+      "learning_rate": 1.1795568340679704e-06,
+      "loss": 0.4096,
+      "step": 19754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9760158204146328,
+      "learning_rate": 1.1793747934351465e-06,
+      "loss": 0.5083,
+      "step": 19755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5423217213566498,
+      "learning_rate": 1.179192843251927e-06,
+      "loss": 0.4731,
+      "step": 19756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8083464511246992,
+      "learning_rate": 1.1790109835221225e-06,
+      "loss": 0.5185,
+      "step": 19757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0390893428270975,
+      "learning_rate": 1.1788292142495384e-06,
+      "loss": 0.4849,
+      "step": 19758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9321146599085461,
+      "learning_rate": 1.178647535437981e-06,
+      "loss": 0.5401,
+      "step": 19759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7924718471863037,
+      "learning_rate": 1.1784659470912537e-06,
+      "loss": 0.4213,
+      "step": 19760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.769783956743615,
+      "learning_rate": 1.178284449213157e-06,
+      "loss": 0.4453,
+      "step": 19761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8345418897144488,
+      "learning_rate": 1.1781030418074912e-06,
+      "loss": 0.5023,
+      "step": 19762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7508481892757037,
+      "learning_rate": 1.177921724878054e-06,
+      "loss": 0.4676,
+      "step": 19763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9102109039141681,
+      "learning_rate": 1.177740498428641e-06,
+      "loss": 0.5784,
+      "step": 19764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6887383909405147,
+      "learning_rate": 1.1775593624630455e-06,
+      "loss": 0.442,
+      "step": 19765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8817819296608156,
+      "learning_rate": 1.1773783169850602e-06,
+      "loss": 0.4365,
+      "step": 19766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.94485499657731,
+      "learning_rate": 1.1771973619984752e-06,
+      "loss": 0.5495,
+      "step": 19767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5971247686204757,
+      "learning_rate": 1.1770164975070781e-06,
+      "loss": 0.3981,
+      "step": 19768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.667859742824532,
+      "learning_rate": 1.1768357235146557e-06,
+      "loss": 0.5664,
+      "step": 19769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3781367092943038,
+      "learning_rate": 1.1766550400249922e-06,
+      "loss": 0.3732,
+      "step": 19770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8165689744464555,
+      "learning_rate": 1.17647444704187e-06,
+      "loss": 0.4371,
+      "step": 19771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6452408483993464,
+      "learning_rate": 1.1762939445690703e-06,
+      "loss": 0.4082,
+      "step": 19772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8239108340054675,
+      "learning_rate": 1.1761135326103705e-06,
+      "loss": 0.6101,
+      "step": 19773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9570919132258435,
+      "learning_rate": 1.1759332111695488e-06,
+      "loss": 0.6087,
+      "step": 19774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.834768390663762,
+      "learning_rate": 1.1757529802503788e-06,
+      "loss": 0.4238,
+      "step": 19775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7449487993391384,
+      "learning_rate": 1.1755728398566356e-06,
+      "loss": 0.4505,
+      "step": 19776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.590663058407217,
+      "learning_rate": 1.175392789992088e-06,
+      "loss": 0.4878,
+      "step": 19777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8510301766837631,
+      "learning_rate": 1.1752128306605063e-06,
+      "loss": 0.3708,
+      "step": 19778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4098471035278481,
+      "learning_rate": 1.1750329618656586e-06,
+      "loss": 0.3712,
+      "step": 19779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7448506750335484,
+      "learning_rate": 1.1748531836113086e-06,
+      "loss": 0.5788,
+      "step": 19780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7581328226509039,
+      "learning_rate": 1.1746734959012209e-06,
+      "loss": 0.6265,
+      "step": 19781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7347191879502961,
+      "learning_rate": 1.1744938987391576e-06,
+      "loss": 0.3703,
+      "step": 19782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7056942016446546,
+      "learning_rate": 1.1743143921288776e-06,
+      "loss": 0.4431,
+      "step": 19783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9550902469120495,
+      "learning_rate": 1.1741349760741386e-06,
+      "loss": 0.47,
+      "step": 19784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.942707223495989,
+      "learning_rate": 1.1739556505786975e-06,
+      "loss": 0.5278,
+      "step": 19785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7303156850339023,
+      "learning_rate": 1.173776415646307e-06,
+      "loss": 0.449,
+      "step": 19786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4533748837937526,
+      "learning_rate": 1.173597271280721e-06,
+      "loss": 0.6962,
+      "step": 19787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9368536073135783,
+      "learning_rate": 1.1734182174856885e-06,
+      "loss": 0.5274,
+      "step": 19788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5463534615045584,
+      "learning_rate": 1.1732392542649586e-06,
+      "loss": 0.428,
+      "step": 19789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7842734098191493,
+      "learning_rate": 1.1730603816222772e-06,
+      "loss": 0.5286,
+      "step": 19790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.513093477674893,
+      "learning_rate": 1.172881599561389e-06,
+      "loss": 0.4476,
+      "step": 19791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.798877499293191,
+      "learning_rate": 1.1727029080860376e-06,
+      "loss": 0.4854,
+      "step": 19792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7327043440391936,
+      "learning_rate": 1.172524307199962e-06,
+      "loss": 0.5044,
+      "step": 19793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6479117265999934,
+      "learning_rate": 1.1723457969069026e-06,
+      "loss": 0.3741,
+      "step": 19794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4668108505683006,
+      "learning_rate": 1.1721673772105962e-06,
+      "loss": 0.3905,
+      "step": 19795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.487664322215532,
+      "learning_rate": 1.1719890481147772e-06,
+      "loss": 0.2969,
+      "step": 19796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6901891475902047,
+      "learning_rate": 1.1718108096231792e-06,
+      "loss": 0.5138,
+      "step": 19797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5931690418772646,
+      "learning_rate": 1.1716326617395342e-06,
+      "loss": 0.4848,
+      "step": 19798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6027498546614325,
+      "learning_rate": 1.1714546044675704e-06,
+      "loss": 0.372,
+      "step": 19799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4248745080574086,
+      "learning_rate": 1.1712766378110158e-06,
+      "loss": 0.5488,
+      "step": 19800
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.48337623476982117,
+      "eval_runtime": 67.8299,
+      "eval_samples_per_second": 50.273,
+      "eval_steps_per_second": 1.577,
+      "step": 19800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.687091498540547,
+      "learning_rate": 1.1710987617735962e-06,
+      "loss": 0.4049,
+      "step": 19801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8612621451427334,
+      "learning_rate": 1.170920976359036e-06,
+      "loss": 0.542,
+      "step": 19802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3153699695664496,
+      "learning_rate": 1.1707432815710555e-06,
+      "loss": 0.3239,
+      "step": 19803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.752405922021592,
+      "learning_rate": 1.1705656774133753e-06,
+      "loss": 0.4603,
+      "step": 19804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7984944150021351,
+      "learning_rate": 1.1703881638897139e-06,
+      "loss": 0.3839,
+      "step": 19805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.586996539310126,
+      "learning_rate": 1.1702107410037868e-06,
+      "loss": 0.381,
+      "step": 19806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5065689713259929,
+      "learning_rate": 1.1700334087593088e-06,
+      "loss": 0.3194,
+      "step": 19807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2992369154152663,
+      "learning_rate": 1.1698561671599924e-06,
+      "loss": 0.5732,
+      "step": 19808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6355202416160461,
+      "learning_rate": 1.1696790162095465e-06,
+      "loss": 0.3829,
+      "step": 19809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7155111330091757,
+      "learning_rate": 1.169501955911681e-06,
+      "loss": 0.4274,
+      "step": 19810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6505652561470128,
+      "learning_rate": 1.1693249862701031e-06,
+      "loss": 0.432,
+      "step": 19811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4522645314965248,
+      "learning_rate": 1.169148107288516e-06,
+      "loss": 0.3342,
+      "step": 19812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7935379071557256,
+      "learning_rate": 1.1689713189706232e-06,
+      "loss": 0.4249,
+      "step": 19813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.617644794208625,
+      "learning_rate": 1.168794621320126e-06,
+      "loss": 0.546,
+      "step": 19814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1502300307792206,
+      "learning_rate": 1.1686180143407237e-06,
+      "loss": 0.6354,
+      "step": 19815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8075182970164116,
+      "learning_rate": 1.1684414980361123e-06,
+      "loss": 0.485,
+      "step": 19816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7966147378012447,
+      "learning_rate": 1.1682650724099875e-06,
+      "loss": 0.5448,
+      "step": 19817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4997364235670523,
+      "learning_rate": 1.1680887374660428e-06,
+      "loss": 0.3429,
+      "step": 19818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5158522529028076,
+      "learning_rate": 1.16791249320797e-06,
+      "loss": 0.3888,
+      "step": 19819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7513901326337595,
+      "learning_rate": 1.1677363396394587e-06,
+      "loss": 0.5697,
+      "step": 19820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8297169080513047,
+      "learning_rate": 1.1675602767641965e-06,
+      "loss": 0.5005,
+      "step": 19821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9301604957091822,
+      "learning_rate": 1.1673843045858684e-06,
+      "loss": 0.4477,
+      "step": 19822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5139233577025952,
+      "learning_rate": 1.1672084231081589e-06,
+      "loss": 0.4208,
+      "step": 19823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.947813592516368,
+      "learning_rate": 1.16703263233475e-06,
+      "loss": 0.4039,
+      "step": 19824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8891904560560695,
+      "learning_rate": 1.1668569322693212e-06,
+      "loss": 0.5738,
+      "step": 19825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7009201624617398,
+      "learning_rate": 1.1666813229155516e-06,
+      "loss": 0.4768,
+      "step": 19826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2558230861553508,
+      "learning_rate": 1.1665058042771173e-06,
+      "loss": 0.4473,
+      "step": 19827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0178092633134783,
+      "learning_rate": 1.1663303763576912e-06,
+      "loss": 0.6872,
+      "step": 19828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9970104713496093,
+      "learning_rate": 1.1661550391609475e-06,
+      "loss": 0.4425,
+      "step": 19829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.773108788738085,
+      "learning_rate": 1.1659797926905568e-06,
+      "loss": 0.4625,
+      "step": 19830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.088064660721392,
+      "learning_rate": 1.1658046369501865e-06,
+      "loss": 0.5664,
+      "step": 19831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0638344122042187,
+      "learning_rate": 1.1656295719435042e-06,
+      "loss": 0.6525,
+      "step": 19832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.003120414010613,
+      "learning_rate": 1.1654545976741743e-06,
+      "loss": 0.4989,
+      "step": 19833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9345452727336356,
+      "learning_rate": 1.165279714145861e-06,
+      "loss": 0.4614,
+      "step": 19834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6555806086073535,
+      "learning_rate": 1.165104921362224e-06,
+      "loss": 0.4196,
+      "step": 19835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4819836876224426,
+      "learning_rate": 1.1649302193269226e-06,
+      "loss": 0.3669,
+      "step": 19836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9499311437410374,
+      "learning_rate": 1.1647556080436154e-06,
+      "loss": 0.5093,
+      "step": 19837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8395896620159877,
+      "learning_rate": 1.164581087515956e-06,
+      "loss": 0.4463,
+      "step": 19838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4259438864699745,
+      "learning_rate": 1.1644066577475988e-06,
+      "loss": 0.4208,
+      "step": 19839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0574611280614583,
+      "learning_rate": 1.1642323187421956e-06,
+      "loss": 0.4869,
+      "step": 19840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6006085962334267,
+      "learning_rate": 1.1640580705033954e-06,
+      "loss": 0.4204,
+      "step": 19841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8462033252280508,
+      "learning_rate": 1.1638839130348463e-06,
+      "loss": 0.4485,
+      "step": 19842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6230187665646636,
+      "learning_rate": 1.1637098463401946e-06,
+      "loss": 0.4244,
+      "step": 19843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7240384885272022,
+      "learning_rate": 1.1635358704230835e-06,
+      "loss": 0.4547,
+      "step": 19844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.834389185113113,
+      "learning_rate": 1.1633619852871558e-06,
+      "loss": 0.5066,
+      "step": 19845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2645071739700504,
+      "learning_rate": 1.163188190936051e-06,
+      "loss": 0.3616,
+      "step": 19846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.783347308286588,
+      "learning_rate": 1.1630144873734082e-06,
+      "loss": 0.3984,
+      "step": 19847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9071956771333365,
+      "learning_rate": 1.162840874602863e-06,
+      "loss": 0.5126,
+      "step": 19848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9844774332210522,
+      "learning_rate": 1.16266735262805e-06,
+      "loss": 0.6371,
+      "step": 19849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8708361956300745,
+      "learning_rate": 1.1624939214526027e-06,
+      "loss": 0.4364,
+      "step": 19850
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.48300668597221375,
+      "eval_runtime": 67.8439,
+      "eval_samples_per_second": 50.262,
+      "eval_steps_per_second": 1.577,
+      "step": 19850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3713903826682452,
+      "learning_rate": 1.1623205810801504e-06,
+      "loss": 0.3572,
+      "step": 19851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9983145497167183,
+      "learning_rate": 1.1621473315143223e-06,
+      "loss": 0.5747,
+      "step": 19852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7213279352541735,
+      "learning_rate": 1.1619741727587467e-06,
+      "loss": 0.3664,
+      "step": 19853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7255077357800928,
+      "learning_rate": 1.1618011048170463e-06,
+      "loss": 0.5294,
+      "step": 19854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.903627070987335,
+      "learning_rate": 1.161628127692845e-06,
+      "loss": 0.6452,
+      "step": 19855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4041756434812194,
+      "learning_rate": 1.161455241389765e-06,
+      "loss": 0.3066,
+      "step": 19856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9545952165440188,
+      "learning_rate": 1.161282445911424e-06,
+      "loss": 0.5935,
+      "step": 19857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9509959019579335,
+      "learning_rate": 1.1611097412614404e-06,
+      "loss": 0.5022,
+      "step": 19858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0607603286298386,
+      "learning_rate": 1.1609371274434292e-06,
+      "loss": 0.4254,
+      "step": 19859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5809121985400862,
+      "learning_rate": 1.1607646044610046e-06,
+      "loss": 0.3776,
+      "step": 19860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.096089919206975,
+      "learning_rate": 1.1605921723177777e-06,
+      "loss": 0.4697,
+      "step": 19861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7635758428137638,
+      "learning_rate": 1.160419831017358e-06,
+      "loss": 0.4016,
+      "step": 19862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.630888449601496,
+      "learning_rate": 1.1602475805633541e-06,
+      "loss": 0.3876,
+      "step": 19863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7850939221745694,
+      "learning_rate": 1.1600754209593714e-06,
+      "loss": 0.4748,
+      "step": 19864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1379825986418326,
+      "learning_rate": 1.1599033522090145e-06,
+      "loss": 0.5487,
+      "step": 19865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.090742661072669,
+      "learning_rate": 1.159731374315885e-06,
+      "loss": 0.4864,
+      "step": 19866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.690536493918772,
+      "learning_rate": 1.1595594872835824e-06,
+      "loss": 0.3585,
+      "step": 19867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7882602705296013,
+      "learning_rate": 1.1593876911157068e-06,
+      "loss": 0.4528,
+      "step": 19868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6223470838819207,
+      "learning_rate": 1.159215985815854e-06,
+      "loss": 0.4333,
+      "step": 19869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6793397961192085,
+      "learning_rate": 1.1590443713876183e-06,
+      "loss": 0.418,
+      "step": 19870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8496244169429559,
+      "learning_rate": 1.1588728478345918e-06,
+      "loss": 0.491,
+      "step": 19871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6844906329079892,
+      "learning_rate": 1.158701415160367e-06,
+      "loss": 0.4339,
+      "step": 19872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9642316928871766,
+      "learning_rate": 1.1585300733685308e-06,
+      "loss": 0.5282,
+      "step": 19873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5719925704150508,
+      "learning_rate": 1.1583588224626707e-06,
+      "loss": 0.4029,
+      "step": 19874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8251903075314189,
+      "learning_rate": 1.1581876624463731e-06,
+      "loss": 0.5182,
+      "step": 19875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8197496566809457,
+      "learning_rate": 1.158016593323219e-06,
+      "loss": 0.4696,
+      "step": 19876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3982184814574705,
+      "learning_rate": 1.1578456150967906e-06,
+      "loss": 0.3887,
+      "step": 19877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7822070084243629,
+      "learning_rate": 1.157674727770668e-06,
+      "loss": 0.3739,
+      "step": 19878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8771691522871823,
+      "learning_rate": 1.1575039313484271e-06,
+      "loss": 0.3899,
+      "step": 19879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8061331002785384,
+      "learning_rate": 1.1573332258336444e-06,
+      "loss": 0.53,
+      "step": 19880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9941319536013036,
+      "learning_rate": 1.157162611229893e-06,
+      "loss": 0.4625,
+      "step": 19881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8237042075298286,
+      "learning_rate": 1.1569920875407457e-06,
+      "loss": 0.4263,
+      "step": 19882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.62346171317856,
+      "learning_rate": 1.1568216547697705e-06,
+      "loss": 0.3837,
+      "step": 19883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9816197819728492,
+      "learning_rate": 1.156651312920537e-06,
+      "loss": 0.6113,
+      "step": 19884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8914078579170026,
+      "learning_rate": 1.1564810619966108e-06,
+      "loss": 0.5478,
+      "step": 19885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5142944855418936,
+      "learning_rate": 1.156310902001555e-06,
+      "loss": 0.3707,
+      "step": 19886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8646616079612046,
+      "learning_rate": 1.1561408329389328e-06,
+      "loss": 0.5841,
+      "step": 19887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.801696635817066,
+      "learning_rate": 1.1559708548123041e-06,
+      "loss": 0.5945,
+      "step": 19888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7054902266124317,
+      "learning_rate": 1.1558009676252272e-06,
+      "loss": 0.4517,
+      "step": 19889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3578797578167718,
+      "learning_rate": 1.155631171381259e-06,
+      "loss": 0.4191,
+      "step": 19890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7103704515539009,
+      "learning_rate": 1.1554614660839536e-06,
+      "loss": 0.4189,
+      "step": 19891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6864819257788368,
+      "learning_rate": 1.1552918517368646e-06,
+      "loss": 0.4578,
+      "step": 19892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0416930169767427,
+      "learning_rate": 1.1551223283435412e-06,
+      "loss": 0.4807,
+      "step": 19893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7941152604945132,
+      "learning_rate": 1.1549528959075334e-06,
+      "loss": 0.4316,
+      "step": 19894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6909355982494638,
+      "learning_rate": 1.1547835544323883e-06,
+      "loss": 0.5526,
+      "step": 19895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8822680906763158,
+      "learning_rate": 1.1546143039216504e-06,
+      "loss": 0.6215,
+      "step": 19896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7007010241268659,
+      "learning_rate": 1.1544451443788632e-06,
+      "loss": 0.5905,
+      "step": 19897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.580889663572526,
+      "learning_rate": 1.1542760758075678e-06,
+      "loss": 0.3486,
+      "step": 19898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.136767093860497,
+      "learning_rate": 1.1541070982113032e-06,
+      "loss": 0.6074,
+      "step": 19899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.640896402681907,
+      "learning_rate": 1.153938211593607e-06,
+      "loss": 0.3056,
+      "step": 19900
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.48362571001052856,
+      "eval_runtime": 67.9127,
+      "eval_samples_per_second": 50.212,
+      "eval_steps_per_second": 1.576,
+      "step": 19900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.066285283489806,
+      "learning_rate": 1.1537694159580156e-06,
+      "loss": 0.518,
+      "step": 19901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8279359908778101,
+      "learning_rate": 1.1536007113080612e-06,
+      "loss": 0.44,
+      "step": 19902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7903955279314923,
+      "learning_rate": 1.1534320976472764e-06,
+      "loss": 0.5111,
+      "step": 19903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.610214014293825,
+      "learning_rate": 1.1532635749791907e-06,
+      "loss": 0.5084,
+      "step": 19904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8592449453901163,
+      "learning_rate": 1.1530951433073329e-06,
+      "loss": 0.4321,
+      "step": 19905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5889398911466504,
+      "learning_rate": 1.152926802635228e-06,
+      "loss": 0.5063,
+      "step": 19906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.226299763079207,
+      "learning_rate": 1.1527585529664002e-06,
+      "loss": 0.5053,
+      "step": 19907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4473511880304062,
+      "learning_rate": 1.1525903943043723e-06,
+      "loss": 0.47,
+      "step": 19908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9517335777263058,
+      "learning_rate": 1.1524223266526642e-06,
+      "loss": 0.5371,
+      "step": 19909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7227567204607088,
+      "learning_rate": 1.1522543500147936e-06,
+      "loss": 0.5141,
+      "step": 19910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7761431094312625,
+      "learning_rate": 1.1520864643942786e-06,
+      "loss": 0.4369,
+      "step": 19911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5252452923958832,
+      "learning_rate": 1.1519186697946323e-06,
+      "loss": 0.4145,
+      "step": 19912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3372770829991847,
+      "learning_rate": 1.1517509662193681e-06,
+      "loss": 0.5214,
+      "step": 19913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8335751095416406,
+      "learning_rate": 1.1515833536719972e-06,
+      "loss": 0.4024,
+      "step": 19914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8371727249181344,
+      "learning_rate": 1.151415832156027e-06,
+      "loss": 0.4695,
+      "step": 19915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9714364236974413,
+      "learning_rate": 1.151248401674965e-06,
+      "loss": 0.4783,
+      "step": 19916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5030356950001653,
+      "learning_rate": 1.1510810622323178e-06,
+      "loss": 0.3353,
+      "step": 19917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9080789124616455,
+      "learning_rate": 1.1509138138315865e-06,
+      "loss": 0.5181,
+      "step": 19918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.246210327655902,
+      "learning_rate": 1.1507466564762736e-06,
+      "loss": 0.6196,
+      "step": 19919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4979145945804755,
+      "learning_rate": 1.150579590169878e-06,
+      "loss": 0.4147,
+      "step": 19920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7078445399736486,
+      "learning_rate": 1.1504126149158972e-06,
+      "loss": 0.4401,
+      "step": 19921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6187352173121046,
+      "learning_rate": 1.1502457307178264e-06,
+      "loss": 0.5087,
+      "step": 19922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.780483861354325,
+      "learning_rate": 1.1500789375791604e-06,
+      "loss": 0.5139,
+      "step": 19923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1341231970279306,
+      "learning_rate": 1.149912235503389e-06,
+      "loss": 0.4871,
+      "step": 19924
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0452347082669657,
+      "learning_rate": 1.1497456244940028e-06,
+      "loss": 0.5485,
+      "step": 19925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8143218704350634,
+      "learning_rate": 1.1495791045544906e-06,
+      "loss": 0.4685,
+      "step": 19926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5992583323352756,
+      "learning_rate": 1.1494126756883376e-06,
+      "loss": 0.4383,
+      "step": 19927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7344067411888546,
+      "learning_rate": 1.1492463378990279e-06,
+      "loss": 0.4527,
+      "step": 19928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7478016696897127,
+      "learning_rate": 1.1490800911900437e-06,
+      "loss": 0.5056,
+      "step": 19929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.091370424604415,
+      "learning_rate": 1.1489139355648655e-06,
+      "loss": 0.2685,
+      "step": 19930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8555197761923483,
+      "learning_rate": 1.1487478710269717e-06,
+      "loss": 0.6049,
+      "step": 19931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4267069379744846,
+      "learning_rate": 1.148581897579838e-06,
+      "loss": 0.4038,
+      "step": 19932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5832529343048942,
+      "learning_rate": 1.1484160152269403e-06,
+      "loss": 0.3381,
+      "step": 19933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.873270391601112,
+      "learning_rate": 1.1482502239717493e-06,
+      "loss": 0.5013,
+      "step": 19934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3870887010858324,
+      "learning_rate": 1.1480845238177374e-06,
+      "loss": 0.3563,
+      "step": 19935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.518011118633213,
+      "learning_rate": 1.1479189147683733e-06,
+      "loss": 0.4621,
+      "step": 19936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5881802306550636,
+      "learning_rate": 1.1477533968271234e-06,
+      "loss": 0.4193,
+      "step": 19937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9087407833775731,
+      "learning_rate": 1.1475879699974528e-06,
+      "loss": 0.4427,
+      "step": 19938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5701528371566231,
+      "learning_rate": 1.147422634282824e-06,
+      "loss": 0.4692,
+      "step": 19939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5195448621634073,
+      "learning_rate": 1.1472573896867001e-06,
+      "loss": 0.3272,
+      "step": 19940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8165521303853036,
+      "learning_rate": 1.1470922362125381e-06,
+      "loss": 0.6239,
+      "step": 19941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0584516822888412,
+      "learning_rate": 1.146927173863797e-06,
+      "loss": 0.4898,
+      "step": 19942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8373548575512568,
+      "learning_rate": 1.1467622026439317e-06,
+      "loss": 0.5731,
+      "step": 19943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5525685434661314,
+      "learning_rate": 1.1465973225563954e-06,
+      "loss": 0.3243,
+      "step": 19944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6271039568333796,
+      "learning_rate": 1.1464325336046404e-06,
+      "loss": 0.4342,
+      "step": 19945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7090846825929706,
+      "learning_rate": 1.1462678357921167e-06,
+      "loss": 0.5536,
+      "step": 19946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.622376349463277,
+      "learning_rate": 1.146103229122271e-06,
+      "loss": 0.3944,
+      "step": 19947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5117639233505658,
+      "learning_rate": 1.1459387135985502e-06,
+      "loss": 0.3738,
+      "step": 19948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7529217228978893,
+      "learning_rate": 1.145774289224398e-06,
+      "loss": 0.3689,
+      "step": 19949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.931790013087106,
+      "learning_rate": 1.1456099560032566e-06,
+      "loss": 0.5507,
+      "step": 19950
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.48329904675483704,
+      "eval_runtime": 67.8703,
+      "eval_samples_per_second": 50.243,
+      "eval_steps_per_second": 1.577,
+      "step": 19950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7568668569443133,
+      "learning_rate": 1.1454457139385654e-06,
+      "loss": 0.4837,
+      "step": 19951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.11127882396358,
+      "learning_rate": 1.145281563033764e-06,
+      "loss": 0.5082,
+      "step": 19952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.498314707923552,
+      "learning_rate": 1.145117503292289e-06,
+      "loss": 0.3724,
+      "step": 19953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2931821469848275,
+      "learning_rate": 1.1449535347175731e-06,
+      "loss": 0.3163,
+      "step": 19954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3343852489872825,
+      "learning_rate": 1.1447896573130501e-06,
+      "loss": 0.591,
+      "step": 19955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6169776164673835,
+      "learning_rate": 1.1446258710821513e-06,
+      "loss": 0.4118,
+      "step": 19956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6192082489283006,
+      "learning_rate": 1.1444621760283038e-06,
+      "loss": 0.4693,
+      "step": 19957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7149010022869093,
+      "learning_rate": 1.1442985721549356e-06,
+      "loss": 0.4781,
+      "step": 19958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7374436088384753,
+      "learning_rate": 1.1441350594654717e-06,
+      "loss": 0.5203,
+      "step": 19959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6685853251714389,
+      "learning_rate": 1.1439716379633345e-06,
+      "loss": 0.4433,
+      "step": 19960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5820459652304049,
+      "learning_rate": 1.1438083076519453e-06,
+      "loss": 0.4322,
+      "step": 19961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7464793052752594,
+      "learning_rate": 1.1436450685347238e-06,
+      "loss": 0.6301,
+      "step": 19962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5307311174045095,
+      "learning_rate": 1.1434819206150863e-06,
+      "loss": 0.4028,
+      "step": 19963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8923106758272106,
+      "learning_rate": 1.1433188638964491e-06,
+      "loss": 0.3896,
+      "step": 19964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7796455653915175,
+      "learning_rate": 1.1431558983822258e-06,
+      "loss": 0.4779,
+      "step": 19965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4510308028848822,
+      "learning_rate": 1.1429930240758275e-06,
+      "loss": 0.3467,
+      "step": 19966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6116064597885422,
+      "learning_rate": 1.1428302409806635e-06,
+      "loss": 0.5234,
+      "step": 19967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9429905255453697,
+      "learning_rate": 1.142667549100143e-06,
+      "loss": 0.5992,
+      "step": 19968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9081852806083641,
+      "learning_rate": 1.14250494843767e-06,
+      "loss": 0.4123,
+      "step": 19969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5187410002383253,
+      "learning_rate": 1.14234243899665e-06,
+      "loss": 0.3826,
+      "step": 19970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1761672964154473,
+      "learning_rate": 1.1421800207804837e-06,
+      "loss": 0.5644,
+      "step": 19971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9439306025164746,
+      "learning_rate": 1.142017693792573e-06,
+      "loss": 0.5719,
+      "step": 19972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6659160496967118,
+      "learning_rate": 1.141855458036314e-06,
+      "loss": 0.4231,
+      "step": 19973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1636372731986295,
+      "learning_rate": 1.1416933135151045e-06,
+      "loss": 0.5307,
+      "step": 19974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9119995941996049,
+      "learning_rate": 1.1415312602323391e-06,
+      "loss": 0.5266,
+      "step": 19975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7058359289795058,
+      "learning_rate": 1.1413692981914087e-06,
+      "loss": 0.6383,
+      "step": 19976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7373205153661968,
+      "learning_rate": 1.141207427395705e-06,
+      "loss": 0.5664,
+      "step": 19977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5205718082923885,
+      "learning_rate": 1.1410456478486165e-06,
+      "loss": 0.3147,
+      "step": 19978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.597285600072023,
+      "learning_rate": 1.1408839595535298e-06,
+      "loss": 0.4106,
+      "step": 19979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6350071241916777,
+      "learning_rate": 1.1407223625138303e-06,
+      "loss": 0.5021,
+      "step": 19980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8995540082099944,
+      "learning_rate": 1.1405608567329004e-06,
+      "loss": 0.6005,
+      "step": 19981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.852807454990353,
+      "learning_rate": 1.1403994422141206e-06,
+      "loss": 0.5276,
+      "step": 19982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.71982836841995,
+      "learning_rate": 1.140238118960871e-06,
+      "loss": 0.4062,
+      "step": 19983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9507927321455054,
+      "learning_rate": 1.1400768869765285e-06,
+      "loss": 0.5081,
+      "step": 19984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5097887905626544,
+      "learning_rate": 1.1399157462644687e-06,
+      "loss": 0.4308,
+      "step": 19985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8106308915057228,
+      "learning_rate": 1.139754696828064e-06,
+      "loss": 0.52,
+      "step": 19986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9039370446858555,
+      "learning_rate": 1.1395937386706864e-06,
+      "loss": 0.6457,
+      "step": 19987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7009862555137776,
+      "learning_rate": 1.139432871795706e-06,
+      "loss": 0.5082,
+      "step": 19988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.658335787840145,
+      "learning_rate": 1.1392720962064902e-06,
+      "loss": 0.3964,
+      "step": 19989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7342362937723177,
+      "learning_rate": 1.1391114119064036e-06,
+      "loss": 0.4292,
+      "step": 19990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7889357544940636,
+      "learning_rate": 1.1389508188988118e-06,
+      "loss": 0.4998,
+      "step": 19991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.034511966127829,
+      "learning_rate": 1.1387903171870752e-06,
+      "loss": 0.4953,
+      "step": 19992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6350081867683683,
+      "learning_rate": 1.1386299067745547e-06,
+      "loss": 0.3677,
+      "step": 19993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6441397220812983,
+      "learning_rate": 1.1384695876646088e-06,
+      "loss": 0.5288,
+      "step": 19994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.30278250645473,
+      "learning_rate": 1.138309359860592e-06,
+      "loss": 0.482,
+      "step": 19995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6476029942853956,
+      "learning_rate": 1.1381492233658599e-06,
+      "loss": 0.4122,
+      "step": 19996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8219946852603812,
+      "learning_rate": 1.137989178183765e-06,
+      "loss": 0.5474,
+      "step": 19997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6661298028329037,
+      "learning_rate": 1.137829224317657e-06,
+      "loss": 0.4154,
+      "step": 19998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7302092578554251,
+      "learning_rate": 1.1376693617708847e-06,
+      "loss": 0.5195,
+      "step": 19999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0987956446001412,
+      "learning_rate": 1.1375095905467949e-06,
+      "loss": 0.5464,
+      "step": 20000
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.483070433139801,
+      "eval_runtime": 67.869,
+      "eval_samples_per_second": 50.244,
+      "eval_steps_per_second": 1.577,
+      "step": 20000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8225703321756628,
+      "learning_rate": 1.1373499106487323e-06,
+      "loss": 0.4655,
+      "step": 20001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8772911638343917,
+      "learning_rate": 1.1371903220800393e-06,
+      "loss": 0.5664,
+      "step": 20002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4945623277183107,
+      "learning_rate": 1.1370308248440572e-06,
+      "loss": 0.3506,
+      "step": 20003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0503031512067413,
+      "learning_rate": 1.136871418944125e-06,
+      "loss": 0.5059,
+      "step": 20004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5082464340133783,
+      "learning_rate": 1.1367121043835794e-06,
+      "loss": 0.4214,
+      "step": 20005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8638232659133462,
+      "learning_rate": 1.1365528811657554e-06,
+      "loss": 0.4171,
+      "step": 20006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.906832867798314,
+      "learning_rate": 1.1363937492939877e-06,
+      "loss": 0.5663,
+      "step": 20007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6121401283654926,
+      "learning_rate": 1.1362347087716054e-06,
+      "loss": 0.598,
+      "step": 20008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9638908917863762,
+      "learning_rate": 1.1360757596019392e-06,
+      "loss": 0.5013,
+      "step": 20009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8395429142323985,
+      "learning_rate": 1.1359169017883166e-06,
+      "loss": 0.3776,
+      "step": 20010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8761100979766252,
+      "learning_rate": 1.1357581353340635e-06,
+      "loss": 0.5152,
+      "step": 20011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.960942302961937,
+      "learning_rate": 1.1355994602425025e-06,
+      "loss": 0.4473,
+      "step": 20012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8479668867148586,
+      "learning_rate": 1.135440876516956e-06,
+      "loss": 0.4041,
+      "step": 20013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.449578982555831,
+      "learning_rate": 1.1352823841607444e-06,
+      "loss": 0.2883,
+      "step": 20014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.709766606759389,
+      "learning_rate": 1.1351239831771843e-06,
+      "loss": 0.4303,
+      "step": 20015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5786793817647566,
+      "learning_rate": 1.134965673569593e-06,
+      "loss": 0.6074,
+      "step": 20016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7434338881368192,
+      "learning_rate": 1.134807455341284e-06,
+      "loss": 0.4157,
+      "step": 20017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4178647998244582,
+      "learning_rate": 1.134649328495569e-06,
+      "loss": 0.4288,
+      "step": 20018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3996383751040953,
+      "learning_rate": 1.1344912930357594e-06,
+      "loss": 0.5439,
+      "step": 20019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7724283728465384,
+      "learning_rate": 1.1343333489651635e-06,
+      "loss": 0.454,
+      "step": 20020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.831499007269984,
+      "learning_rate": 1.134175496287087e-06,
+      "loss": 0.4614,
+      "step": 20021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8171119256604777,
+      "learning_rate": 1.1340177350048344e-06,
+      "loss": 0.4107,
+      "step": 20022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8643369854982874,
+      "learning_rate": 1.1338600651217095e-06,
+      "loss": 0.5802,
+      "step": 20023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.47897421094743,
+      "learning_rate": 1.1337024866410114e-06,
+      "loss": 0.4607,
+      "step": 20024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6088412410369197,
+      "learning_rate": 1.1335449995660398e-06,
+      "loss": 0.4712,
+      "step": 20025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7557127020556818,
+      "learning_rate": 1.1333876039000922e-06,
+      "loss": 0.3315,
+      "step": 20026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.027870719701707,
+      "learning_rate": 1.1332302996464622e-06,
+      "loss": 0.6847,
+      "step": 20027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6499628470776473,
+      "learning_rate": 1.133073086808444e-06,
+      "loss": 0.4565,
+      "step": 20028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0799521203940214,
+      "learning_rate": 1.1329159653893285e-06,
+      "loss": 0.7122,
+      "step": 20029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9909661914280974,
+      "learning_rate": 1.132758935392404e-06,
+      "loss": 0.5601,
+      "step": 20030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9673386344065147,
+      "learning_rate": 1.1326019968209592e-06,
+      "loss": 0.4887,
+      "step": 20031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9628282552635303,
+      "learning_rate": 1.1324451496782783e-06,
+      "loss": 0.6088,
+      "step": 20032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9210255584526246,
+      "learning_rate": 1.1322883939676465e-06,
+      "loss": 0.4853,
+      "step": 20033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.118456306908489,
+      "learning_rate": 1.1321317296923433e-06,
+      "loss": 0.5895,
+      "step": 20034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7411532963764893,
+      "learning_rate": 1.1319751568556498e-06,
+      "loss": 0.5585,
+      "step": 20035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7046278105285784,
+      "learning_rate": 1.1318186754608435e-06,
+      "loss": 0.5296,
+      "step": 20036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.484695156575071,
+      "learning_rate": 1.1316622855111996e-06,
+      "loss": 0.3425,
+      "step": 20037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7805426828900808,
+      "learning_rate": 1.1315059870099926e-06,
+      "loss": 0.3981,
+      "step": 20038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8725335050580223,
+      "learning_rate": 1.1313497799604946e-06,
+      "loss": 0.4239,
+      "step": 20039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.827114429908823,
+      "learning_rate": 1.1311936643659754e-06,
+      "loss": 0.4174,
+      "step": 20040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6108779138548446,
+      "learning_rate": 1.131037640229703e-06,
+      "loss": 0.3797,
+      "step": 20041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9436580365650669,
+      "learning_rate": 1.130881707554944e-06,
+      "loss": 0.4512,
+      "step": 20042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9189319580102353,
+      "learning_rate": 1.130725866344963e-06,
+      "loss": 0.5024,
+      "step": 20043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.982343394204548,
+      "learning_rate": 1.1305701166030219e-06,
+      "loss": 0.4656,
+      "step": 20044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4287368069990933,
+      "learning_rate": 1.1304144583323814e-06,
+      "loss": 0.4002,
+      "step": 20045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9444264156137023,
+      "learning_rate": 1.1302588915363009e-06,
+      "loss": 0.4059,
+      "step": 20046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7462756370800916,
+      "learning_rate": 1.130103416218035e-06,
+      "loss": 0.4908,
+      "step": 20047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4599091850962278,
+      "learning_rate": 1.1299480323808406e-06,
+      "loss": 0.4254,
+      "step": 20048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9087941762896823,
+      "learning_rate": 1.1297927400279702e-06,
+      "loss": 0.5511,
+      "step": 20049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.023571032717716,
+      "learning_rate": 1.1296375391626738e-06,
+      "loss": 0.3939,
+      "step": 20050
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.48261499404907227,
+      "eval_runtime": 67.9214,
+      "eval_samples_per_second": 50.205,
+      "eval_steps_per_second": 1.575,
+      "step": 20050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6575708825296278,
+      "learning_rate": 1.1294824297882013e-06,
+      "loss": 0.3547,
+      "step": 20051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4710736158616895,
+      "learning_rate": 1.1293274119077997e-06,
+      "loss": 0.4053,
+      "step": 20052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1969374675406974,
+      "learning_rate": 1.1291724855247131e-06,
+      "loss": 0.6686,
+      "step": 20053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.115867302494562,
+      "learning_rate": 1.1290176506421863e-06,
+      "loss": 0.4541,
+      "step": 20054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6911219335679715,
+      "learning_rate": 1.12886290726346e-06,
+      "loss": 0.4268,
+      "step": 20055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9459830580121944,
+      "learning_rate": 1.1287082553917736e-06,
+      "loss": 0.4801,
+      "step": 20056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.952027178050598,
+      "learning_rate": 1.1285536950303646e-06,
+      "loss": 0.434,
+      "step": 20057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6006990792479374,
+      "learning_rate": 1.1283992261824693e-06,
+      "loss": 0.4134,
+      "step": 20058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8996983131033767,
+      "learning_rate": 1.128244848851321e-06,
+      "loss": 0.4351,
+      "step": 20059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6864732019784958,
+      "learning_rate": 1.1280905630401509e-06,
+      "loss": 0.3639,
+      "step": 20060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1368029196538427,
+      "learning_rate": 1.1279363687521894e-06,
+      "loss": 0.569,
+      "step": 20061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3123513047529793,
+      "learning_rate": 1.1277822659906651e-06,
+      "loss": 0.5368,
+      "step": 20062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.87988746643565,
+      "learning_rate": 1.1276282547588024e-06,
+      "loss": 0.4866,
+      "step": 20063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9926733134072254,
+      "learning_rate": 1.1274743350598261e-06,
+      "loss": 0.3964,
+      "step": 20064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0189802034177133,
+      "learning_rate": 1.1273205068969599e-06,
+      "loss": 0.4784,
+      "step": 20065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6655632358952122,
+      "learning_rate": 1.1271667702734222e-06,
+      "loss": 0.4341,
+      "step": 20066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7244186593611264,
+      "learning_rate": 1.1270131251924322e-06,
+      "loss": 0.4207,
+      "step": 20067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3344276923672023,
+      "learning_rate": 1.126859571657207e-06,
+      "loss": 0.286,
+      "step": 20068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7539355775877872,
+      "learning_rate": 1.1267061096709595e-06,
+      "loss": 0.4127,
+      "step": 20069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5967802666113649,
+      "learning_rate": 1.1265527392369036e-06,
+      "loss": 0.3511,
+      "step": 20070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.040314457396979,
+      "learning_rate": 1.1263994603582501e-06,
+      "loss": 0.5244,
+      "step": 20071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.733930144720349,
+      "learning_rate": 1.1262462730382068e-06,
+      "loss": 0.4621,
+      "step": 20072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0488746615207702,
+      "learning_rate": 1.126093177279981e-06,
+      "loss": 0.5902,
+      "step": 20073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0742888759829543,
+      "learning_rate": 1.1259401730867783e-06,
+      "loss": 0.4908,
+      "step": 20074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6887567042132616,
+      "learning_rate": 1.1257872604618005e-06,
+      "loss": 0.3884,
+      "step": 20075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.234707409048664,
+      "learning_rate": 1.1256344394082504e-06,
+      "loss": 0.5283,
+      "step": 20076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.573329365304706,
+      "learning_rate": 1.1254817099293255e-06,
+      "loss": 0.3632,
+      "step": 20077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8511082107654175,
+      "learning_rate": 1.1253290720282245e-06,
+      "loss": 0.518,
+      "step": 20078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.565551030333989,
+      "learning_rate": 1.1251765257081417e-06,
+      "loss": 0.4865,
+      "step": 20079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.738852498380287,
+      "learning_rate": 1.125024070972271e-06,
+      "loss": 0.4511,
+      "step": 20080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6417876386671864,
+      "learning_rate": 1.1248717078238047e-06,
+      "loss": 0.4257,
+      "step": 20081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8753413675462718,
+      "learning_rate": 1.124719436265931e-06,
+      "loss": 0.4353,
+      "step": 20082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.894815938214157,
+      "learning_rate": 1.1245672563018383e-06,
+      "loss": 0.4222,
+      "step": 20083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.764848730723268,
+      "learning_rate": 1.124415167934713e-06,
+      "loss": 0.4357,
+      "step": 20084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.55796697917701,
+      "learning_rate": 1.1242631711677378e-06,
+      "loss": 0.3559,
+      "step": 20085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7284648652237589,
+      "learning_rate": 1.1241112660040954e-06,
+      "loss": 0.4478,
+      "step": 20086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6524806435367414,
+      "learning_rate": 1.123959452446966e-06,
+      "loss": 0.4686,
+      "step": 20087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.344304567487665,
+      "learning_rate": 1.1238077304995273e-06,
+      "loss": 0.7073,
+      "step": 20088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.116729676293656,
+      "learning_rate": 1.123656100164955e-06,
+      "loss": 0.3616,
+      "step": 20089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.079401388527462,
+      "learning_rate": 1.1235045614464245e-06,
+      "loss": 0.473,
+      "step": 20090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7938041119424024,
+      "learning_rate": 1.123353114347108e-06,
+      "loss": 0.4262,
+      "step": 20091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.626732220996223,
+      "learning_rate": 1.1232017588701749e-06,
+      "loss": 0.5132,
+      "step": 20092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9880902655544843,
+      "learning_rate": 1.1230504950187948e-06,
+      "loss": 0.4102,
+      "step": 20093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7671526033168516,
+      "learning_rate": 1.1228993227961338e-06,
+      "loss": 0.4032,
+      "step": 20094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6966091340412466,
+      "learning_rate": 1.1227482422053571e-06,
+      "loss": 0.5749,
+      "step": 20095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6939617086724919,
+      "learning_rate": 1.122597253249627e-06,
+      "loss": 0.5668,
+      "step": 20096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.021507582028939,
+      "learning_rate": 1.1224463559321047e-06,
+      "loss": 0.5239,
+      "step": 20097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.956193777655933,
+      "learning_rate": 1.1222955502559481e-06,
+      "loss": 0.5215,
+      "step": 20098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7493330115908248,
+      "learning_rate": 1.122144836224316e-06,
+      "loss": 0.4463,
+      "step": 20099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8363517501430138,
+      "learning_rate": 1.121994213840362e-06,
+      "loss": 0.491,
+      "step": 20100
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.4826020300388336,
+      "eval_runtime": 67.8964,
+      "eval_samples_per_second": 50.224,
+      "eval_steps_per_second": 1.576,
+      "step": 20100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6867866483440164,
+      "learning_rate": 1.1218436831072403e-06,
+      "loss": 0.4125,
+      "step": 20101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9348791762340398,
+      "learning_rate": 1.1216932440281012e-06,
+      "loss": 0.5335,
+      "step": 20102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5923628388900828,
+      "learning_rate": 1.121542896606095e-06,
+      "loss": 0.3827,
+      "step": 20103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9575562763495946,
+      "learning_rate": 1.1213926408443688e-06,
+      "loss": 0.4311,
+      "step": 20104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7264921014343755,
+      "learning_rate": 1.1212424767460671e-06,
+      "loss": 0.4297,
+      "step": 20105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6824330311009672,
+      "learning_rate": 1.1210924043143351e-06,
+      "loss": 0.3958,
+      "step": 20106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5141934472279686,
+      "learning_rate": 1.1209424235523144e-06,
+      "loss": 0.3181,
+      "step": 20107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.640372789017428,
+      "learning_rate": 1.120792534463143e-06,
+      "loss": 0.4662,
+      "step": 20108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.701625073929639,
+      "learning_rate": 1.1206427370499604e-06,
+      "loss": 0.3153,
+      "step": 20109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.405905261354233,
+      "learning_rate": 1.1204930313159023e-06,
+      "loss": 0.5624,
+      "step": 20110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0896321584852777,
+      "learning_rate": 1.1203434172641018e-06,
+      "loss": 0.5339,
+      "step": 20111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6331606483481218,
+      "learning_rate": 1.1201938948976914e-06,
+      "loss": 0.3736,
+      "step": 20112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7190036423230188,
+      "learning_rate": 1.120044464219802e-06,
+      "loss": 0.4238,
+      "step": 20113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6043229055241077,
+      "learning_rate": 1.1198951252335607e-06,
+      "loss": 0.4437,
+      "step": 20114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6520852296294828,
+      "learning_rate": 1.1197458779420943e-06,
+      "loss": 0.2977,
+      "step": 20115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5726084344010185,
+      "learning_rate": 1.119596722348528e-06,
+      "loss": 0.4054,
+      "step": 20116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5016514325141632,
+      "learning_rate": 1.1194476584559832e-06,
+      "loss": 0.3985,
+      "step": 20117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9256549284305766,
+      "learning_rate": 1.1192986862675804e-06,
+      "loss": 0.4263,
+      "step": 20118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4727078195642114,
+      "learning_rate": 1.1191498057864395e-06,
+      "loss": 0.3205,
+      "step": 20119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7736463805881708,
+      "learning_rate": 1.119001017015676e-06,
+      "loss": 0.4953,
+      "step": 20120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6320738383921418,
+      "learning_rate": 1.1188523199584048e-06,
+      "loss": 0.4547,
+      "step": 20121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6497402801331007,
+      "learning_rate": 1.1187037146177393e-06,
+      "loss": 0.3162,
+      "step": 20122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.967277463475165,
+      "learning_rate": 1.1185552009967902e-06,
+      "loss": 0.5397,
+      "step": 20123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.036884177750015,
+      "learning_rate": 1.1184067790986663e-06,
+      "loss": 0.5483,
+      "step": 20124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5802220338783024,
+      "learning_rate": 1.1182584489264749e-06,
+      "loss": 0.3431,
+      "step": 20125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7593651111823352,
+      "learning_rate": 1.1181102104833218e-06,
+      "loss": 0.4741,
+      "step": 20126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5243055316309333,
+      "learning_rate": 1.1179620637723097e-06,
+      "loss": 0.3421,
+      "step": 20127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6728476662209792,
+      "learning_rate": 1.1178140087965397e-06,
+      "loss": 0.3827,
+      "step": 20128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1144829500636075,
+      "learning_rate": 1.117666045559112e-06,
+      "loss": 0.6672,
+      "step": 20129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.029670382887337,
+      "learning_rate": 1.117518174063123e-06,
+      "loss": 0.4722,
+      "step": 20130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.211604846423541,
+      "learning_rate": 1.117370394311669e-06,
+      "loss": 0.4436,
+      "step": 20131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6431300225906191,
+      "learning_rate": 1.1172227063078445e-06,
+      "loss": 0.4216,
+      "step": 20132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9710576148077097,
+      "learning_rate": 1.1170751100547395e-06,
+      "loss": 0.4286,
+      "step": 20133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2453637074332864,
+      "learning_rate": 1.1169276055554452e-06,
+      "loss": 0.594,
+      "step": 20134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8290098562084913,
+      "learning_rate": 1.116780192813049e-06,
+      "loss": 0.4331,
+      "step": 20135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0925497690053,
+      "learning_rate": 1.1166328718306367e-06,
+      "loss": 0.4817,
+      "step": 20136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.834649642384891,
+      "learning_rate": 1.1164856426112925e-06,
+      "loss": 0.4823,
+      "step": 20137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7379752576302432,
+      "learning_rate": 1.1163385051580992e-06,
+      "loss": 0.4876,
+      "step": 20138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2400625350733203,
+      "learning_rate": 1.1161914594741365e-06,
+      "loss": 0.3979,
+      "step": 20139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8760218441853929,
+      "learning_rate": 1.116044505562482e-06,
+      "loss": 0.4716,
+      "step": 20140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3106784157104858,
+      "learning_rate": 1.1158976434262134e-06,
+      "loss": 0.3116,
+      "step": 20141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9758698649184336,
+      "learning_rate": 1.1157508730684046e-06,
+      "loss": 0.5026,
+      "step": 20142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0894978326237488,
+      "learning_rate": 1.1156041944921278e-06,
+      "loss": 0.7581,
+      "step": 20143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5652848663914598,
+      "learning_rate": 1.1154576077004541e-06,
+      "loss": 0.4327,
+      "step": 20144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7420257213816057,
+      "learning_rate": 1.115311112696452e-06,
+      "loss": 0.5908,
+      "step": 20145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.925382527054629,
+      "learning_rate": 1.1151647094831886e-06,
+      "loss": 0.6525,
+      "step": 20146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.684650726872088,
+      "learning_rate": 1.1150183980637282e-06,
+      "loss": 0.3443,
+      "step": 20147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4322858042904163,
+      "learning_rate": 1.1148721784411337e-06,
+      "loss": 0.3637,
+      "step": 20148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7585576307317354,
+      "learning_rate": 1.114726050618467e-06,
+      "loss": 0.4453,
+      "step": 20149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6429751382182134,
+      "learning_rate": 1.1145800145987867e-06,
+      "loss": 0.3226,
+      "step": 20150
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.4825827479362488,
+      "eval_runtime": 67.9304,
+      "eval_samples_per_second": 50.198,
+      "eval_steps_per_second": 1.575,
+      "step": 20150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9317300811369607,
+      "learning_rate": 1.1144340703851493e-06,
+      "loss": 0.4505,
+      "step": 20151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8024668772118908,
+      "learning_rate": 1.1142882179806113e-06,
+      "loss": 0.4567,
+      "step": 20152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7678311397105306,
+      "learning_rate": 1.1141424573882249e-06,
+      "loss": 0.4622,
+      "step": 20153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6258253822914845,
+      "learning_rate": 1.1139967886110418e-06,
+      "loss": 0.4243,
+      "step": 20154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6100267032846802,
+      "learning_rate": 1.1138512116521127e-06,
+      "loss": 0.3563,
+      "step": 20155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7177214888346723,
+      "learning_rate": 1.1137057265144832e-06,
+      "loss": 0.3529,
+      "step": 20156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.786669098602353,
+      "learning_rate": 1.1135603332012002e-06,
+      "loss": 0.4304,
+      "step": 20157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7585013888397139,
+      "learning_rate": 1.1134150317153077e-06,
+      "loss": 0.4558,
+      "step": 20158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7613552662098022,
+      "learning_rate": 1.1132698220598456e-06,
+      "loss": 0.4048,
+      "step": 20159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.89717397887399,
+      "learning_rate": 1.1131247042378562e-06,
+      "loss": 0.5083,
+      "step": 20160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8945761601399582,
+      "learning_rate": 1.1129796782523753e-06,
+      "loss": 0.6284,
+      "step": 20161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.130688062831196,
+      "learning_rate": 1.1128347441064411e-06,
+      "loss": 0.6136,
+      "step": 20162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.664944235804186,
+      "learning_rate": 1.1126899018030853e-06,
+      "loss": 0.564,
+      "step": 20163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9494424659961636,
+      "learning_rate": 1.1125451513453429e-06,
+      "loss": 0.5324,
+      "step": 20164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5425111866420336,
+      "learning_rate": 1.1124004927362415e-06,
+      "loss": 0.3539,
+      "step": 20165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0157764327751533,
+      "learning_rate": 1.1122559259788108e-06,
+      "loss": 0.4089,
+      "step": 20166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6112508754847772,
+      "learning_rate": 1.112111451076077e-06,
+      "loss": 0.3868,
+      "step": 20167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9009875433340755,
+      "learning_rate": 1.1119670680310646e-06,
+      "loss": 0.4812,
+      "step": 20168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.065424704863185,
+      "learning_rate": 1.111822776846796e-06,
+      "loss": 0.6021,
+      "step": 20169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.919284706609974,
+      "learning_rate": 1.1116785775262916e-06,
+      "loss": 0.4574,
+      "step": 20170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2854749211957155,
+      "learning_rate": 1.1115344700725714e-06,
+      "loss": 0.558,
+      "step": 20171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8150797885269454,
+      "learning_rate": 1.1113904544886507e-06,
+      "loss": 0.4881,
+      "step": 20172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6885925301353029,
+      "learning_rate": 1.1112465307775448e-06,
+      "loss": 0.4559,
+      "step": 20173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.641077277228619,
+      "learning_rate": 1.1111026989422677e-06,
+      "loss": 0.4504,
+      "step": 20174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.483190436852633,
+      "learning_rate": 1.1109589589858288e-06,
+      "loss": 0.4772,
+      "step": 20175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5227009234971256,
+      "learning_rate": 1.110815310911238e-06,
+      "loss": 0.3882,
+      "step": 20176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.701354974243467,
+      "learning_rate": 1.110671754721503e-06,
+      "loss": 0.4567,
+      "step": 20177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6702122308158918,
+      "learning_rate": 1.1105282904196275e-06,
+      "loss": 0.4161,
+      "step": 20178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8816912490885462,
+      "learning_rate": 1.1103849180086166e-06,
+      "loss": 0.4668,
+      "step": 20179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0071326851954328,
+      "learning_rate": 1.1102416374914709e-06,
+      "loss": 0.6195,
+      "step": 20180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4401602194058638,
+      "learning_rate": 1.1100984488711897e-06,
+      "loss": 0.378,
+      "step": 20181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9081813373221868,
+      "learning_rate": 1.109955352150771e-06,
+      "loss": 0.5869,
+      "step": 20182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.980619491989593,
+      "learning_rate": 1.10981234733321e-06,
+      "loss": 0.5689,
+      "step": 20183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9159758742682553,
+      "learning_rate": 1.1096694344215012e-06,
+      "loss": 0.4728,
+      "step": 20184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.522763912937661,
+      "learning_rate": 1.109526613418635e-06,
+      "loss": 0.2775,
+      "step": 20185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5205637529662466,
+      "learning_rate": 1.1093838843276028e-06,
+      "loss": 0.3476,
+      "step": 20186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.921317255893538,
+      "learning_rate": 1.109241247151392e-06,
+      "loss": 0.5817,
+      "step": 20187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6034833945432598,
+      "learning_rate": 1.109098701892988e-06,
+      "loss": 0.4333,
+      "step": 20188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.041660196637727,
+      "learning_rate": 1.1089562485553751e-06,
+      "loss": 0.7153,
+      "step": 20189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4739591101245766,
+      "learning_rate": 1.1088138871415368e-06,
+      "loss": 0.4025,
+      "step": 20190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3282179515846955,
+      "learning_rate": 1.1086716176544515e-06,
+      "loss": 0.3364,
+      "step": 20191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5247853463050316,
+      "learning_rate": 1.1085294400970983e-06,
+      "loss": 0.3909,
+      "step": 20192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5594069748475372,
+      "learning_rate": 1.1083873544724541e-06,
+      "loss": 0.3338,
+      "step": 20193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.931165414014119,
+      "learning_rate": 1.1082453607834924e-06,
+      "loss": 0.5195,
+      "step": 20194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5921158324086848,
+      "learning_rate": 1.108103459033187e-06,
+      "loss": 0.4128,
+      "step": 20195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1780072108475714,
+      "learning_rate": 1.107961649224507e-06,
+      "loss": 0.6959,
+      "step": 20196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.980895213825296,
+      "learning_rate": 1.107819931360422e-06,
+      "loss": 0.4923,
+      "step": 20197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.669246875835409,
+      "learning_rate": 1.1076783054438992e-06,
+      "loss": 0.3602,
+      "step": 20198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8188910875897792,
+      "learning_rate": 1.107536771477902e-06,
+      "loss": 0.4416,
+      "step": 20199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4928797506936147,
+      "learning_rate": 1.1073953294653953e-06,
+      "loss": 0.3811,
+      "step": 20200
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.48247817158699036,
+      "eval_runtime": 67.9071,
+      "eval_samples_per_second": 50.216,
+      "eval_steps_per_second": 1.576,
+      "step": 20200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.084952972146676,
+      "learning_rate": 1.1072539794093384e-06,
+      "loss": 0.5735,
+      "step": 20201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.979569907928159,
+      "learning_rate": 1.1071127213126911e-06,
+      "loss": 0.4938,
+      "step": 20202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8320155629012025,
+      "learning_rate": 1.1069715551784105e-06,
+      "loss": 0.4559,
+      "step": 20203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.785809341368306,
+      "learning_rate": 1.106830481009452e-06,
+      "loss": 0.3437,
+      "step": 20204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.819171786390754,
+      "learning_rate": 1.1066894988087682e-06,
+      "loss": 0.6552,
+      "step": 20205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5093967468369096,
+      "learning_rate": 1.106548608579311e-06,
+      "loss": 0.3827,
+      "step": 20206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5993203183056417,
+      "learning_rate": 1.1064078103240298e-06,
+      "loss": 0.4045,
+      "step": 20207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5666338847310144,
+      "learning_rate": 1.1062671040458723e-06,
+      "loss": 0.4647,
+      "step": 20208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7570777452508284,
+      "learning_rate": 1.1061264897477841e-06,
+      "loss": 0.4016,
+      "step": 20209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9997103387054183,
+      "learning_rate": 1.1059859674327086e-06,
+      "loss": 0.4415,
+      "step": 20210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8208911409207782,
+      "learning_rate": 1.1058455371035877e-06,
+      "loss": 0.5141,
+      "step": 20211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7935931248515082,
+      "learning_rate": 1.1057051987633606e-06,
+      "loss": 0.5133,
+      "step": 20212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3161826318859453,
+      "learning_rate": 1.1055649524149665e-06,
+      "loss": 0.3423,
+      "step": 20213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.580647701723379,
+      "learning_rate": 1.10542479806134e-06,
+      "loss": 0.4387,
+      "step": 20214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3747085589550243,
+      "learning_rate": 1.1052847357054163e-06,
+      "loss": 0.2858,
+      "step": 20215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.932709411417542,
+      "learning_rate": 1.1051447653501269e-06,
+      "loss": 0.471,
+      "step": 20216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.815822093787795,
+      "learning_rate": 1.1050048869984021e-06,
+      "loss": 0.4688,
+      "step": 20217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7288398906128042,
+      "learning_rate": 1.1048651006531696e-06,
+      "loss": 0.4296,
+      "step": 20218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9629588620096277,
+      "learning_rate": 1.104725406317357e-06,
+      "loss": 0.4641,
+      "step": 20219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6910699880649358,
+      "learning_rate": 1.1045858039938877e-06,
+      "loss": 0.4076,
+      "step": 20220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9246615286591993,
+      "learning_rate": 1.1044462936856845e-06,
+      "loss": 0.4068,
+      "step": 20221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6010346763926135,
+      "learning_rate": 1.1043068753956684e-06,
+      "loss": 0.4202,
+      "step": 20222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.646701235773373,
+      "learning_rate": 1.1041675491267573e-06,
+      "loss": 0.4469,
+      "step": 20223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6683120713730175,
+      "learning_rate": 1.1040283148818678e-06,
+      "loss": 0.388,
+      "step": 20224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7720609078714138,
+      "learning_rate": 1.1038891726639157e-06,
+      "loss": 0.4361,
+      "step": 20225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8684991606330066,
+      "learning_rate": 1.103750122475813e-06,
+      "loss": 0.582,
+      "step": 20226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8734655028517928,
+      "learning_rate": 1.103611164320471e-06,
+      "loss": 0.4461,
+      "step": 20227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8458191075844113,
+      "learning_rate": 1.1034722982007983e-06,
+      "loss": 0.4962,
+      "step": 20228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7052271420723022,
+      "learning_rate": 1.1033335241197026e-06,
+      "loss": 0.5289,
+      "step": 20229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6947778691128166,
+      "learning_rate": 1.103194842080088e-06,
+      "loss": 0.4229,
+      "step": 20230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.071183059180038,
+      "learning_rate": 1.1030562520848591e-06,
+      "loss": 0.4561,
+      "step": 20231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5772761518668785,
+      "learning_rate": 1.1029177541369164e-06,
+      "loss": 0.645,
+      "step": 20232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.872335741626557,
+      "learning_rate": 1.1027793482391592e-06,
+      "loss": 0.4639,
+      "step": 20233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7318488560699443,
+      "learning_rate": 1.102641034394485e-06,
+      "loss": 0.3766,
+      "step": 20234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7847859626409597,
+      "learning_rate": 1.1025028126057904e-06,
+      "loss": 0.4882,
+      "step": 20235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7309320306047673,
+      "learning_rate": 1.1023646828759672e-06,
+      "loss": 0.4873,
+      "step": 20236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.003190110987445,
+      "learning_rate": 1.1022266452079078e-06,
+      "loss": 0.5676,
+      "step": 20237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4743269995574433,
+      "learning_rate": 1.1020886996045026e-06,
+      "loss": 0.3398,
+      "step": 20238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5434850297294687,
+      "learning_rate": 1.1019508460686382e-06,
+      "loss": 0.4171,
+      "step": 20239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8512928181777724,
+      "learning_rate": 1.1018130846032016e-06,
+      "loss": 0.4144,
+      "step": 20240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9368875115323858,
+      "learning_rate": 1.1016754152110758e-06,
+      "loss": 0.4254,
+      "step": 20241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.139782295690954,
+      "learning_rate": 1.1015378378951437e-06,
+      "loss": 0.3875,
+      "step": 20242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7207905928034808,
+      "learning_rate": 1.101400352658285e-06,
+      "loss": 0.3098,
+      "step": 20243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.741670126991393,
+      "learning_rate": 1.1012629595033774e-06,
+      "loss": 0.4531,
+      "step": 20244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9968694758094467,
+      "learning_rate": 1.101125658433298e-06,
+      "loss": 0.5894,
+      "step": 20245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3575815801026547,
+      "learning_rate": 1.1009884494509208e-06,
+      "loss": 0.3917,
+      "step": 20246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7877872752957664,
+      "learning_rate": 1.1008513325591175e-06,
+      "loss": 0.4209,
+      "step": 20247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9229124444153185,
+      "learning_rate": 1.1007143077607597e-06,
+      "loss": 0.4428,
+      "step": 20248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.114895387745273,
+      "learning_rate": 1.1005773750587151e-06,
+      "loss": 0.3994,
+      "step": 20249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.793548298950796,
+      "learning_rate": 1.1004405344558502e-06,
+      "loss": 0.432,
+      "step": 20250
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.4823748469352722,
+      "eval_runtime": 67.914,
+      "eval_samples_per_second": 50.211,
+      "eval_steps_per_second": 1.576,
+      "step": 20250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5608881790953983,
+      "learning_rate": 1.1003037859550306e-06,
+      "loss": 0.4352,
+      "step": 20251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8806949682250635,
+      "learning_rate": 1.1001671295591186e-06,
+      "loss": 0.4816,
+      "step": 20252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6928679397666455,
+      "learning_rate": 1.1000305652709744e-06,
+      "loss": 0.354,
+      "step": 20253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.544407802218005,
+      "learning_rate": 1.0998940930934577e-06,
+      "loss": 0.3822,
+      "step": 20254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.927279645539715,
+      "learning_rate": 1.0997577130294254e-06,
+      "loss": 0.5637,
+      "step": 20255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0134994011207694,
+      "learning_rate": 1.0996214250817318e-06,
+      "loss": 0.6405,
+      "step": 20256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.806554759136786,
+      "learning_rate": 1.0994852292532304e-06,
+      "loss": 0.4949,
+      "step": 20257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.900477537828466,
+      "learning_rate": 1.0993491255467733e-06,
+      "loss": 0.5362,
+      "step": 20258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6259747440549777,
+      "learning_rate": 1.0992131139652083e-06,
+      "loss": 0.4127,
+      "step": 20259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4997683789847305,
+      "learning_rate": 1.0990771945113829e-06,
+      "loss": 0.3912,
+      "step": 20260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.792620252051775,
+      "learning_rate": 1.0989413671881436e-06,
+      "loss": 0.5199,
+      "step": 20261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7295639156535736,
+      "learning_rate": 1.0988056319983325e-06,
+      "loss": 0.4477,
+      "step": 20262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7083205798985102,
+      "learning_rate": 1.0986699889447923e-06,
+      "loss": 0.4863,
+      "step": 20263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.376082257700868,
+      "learning_rate": 1.0985344380303625e-06,
+      "loss": 0.3382,
+      "step": 20264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6230242968742323,
+      "learning_rate": 1.09839897925788e-06,
+      "loss": 0.5148,
+      "step": 20265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.046475476193073,
+      "learning_rate": 1.0982636126301805e-06,
+      "loss": 0.5544,
+      "step": 20266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5686566176164418,
+      "learning_rate": 1.098128338150099e-06,
+      "loss": 0.4303,
+      "step": 20267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7987081479500295,
+      "learning_rate": 1.0979931558204661e-06,
+      "loss": 0.4561,
+      "step": 20268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7661830147774025,
+      "learning_rate": 1.0978580656441124e-06,
+      "loss": 0.4718,
+      "step": 20269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7489896629998691,
+      "learning_rate": 1.0977230676238663e-06,
+      "loss": 0.4846,
+      "step": 20270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.571714721885249,
+      "learning_rate": 1.0975881617625529e-06,
+      "loss": 0.3907,
+      "step": 20271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0095674716457608,
+      "learning_rate": 1.0974533480629968e-06,
+      "loss": 0.5665,
+      "step": 20272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5736185533059892,
+      "learning_rate": 1.0973186265280211e-06,
+      "loss": 0.3671,
+      "step": 20273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6144178665380835,
+      "learning_rate": 1.0971839971604445e-06,
+      "loss": 0.4619,
+      "step": 20274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8061684955479422,
+      "learning_rate": 1.0970494599630868e-06,
+      "loss": 0.446,
+      "step": 20275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.096365338303489,
+      "learning_rate": 1.0969150149387633e-06,
+      "loss": 0.6541,
+      "step": 20276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8980494194442707,
+      "learning_rate": 1.0967806620902897e-06,
+      "loss": 0.4594,
+      "step": 20277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6268884141820101,
+      "learning_rate": 1.0966464014204774e-06,
+      "loss": 0.467,
+      "step": 20278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8078824205831796,
+      "learning_rate": 1.096512232932138e-06,
+      "loss": 0.4144,
+      "step": 20279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7198001455952534,
+      "learning_rate": 1.0963781566280804e-06,
+      "loss": 0.4152,
+      "step": 20280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1014510445401484,
+      "learning_rate": 1.0962441725111104e-06,
+      "loss": 0.5177,
+      "step": 20281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7276672267255275,
+      "learning_rate": 1.0961102805840332e-06,
+      "loss": 0.5933,
+      "step": 20282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4631778145413021,
+      "learning_rate": 1.0959764808496524e-06,
+      "loss": 0.3417,
+      "step": 20283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.080130496549302,
+      "learning_rate": 1.0958427733107681e-06,
+      "loss": 0.4956,
+      "step": 20284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0638426300414845,
+      "learning_rate": 1.0957091579701797e-06,
+      "loss": 0.5902,
+      "step": 20285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2388015939776884,
+      "learning_rate": 1.0955756348306849e-06,
+      "loss": 0.7562,
+      "step": 20286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.707407084832089,
+      "learning_rate": 1.0954422038950784e-06,
+      "loss": 0.4353,
+      "step": 20287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.147754637000685,
+      "learning_rate": 1.0953088651661531e-06,
+      "loss": 0.6578,
+      "step": 20288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.265402818049671,
+      "learning_rate": 1.0951756186467016e-06,
+      "loss": 0.6301,
+      "step": 20289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8945684467220931,
+      "learning_rate": 1.0950424643395122e-06,
+      "loss": 0.4138,
+      "step": 20290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8682729733314434,
+      "learning_rate": 1.0949094022473731e-06,
+      "loss": 0.5529,
+      "step": 20291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.814278660477831,
+      "learning_rate": 1.0947764323730686e-06,
+      "loss": 0.4762,
+      "step": 20292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9557903086125112,
+      "learning_rate": 1.0946435547193845e-06,
+      "loss": 0.5009,
+      "step": 20293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5928520140701488,
+      "learning_rate": 1.0945107692891009e-06,
+      "loss": 0.3975,
+      "step": 20294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.850221733857497,
+      "learning_rate": 1.0943780760849976e-06,
+      "loss": 0.4267,
+      "step": 20295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7311181796129558,
+      "learning_rate": 1.0942454751098535e-06,
+      "loss": 0.4026,
+      "step": 20296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9960964950790725,
+      "learning_rate": 1.0941129663664432e-06,
+      "loss": 0.5316,
+      "step": 20297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.739564032809371,
+      "learning_rate": 1.0939805498575418e-06,
+      "loss": 0.3542,
+      "step": 20298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8717982980589225,
+      "learning_rate": 1.0938482255859208e-06,
+      "loss": 0.3253,
+      "step": 20299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4040250751769507,
+      "learning_rate": 1.0937159935543503e-06,
+      "loss": 0.3644,
+      "step": 20300
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.4820892810821533,
+      "eval_runtime": 67.8987,
+      "eval_samples_per_second": 50.222,
+      "eval_steps_per_second": 1.576,
+      "step": 20300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2179917433018517,
+      "learning_rate": 1.0935838537655988e-06,
+      "loss": 0.5586,
+      "step": 20301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8162793616699173,
+      "learning_rate": 1.093451806222432e-06,
+      "loss": 0.4443,
+      "step": 20302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9577586662204511,
+      "learning_rate": 1.0933198509276154e-06,
+      "loss": 0.7362,
+      "step": 20303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.023099012449903,
+      "learning_rate": 1.09318798788391e-06,
+      "loss": 0.5121,
+      "step": 20304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8000245389631886,
+      "learning_rate": 1.0930562170940766e-06,
+      "loss": 0.4577,
+      "step": 20305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5078005451278327,
+      "learning_rate": 1.092924538560875e-06,
+      "loss": 0.4076,
+      "step": 20306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8638481450746567,
+      "learning_rate": 1.0927929522870603e-06,
+      "loss": 0.3226,
+      "step": 20307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7737362324733408,
+      "learning_rate": 1.092661458275388e-06,
+      "loss": 0.3845,
+      "step": 20308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5957063791017663,
+      "learning_rate": 1.0925300565286107e-06,
+      "loss": 0.6954,
+      "step": 20309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9986859155581616,
+      "learning_rate": 1.092398747049478e-06,
+      "loss": 0.7193,
+      "step": 20310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6821349641089947,
+      "learning_rate": 1.0922675298407405e-06,
+      "loss": 0.492,
+      "step": 20311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7496321388731695,
+      "learning_rate": 1.092136404905145e-06,
+      "loss": 0.4971,
+      "step": 20312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4682040253276565,
+      "learning_rate": 1.092005372245436e-06,
+      "loss": 0.2919,
+      "step": 20313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.396451508124207,
+      "learning_rate": 1.091874431864356e-06,
+      "loss": 0.5059,
+      "step": 20314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9409658642230496,
+      "learning_rate": 1.0917435837646472e-06,
+      "loss": 0.606,
+      "step": 20315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9150650466922658,
+      "learning_rate": 1.0916128279490482e-06,
+      "loss": 0.5203,
+      "step": 20316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4890712457022628,
+      "learning_rate": 1.0914821644202964e-06,
+      "loss": 0.3902,
+      "step": 20317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7256835991361938,
+      "learning_rate": 1.0913515931811277e-06,
+      "loss": 0.4002,
+      "step": 20318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6764119701791835,
+      "learning_rate": 1.0912211142342747e-06,
+      "loss": 0.4321,
+      "step": 20319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8189856519912784,
+      "learning_rate": 1.0910907275824696e-06,
+      "loss": 0.5846,
+      "step": 20320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6334041827869195,
+      "learning_rate": 1.0909604332284411e-06,
+      "loss": 0.3732,
+      "step": 20321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.227186446655387,
+      "learning_rate": 1.0908302311749179e-06,
+      "loss": 0.5682,
+      "step": 20322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.438957305080292,
+      "learning_rate": 1.0907001214246251e-06,
+      "loss": 0.3321,
+      "step": 20323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7945014477990606,
+      "learning_rate": 1.0905701039802863e-06,
+      "loss": 0.4151,
+      "step": 20324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.793721545557173,
+      "learning_rate": 1.0904401788446242e-06,
+      "loss": 0.4944,
+      "step": 20325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7666395161576296,
+      "learning_rate": 1.090310346020357e-06,
+      "loss": 0.5045,
+      "step": 20326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3805454927935528,
+      "learning_rate": 1.0901806055102044e-06,
+      "loss": 0.3467,
+      "step": 20327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.436690326773868,
+      "learning_rate": 1.0900509573168823e-06,
+      "loss": 0.2745,
+      "step": 20328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0198381873900084,
+      "learning_rate": 1.0899214014431032e-06,
+      "loss": 0.4523,
+      "step": 20329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7786353744159453,
+      "learning_rate": 1.0897919378915812e-06,
+      "loss": 0.4293,
+      "step": 20330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7200299322586143,
+      "learning_rate": 1.0896625666650255e-06,
+      "loss": 0.4086,
+      "step": 20331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7899707131464968,
+      "learning_rate": 1.0895332877661445e-06,
+      "loss": 0.5105,
+      "step": 20332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8976940896623054,
+      "learning_rate": 1.0894041011976444e-06,
+      "loss": 0.3304,
+      "step": 20333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.164419261106154,
+      "learning_rate": 1.0892750069622297e-06,
+      "loss": 0.4649,
+      "step": 20334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4088777729958495,
+      "learning_rate": 1.0891460050626044e-06,
+      "loss": 0.3285,
+      "step": 20335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0554821050618686,
+      "learning_rate": 1.0890170955014665e-06,
+      "loss": 0.4654,
+      "step": 20336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5704406644200706,
+      "learning_rate": 1.0888882782815166e-06,
+      "loss": 0.335,
+      "step": 20337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8810803524525739,
+      "learning_rate": 1.0887595534054507e-06,
+      "loss": 0.5795,
+      "step": 20338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.300866257335257,
+      "learning_rate": 1.0886309208759638e-06,
+      "loss": 0.3857,
+      "step": 20339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0037402268582523,
+      "learning_rate": 1.0885023806957478e-06,
+      "loss": 0.3813,
+      "step": 20340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9641658077223931,
+      "learning_rate": 1.0883739328674953e-06,
+      "loss": 0.4941,
+      "step": 20341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5715839101685347,
+      "learning_rate": 1.088245577393894e-06,
+      "loss": 0.3601,
+      "step": 20342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9948138614515865,
+      "learning_rate": 1.0881173142776309e-06,
+      "loss": 0.4912,
+      "step": 20343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.014815426217299,
+      "learning_rate": 1.087989143521392e-06,
+      "loss": 0.6267,
+      "step": 20344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9028893095620778,
+      "learning_rate": 1.0878610651278595e-06,
+      "loss": 0.5657,
+      "step": 20345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5848652492550659,
+      "learning_rate": 1.0877330790997155e-06,
+      "loss": 0.4352,
+      "step": 20346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2633785381096891,
+      "learning_rate": 1.0876051854396387e-06,
+      "loss": 0.2859,
+      "step": 20347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8993641820082388,
+      "learning_rate": 1.0874773841503074e-06,
+      "loss": 0.5374,
+      "step": 20348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8480177173767116,
+      "learning_rate": 1.0873496752343954e-06,
+      "loss": 0.5293,
+      "step": 20349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7755370154010586,
+      "learning_rate": 1.0872220586945773e-06,
+      "loss": 0.5376,
+      "step": 20350
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.4822619557380676,
+      "eval_runtime": 67.8758,
+      "eval_samples_per_second": 50.239,
+      "eval_steps_per_second": 1.576,
+      "step": 20350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5401699633918333,
+      "learning_rate": 1.0870945345335253e-06,
+      "loss": 0.3107,
+      "step": 20351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8548887442940654,
+      "learning_rate": 1.0869671027539078e-06,
+      "loss": 0.3945,
+      "step": 20352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.364230282791656,
+      "learning_rate": 1.0868397633583926e-06,
+      "loss": 0.4318,
+      "step": 20353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5130608655091338,
+      "learning_rate": 1.0867125163496463e-06,
+      "loss": 0.507,
+      "step": 20354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.29616124913101,
+      "learning_rate": 1.086585361730332e-06,
+      "loss": 0.3539,
+      "step": 20355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6871636791547997,
+      "learning_rate": 1.0864582995031123e-06,
+      "loss": 0.421,
+      "step": 20356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.060584032889151,
+      "learning_rate": 1.0863313296706463e-06,
+      "loss": 0.4834,
+      "step": 20357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6897514650828864,
+      "learning_rate": 1.0862044522355924e-06,
+      "loss": 0.475,
+      "step": 20358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7044130592259465,
+      "learning_rate": 1.0860776672006073e-06,
+      "loss": 0.4122,
+      "step": 20359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6730575922738922,
+      "learning_rate": 1.0859509745683442e-06,
+      "loss": 0.451,
+      "step": 20360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.083592754938782,
+      "learning_rate": 1.0858243743414559e-06,
+      "loss": 0.4803,
+      "step": 20361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4215224922038476,
+      "learning_rate": 1.0856978665225924e-06,
+      "loss": 0.3085,
+      "step": 20362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.95994260157662,
+      "learning_rate": 1.0855714511144027e-06,
+      "loss": 0.5545,
+      "step": 20363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8620883824731689,
+      "learning_rate": 1.0854451281195323e-06,
+      "loss": 0.4334,
+      "step": 20364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5210976128629836,
+      "learning_rate": 1.0853188975406262e-06,
+      "loss": 0.5202,
+      "step": 20365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4060889976775515,
+      "learning_rate": 1.0851927593803274e-06,
+      "loss": 0.2766,
+      "step": 20366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.502451762408252,
+      "learning_rate": 1.0850667136412758e-06,
+      "loss": 0.3548,
+      "step": 20367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.925460961057095,
+      "learning_rate": 1.08494076032611e-06,
+      "loss": 0.4509,
+      "step": 20368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7483395748939086,
+      "learning_rate": 1.0848148994374674e-06,
+      "loss": 0.3999,
+      "step": 20369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.048423503610214,
+      "learning_rate": 1.0846891309779825e-06,
+      "loss": 0.5972,
+      "step": 20370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3058116415650527,
+      "learning_rate": 1.0845634549502883e-06,
+      "loss": 0.3277,
+      "step": 20371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8809268663969867,
+      "learning_rate": 1.0844378713570155e-06,
+      "loss": 0.3975,
+      "step": 20372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.697262235833605,
+      "learning_rate": 1.0843123802007933e-06,
+      "loss": 0.565,
+      "step": 20373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8210045606533423,
+      "learning_rate": 1.0841869814842488e-06,
+      "loss": 0.5285,
+      "step": 20374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4476388163433727,
+      "learning_rate": 1.084061675210007e-06,
+      "loss": 0.4426,
+      "step": 20375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3111732166487506,
+      "learning_rate": 1.0839364613806915e-06,
+      "loss": 0.474,
+      "step": 20376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5604157757363928,
+      "learning_rate": 1.0838113399989225e-06,
+      "loss": 0.4339,
+      "step": 20377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.743450325799422,
+      "learning_rate": 1.0836863110673206e-06,
+      "loss": 0.3762,
+      "step": 20378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6178886888022899,
+      "learning_rate": 1.0835613745885026e-06,
+      "loss": 0.3568,
+      "step": 20379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4461761503565147,
+      "learning_rate": 1.0834365305650843e-06,
+      "loss": 0.4592,
+      "step": 20380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7391428145386356,
+      "learning_rate": 1.0833117789996785e-06,
+      "loss": 0.5753,
+      "step": 20381
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.628092285002639,
+      "learning_rate": 1.0831871198948975e-06,
+      "loss": 0.4496,
+      "step": 20382
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6442121230469022,
+      "learning_rate": 1.083062553253351e-06,
+      "loss": 0.3687,
+      "step": 20383
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8198191379864308,
+      "learning_rate": 1.0829380790776462e-06,
+      "loss": 0.415,
+      "step": 20384
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.771732130686991,
+      "learning_rate": 1.0828136973703894e-06,
+      "loss": 0.4489,
+      "step": 20385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7158380947874066,
+      "learning_rate": 1.0826894081341846e-06,
+      "loss": 0.4918,
+      "step": 20386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5946808284622138,
+      "learning_rate": 1.0825652113716327e-06,
+      "loss": 0.3853,
+      "step": 20387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6213211877259928,
+      "learning_rate": 1.0824411070853345e-06,
+      "loss": 0.4682,
+      "step": 20388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5844205488554208,
+      "learning_rate": 1.082317095277888e-06,
+      "loss": 0.4184,
+      "step": 20389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6689722741006887,
+      "learning_rate": 1.082193175951889e-06,
+      "loss": 0.432,
+      "step": 20390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8331891741882773,
+      "learning_rate": 1.0820693491099318e-06,
+      "loss": 0.3817,
+      "step": 20391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6758395990226032,
+      "learning_rate": 1.0819456147546092e-06,
+      "loss": 0.4001,
+      "step": 20392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.730603627774755,
+      "learning_rate": 1.0818219728885107e-06,
+      "loss": 0.4545,
+      "step": 20393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0400922252874993,
+      "learning_rate": 1.0816984235142249e-06,
+      "loss": 0.5764,
+      "step": 20394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.924216987113337,
+      "learning_rate": 1.0815749666343383e-06,
+      "loss": 0.6738,
+      "step": 20395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6533592340359173,
+      "learning_rate": 1.0814516022514357e-06,
+      "loss": 0.4474,
+      "step": 20396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8623825446688016,
+      "learning_rate": 1.0813283303680987e-06,
+      "loss": 0.4425,
+      "step": 20397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7807378057490764,
+      "learning_rate": 1.081205150986909e-06,
+      "loss": 0.4993,
+      "step": 20398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.760354796556047,
+      "learning_rate": 1.081082064110445e-06,
+      "loss": 0.4432,
+      "step": 20399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1665291579294466,
+      "learning_rate": 1.080959069741283e-06,
+      "loss": 0.4826,
+      "step": 20400
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.4822576642036438,
+      "eval_runtime": 67.9279,
+      "eval_samples_per_second": 50.2,
+      "eval_steps_per_second": 1.575,
+      "step": 20400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9394254478668516,
+      "learning_rate": 1.0808361678819977e-06,
+      "loss": 0.4889,
+      "step": 20401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0424068921476954,
+      "learning_rate": 1.0807133585351633e-06,
+      "loss": 0.4529,
+      "step": 20402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1539072316078727,
+      "learning_rate": 1.0805906417033494e-06,
+      "loss": 0.3917,
+      "step": 20403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0848860537153024,
+      "learning_rate": 1.0804680173891252e-06,
+      "loss": 0.532,
+      "step": 20404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0053534094165246,
+      "learning_rate": 1.0803454855950586e-06,
+      "loss": 0.5769,
+      "step": 20405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5744723224512698,
+      "learning_rate": 1.0802230463237133e-06,
+      "loss": 0.4485,
+      "step": 20406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.57564008093095,
+      "learning_rate": 1.0801006995776535e-06,
+      "loss": 0.3761,
+      "step": 20407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8005772573876262,
+      "learning_rate": 1.0799784453594402e-06,
+      "loss": 0.4292,
+      "step": 20408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.885265820216119,
+      "learning_rate": 1.0798562836716331e-06,
+      "loss": 0.4314,
+      "step": 20409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4349124126939108,
+      "learning_rate": 1.079734214516789e-06,
+      "loss": 0.3611,
+      "step": 20410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.211000173599419,
+      "learning_rate": 1.079612237897464e-06,
+      "loss": 0.3142,
+      "step": 20411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6330745326842275,
+      "learning_rate": 1.0794903538162112e-06,
+      "loss": 0.4492,
+      "step": 20412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.062098747366563,
+      "learning_rate": 1.0793685622755814e-06,
+      "loss": 0.7221,
+      "step": 20413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9452147615270745,
+      "learning_rate": 1.0792468632781256e-06,
+      "loss": 0.4152,
+      "step": 20414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.03514801029163,
+      "learning_rate": 1.079125256826391e-06,
+      "loss": 0.3479,
+      "step": 20415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7991024084557439,
+      "learning_rate": 1.0790037429229228e-06,
+      "loss": 0.4106,
+      "step": 20416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.550912015822217,
+      "learning_rate": 1.0788823215702654e-06,
+      "loss": 0.3882,
+      "step": 20417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.351178882352683,
+      "learning_rate": 1.078760992770961e-06,
+      "loss": 0.48,
+      "step": 20418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0872835682058177,
+      "learning_rate": 1.0786397565275486e-06,
+      "loss": 0.6539,
+      "step": 20419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9169118337097737,
+      "learning_rate": 1.0785186128425664e-06,
+      "loss": 0.5046,
+      "step": 20420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1428670762094075,
+      "learning_rate": 1.0783975617185519e-06,
+      "loss": 0.4589,
+      "step": 20421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.67985871280005,
+      "learning_rate": 1.0782766031580371e-06,
+      "loss": 0.5414,
+      "step": 20422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8357305416774536,
+      "learning_rate": 1.0781557371635552e-06,
+      "loss": 0.5158,
+      "step": 20423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.651578773077474,
+      "learning_rate": 1.0780349637376371e-06,
+      "loss": 0.5273,
+      "step": 20424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0421180021874776,
+      "learning_rate": 1.07791428288281e-06,
+      "loss": 0.6868,
+      "step": 20425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6361586771453809,
+      "learning_rate": 1.0777936946016004e-06,
+      "loss": 0.3465,
+      "step": 20426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.774994099385585,
+      "learning_rate": 1.077673198896533e-06,
+      "loss": 0.4732,
+      "step": 20427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4368078654469199,
+      "learning_rate": 1.0775527957701315e-06,
+      "loss": 0.3996,
+      "step": 20428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2852856183614358,
+      "learning_rate": 1.0774324852249142e-06,
+      "loss": 0.3964,
+      "step": 20429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7536495907500114,
+      "learning_rate": 1.0773122672634012e-06,
+      "loss": 0.4692,
+      "step": 20430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.826583354570905,
+      "learning_rate": 1.0771921418881092e-06,
+      "loss": 0.3636,
+      "step": 20431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8213856881915522,
+      "learning_rate": 1.0770721091015524e-06,
+      "loss": 0.5629,
+      "step": 20432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9558798445659886,
+      "learning_rate": 1.0769521689062437e-06,
+      "loss": 0.4682,
+      "step": 20433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6474980868486113,
+      "learning_rate": 1.0768323213046943e-06,
+      "loss": 0.4274,
+      "step": 20434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8899319156972445,
+      "learning_rate": 1.0767125662994124e-06,
+      "loss": 0.4768,
+      "step": 20435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2856454537029045,
+      "learning_rate": 1.076592903892906e-06,
+      "loss": 0.3383,
+      "step": 20436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3818239657055693,
+      "learning_rate": 1.07647333408768e-06,
+      "loss": 0.27,
+      "step": 20437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9958117253572274,
+      "learning_rate": 1.0763538568862367e-06,
+      "loss": 0.4486,
+      "step": 20438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5828399481111923,
+      "learning_rate": 1.0762344722910774e-06,
+      "loss": 0.408,
+      "step": 20439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.215242543737064,
+      "learning_rate": 1.0761151803047019e-06,
+      "loss": 0.4238,
+      "step": 20440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4762600318313812,
+      "learning_rate": 1.0759959809296078e-06,
+      "loss": 0.3991,
+      "step": 20441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.478842906116091,
+      "learning_rate": 1.0758768741682897e-06,
+      "loss": 0.3706,
+      "step": 20442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5161424876124465,
+      "learning_rate": 1.0757578600232415e-06,
+      "loss": 0.3804,
+      "step": 20443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6419021269932914,
+      "learning_rate": 1.0756389384969546e-06,
+      "loss": 0.4093,
+      "step": 20444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.380472398752897,
+      "learning_rate": 1.0755201095919178e-06,
+      "loss": 0.2736,
+      "step": 20445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8877704759885443,
+      "learning_rate": 1.0754013733106198e-06,
+      "loss": 0.5052,
+      "step": 20446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.845294607058835,
+      "learning_rate": 1.0752827296555464e-06,
+      "loss": 0.4951,
+      "step": 20447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7457279157455614,
+      "learning_rate": 1.0751641786291796e-06,
+      "loss": 0.4504,
+      "step": 20448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8078529825903171,
+      "learning_rate": 1.0750457202340031e-06,
+      "loss": 0.454,
+      "step": 20449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8286076263814257,
+      "learning_rate": 1.074927354472496e-06,
+      "loss": 0.4228,
+      "step": 20450
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.4823189973831177,
+      "eval_runtime": 67.8931,
+      "eval_samples_per_second": 50.226,
+      "eval_steps_per_second": 1.576,
+      "step": 20450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1550383104341404,
+      "learning_rate": 1.0748090813471358e-06,
+      "loss": 0.5527,
+      "step": 20451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1296588074190947,
+      "learning_rate": 1.074690900860399e-06,
+      "loss": 0.4623,
+      "step": 20452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8417361950053943,
+      "learning_rate": 1.0745728130147596e-06,
+      "loss": 0.4407,
+      "step": 20453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7683090504544936,
+      "learning_rate": 1.0744548178126901e-06,
+      "loss": 0.4598,
+      "step": 20454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.577761033206876,
+      "learning_rate": 1.0743369152566597e-06,
+      "loss": 0.5083,
+      "step": 20455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5672776153059176,
+      "learning_rate": 1.0742191053491368e-06,
+      "loss": 0.4486,
+      "step": 20456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9348248022582124,
+      "learning_rate": 1.0741013880925887e-06,
+      "loss": 0.4427,
+      "step": 20457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0273424848251893,
+      "learning_rate": 1.0739837634894782e-06,
+      "loss": 0.6423,
+      "step": 20458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7215824599424236,
+      "learning_rate": 1.0738662315422697e-06,
+      "loss": 0.393,
+      "step": 20459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.093002806095957,
+      "learning_rate": 1.0737487922534214e-06,
+      "loss": 0.6012,
+      "step": 20460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.802874678376539,
+      "learning_rate": 1.0736314456253932e-06,
+      "loss": 0.4847,
+      "step": 20461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.855146280712206,
+      "learning_rate": 1.0735141916606411e-06,
+      "loss": 0.564,
+      "step": 20462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7898021900919714,
+      "learning_rate": 1.0733970303616206e-06,
+      "loss": 0.5164,
+      "step": 20463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5368968958436962,
+      "learning_rate": 1.0732799617307834e-06,
+      "loss": 0.3159,
+      "step": 20464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7534983750007562,
+      "learning_rate": 1.0731629857705806e-06,
+      "loss": 0.4501,
+      "step": 20465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1329565480582566,
+      "learning_rate": 1.0730461024834615e-06,
+      "loss": 0.4351,
+      "step": 20466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5619971455018715,
+      "learning_rate": 1.0729293118718722e-06,
+      "loss": 0.448,
+      "step": 20467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.000581087106719,
+      "learning_rate": 1.0728126139382582e-06,
+      "loss": 0.5544,
+      "step": 20468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1031862268207706,
+      "learning_rate": 1.0726960086850623e-06,
+      "loss": 0.6063,
+      "step": 20469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6338969308342073,
+      "learning_rate": 1.0725794961147256e-06,
+      "loss": 0.4447,
+      "step": 20470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7676793484973372,
+      "learning_rate": 1.0724630762296867e-06,
+      "loss": 0.4608,
+      "step": 20471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9421806287529335,
+      "learning_rate": 1.072346749032384e-06,
+      "loss": 0.5462,
+      "step": 20472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3110730688784296,
+      "learning_rate": 1.0722305145252517e-06,
+      "loss": 0.3302,
+      "step": 20473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3552402596936117,
+      "learning_rate": 1.0721143727107228e-06,
+      "loss": 0.344,
+      "step": 20474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7946335753302642,
+      "learning_rate": 1.0719983235912299e-06,
+      "loss": 0.512,
+      "step": 20475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4438804544258408,
+      "learning_rate": 1.071882367169202e-06,
+      "loss": 0.4272,
+      "step": 20476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7810911918233643,
+      "learning_rate": 1.0717665034470655e-06,
+      "loss": 0.4813,
+      "step": 20477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4296128195993245,
+      "learning_rate": 1.0716507324272473e-06,
+      "loss": 0.435,
+      "step": 20478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7989651177323094,
+      "learning_rate": 1.0715350541121704e-06,
+      "loss": 0.5137,
+      "step": 20479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6758068328729403,
+      "learning_rate": 1.0714194685042564e-06,
+      "loss": 0.4817,
+      "step": 20480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7173891022377425,
+      "learning_rate": 1.0713039756059252e-06,
+      "loss": 0.5442,
+      "step": 20481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5733308692455075,
+      "learning_rate": 1.0711885754195947e-06,
+      "loss": 0.416,
+      "step": 20482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0538166611482542,
+      "learning_rate": 1.0710732679476805e-06,
+      "loss": 0.6585,
+      "step": 20483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6506843787245136,
+      "learning_rate": 1.0709580531925962e-06,
+      "loss": 0.5079,
+      "step": 20484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7367167716255356,
+      "learning_rate": 1.0708429311567544e-06,
+      "loss": 0.4639,
+      "step": 20485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.89451410288874,
+      "learning_rate": 1.0707279018425645e-06,
+      "loss": 0.4314,
+      "step": 20486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.585842934092112,
+      "learning_rate": 1.0706129652524352e-06,
+      "loss": 0.4309,
+      "step": 20487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5922241918158972,
+      "learning_rate": 1.0704981213887718e-06,
+      "loss": 0.4484,
+      "step": 20488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7857420171083682,
+      "learning_rate": 1.0703833702539796e-06,
+      "loss": 0.5794,
+      "step": 20489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8918493270787036,
+      "learning_rate": 1.0702687118504599e-06,
+      "loss": 0.4738,
+      "step": 20490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5206755773507408,
+      "learning_rate": 1.0701541461806129e-06,
+      "loss": 0.4323,
+      "step": 20491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5328830414206243,
+      "learning_rate": 1.0700396732468383e-06,
+      "loss": 0.4179,
+      "step": 20492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6209601659022972,
+      "learning_rate": 1.0699252930515306e-06,
+      "loss": 0.5338,
+      "step": 20493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5836209764992493,
+      "learning_rate": 1.069811005597086e-06,
+      "loss": 0.3668,
+      "step": 20494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5188548543690532,
+      "learning_rate": 1.0696968108858958e-06,
+      "loss": 0.42,
+      "step": 20495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8615782610248857,
+      "learning_rate": 1.0695827089203513e-06,
+      "loss": 0.5621,
+      "step": 20496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7725712366535415,
+      "learning_rate": 1.0694686997028412e-06,
+      "loss": 0.4055,
+      "step": 20497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.809331448184698,
+      "learning_rate": 1.0693547832357514e-06,
+      "loss": 0.5072,
+      "step": 20498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8014211365555288,
+      "learning_rate": 1.0692409595214679e-06,
+      "loss": 0.4662,
+      "step": 20499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.606813483080893,
+      "learning_rate": 1.0691272285623724e-06,
+      "loss": 0.3133,
+      "step": 20500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.4820255637168884,
+      "eval_runtime": 67.9574,
+      "eval_samples_per_second": 50.178,
+      "eval_steps_per_second": 1.575,
+      "step": 20500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7846577249021969,
+      "learning_rate": 1.0690135903608462e-06,
+      "loss": 0.407,
+      "step": 20501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7126424885654243,
+      "learning_rate": 1.0689000449192688e-06,
+      "loss": 0.5074,
+      "step": 20502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.363492422837173,
+      "learning_rate": 1.0687865922400164e-06,
+      "loss": 0.3231,
+      "step": 20503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1622699820328406,
+      "learning_rate": 1.0686732323254643e-06,
+      "loss": 0.5865,
+      "step": 20504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.817694085002479,
+      "learning_rate": 1.068559965177986e-06,
+      "loss": 0.5832,
+      "step": 20505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6718417493439157,
+      "learning_rate": 1.0684467907999518e-06,
+      "loss": 0.471,
+      "step": 20506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9841227822077576,
+      "learning_rate": 1.0683337091937313e-06,
+      "loss": 0.5209,
+      "step": 20507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5094497145902996,
+      "learning_rate": 1.0682207203616928e-06,
+      "loss": 0.3042,
+      "step": 20508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8163956895139208,
+      "learning_rate": 1.0681078243062005e-06,
+      "loss": 0.5605,
+      "step": 20509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4816134444267814,
+      "learning_rate": 1.067995021029618e-06,
+      "loss": 0.5546,
+      "step": 20510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6711123146199298,
+      "learning_rate": 1.067882310534307e-06,
+      "loss": 0.4591,
+      "step": 20511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8533920933051673,
+      "learning_rate": 1.0677696928226269e-06,
+      "loss": 0.5393,
+      "step": 20512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8905889427342315,
+      "learning_rate": 1.0676571678969356e-06,
+      "loss": 0.3101,
+      "step": 20513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.400189374932305,
+      "learning_rate": 1.0675447357595886e-06,
+      "loss": 0.4882,
+      "step": 20514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6915325702913666,
+      "learning_rate": 1.0674323964129389e-06,
+      "loss": 0.4939,
+      "step": 20515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.042507940628582,
+      "learning_rate": 1.067320149859339e-06,
+      "loss": 0.6835,
+      "step": 20516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3208662293631557,
+      "learning_rate": 1.0672079961011388e-06,
+      "loss": 0.2885,
+      "step": 20517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8089274979006116,
+      "learning_rate": 1.067095935140686e-06,
+      "loss": 0.3451,
+      "step": 20518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.64575664296587,
+      "learning_rate": 1.0669839669803256e-06,
+      "loss": 0.4885,
+      "step": 20519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.753704908358091,
+      "learning_rate": 1.066872091622403e-06,
+      "loss": 0.4274,
+      "step": 20520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3450571528801643,
+      "learning_rate": 1.0667603090692596e-06,
+      "loss": 0.3406,
+      "step": 20521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7501769755803436,
+      "learning_rate": 1.0666486193232358e-06,
+      "loss": 0.3568,
+      "step": 20522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7635185516243346,
+      "learning_rate": 1.066537022386669e-06,
+      "loss": 0.4231,
+      "step": 20523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9873204584168727,
+      "learning_rate": 1.066425518261896e-06,
+      "loss": 0.6127,
+      "step": 20524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8718887197409586,
+      "learning_rate": 1.066314106951251e-06,
+      "loss": 0.5846,
+      "step": 20525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1592233374334677,
+      "learning_rate": 1.0662027884570664e-06,
+      "loss": 0.6237,
+      "step": 20526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8867413880339734,
+      "learning_rate": 1.0660915627816726e-06,
+      "loss": 0.4952,
+      "step": 20527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4209847720879816,
+      "learning_rate": 1.0659804299273978e-06,
+      "loss": 0.3306,
+      "step": 20528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9155893372735708,
+      "learning_rate": 1.0658693898965688e-06,
+      "loss": 0.4796,
+      "step": 20529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0520993108401737,
+      "learning_rate": 1.0657584426915097e-06,
+      "loss": 0.4892,
+      "step": 20530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.040956493133499,
+      "learning_rate": 1.0656475883145436e-06,
+      "loss": 0.6436,
+      "step": 20531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9537806637751878,
+      "learning_rate": 1.0655368267679908e-06,
+      "loss": 0.63,
+      "step": 20532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8302330893001764,
+      "learning_rate": 1.06542615805417e-06,
+      "loss": 0.6627,
+      "step": 20533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9905042721782595,
+      "learning_rate": 1.065315582175399e-06,
+      "loss": 0.5623,
+      "step": 20534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7642513539140225,
+      "learning_rate": 1.0652050991339908e-06,
+      "loss": 0.348,
+      "step": 20535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1475192031393977,
+      "learning_rate": 1.0650947089322595e-06,
+      "loss": 0.8323,
+      "step": 20536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8084566810241567,
+      "learning_rate": 1.0649844115725164e-06,
+      "loss": 0.5753,
+      "step": 20537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7357938754935134,
+      "learning_rate": 1.0648742070570698e-06,
+      "loss": 0.4036,
+      "step": 20538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9353406285187333,
+      "learning_rate": 1.0647640953882264e-06,
+      "loss": 0.5596,
+      "step": 20539
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1695278566457503,
+      "learning_rate": 1.0646540765682924e-06,
+      "loss": 0.6324,
+      "step": 20540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7626422128786694,
+      "learning_rate": 1.0645441505995704e-06,
+      "loss": 0.5205,
+      "step": 20541
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9157915843401205,
+      "learning_rate": 1.064434317484361e-06,
+      "loss": 0.4365,
+      "step": 20542
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7266941775014655,
+      "learning_rate": 1.064324577224965e-06,
+      "loss": 0.4168,
+      "step": 20543
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7732721092457864,
+      "learning_rate": 1.0642149298236779e-06,
+      "loss": 0.4958,
+      "step": 20544
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9087274795226952,
+      "learning_rate": 1.0641053752827968e-06,
+      "loss": 0.5258,
+      "step": 20545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8593818701416391,
+      "learning_rate": 1.0639959136046142e-06,
+      "loss": 0.4728,
+      "step": 20546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9682442088645171,
+      "learning_rate": 1.0638865447914223e-06,
+      "loss": 0.4843,
+      "step": 20547
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7993796634754062,
+      "learning_rate": 1.0637772688455091e-06,
+      "loss": 0.525,
+      "step": 20548
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7636604467625296,
+      "learning_rate": 1.0636680857691642e-06,
+      "loss": 0.3754,
+      "step": 20549
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.400128561296626,
+      "learning_rate": 1.0635589955646725e-06,
+      "loss": 0.3849,
+      "step": 20550
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.4816506505012512,
+      "eval_runtime": 67.9715,
+      "eval_samples_per_second": 50.168,
+      "eval_steps_per_second": 1.574,
+      "step": 20550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8579905714442466,
+      "learning_rate": 1.0634499982343176e-06,
+      "loss": 0.4862,
+      "step": 20551
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8693567870996979,
+      "learning_rate": 1.0633410937803807e-06,
+      "loss": 0.4539,
+      "step": 20552
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.058698282597655,
+      "learning_rate": 1.063232282205143e-06,
+      "loss": 0.5627,
+      "step": 20553
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2820845431410723,
+      "learning_rate": 1.0631235635108814e-06,
+      "loss": 0.4207,
+      "step": 20554
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.820521661161671,
+      "learning_rate": 1.063014937699872e-06,
+      "loss": 0.5268,
+      "step": 20555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.492182640837674,
+      "learning_rate": 1.0629064047743895e-06,
+      "loss": 0.4671,
+      "step": 20556
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5885360405079807,
+      "learning_rate": 1.0627979647367046e-06,
+      "loss": 0.4194,
+      "step": 20557
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9457756396470403,
+      "learning_rate": 1.0626896175890887e-06,
+      "loss": 0.5294,
+      "step": 20558
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6427304674728769,
+      "learning_rate": 1.0625813633338098e-06,
+      "loss": 0.4669,
+      "step": 20559
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0437728206831447,
+      "learning_rate": 1.0624732019731338e-06,
+      "loss": 0.5843,
+      "step": 20560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6822861984784248,
+      "learning_rate": 1.0623651335093245e-06,
+      "loss": 0.663,
+      "step": 20561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6474906716159174,
+      "learning_rate": 1.0622571579446458e-06,
+      "loss": 0.3738,
+      "step": 20562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.963567767261639,
+      "learning_rate": 1.0621492752813565e-06,
+      "loss": 0.5453,
+      "step": 20563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6320501360012127,
+      "learning_rate": 1.0620414855217155e-06,
+      "loss": 0.4385,
+      "step": 20564
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.761161006310846,
+      "learning_rate": 1.0619337886679796e-06,
+      "loss": 0.4118,
+      "step": 20565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.606048814346291,
+      "learning_rate": 1.0618261847224037e-06,
+      "loss": 0.5003,
+      "step": 20566
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6034073044064459,
+      "learning_rate": 1.0617186736872393e-06,
+      "loss": 0.4669,
+      "step": 20567
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.686382681971954,
+      "learning_rate": 1.0616112555647382e-06,
+      "loss": 0.4854,
+      "step": 20568
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8247899046966312,
+      "learning_rate": 1.0615039303571487e-06,
+      "loss": 0.5485,
+      "step": 20569
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7476741860452847,
+      "learning_rate": 1.0613966980667174e-06,
+      "loss": 0.6372,
+      "step": 20570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.651010924376149,
+      "learning_rate": 1.0612895586956893e-06,
+      "loss": 0.6522,
+      "step": 20571
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.707127286870436,
+      "learning_rate": 1.0611825122463072e-06,
+      "loss": 0.472,
+      "step": 20572
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.855201506787421,
+      "learning_rate": 1.061075558720812e-06,
+      "loss": 0.5704,
+      "step": 20573
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2515717184457342,
+      "learning_rate": 1.0609686981214434e-06,
+      "loss": 0.5215,
+      "step": 20574
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7779001204625837,
+      "learning_rate": 1.0608619304504382e-06,
+      "loss": 0.4853,
+      "step": 20575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2028133353864026,
+      "learning_rate": 1.0607552557100305e-06,
+      "loss": 0.5049,
+      "step": 20576
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6925266269014456,
+      "learning_rate": 1.060648673902454e-06,
+      "loss": 0.4863,
+      "step": 20577
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7206273470202,
+      "learning_rate": 1.0605421850299406e-06,
+      "loss": 0.6465,
+      "step": 20578
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.03457722745149,
+      "learning_rate": 1.0604357890947194e-06,
+      "loss": 0.5404,
+      "step": 20579
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8474431101142412,
+      "learning_rate": 1.060329486099017e-06,
+      "loss": 0.4711,
+      "step": 20580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6933287874613456,
+      "learning_rate": 1.0602232760450592e-06,
+      "loss": 0.4329,
+      "step": 20581
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8678595747328381,
+      "learning_rate": 1.06011715893507e-06,
+      "loss": 0.7058,
+      "step": 20582
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9077203213270024,
+      "learning_rate": 1.0600111347712696e-06,
+      "loss": 0.4378,
+      "step": 20583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7582215017514564,
+      "learning_rate": 1.0599052035558784e-06,
+      "loss": 0.6121,
+      "step": 20584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8190893560212076,
+      "learning_rate": 1.0597993652911148e-06,
+      "loss": 0.4808,
+      "step": 20585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7830356637304112,
+      "learning_rate": 1.059693619979193e-06,
+      "loss": 0.5074,
+      "step": 20586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.56274318001191,
+      "learning_rate": 1.0595879676223266e-06,
+      "loss": 0.3499,
+      "step": 20587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9019831311641693,
+      "learning_rate": 1.0594824082227293e-06,
+      "loss": 0.5488,
+      "step": 20588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8954757504801258,
+      "learning_rate": 1.0593769417826085e-06,
+      "loss": 0.5052,
+      "step": 20589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9391344733591014,
+      "learning_rate": 1.0592715683041744e-06,
+      "loss": 0.4627,
+      "step": 20590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.71657828804923,
+      "learning_rate": 1.059166287789631e-06,
+      "loss": 0.5115,
+      "step": 20591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8036129701063832,
+      "learning_rate": 1.0590611002411836e-06,
+      "loss": 0.4613,
+      "step": 20592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8913840588130215,
+      "learning_rate": 1.058956005661033e-06,
+      "loss": 0.4747,
+      "step": 20593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.643988840897346,
+      "learning_rate": 1.05885100405138e-06,
+      "loss": 0.368,
+      "step": 20594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9897001110041996,
+      "learning_rate": 1.0587460954144235e-06,
+      "loss": 0.483,
+      "step": 20595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7705960176405375,
+      "learning_rate": 1.0586412797523588e-06,
+      "loss": 0.3915,
+      "step": 20596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9362718358630246,
+      "learning_rate": 1.0585365570673794e-06,
+      "loss": 0.4585,
+      "step": 20597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9632123833842838,
+      "learning_rate": 1.0584319273616795e-06,
+      "loss": 0.4463,
+      "step": 20598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.491156664694872,
+      "learning_rate": 1.0583273906374481e-06,
+      "loss": 0.5038,
+      "step": 20599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7456563511908465,
+      "learning_rate": 1.0582229468968733e-06,
+      "loss": 0.5087,
+      "step": 20600
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.4815196692943573,
+      "eval_runtime": 68.0102,
+      "eval_samples_per_second": 50.14,
+      "eval_steps_per_second": 1.573,
+      "step": 20600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6522757796293208,
+      "learning_rate": 1.0581185961421431e-06,
+      "loss": 0.4008,
+      "step": 20601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.610563308550339,
+      "learning_rate": 1.0580143383754408e-06,
+      "loss": 0.4213,
+      "step": 20602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.864634174904447,
+      "learning_rate": 1.0579101735989495e-06,
+      "loss": 0.5932,
+      "step": 20603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5723177785269917,
+      "learning_rate": 1.057806101814849e-06,
+      "loss": 0.3705,
+      "step": 20604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4923883769088704,
+      "learning_rate": 1.0577021230253195e-06,
+      "loss": 0.3887,
+      "step": 20605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9058580646014824,
+      "learning_rate": 1.0575982372325359e-06,
+      "loss": 0.5702,
+      "step": 20606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4586747323775398,
+      "learning_rate": 1.0574944444386746e-06,
+      "loss": 0.3198,
+      "step": 20607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0263502325109863,
+      "learning_rate": 1.0573907446459079e-06,
+      "loss": 0.5742,
+      "step": 20608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8124080717145425,
+      "learning_rate": 1.0572871378564063e-06,
+      "loss": 0.447,
+      "step": 20609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1759236797513553,
+      "learning_rate": 1.0571836240723388e-06,
+      "loss": 0.7603,
+      "step": 20610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.377161032117264,
+      "learning_rate": 1.0570802032958725e-06,
+      "loss": 0.6206,
+      "step": 20611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.549883317314455,
+      "learning_rate": 1.0569768755291727e-06,
+      "loss": 0.4106,
+      "step": 20612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8416385400678514,
+      "learning_rate": 1.0568736407744025e-06,
+      "loss": 0.3811,
+      "step": 20613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6216024413532706,
+      "learning_rate": 1.0567704990337232e-06,
+      "loss": 0.4713,
+      "step": 20614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6921173678627177,
+      "learning_rate": 1.056667450309293e-06,
+      "loss": 0.5117,
+      "step": 20615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6944372673877177,
+      "learning_rate": 1.0565644946032701e-06,
+      "loss": 0.6164,
+      "step": 20616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6616229262517406,
+      "learning_rate": 1.0564616319178102e-06,
+      "loss": 0.4319,
+      "step": 20617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6947581665674507,
+      "learning_rate": 1.0563588622550654e-06,
+      "loss": 0.4331,
+      "step": 20618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.558616260594567,
+      "learning_rate": 1.056256185617188e-06,
+      "loss": 0.3001,
+      "step": 20619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7172186239331193,
+      "learning_rate": 1.056153602006327e-06,
+      "loss": 0.5112,
+      "step": 20620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9970602064275185,
+      "learning_rate": 1.0560511114246305e-06,
+      "loss": 0.5569,
+      "step": 20621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6910124553680481,
+      "learning_rate": 1.0559487138742436e-06,
+      "loss": 0.4266,
+      "step": 20622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.725674259756756,
+      "learning_rate": 1.05584640935731e-06,
+      "loss": 0.4225,
+      "step": 20623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2596385386361386,
+      "learning_rate": 1.0557441978759714e-06,
+      "loss": 0.5691,
+      "step": 20624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.186875809017399,
+      "learning_rate": 1.0556420794323677e-06,
+      "loss": 0.5092,
+      "step": 20625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.607224589999891,
+      "learning_rate": 1.0555400540286364e-06,
+      "loss": 0.4831,
+      "step": 20626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0453700761828473,
+      "learning_rate": 1.0554381216669141e-06,
+      "loss": 0.5823,
+      "step": 20627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9443897676672155,
+      "learning_rate": 1.0553362823493337e-06,
+      "loss": 0.5796,
+      "step": 20628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8452497962316736,
+      "learning_rate": 1.0552345360780274e-06,
+      "loss": 0.5817,
+      "step": 20629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8575997397938437,
+      "learning_rate": 1.0551328828551258e-06,
+      "loss": 0.4021,
+      "step": 20630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8840865487388918,
+      "learning_rate": 1.0550313226827557e-06,
+      "loss": 0.5058,
+      "step": 20631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4964115394875246,
+      "learning_rate": 1.0549298555630444e-06,
+      "loss": 0.3362,
+      "step": 20632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7220659300277683,
+      "learning_rate": 1.0548284814981159e-06,
+      "loss": 0.4454,
+      "step": 20633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9365761957069485,
+      "learning_rate": 1.0547272004900922e-06,
+      "loss": 0.5064,
+      "step": 20634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.961490957517545,
+      "learning_rate": 1.0546260125410926e-06,
+      "loss": 0.5344,
+      "step": 20635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5483290378286276,
+      "learning_rate": 1.0545249176532372e-06,
+      "loss": 0.499,
+      "step": 20636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.770704881872255,
+      "learning_rate": 1.0544239158286412e-06,
+      "loss": 0.4072,
+      "step": 20637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6454296793896177,
+      "learning_rate": 1.0543230070694192e-06,
+      "loss": 0.3544,
+      "step": 20638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9149379131759159,
+      "learning_rate": 1.054222191377684e-06,
+      "loss": 0.394,
+      "step": 20639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6364334703239394,
+      "learning_rate": 1.0541214687555457e-06,
+      "loss": 0.476,
+      "step": 20640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0063182937571304,
+      "learning_rate": 1.0540208392051127e-06,
+      "loss": 0.4959,
+      "step": 20641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6861299891790913,
+      "learning_rate": 1.0539203027284925e-06,
+      "loss": 0.511,
+      "step": 20642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8268082875247014,
+      "learning_rate": 1.0538198593277892e-06,
+      "loss": 0.5688,
+      "step": 20643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9098248185276998,
+      "learning_rate": 1.0537195090051045e-06,
+      "loss": 0.446,
+      "step": 20644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8180318151060946,
+      "learning_rate": 1.0536192517625412e-06,
+      "loss": 0.4849,
+      "step": 20645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3510725619999977,
+      "learning_rate": 1.0535190876021966e-06,
+      "loss": 0.307,
+      "step": 20646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4333018050821738,
+      "learning_rate": 1.0534190165261685e-06,
+      "loss": 0.3153,
+      "step": 20647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8203944926957885,
+      "learning_rate": 1.0533190385365513e-06,
+      "loss": 0.4545,
+      "step": 20648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8905301185202086,
+      "learning_rate": 1.0532191536354382e-06,
+      "loss": 0.4296,
+      "step": 20649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.838442695405176,
+      "learning_rate": 1.0531193618249196e-06,
+      "loss": 0.5841,
+      "step": 20650
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.4816315174102783,
+      "eval_runtime": 68.0289,
+      "eval_samples_per_second": 50.126,
+      "eval_steps_per_second": 1.573,
+      "step": 20650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0067955999001112,
+      "learning_rate": 1.0530196631070857e-06,
+      "loss": 0.4821,
+      "step": 20651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0242196366296765,
+      "learning_rate": 1.0529200574840225e-06,
+      "loss": 0.5225,
+      "step": 20652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1698133054435753,
+      "learning_rate": 1.0528205449578163e-06,
+      "loss": 0.4133,
+      "step": 20653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8905685917503556,
+      "learning_rate": 1.0527211255305496e-06,
+      "loss": 0.4267,
+      "step": 20654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9825960018645803,
+      "learning_rate": 1.0526217992043032e-06,
+      "loss": 0.433,
+      "step": 20655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7031206025470578,
+      "learning_rate": 1.0525225659811575e-06,
+      "loss": 0.4897,
+      "step": 20656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7082940740962782,
+      "learning_rate": 1.0524234258631895e-06,
+      "loss": 0.4317,
+      "step": 20657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9825960831703635,
+      "learning_rate": 1.0523243788524743e-06,
+      "loss": 0.465,
+      "step": 20658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.539128748729058,
+      "learning_rate": 1.0522254249510862e-06,
+      "loss": 0.3767,
+      "step": 20659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.651862086902923,
+      "learning_rate": 1.052126564161096e-06,
+      "loss": 0.519,
+      "step": 20660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.753152715274175,
+      "learning_rate": 1.0520277964845732e-06,
+      "loss": 0.3558,
+      "step": 20661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5221187530087255,
+      "learning_rate": 1.0519291219235864e-06,
+      "loss": 0.3677,
+      "step": 20662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.876911127316757,
+      "learning_rate": 1.0518305404802002e-06,
+      "loss": 0.4349,
+      "step": 20663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9868567823875327,
+      "learning_rate": 1.0517320521564786e-06,
+      "loss": 0.5106,
+      "step": 20664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9393989417824509,
+      "learning_rate": 1.0516336569544839e-06,
+      "loss": 0.4155,
+      "step": 20665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.041063338195175,
+      "learning_rate": 1.0515353548762754e-06,
+      "loss": 0.4758,
+      "step": 20666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7879599644447985,
+      "learning_rate": 1.0514371459239115e-06,
+      "loss": 0.5047,
+      "step": 20667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1068735149860867,
+      "learning_rate": 1.0513390300994473e-06,
+      "loss": 0.2173,
+      "step": 20668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.645461124901963,
+      "learning_rate": 1.051241007404938e-06,
+      "loss": 0.4329,
+      "step": 20669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4698777171556523,
+      "learning_rate": 1.0511430778424343e-06,
+      "loss": 0.4328,
+      "step": 20670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3984678056199928,
+      "learning_rate": 1.0510452414139876e-06,
+      "loss": 0.3168,
+      "step": 20671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7238769232718067,
+      "learning_rate": 1.0509474981216453e-06,
+      "loss": 0.4632,
+      "step": 20672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8299475156004368,
+      "learning_rate": 1.0508498479674535e-06,
+      "loss": 0.5657,
+      "step": 20673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.88053356503194,
+      "learning_rate": 1.0507522909534564e-06,
+      "loss": 0.438,
+      "step": 20674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.847773085964153,
+      "learning_rate": 1.050654827081697e-06,
+      "loss": 0.5255,
+      "step": 20675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0571212123284774,
+      "learning_rate": 1.050557456354215e-06,
+      "loss": 0.5082,
+      "step": 20676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.005930595694967,
+      "learning_rate": 1.0504601787730494e-06,
+      "loss": 0.4845,
+      "step": 20677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.020881703303283,
+      "learning_rate": 1.050362994340236e-06,
+      "loss": 0.4737,
+      "step": 20678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8082168145378001,
+      "learning_rate": 1.0502659030578092e-06,
+      "loss": 0.5144,
+      "step": 20679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6657326807111792,
+      "learning_rate": 1.0501689049278022e-06,
+      "loss": 0.4753,
+      "step": 20680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.071673145598803,
+      "learning_rate": 1.0500719999522457e-06,
+      "loss": 0.4899,
+      "step": 20681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.968863152639135,
+      "learning_rate": 1.0499751881331674e-06,
+      "loss": 0.5694,
+      "step": 20682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8006374645873626,
+      "learning_rate": 1.0498784694725944e-06,
+      "loss": 0.5365,
+      "step": 20683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4749020269191875,
+      "learning_rate": 1.0497818439725518e-06,
+      "loss": 0.3668,
+      "step": 20684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7926476614001101,
+      "learning_rate": 1.0496853116350622e-06,
+      "loss": 0.5162,
+      "step": 20685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.645276006234231,
+      "learning_rate": 1.0495888724621462e-06,
+      "loss": 0.4292,
+      "step": 20686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9575136105067092,
+      "learning_rate": 1.049492526455823e-06,
+      "loss": 0.3693,
+      "step": 20687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6786865361950771,
+      "learning_rate": 1.04939627361811e-06,
+      "loss": 0.4536,
+      "step": 20688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9161492966892253,
+      "learning_rate": 1.049300113951021e-06,
+      "loss": 0.4282,
+      "step": 20689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0236787016776976,
+      "learning_rate": 1.0492040474565696e-06,
+      "loss": 0.5015,
+      "step": 20690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8038135706645855,
+      "learning_rate": 1.0491080741367671e-06,
+      "loss": 0.605,
+      "step": 20691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6734181563652128,
+      "learning_rate": 1.0490121939936227e-06,
+      "loss": 0.4579,
+      "step": 20692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.731345359927361,
+      "learning_rate": 1.0489164070291436e-06,
+      "loss": 0.3989,
+      "step": 20693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8107880314434464,
+      "learning_rate": 1.0488207132453345e-06,
+      "loss": 0.4069,
+      "step": 20694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8867918132117634,
+      "learning_rate": 1.048725112644199e-06,
+      "loss": 0.4899,
+      "step": 20695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8081852556575093,
+      "learning_rate": 1.0486296052277387e-06,
+      "loss": 0.4773,
+      "step": 20696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8460802360011774,
+      "learning_rate": 1.0485341909979526e-06,
+      "loss": 0.6224,
+      "step": 20697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8065842100532787,
+      "learning_rate": 1.0484388699568381e-06,
+      "loss": 0.4268,
+      "step": 20698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8055012385946347,
+      "learning_rate": 1.0483436421063915e-06,
+      "loss": 0.4616,
+      "step": 20699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8829708417061302,
+      "learning_rate": 1.048248507448605e-06,
+      "loss": 0.5272,
+      "step": 20700
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.48141786456108093,
+      "eval_runtime": 67.9907,
+      "eval_samples_per_second": 50.154,
+      "eval_steps_per_second": 1.574,
+      "step": 20700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.752170716780441,
+      "learning_rate": 1.0481534659854714e-06,
+      "loss": 0.5056,
+      "step": 20701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6819377079690194,
+      "learning_rate": 1.0480585177189797e-06,
+      "loss": 0.3441,
+      "step": 20702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6046412546877884,
+      "learning_rate": 1.0479636626511181e-06,
+      "loss": 0.3335,
+      "step": 20703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7194945404140474,
+      "learning_rate": 1.0478689007838714e-06,
+      "loss": 0.4003,
+      "step": 20704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6660977753091808,
+      "learning_rate": 1.0477742321192238e-06,
+      "loss": 0.4344,
+      "step": 20705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9113748535703623,
+      "learning_rate": 1.0476796566591575e-06,
+      "loss": 0.4539,
+      "step": 20706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8087558192103423,
+      "learning_rate": 1.0475851744056525e-06,
+      "loss": 0.4284,
+      "step": 20707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7636040655850485,
+      "learning_rate": 1.047490785360686e-06,
+      "loss": 0.4446,
+      "step": 20708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.518766154522957,
+      "learning_rate": 1.0473964895262349e-06,
+      "loss": 0.4636,
+      "step": 20709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1365380170442463,
+      "learning_rate": 1.0473022869042722e-06,
+      "loss": 0.4746,
+      "step": 20710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9949278857601536,
+      "learning_rate": 1.0472081774967702e-06,
+      "loss": 0.4513,
+      "step": 20711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7942932629412867,
+      "learning_rate": 1.0471141613057e-06,
+      "loss": 0.5582,
+      "step": 20712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5608316133317341,
+      "learning_rate": 1.0470202383330286e-06,
+      "loss": 0.403,
+      "step": 20713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0329098461777892,
+      "learning_rate": 1.046926408580723e-06,
+      "loss": 0.7448,
+      "step": 20714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7330731558565045,
+      "learning_rate": 1.0468326720507464e-06,
+      "loss": 0.6191,
+      "step": 20715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.816242818881568,
+      "learning_rate": 1.0467390287450627e-06,
+      "loss": 0.5751,
+      "step": 20716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.028589811104692,
+      "learning_rate": 1.0466454786656305e-06,
+      "loss": 0.4995,
+      "step": 20717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8603853228554366,
+      "learning_rate": 1.0465520218144097e-06,
+      "loss": 0.4638,
+      "step": 20718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9147691122153914,
+      "learning_rate": 1.046458658193356e-06,
+      "loss": 0.4528,
+      "step": 20719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.86260653273426,
+      "learning_rate": 1.0463653878044246e-06,
+      "loss": 0.4137,
+      "step": 20720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4884386010362827,
+      "learning_rate": 1.046272210649567e-06,
+      "loss": 0.4308,
+      "step": 20721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4058270002368227,
+      "learning_rate": 1.0461791267307344e-06,
+      "loss": 0.4372,
+      "step": 20722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8101431821040836,
+      "learning_rate": 1.046086136049876e-06,
+      "loss": 0.5083,
+      "step": 20723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2278532911417033,
+      "learning_rate": 1.0459932386089371e-06,
+      "loss": 0.2698,
+      "step": 20724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7855947853596388,
+      "learning_rate": 1.0459004344098637e-06,
+      "loss": 0.4747,
+      "step": 20725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6837852741802017,
+      "learning_rate": 1.0458077234545983e-06,
+      "loss": 0.4362,
+      "step": 20726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7817193684818635,
+      "learning_rate": 1.0457151057450812e-06,
+      "loss": 0.4159,
+      "step": 20727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7998294271972237,
+      "learning_rate": 1.045622581283252e-06,
+      "loss": 0.4796,
+      "step": 20728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.228419349002652,
+      "learning_rate": 1.0455301500710473e-06,
+      "loss": 0.4848,
+      "step": 20729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0974858691616984,
+      "learning_rate": 1.0454378121104016e-06,
+      "loss": 0.5358,
+      "step": 20730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8006032233572593,
+      "learning_rate": 1.0453455674032489e-06,
+      "loss": 0.6706,
+      "step": 20731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7878480933313372,
+      "learning_rate": 1.04525341595152e-06,
+      "loss": 0.4788,
+      "step": 20732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7070162701285172,
+      "learning_rate": 1.045161357757144e-06,
+      "loss": 0.3885,
+      "step": 20733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8247554742355263,
+      "learning_rate": 1.0450693928220477e-06,
+      "loss": 0.4776,
+      "step": 20734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.158019367180059,
+      "learning_rate": 1.0449775211481563e-06,
+      "loss": 0.4227,
+      "step": 20735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9202012437437384,
+      "learning_rate": 1.044885742737394e-06,
+      "loss": 0.5115,
+      "step": 20736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4576765269009557,
+      "learning_rate": 1.0447940575916816e-06,
+      "loss": 0.3661,
+      "step": 20737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9251383381435634,
+      "learning_rate": 1.0447024657129375e-06,
+      "loss": 0.5936,
+      "step": 20738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6049741974812168,
+      "learning_rate": 1.0446109671030807e-06,
+      "loss": 0.4659,
+      "step": 20739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4942803321015756,
+      "learning_rate": 1.0445195617640254e-06,
+      "loss": 0.3612,
+      "step": 20740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7291103602429625,
+      "learning_rate": 1.0444282496976864e-06,
+      "loss": 0.4684,
+      "step": 20741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.802977973568266,
+      "learning_rate": 1.0443370309059741e-06,
+      "loss": 0.371,
+      "step": 20742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9841601999751468,
+      "learning_rate": 1.0442459053907986e-06,
+      "loss": 0.5804,
+      "step": 20743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.910732269386776,
+      "learning_rate": 1.0441548731540674e-06,
+      "loss": 0.5664,
+      "step": 20744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8633070574448203,
+      "learning_rate": 1.0440639341976866e-06,
+      "loss": 0.591,
+      "step": 20745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0320980234551964,
+      "learning_rate": 1.0439730885235594e-06,
+      "loss": 0.4556,
+      "step": 20746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.764842731351679,
+      "learning_rate": 1.0438823361335878e-06,
+      "loss": 0.4632,
+      "step": 20747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.738648633879679,
+      "learning_rate": 1.0437916770296718e-06,
+      "loss": 0.5407,
+      "step": 20748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9816629466849143,
+      "learning_rate": 1.0437011112137093e-06,
+      "loss": 0.5661,
+      "step": 20749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7021146673215277,
+      "learning_rate": 1.0436106386875958e-06,
+      "loss": 0.3297,
+      "step": 20750
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.4810793101787567,
+      "eval_runtime": 67.966,
+      "eval_samples_per_second": 50.172,
+      "eval_steps_per_second": 1.574,
+      "step": 20750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6137278497738403,
+      "learning_rate": 1.0435202594532259e-06,
+      "loss": 0.4276,
+      "step": 20751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.887001047111621,
+      "learning_rate": 1.0434299735124916e-06,
+      "loss": 0.5111,
+      "step": 20752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6854653163699889,
+      "learning_rate": 1.0433397808672818e-06,
+      "loss": 0.4942,
+      "step": 20753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9070650249762189,
+      "learning_rate": 1.0432496815194863e-06,
+      "loss": 0.492,
+      "step": 20754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8016101946011907,
+      "learning_rate": 1.0431596754709905e-06,
+      "loss": 0.4853,
+      "step": 20755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7363147290489527,
+      "learning_rate": 1.0430697627236785e-06,
+      "loss": 0.4303,
+      "step": 20756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.544437473084116,
+      "learning_rate": 1.0429799432794328e-06,
+      "loss": 0.4347,
+      "step": 20757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.882952564628365,
+      "learning_rate": 1.0428902171401337e-06,
+      "loss": 0.4819,
+      "step": 20758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2849380074037833,
+      "learning_rate": 1.0428005843076595e-06,
+      "loss": 0.3371,
+      "step": 20759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2873402570495127,
+      "learning_rate": 1.042711044783886e-06,
+      "loss": 0.5309,
+      "step": 20760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8853483850294377,
+      "learning_rate": 1.0426215985706888e-06,
+      "loss": 0.5574,
+      "step": 20761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4513696871965944,
+      "learning_rate": 1.04253224566994e-06,
+      "loss": 0.3807,
+      "step": 20762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.645454091261572,
+      "learning_rate": 1.0424429860835099e-06,
+      "loss": 0.3909,
+      "step": 20763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6458475955823528,
+      "learning_rate": 1.0423538198132669e-06,
+      "loss": 0.3828,
+      "step": 20764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8224618052225876,
+      "learning_rate": 1.0422647468610784e-06,
+      "loss": 0.5935,
+      "step": 20765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8517488389029528,
+      "learning_rate": 1.0421757672288082e-06,
+      "loss": 0.4429,
+      "step": 20766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7163421196041901,
+      "learning_rate": 1.0420868809183197e-06,
+      "loss": 0.4282,
+      "step": 20767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5850121033210485,
+      "learning_rate": 1.041998087931473e-06,
+      "loss": 0.3746,
+      "step": 20768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8033561648664345,
+      "learning_rate": 1.0419093882701276e-06,
+      "loss": 0.4025,
+      "step": 20769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7983003832926703,
+      "learning_rate": 1.0418207819361402e-06,
+      "loss": 0.5348,
+      "step": 20770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6534167017378543,
+      "learning_rate": 1.0417322689313662e-06,
+      "loss": 0.4377,
+      "step": 20771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.16751991649017,
+      "learning_rate": 1.0416438492576573e-06,
+      "loss": 0.4192,
+      "step": 20772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6534703835216815,
+      "learning_rate": 1.0415555229168654e-06,
+      "loss": 0.4064,
+      "step": 20773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0039131600861535,
+      "learning_rate": 1.0414672899108394e-06,
+      "loss": 0.4714,
+      "step": 20774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6908088750622472,
+      "learning_rate": 1.0413791502414265e-06,
+      "loss": 0.4563,
+      "step": 20775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.01399254139174,
+      "learning_rate": 1.0412911039104714e-06,
+      "loss": 0.4385,
+      "step": 20776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.644756040225722,
+      "learning_rate": 1.041203150919818e-06,
+      "loss": 0.4132,
+      "step": 20777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.037174233174046,
+      "learning_rate": 1.0411152912713072e-06,
+      "loss": 0.6155,
+      "step": 20778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7929990687419488,
+      "learning_rate": 1.0410275249667777e-06,
+      "loss": 0.354,
+      "step": 20779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6763073570342566,
+      "learning_rate": 1.0409398520080678e-06,
+      "loss": 0.528,
+      "step": 20780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.434903324822957,
+      "learning_rate": 1.040852272397013e-06,
+      "loss": 0.5563,
+      "step": 20781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.820265308893902,
+      "learning_rate": 1.0407647861354452e-06,
+      "loss": 0.5808,
+      "step": 20782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1221885539786047,
+      "learning_rate": 1.0406773932251974e-06,
+      "loss": 0.5584,
+      "step": 20783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4213354934914597,
+      "learning_rate": 1.0405900936680985e-06,
+      "loss": 0.6564,
+      "step": 20784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7412550180969684,
+      "learning_rate": 1.0405028874659764e-06,
+      "loss": 0.4587,
+      "step": 20785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0786804565162558,
+      "learning_rate": 1.0404157746206563e-06,
+      "loss": 0.4513,
+      "step": 20786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6451039972641475,
+      "learning_rate": 1.0403287551339617e-06,
+      "loss": 0.5188,
+      "step": 20787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9871493172659094,
+      "learning_rate": 1.0402418290077148e-06,
+      "loss": 0.5343,
+      "step": 20788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8104700359813612,
+      "learning_rate": 1.0401549962437352e-06,
+      "loss": 0.5587,
+      "step": 20789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4931558815774453,
+      "learning_rate": 1.0400682568438403e-06,
+      "loss": 0.4671,
+      "step": 20790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5702382536958936,
+      "learning_rate": 1.0399816108098473e-06,
+      "loss": 0.3715,
+      "step": 20791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2508167404180415,
+      "learning_rate": 1.0398950581435681e-06,
+      "loss": 0.5216,
+      "step": 20792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6004643548483748,
+      "learning_rate": 1.0398085988468157e-06,
+      "loss": 0.4417,
+      "step": 20793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.862875895305218,
+      "learning_rate": 1.0397222329214003e-06,
+      "loss": 0.4349,
+      "step": 20794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5258755445074783,
+      "learning_rate": 1.039635960369129e-06,
+      "loss": 0.3955,
+      "step": 20795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9828458403982783,
+      "learning_rate": 1.039549781191809e-06,
+      "loss": 0.4971,
+      "step": 20796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8241510061657562,
+      "learning_rate": 1.039463695391244e-06,
+      "loss": 0.5299,
+      "step": 20797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9369786944458272,
+      "learning_rate": 1.0393777029692358e-06,
+      "loss": 0.496,
+      "step": 20798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4283732990429492,
+      "learning_rate": 1.0392918039275845e-06,
+      "loss": 0.3606,
+      "step": 20799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.533552117789291,
+      "learning_rate": 1.0392059982680888e-06,
+      "loss": 0.3929,
+      "step": 20800
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.4808432161808014,
+      "eval_runtime": 68.0732,
+      "eval_samples_per_second": 50.093,
+      "eval_steps_per_second": 1.572,
+      "step": 20800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6788254615811566,
+      "learning_rate": 1.0391202859925449e-06,
+      "loss": 0.3125,
+      "step": 20801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8008020944219276,
+      "learning_rate": 1.0390346671027468e-06,
+      "loss": 0.4771,
+      "step": 20802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6647666704196833,
+      "learning_rate": 1.038949141600487e-06,
+      "loss": 0.3663,
+      "step": 20803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8544320095380817,
+      "learning_rate": 1.0388637094875565e-06,
+      "loss": 0.3952,
+      "step": 20804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7876099049748182,
+      "learning_rate": 1.0387783707657428e-06,
+      "loss": 0.519,
+      "step": 20805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8871800359628492,
+      "learning_rate": 1.0386931254368336e-06,
+      "loss": 0.4305,
+      "step": 20806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0203204029646327,
+      "learning_rate": 1.0386079735026123e-06,
+      "loss": 0.5082,
+      "step": 20807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.831177131070864,
+      "learning_rate": 1.0385229149648622e-06,
+      "loss": 0.62,
+      "step": 20808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.792350629578783,
+      "learning_rate": 1.0384379498253641e-06,
+      "loss": 0.5423,
+      "step": 20809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9318754414803163,
+      "learning_rate": 1.0383530780858963e-06,
+      "loss": 0.5375,
+      "step": 20810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5680647738598066,
+      "learning_rate": 1.0382682997482352e-06,
+      "loss": 0.3013,
+      "step": 20811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1013688521264378,
+      "learning_rate": 1.038183614814156e-06,
+      "loss": 0.5648,
+      "step": 20812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.540308458624272,
+      "learning_rate": 1.0380990232854313e-06,
+      "loss": 0.3978,
+      "step": 20813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.752311465085952,
+      "learning_rate": 1.0380145251638326e-06,
+      "loss": 0.3544,
+      "step": 20814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7706297634428103,
+      "learning_rate": 1.0379301204511279e-06,
+      "loss": 0.4357,
+      "step": 20815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9643558894343227,
+      "learning_rate": 1.037845809149085e-06,
+      "loss": 0.5117,
+      "step": 20816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.571511108541465,
+      "learning_rate": 1.0377615912594684e-06,
+      "loss": 0.356,
+      "step": 20817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0081041358369593,
+      "learning_rate": 1.037677466784041e-06,
+      "loss": 0.5115,
+      "step": 20818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.590341202458246,
+      "learning_rate": 1.037593435724565e-06,
+      "loss": 0.3937,
+      "step": 20819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0124818211161744,
+      "learning_rate": 1.0375094980827983e-06,
+      "loss": 0.5299,
+      "step": 20820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3535190991880983,
+      "learning_rate": 1.0374256538604982e-06,
+      "loss": 0.4098,
+      "step": 20821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6156999876606073,
+      "learning_rate": 1.0373419030594205e-06,
+      "loss": 0.3774,
+      "step": 20822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8334862710404114,
+      "learning_rate": 1.0372582456813186e-06,
+      "loss": 0.381,
+      "step": 20823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5234675897681305,
+      "learning_rate": 1.0371746817279426e-06,
+      "loss": 0.3662,
+      "step": 20824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9250005110907478,
+      "learning_rate": 1.0370912112010432e-06,
+      "loss": 0.4,
+      "step": 20825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.822166738675792,
+      "learning_rate": 1.0370078341023671e-06,
+      "loss": 0.4729,
+      "step": 20826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.189175435961895,
+      "learning_rate": 1.03692455043366e-06,
+      "loss": 0.5441,
+      "step": 20827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7188065410784816,
+      "learning_rate": 1.0368413601966655e-06,
+      "loss": 0.5123,
+      "step": 20828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8962340062760268,
+      "learning_rate": 1.0367582633931248e-06,
+      "loss": 0.4693,
+      "step": 20829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7100824422846577,
+      "learning_rate": 1.0366752600247774e-06,
+      "loss": 0.4363,
+      "step": 20830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6276368040490758,
+      "learning_rate": 1.0365923500933614e-06,
+      "loss": 0.5669,
+      "step": 20831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0137002940981206,
+      "learning_rate": 1.036509533600612e-06,
+      "loss": 0.6309,
+      "step": 20832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5776783098384384,
+      "learning_rate": 1.0364268105482632e-06,
+      "loss": 0.4821,
+      "step": 20833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.849434503311982,
+      "learning_rate": 1.0363441809380467e-06,
+      "loss": 0.5892,
+      "step": 20834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9100293315655548,
+      "learning_rate": 1.0362616447716927e-06,
+      "loss": 0.7133,
+      "step": 20835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6837609194330374,
+      "learning_rate": 1.036179202050928e-06,
+      "loss": 0.4258,
+      "step": 20836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2412337745015773,
+      "learning_rate": 1.0360968527774798e-06,
+      "loss": 0.6656,
+      "step": 20837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8996480153685058,
+      "learning_rate": 1.036014596953071e-06,
+      "loss": 0.4613,
+      "step": 20838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6055961091768784,
+      "learning_rate": 1.0359324345794237e-06,
+      "loss": 0.4988,
+      "step": 20839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6655676047253707,
+      "learning_rate": 1.0358503656582583e-06,
+      "loss": 0.3954,
+      "step": 20840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5641387648098855,
+      "learning_rate": 1.0357683901912929e-06,
+      "loss": 0.2657,
+      "step": 20841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.60801713802499,
+      "learning_rate": 1.0356865081802434e-06,
+      "loss": 0.4832,
+      "step": 20842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5535020300266875,
+      "learning_rate": 1.0356047196268238e-06,
+      "loss": 0.4773,
+      "step": 20843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.594857078089441,
+      "learning_rate": 1.0355230245327464e-06,
+      "loss": 0.4134,
+      "step": 20844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.010323628094032,
+      "learning_rate": 1.035441422899722e-06,
+      "loss": 0.4446,
+      "step": 20845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6328726140289938,
+      "learning_rate": 1.035359914729458e-06,
+      "loss": 0.4662,
+      "step": 20846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.743809471899611,
+      "learning_rate": 1.0352785000236608e-06,
+      "loss": 0.5183,
+      "step": 20847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7213629398751333,
+      "learning_rate": 1.035197178784035e-06,
+      "loss": 0.414,
+      "step": 20848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7864690101832026,
+      "learning_rate": 1.0351159510122839e-06,
+      "loss": 0.4875,
+      "step": 20849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8582312983063554,
+      "learning_rate": 1.0350348167101066e-06,
+      "loss": 0.5048,
+      "step": 20850
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.4811786413192749,
+      "eval_runtime": 68.0225,
+      "eval_samples_per_second": 50.13,
+      "eval_steps_per_second": 1.573,
+      "step": 20850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6348805027946327,
+      "learning_rate": 1.0349537758792022e-06,
+      "loss": 0.4477,
+      "step": 20851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0814317336719568,
+      "learning_rate": 1.0348728285212663e-06,
+      "loss": 0.521,
+      "step": 20852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6177787333068256,
+      "learning_rate": 1.0347919746379953e-06,
+      "loss": 0.4166,
+      "step": 20853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.089230150574542,
+      "learning_rate": 1.0347112142310808e-06,
+      "loss": 0.7241,
+      "step": 20854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3989340949635005,
+      "learning_rate": 1.0346305473022137e-06,
+      "loss": 0.387,
+      "step": 20855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.040308819827654,
+      "learning_rate": 1.034549973853082e-06,
+      "loss": 0.5281,
+      "step": 20856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4750603136034957,
+      "learning_rate": 1.0344694938853735e-06,
+      "loss": 0.648,
+      "step": 20857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6145907080833335,
+      "learning_rate": 1.0343891074007726e-06,
+      "loss": 0.5613,
+      "step": 20858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1557174757566364,
+      "learning_rate": 1.0343088144009617e-06,
+      "loss": 0.5128,
+      "step": 20859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.527438113127247,
+      "learning_rate": 1.0342286148876217e-06,
+      "loss": 0.3669,
+      "step": 20860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.41577305970425,
+      "learning_rate": 1.0341485088624326e-06,
+      "loss": 0.4405,
+      "step": 20861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.760877116194124,
+      "learning_rate": 1.03406849632707e-06,
+      "loss": 0.4662,
+      "step": 20862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7645516488852764,
+      "learning_rate": 1.0339885772832096e-06,
+      "loss": 0.5269,
+      "step": 20863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.184838898435426,
+      "learning_rate": 1.0339087517325252e-06,
+      "loss": 0.5754,
+      "step": 20864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1490150229501195,
+      "learning_rate": 1.033829019676686e-06,
+      "loss": 0.5776,
+      "step": 20865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.863236741512616,
+      "learning_rate": 1.0337493811173628e-06,
+      "loss": 0.5655,
+      "step": 20866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7493850778207933,
+      "learning_rate": 1.0336698360562225e-06,
+      "loss": 0.5543,
+      "step": 20867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7378015386979118,
+      "learning_rate": 1.0335903844949292e-06,
+      "loss": 0.4376,
+      "step": 20868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.429563751332662,
+      "learning_rate": 1.0335110264351478e-06,
+      "loss": 0.5643,
+      "step": 20869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.243938636712895,
+      "learning_rate": 1.0334317618785384e-06,
+      "loss": 0.5608,
+      "step": 20870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.660194020007436,
+      "learning_rate": 1.033352590826761e-06,
+      "loss": 0.3612,
+      "step": 20871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.221205431726421,
+      "learning_rate": 1.0332735132814725e-06,
+      "loss": 0.5418,
+      "step": 20872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4952163410401746,
+      "learning_rate": 1.0331945292443284e-06,
+      "loss": 0.4577,
+      "step": 20873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.632117263549485,
+      "learning_rate": 1.033115638716983e-06,
+      "loss": 0.5063,
+      "step": 20874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6609201309749317,
+      "learning_rate": 1.0330368417010869e-06,
+      "loss": 0.5165,
+      "step": 20875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6158220609674059,
+      "learning_rate": 1.0329581381982899e-06,
+      "loss": 0.374,
+      "step": 20876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1599397648114538,
+      "learning_rate": 1.0328795282102405e-06,
+      "loss": 0.4443,
+      "step": 20877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9804696625426554,
+      "learning_rate": 1.0328010117385825e-06,
+      "loss": 0.471,
+      "step": 20878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8925258052439906,
+      "learning_rate": 1.0327225887849612e-06,
+      "loss": 0.5096,
+      "step": 20879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6562011972617894,
+      "learning_rate": 1.0326442593510174e-06,
+      "loss": 0.5156,
+      "step": 20880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1505442313524257,
+      "learning_rate": 1.0325660234383916e-06,
+      "loss": 0.5552,
+      "step": 20881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1869693996846755,
+      "learning_rate": 1.0324878810487212e-06,
+      "loss": 0.5176,
+      "step": 20882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7931768859064732,
+      "learning_rate": 1.0324098321836418e-06,
+      "loss": 0.4384,
+      "step": 20883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1031717004272275,
+      "learning_rate": 1.0323318768447882e-06,
+      "loss": 0.4702,
+      "step": 20884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5184410739665368,
+      "learning_rate": 1.0322540150337915e-06,
+      "loss": 0.4006,
+      "step": 20885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6861193007943143,
+      "learning_rate": 1.032176246752282e-06,
+      "loss": 0.4661,
+      "step": 20886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0814188242461835,
+      "learning_rate": 1.0320985720018878e-06,
+      "loss": 0.6412,
+      "step": 20887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9075528519498375,
+      "learning_rate": 1.0320209907842347e-06,
+      "loss": 0.4918,
+      "step": 20888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.46807693140752,
+      "learning_rate": 1.0319435031009468e-06,
+      "loss": 0.2942,
+      "step": 20889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.108116995589857,
+      "learning_rate": 1.031866108953647e-06,
+      "loss": 0.6236,
+      "step": 20890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5668802742089742,
+      "learning_rate": 1.0317888083439543e-06,
+      "loss": 0.4753,
+      "step": 20891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.213717758593989,
+      "learning_rate": 1.031711601273488e-06,
+      "loss": 0.6801,
+      "step": 20892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.476222439792471,
+      "learning_rate": 1.031634487743864e-06,
+      "loss": 0.3797,
+      "step": 20893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3243927383875804,
+      "learning_rate": 1.0315574677566966e-06,
+      "loss": 0.5911,
+      "step": 20894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0033322887746405,
+      "learning_rate": 1.0314805413135979e-06,
+      "loss": 0.5398,
+      "step": 20895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7283656689954547,
+      "learning_rate": 1.0314037084161787e-06,
+      "loss": 0.3766,
+      "step": 20896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8134683823299176,
+      "learning_rate": 1.0313269690660477e-06,
+      "loss": 0.3636,
+      "step": 20897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9608535226660624,
+      "learning_rate": 1.0312503232648105e-06,
+      "loss": 0.586,
+      "step": 20898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4419248057846992,
+      "learning_rate": 1.031173771014072e-06,
+      "loss": 0.3841,
+      "step": 20899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.917420163605719,
+      "learning_rate": 1.0310973123154357e-06,
+      "loss": 0.5819,
+      "step": 20900
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.4810943007469177,
+      "eval_runtime": 68.0205,
+      "eval_samples_per_second": 50.132,
+      "eval_steps_per_second": 1.573,
+      "step": 20900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.49824631506257,
+      "learning_rate": 1.0310209471705006e-06,
+      "loss": 0.4,
+      "step": 20901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.663218270123576,
+      "learning_rate": 1.0309446755808665e-06,
+      "loss": 0.4402,
+      "step": 20902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7028921078689774,
+      "learning_rate": 1.0308684975481295e-06,
+      "loss": 0.4323,
+      "step": 20903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8786463444643133,
+      "learning_rate": 1.0307924130738853e-06,
+      "loss": 0.4409,
+      "step": 20904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7509108573984722,
+      "learning_rate": 1.0307164221597254e-06,
+      "loss": 0.438,
+      "step": 20905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.657406650415388,
+      "learning_rate": 1.0306405248072416e-06,
+      "loss": 0.4703,
+      "step": 20906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0235335484049486,
+      "learning_rate": 1.030564721018022e-06,
+      "loss": 0.4309,
+      "step": 20907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6890409751748712,
+      "learning_rate": 1.0304890107936543e-06,
+      "loss": 0.538,
+      "step": 20908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6677603101470473,
+      "learning_rate": 1.030413394135723e-06,
+      "loss": 0.5374,
+      "step": 20909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4315049411351395,
+      "learning_rate": 1.030337871045811e-06,
+      "loss": 0.35,
+      "step": 20910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9087481417831043,
+      "learning_rate": 1.0302624415254994e-06,
+      "loss": 0.4897,
+      "step": 20911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5287141371960675,
+      "learning_rate": 1.0301871055763678e-06,
+      "loss": 0.4424,
+      "step": 20912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5706131248676307,
+      "learning_rate": 1.0301118631999924e-06,
+      "loss": 0.4521,
+      "step": 20913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8335340469468266,
+      "learning_rate": 1.0300367143979486e-06,
+      "loss": 0.3885,
+      "step": 20914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4847107028320508,
+      "learning_rate": 1.02996165917181e-06,
+      "loss": 0.4378,
+      "step": 20915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1410112730142883,
+      "learning_rate": 1.0298866975231477e-06,
+      "loss": 0.5047,
+      "step": 20916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5634229046814963,
+      "learning_rate": 1.0298118294535308e-06,
+      "loss": 0.4332,
+      "step": 20917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.824235504327014,
+      "learning_rate": 1.0297370549645266e-06,
+      "loss": 0.4485,
+      "step": 20918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8026025531807328,
+      "learning_rate": 1.0296623740577011e-06,
+      "loss": 0.5447,
+      "step": 20919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.524730526000842,
+      "learning_rate": 1.0295877867346167e-06,
+      "loss": 0.3889,
+      "step": 20920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5979308151272968,
+      "learning_rate": 1.0295132929968354e-06,
+      "loss": 0.3638,
+      "step": 20921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6232853960094435,
+      "learning_rate": 1.0294388928459167e-06,
+      "loss": 0.5782,
+      "step": 20922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4945610089753456,
+      "learning_rate": 1.0293645862834182e-06,
+      "loss": 0.3568,
+      "step": 20923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0530015348074477,
+      "learning_rate": 1.029290373310895e-06,
+      "loss": 0.6249,
+      "step": 20924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5314298246327196,
+      "learning_rate": 1.0292162539299005e-06,
+      "loss": 0.3725,
+      "step": 20925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7354272963329824,
+      "learning_rate": 1.0291422281419876e-06,
+      "loss": 0.5005,
+      "step": 20926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7873816662979933,
+      "learning_rate": 1.0290682959487047e-06,
+      "loss": 0.3613,
+      "step": 20927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7277477495284652,
+      "learning_rate": 1.0289944573516001e-06,
+      "loss": 0.4974,
+      "step": 20928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.095467827931123,
+      "learning_rate": 1.02892071235222e-06,
+      "loss": 0.6036,
+      "step": 20929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7091303437072582,
+      "learning_rate": 1.0288470609521072e-06,
+      "loss": 0.4228,
+      "step": 20930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4406750358821312,
+      "learning_rate": 1.0287735031528042e-06,
+      "loss": 0.3637,
+      "step": 20931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9652809180925956,
+      "learning_rate": 1.0287000389558503e-06,
+      "loss": 0.4409,
+      "step": 20932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6846736469084302,
+      "learning_rate": 1.0286266683627842e-06,
+      "loss": 0.3302,
+      "step": 20933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9099412518259682,
+      "learning_rate": 1.0285533913751416e-06,
+      "loss": 0.5122,
+      "step": 20934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7633976430374134,
+      "learning_rate": 1.0284802079944566e-06,
+      "loss": 0.4016,
+      "step": 20935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0128052928421036,
+      "learning_rate": 1.028407118222261e-06,
+      "loss": 0.5668,
+      "step": 20936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.062401387356963,
+      "learning_rate": 1.0283341220600848e-06,
+      "loss": 0.6655,
+      "step": 20937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7499695872452106,
+      "learning_rate": 1.0282612195094562e-06,
+      "loss": 0.3964,
+      "step": 20938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6857762254395667,
+      "learning_rate": 1.0281884105719019e-06,
+      "loss": 0.4916,
+      "step": 20939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.096832600207777,
+      "learning_rate": 1.028115695248945e-06,
+      "loss": 0.5985,
+      "step": 20940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4938936818824744,
+      "learning_rate": 1.0280430735421087e-06,
+      "loss": 0.4001,
+      "step": 20941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6118079204064126,
+      "learning_rate": 1.0279705454529133e-06,
+      "loss": 0.3249,
+      "step": 20942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.841464230286301,
+      "learning_rate": 1.0278981109828768e-06,
+      "loss": 0.4701,
+      "step": 20943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5179053324447906,
+      "learning_rate": 1.0278257701335155e-06,
+      "loss": 0.3949,
+      "step": 20944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7856878479675178,
+      "learning_rate": 1.0277535229063442e-06,
+      "loss": 0.481,
+      "step": 20945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.021004582997011,
+      "learning_rate": 1.0276813693028748e-06,
+      "loss": 0.4098,
+      "step": 20946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.234668176208467,
+      "learning_rate": 1.0276093093246179e-06,
+      "loss": 0.4056,
+      "step": 20947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.945543137005742,
+      "learning_rate": 1.0275373429730827e-06,
+      "loss": 0.473,
+      "step": 20948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.036948130474932,
+      "learning_rate": 1.0274654702497748e-06,
+      "loss": 0.5104,
+      "step": 20949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8032129511876844,
+      "learning_rate": 1.027393691156199e-06,
+      "loss": 0.5678,
+      "step": 20950
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.4811457395553589,
+      "eval_runtime": 68.0117,
+      "eval_samples_per_second": 50.138,
+      "eval_steps_per_second": 1.573,
+      "step": 20950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6379060118409128,
+      "learning_rate": 1.0273220056938587e-06,
+      "loss": 0.5452,
+      "step": 20951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7802800093211293,
+      "learning_rate": 1.027250413864254e-06,
+      "loss": 0.4337,
+      "step": 20952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6468866878581288,
+      "learning_rate": 1.027178915668884e-06,
+      "loss": 0.4978,
+      "step": 20953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0734520537684547,
+      "learning_rate": 1.0271075111092453e-06,
+      "loss": 0.4667,
+      "step": 20954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7148501896951405,
+      "learning_rate": 1.0270362001868323e-06,
+      "loss": 0.5752,
+      "step": 20955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7068415168231494,
+      "learning_rate": 1.0269649829031384e-06,
+      "loss": 0.4708,
+      "step": 20956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.068856440703704,
+      "learning_rate": 1.0268938592596543e-06,
+      "loss": 0.5829,
+      "step": 20957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5775092110777564,
+      "learning_rate": 1.026822829257869e-06,
+      "loss": 0.372,
+      "step": 20958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9524522062144523,
+      "learning_rate": 1.0267518928992693e-06,
+      "loss": 0.375,
+      "step": 20959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7781259708655812,
+      "learning_rate": 1.0266810501853403e-06,
+      "loss": 0.4337,
+      "step": 20960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.716112734461572,
+      "learning_rate": 1.0266103011175644e-06,
+      "loss": 0.385,
+      "step": 20961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7884621828238798,
+      "learning_rate": 1.0265396456974238e-06,
+      "loss": 0.5713,
+      "step": 20962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.834137632488319,
+      "learning_rate": 1.0264690839263972e-06,
+      "loss": 0.3931,
+      "step": 20963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8858068933055392,
+      "learning_rate": 1.0263986158059622e-06,
+      "loss": 0.508,
+      "step": 20964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.59767620623041,
+      "learning_rate": 1.0263282413375931e-06,
+      "loss": 0.3878,
+      "step": 20965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4359765387250967,
+      "learning_rate": 1.0262579605227633e-06,
+      "loss": 0.5364,
+      "step": 20966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6222199877131567,
+      "learning_rate": 1.0261877733629448e-06,
+      "loss": 0.4762,
+      "step": 20967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8786697747157133,
+      "learning_rate": 1.026117679859606e-06,
+      "loss": 0.4007,
+      "step": 20968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8630459623592752,
+      "learning_rate": 1.0260476800142149e-06,
+      "loss": 0.5608,
+      "step": 20969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7554646108774121,
+      "learning_rate": 1.025977773828237e-06,
+      "loss": 0.5719,
+      "step": 20970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8686854489677855,
+      "learning_rate": 1.025907961303135e-06,
+      "loss": 0.5015,
+      "step": 20971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.789921372343912,
+      "learning_rate": 1.0258382424403712e-06,
+      "loss": 0.5742,
+      "step": 20972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6251381320257807,
+      "learning_rate": 1.025768617241405e-06,
+      "loss": 0.4781,
+      "step": 20973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.824299747306551,
+      "learning_rate": 1.0256990857076932e-06,
+      "loss": 0.4044,
+      "step": 20974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.830060213093062,
+      "learning_rate": 1.0256296478406923e-06,
+      "loss": 0.4557,
+      "step": 20975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7501642817113705,
+      "learning_rate": 1.0255603036418554e-06,
+      "loss": 0.4571,
+      "step": 20976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8888163533223283,
+      "learning_rate": 1.0254910531126349e-06,
+      "loss": 0.6018,
+      "step": 20977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.904209989312124,
+      "learning_rate": 1.0254218962544798e-06,
+      "loss": 0.4977,
+      "step": 20978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7370496122043495,
+      "learning_rate": 1.0253528330688374e-06,
+      "loss": 0.4913,
+      "step": 20979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6245055593104312,
+      "learning_rate": 1.0252838635571547e-06,
+      "loss": 0.3871,
+      "step": 20980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.694791814170628,
+      "learning_rate": 1.0252149877208748e-06,
+      "loss": 0.4372,
+      "step": 20981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.623702403980246,
+      "learning_rate": 1.0251462055614396e-06,
+      "loss": 0.3899,
+      "step": 20982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5643965477011594,
+      "learning_rate": 1.0250775170802896e-06,
+      "loss": 0.644,
+      "step": 20983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5590131221649748,
+      "learning_rate": 1.0250089222788622e-06,
+      "loss": 0.3566,
+      "step": 20984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7303403791806446,
+      "learning_rate": 1.024940421158593e-06,
+      "loss": 0.4112,
+      "step": 20985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.17993971257229,
+      "learning_rate": 1.0248720137209173e-06,
+      "loss": 0.5163,
+      "step": 20986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0047532303675784,
+      "learning_rate": 1.0248036999672658e-06,
+      "loss": 0.5039,
+      "step": 20987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7188407017805818,
+      "learning_rate": 1.0247354798990693e-06,
+      "loss": 0.4302,
+      "step": 20988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4778614886628532,
+      "learning_rate": 1.0246673535177554e-06,
+      "loss": 0.3428,
+      "step": 20989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7362216947178803,
+      "learning_rate": 1.0245993208247513e-06,
+      "loss": 0.4519,
+      "step": 20990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8377316864325255,
+      "learning_rate": 1.0245313818214806e-06,
+      "loss": 0.5266,
+      "step": 20991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9232597320704239,
+      "learning_rate": 1.0244635365093653e-06,
+      "loss": 0.4422,
+      "step": 20992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6967268289590336,
+      "learning_rate": 1.0243957848898268e-06,
+      "loss": 0.4165,
+      "step": 20993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6706649648459793,
+      "learning_rate": 1.024328126964282e-06,
+      "loss": 0.3494,
+      "step": 20994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8176706028617318,
+      "learning_rate": 1.024260562734148e-06,
+      "loss": 0.4916,
+      "step": 20995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.869410405206679,
+      "learning_rate": 1.0241930922008391e-06,
+      "loss": 0.592,
+      "step": 20996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9976988011041614,
+      "learning_rate": 1.024125715365768e-06,
+      "loss": 0.7496,
+      "step": 20997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.234716459845205,
+      "learning_rate": 1.024058432230345e-06,
+      "loss": 0.4698,
+      "step": 20998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.820929780031133,
+      "learning_rate": 1.023991242795978e-06,
+      "loss": 0.5233,
+      "step": 20999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6363909265941763,
+      "learning_rate": 1.0239241470640745e-06,
+      "loss": 0.3846,
+      "step": 21000
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.4810436964035034,
+      "eval_runtime": 68.0006,
+      "eval_samples_per_second": 50.147,
+      "eval_steps_per_second": 1.574,
+      "step": 21000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1219612824471117,
+      "learning_rate": 1.0238571450360388e-06,
+      "loss": 0.4704,
+      "step": 21001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.959613960682617,
+      "learning_rate": 1.0237902367132737e-06,
+      "loss": 0.4661,
+      "step": 21002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.289486388280581,
+      "learning_rate": 1.0237234220971798e-06,
+      "loss": 0.5869,
+      "step": 21003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.911237239345337,
+      "learning_rate": 1.0236567011891554e-06,
+      "loss": 0.5177,
+      "step": 21004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5557353996408896,
+      "learning_rate": 1.023590073990598e-06,
+      "loss": 0.3653,
+      "step": 21005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8690746170695263,
+      "learning_rate": 1.0235235405029018e-06,
+      "loss": 0.4886,
+      "step": 21006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.214012267878639,
+      "learning_rate": 1.0234571007274593e-06,
+      "loss": 0.4872,
+      "step": 21007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0946908473420054,
+      "learning_rate": 1.0233907546656624e-06,
+      "loss": 0.5536,
+      "step": 21008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.887712042852424,
+      "learning_rate": 1.0233245023188998e-06,
+      "loss": 0.6138,
+      "step": 21009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5672615232749016,
+      "learning_rate": 1.0232583436885582e-06,
+      "loss": 0.424,
+      "step": 21010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0631887786155594,
+      "learning_rate": 1.023192278776022e-06,
+      "loss": 0.5041,
+      "step": 21011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8893146522545428,
+      "learning_rate": 1.0231263075826755e-06,
+      "loss": 0.4772,
+      "step": 21012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.116089773087893,
+      "learning_rate": 1.0230604301098985e-06,
+      "loss": 0.5122,
+      "step": 21013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0818440480460354,
+      "learning_rate": 1.0229946463590704e-06,
+      "loss": 0.5603,
+      "step": 21014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.119977578535699,
+      "learning_rate": 1.0229289563315693e-06,
+      "loss": 0.5955,
+      "step": 21015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9687945374162692,
+      "learning_rate": 1.022863360028769e-06,
+      "loss": 0.601,
+      "step": 21016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0414122586407957,
+      "learning_rate": 1.022797857452044e-06,
+      "loss": 0.4522,
+      "step": 21017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9230224680905526,
+      "learning_rate": 1.0227324486027653e-06,
+      "loss": 0.5139,
+      "step": 21018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1471583029228576,
+      "learning_rate": 1.022667133482301e-06,
+      "loss": 0.4616,
+      "step": 21019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1273323461036857,
+      "learning_rate": 1.0226019120920197e-06,
+      "loss": 0.53,
+      "step": 21020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.319294441481313,
+      "learning_rate": 1.0225367844332861e-06,
+      "loss": 0.5886,
+      "step": 21021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9326605994291284,
+      "learning_rate": 1.0224717505074644e-06,
+      "loss": 0.5169,
+      "step": 21022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0122896827596315,
+      "learning_rate": 1.0224068103159148e-06,
+      "loss": 0.5022,
+      "step": 21023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6836526959419353,
+      "learning_rate": 1.0223419638599976e-06,
+      "loss": 0.4056,
+      "step": 21024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5420167152863509,
+      "learning_rate": 1.022277211141071e-06,
+      "loss": 0.3778,
+      "step": 21025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7778246990969795,
+      "learning_rate": 1.0222125521604892e-06,
+      "loss": 0.4778,
+      "step": 21026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8558679641565892,
+      "learning_rate": 1.022147986919606e-06,
+      "loss": 0.5196,
+      "step": 21027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9915491266763914,
+      "learning_rate": 1.0220835154197742e-06,
+      "loss": 0.5191,
+      "step": 21028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3886330360620645,
+      "learning_rate": 1.0220191376623423e-06,
+      "loss": 0.5128,
+      "step": 21029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4819101150392764,
+      "learning_rate": 1.021954853648658e-06,
+      "loss": 0.4131,
+      "step": 21030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5012530043202819,
+      "learning_rate": 1.0218906633800682e-06,
+      "loss": 0.3244,
+      "step": 21031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.120261573356994,
+      "learning_rate": 1.0218265668579151e-06,
+      "loss": 0.551,
+      "step": 21032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.052688717587977,
+      "learning_rate": 1.0217625640835418e-06,
+      "loss": 0.5308,
+      "step": 21033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0042185768988423,
+      "learning_rate": 1.0216986550582875e-06,
+      "loss": 0.3541,
+      "step": 21034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6923763896029753,
+      "learning_rate": 1.0216348397834904e-06,
+      "loss": 0.4993,
+      "step": 21035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8300878708446127,
+      "learning_rate": 1.0215711182604863e-06,
+      "loss": 0.653,
+      "step": 21036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8761213662952165,
+      "learning_rate": 1.0215074904906088e-06,
+      "loss": 0.3917,
+      "step": 21037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5578365078294456,
+      "learning_rate": 1.0214439564751906e-06,
+      "loss": 0.3182,
+      "step": 21038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.416366546891933,
+      "learning_rate": 1.0213805162155615e-06,
+      "loss": 0.3331,
+      "step": 21039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9927650165238597,
+      "learning_rate": 1.0213171697130497e-06,
+      "loss": 0.6466,
+      "step": 21040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.808823891622769,
+      "learning_rate": 1.0212539169689811e-06,
+      "loss": 0.5874,
+      "step": 21041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.925976476807634,
+      "learning_rate": 1.0211907579846795e-06,
+      "loss": 0.5466,
+      "step": 21042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8959154088082235,
+      "learning_rate": 1.0211276927614675e-06,
+      "loss": 0.4235,
+      "step": 21043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1570539665919117,
+      "learning_rate": 1.0210647213006657e-06,
+      "loss": 0.5719,
+      "step": 21044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6437212844703422,
+      "learning_rate": 1.0210018436035921e-06,
+      "loss": 0.5186,
+      "step": 21045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.660170752379411,
+      "learning_rate": 1.0209390596715624e-06,
+      "loss": 0.3483,
+      "step": 21046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8240187576931663,
+      "learning_rate": 1.0208763695058917e-06,
+      "loss": 0.6385,
+      "step": 21047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5994158886081284,
+      "learning_rate": 1.0208137731078922e-06,
+      "loss": 0.4197,
+      "step": 21048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7887068862287288,
+      "learning_rate": 1.0207512704788738e-06,
+      "loss": 0.4797,
+      "step": 21049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.26376920688434,
+      "learning_rate": 1.0206888616201458e-06,
+      "loss": 0.6708,
+      "step": 21050
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.4804990291595459,
+      "eval_runtime": 68.0031,
+      "eval_samples_per_second": 50.145,
+      "eval_steps_per_second": 1.573,
+      "step": 21050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6675620807133,
+      "learning_rate": 1.0206265465330139e-06,
+      "loss": 0.4503,
+      "step": 21051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8142253248635978,
+      "learning_rate": 1.0205643252187836e-06,
+      "loss": 0.5486,
+      "step": 21052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4999199213000423,
+      "learning_rate": 1.0205021976787566e-06,
+      "loss": 0.3489,
+      "step": 21053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6371705039715772,
+      "learning_rate": 1.020440163914234e-06,
+      "loss": 0.4166,
+      "step": 21054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7685786214044574,
+      "learning_rate": 1.0203782239265135e-06,
+      "loss": 0.4174,
+      "step": 21055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6890914305510036,
+      "learning_rate": 1.020316377716893e-06,
+      "loss": 0.3768,
+      "step": 21056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1232776462342073,
+      "learning_rate": 1.0202546252866665e-06,
+      "loss": 0.5038,
+      "step": 21057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7125299555379347,
+      "learning_rate": 1.020192966637127e-06,
+      "loss": 0.5835,
+      "step": 21058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2875201718641778,
+      "learning_rate": 1.0201314017695655e-06,
+      "loss": 0.6814,
+      "step": 21059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.857442803498442,
+      "learning_rate": 1.0200699306852704e-06,
+      "loss": 0.4879,
+      "step": 21060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6350515561491925,
+      "learning_rate": 1.0200085533855287e-06,
+      "loss": 0.405,
+      "step": 21061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7240897027607616,
+      "learning_rate": 1.019947269871625e-06,
+      "loss": 0.4888,
+      "step": 21062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8328298832533976,
+      "learning_rate": 1.0198860801448431e-06,
+      "loss": 0.399,
+      "step": 21063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9785404465129892,
+      "learning_rate": 1.0198249842064632e-06,
+      "loss": 0.5669,
+      "step": 21064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6489432840592668,
+      "learning_rate": 1.0197639820577642e-06,
+      "loss": 0.3974,
+      "step": 21065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8001116649883018,
+      "learning_rate": 1.0197030737000238e-06,
+      "loss": 0.4415,
+      "step": 21066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.715276237308872,
+      "learning_rate": 1.019642259134517e-06,
+      "loss": 0.4662,
+      "step": 21067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6073265910399879,
+      "learning_rate": 1.0195815383625158e-06,
+      "loss": 0.4398,
+      "step": 21068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9289826597450985,
+      "learning_rate": 1.019520911385293e-06,
+      "loss": 0.4292,
+      "step": 21069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8131661015904579,
+      "learning_rate": 1.0194603782041167e-06,
+      "loss": 0.518,
+      "step": 21070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.117727130660412,
+      "learning_rate": 1.0193999388202541e-06,
+      "loss": 0.4693,
+      "step": 21071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8894608815280618,
+      "learning_rate": 1.0193395932349713e-06,
+      "loss": 0.5616,
+      "step": 21072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4964806675016906,
+      "learning_rate": 1.0192793414495308e-06,
+      "loss": 0.3493,
+      "step": 21073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7560239255564503,
+      "learning_rate": 1.0192191834651942e-06,
+      "loss": 0.4155,
+      "step": 21074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8112029693571619,
+      "learning_rate": 1.0191591192832209e-06,
+      "loss": 0.3864,
+      "step": 21075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5844681372814151,
+      "learning_rate": 1.0190991489048686e-06,
+      "loss": 0.5235,
+      "step": 21076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.206554181159005,
+      "learning_rate": 1.0190392723313922e-06,
+      "loss": 0.5982,
+      "step": 21077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4916349176791277,
+      "learning_rate": 1.018979489564045e-06,
+      "loss": 0.4048,
+      "step": 21078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5410890889257216,
+      "learning_rate": 1.0189198006040795e-06,
+      "loss": 0.4714,
+      "step": 21079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0038311718350315,
+      "learning_rate": 1.0188602054527447e-06,
+      "loss": 0.555,
+      "step": 21080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9454356042485077,
+      "learning_rate": 1.0188007041112877e-06,
+      "loss": 0.6211,
+      "step": 21081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.60862678090478,
+      "learning_rate": 1.0187412965809545e-06,
+      "loss": 0.3736,
+      "step": 21082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5029282959617725,
+      "learning_rate": 1.0186819828629895e-06,
+      "loss": 0.3895,
+      "step": 21083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.035240588049639,
+      "learning_rate": 1.0186227629586332e-06,
+      "loss": 0.5588,
+      "step": 21084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.38876901650745,
+      "learning_rate": 1.0185636368691262e-06,
+      "loss": 0.5334,
+      "step": 21085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8727301663341347,
+      "learning_rate": 1.0185046045957056e-06,
+      "loss": 0.5022,
+      "step": 21086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4857131343163215,
+      "learning_rate": 1.018445666139608e-06,
+      "loss": 0.4065,
+      "step": 21087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8807074883542945,
+      "learning_rate": 1.0183868215020663e-06,
+      "loss": 0.4725,
+      "step": 21088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7505587031175591,
+      "learning_rate": 1.018328070684313e-06,
+      "loss": 0.5041,
+      "step": 21089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6393314180462337,
+      "learning_rate": 1.0182694136875783e-06,
+      "loss": 0.4227,
+      "step": 21090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.758187028412827,
+      "learning_rate": 1.018210850513089e-06,
+      "loss": 0.5527,
+      "step": 21091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4494312310535675,
+      "learning_rate": 1.0181523811620721e-06,
+      "loss": 0.3423,
+      "step": 21092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7182892006013066,
+      "learning_rate": 1.0180940056357518e-06,
+      "loss": 0.6411,
+      "step": 21093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9972584752519014,
+      "learning_rate": 1.018035723935349e-06,
+      "loss": 0.5732,
+      "step": 21094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6618949598347912,
+      "learning_rate": 1.0179775360620848e-06,
+      "loss": 0.5646,
+      "step": 21095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.075826394217026,
+      "learning_rate": 1.0179194420171772e-06,
+      "loss": 0.4198,
+      "step": 21096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5149669489739261,
+      "learning_rate": 1.0178614418018416e-06,
+      "loss": 0.3757,
+      "step": 21097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.616956525145367,
+      "learning_rate": 1.0178035354172927e-06,
+      "loss": 0.5459,
+      "step": 21098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5441426540585064,
+      "learning_rate": 1.0177457228647436e-06,
+      "loss": 0.4533,
+      "step": 21099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.698880389626364,
+      "learning_rate": 1.017688004145403e-06,
+      "loss": 0.3731,
+      "step": 21100
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.48068180680274963,
+      "eval_runtime": 67.9388,
+      "eval_samples_per_second": 50.192,
+      "eval_steps_per_second": 1.575,
+      "step": 21100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9453943706164651,
+      "learning_rate": 1.0176303792604802e-06,
+      "loss": 0.5417,
+      "step": 21101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3429358542003766,
+      "learning_rate": 1.0175728482111818e-06,
+      "loss": 0.3729,
+      "step": 21102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8627839891614797,
+      "learning_rate": 1.0175154109987111e-06,
+      "loss": 0.5307,
+      "step": 21103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8239569946685437,
+      "learning_rate": 1.0174580676242712e-06,
+      "loss": 0.6632,
+      "step": 21104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.701506073303426,
+      "learning_rate": 1.0174008180890623e-06,
+      "loss": 0.5182,
+      "step": 21105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9420744898256772,
+      "learning_rate": 1.0173436623942832e-06,
+      "loss": 0.4132,
+      "step": 21106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6122434108612964,
+      "learning_rate": 1.01728660054113e-06,
+      "loss": 0.5236,
+      "step": 21107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2174428037640745,
+      "learning_rate": 1.017229632530798e-06,
+      "loss": 0.5275,
+      "step": 21108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6873756020499262,
+      "learning_rate": 1.017172758364479e-06,
+      "loss": 0.5367,
+      "step": 21109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7677636269679027,
+      "learning_rate": 1.0171159780433642e-06,
+      "loss": 0.4179,
+      "step": 21110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6070471947470282,
+      "learning_rate": 1.017059291568642e-06,
+      "loss": 0.4128,
+      "step": 21111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9123706263324263,
+      "learning_rate": 1.0170026989414985e-06,
+      "loss": 0.4942,
+      "step": 21112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9911764604321494,
+      "learning_rate": 1.0169462001631195e-06,
+      "loss": 0.4791,
+      "step": 21113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.824219135505413,
+      "learning_rate": 1.016889795234687e-06,
+      "loss": 0.5316,
+      "step": 21114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7471934345904372,
+      "learning_rate": 1.0168334841573822e-06,
+      "loss": 0.4774,
+      "step": 21115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7257291172091203,
+      "learning_rate": 1.0167772669323839e-06,
+      "loss": 0.4428,
+      "step": 21116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8937735404225293,
+      "learning_rate": 1.0167211435608687e-06,
+      "loss": 0.5286,
+      "step": 21117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9452846283021732,
+      "learning_rate": 1.0166651140440116e-06,
+      "loss": 0.5191,
+      "step": 21118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5590518075100939,
+      "learning_rate": 1.0166091783829861e-06,
+      "loss": 0.4465,
+      "step": 21119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6251109729781268,
+      "learning_rate": 1.0165533365789623e-06,
+      "loss": 0.489,
+      "step": 21120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0929417883467987,
+      "learning_rate": 1.0164975886331096e-06,
+      "loss": 0.5168,
+      "step": 21121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.00737262557232,
+      "learning_rate": 1.0164419345465954e-06,
+      "loss": 0.4669,
+      "step": 21122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8347323852945123,
+      "learning_rate": 1.0163863743205844e-06,
+      "loss": 0.4478,
+      "step": 21123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7369877534890255,
+      "learning_rate": 1.0163309079562395e-06,
+      "loss": 0.4401,
+      "step": 21124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8001076753289706,
+      "learning_rate": 1.0162755354547224e-06,
+      "loss": 0.4217,
+      "step": 21125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6486346089788588,
+      "learning_rate": 1.0162202568171919e-06,
+      "loss": 0.4849,
+      "step": 21126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0454098984097686,
+      "learning_rate": 1.0161650720448051e-06,
+      "loss": 0.5618,
+      "step": 21127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7730660289556055,
+      "learning_rate": 1.0161099811387182e-06,
+      "loss": 0.4116,
+      "step": 21128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8649814608655184,
+      "learning_rate": 1.016054984100083e-06,
+      "loss": 0.4519,
+      "step": 21129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0164222234547515,
+      "learning_rate": 1.016000080930052e-06,
+      "loss": 0.6404,
+      "step": 21130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6851423497932585,
+      "learning_rate": 1.0159452716297744e-06,
+      "loss": 0.3752,
+      "step": 21131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5356967290119148,
+      "learning_rate": 1.0158905562003971e-06,
+      "loss": 0.4509,
+      "step": 21132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5221834145575823,
+      "learning_rate": 1.015835934643066e-06,
+      "loss": 0.4291,
+      "step": 21133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.22980200546287,
+      "learning_rate": 1.0157814069589239e-06,
+      "loss": 0.4353,
+      "step": 21134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6618556755801805,
+      "learning_rate": 1.0157269731491134e-06,
+      "loss": 0.324,
+      "step": 21135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7250256318695913,
+      "learning_rate": 1.0156726332147733e-06,
+      "loss": 0.4533,
+      "step": 21136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7880965817940546,
+      "learning_rate": 1.0156183871570413e-06,
+      "loss": 0.5313,
+      "step": 21137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0018301664922213,
+      "learning_rate": 1.0155642349770526e-06,
+      "loss": 0.4896,
+      "step": 21138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9590022034052461,
+      "learning_rate": 1.0155101766759417e-06,
+      "loss": 0.426,
+      "step": 21139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8078978270180561,
+      "learning_rate": 1.0154562122548394e-06,
+      "loss": 0.4451,
+      "step": 21140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.546832420651111,
+      "learning_rate": 1.0154023417148765e-06,
+      "loss": 0.4402,
+      "step": 21141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8970414807346179,
+      "learning_rate": 1.0153485650571797e-06,
+      "loss": 0.5318,
+      "step": 21142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5034216996265992,
+      "learning_rate": 1.0152948822828754e-06,
+      "loss": 0.544,
+      "step": 21143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6236821820607654,
+      "learning_rate": 1.0152412933930866e-06,
+      "loss": 0.4645,
+      "step": 21144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6771179616523508,
+      "learning_rate": 1.0151877983889364e-06,
+      "loss": 0.4641,
+      "step": 21145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9240040126079436,
+      "learning_rate": 1.0151343972715437e-06,
+      "loss": 0.4258,
+      "step": 21146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4271761755424235,
+      "learning_rate": 1.0150810900420266e-06,
+      "loss": 0.362,
+      "step": 21147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.950341762716818,
+      "learning_rate": 1.0150278767015012e-06,
+      "loss": 0.4261,
+      "step": 21148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8706652708346838,
+      "learning_rate": 1.0149747572510818e-06,
+      "loss": 0.6065,
+      "step": 21149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.042104686716567,
+      "learning_rate": 1.0149217316918804e-06,
+      "loss": 0.4181,
+      "step": 21150
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.48028114438056946,
+      "eval_runtime": 68.0106,
+      "eval_samples_per_second": 50.139,
+      "eval_steps_per_second": 1.573,
+      "step": 21150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4653937927086627,
+      "learning_rate": 1.0148688000250058e-06,
+      "loss": 0.4023,
+      "step": 21151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.572830592443906,
+      "learning_rate": 1.0148159622515676e-06,
+      "loss": 0.4512,
+      "step": 21152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5164708606372257,
+      "learning_rate": 1.0147632183726718e-06,
+      "loss": 0.4179,
+      "step": 21153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6485536458266943,
+      "learning_rate": 1.0147105683894217e-06,
+      "loss": 0.313,
+      "step": 21154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7383361769541272,
+      "learning_rate": 1.0146580123029199e-06,
+      "loss": 0.4478,
+      "step": 21155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7415487395122344,
+      "learning_rate": 1.0146055501142664e-06,
+      "loss": 0.564,
+      "step": 21156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.65023972293727,
+      "learning_rate": 1.0145531818245604e-06,
+      "loss": 0.4603,
+      "step": 21157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9467137623428001,
+      "learning_rate": 1.0145009074348968e-06,
+      "loss": 0.53,
+      "step": 21158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2122511252792192,
+      "learning_rate": 1.0144487269463708e-06,
+      "loss": 0.2762,
+      "step": 21159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9213807799748885,
+      "learning_rate": 1.0143966403600753e-06,
+      "loss": 0.5665,
+      "step": 21160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8504482209895052,
+      "learning_rate": 1.0143446476770993e-06,
+      "loss": 0.4862,
+      "step": 21161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5549137144751082,
+      "learning_rate": 1.014292748898532e-06,
+      "loss": 0.3855,
+      "step": 21162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8597659881670432,
+      "learning_rate": 1.0142409440254602e-06,
+      "loss": 0.4927,
+      "step": 21163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8094693340017034,
+      "learning_rate": 1.0141892330589679e-06,
+      "loss": 0.3882,
+      "step": 21164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5697386274917886,
+      "learning_rate": 1.0141376160001379e-06,
+      "loss": 0.405,
+      "step": 21165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7968758524461028,
+      "learning_rate": 1.0140860928500504e-06,
+      "loss": 0.4408,
+      "step": 21166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0164993577319508,
+      "learning_rate": 1.0140346636097845e-06,
+      "loss": 0.5616,
+      "step": 21167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5149494871815357,
+      "learning_rate": 1.013983328280416e-06,
+      "loss": 0.4732,
+      "step": 21168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6308318455902227,
+      "learning_rate": 1.013932086863021e-06,
+      "loss": 0.5063,
+      "step": 21169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3038281892966077,
+      "learning_rate": 1.013880939358671e-06,
+      "loss": 0.3111,
+      "step": 21170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6184932042312952,
+      "learning_rate": 1.0138298857684366e-06,
+      "loss": 0.4316,
+      "step": 21171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1196890529815633,
+      "learning_rate": 1.0137789260933874e-06,
+      "loss": 0.532,
+      "step": 21172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9723967121045887,
+      "learning_rate": 1.0137280603345904e-06,
+      "loss": 0.4073,
+      "step": 21173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5258276449942578,
+      "learning_rate": 1.0136772884931092e-06,
+      "loss": 0.4195,
+      "step": 21174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8914167778223623,
+      "learning_rate": 1.0136266105700077e-06,
+      "loss": 0.4239,
+      "step": 21175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8556780453426511,
+      "learning_rate": 1.0135760265663465e-06,
+      "loss": 0.4197,
+      "step": 21176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7775748411856267,
+      "learning_rate": 1.0135255364831847e-06,
+      "loss": 0.5641,
+      "step": 21177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6377072067421157,
+      "learning_rate": 1.0134751403215788e-06,
+      "loss": 0.3668,
+      "step": 21178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5599642343511875,
+      "learning_rate": 1.0134248380825847e-06,
+      "loss": 0.4646,
+      "step": 21179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8143254264250706,
+      "learning_rate": 1.0133746297672548e-06,
+      "loss": 0.574,
+      "step": 21180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7127954883490857,
+      "learning_rate": 1.0133245153766402e-06,
+      "loss": 0.4049,
+      "step": 21181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8184277795526065,
+      "learning_rate": 1.01327449491179e-06,
+      "loss": 0.4888,
+      "step": 21182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5744400714902822,
+      "learning_rate": 1.013224568373751e-06,
+      "loss": 0.4323,
+      "step": 21183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7364770126565632,
+      "learning_rate": 1.0131747357635693e-06,
+      "loss": 0.4185,
+      "step": 21184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9482454151828,
+      "learning_rate": 1.0131249970822876e-06,
+      "loss": 0.4889,
+      "step": 21185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6845013490637777,
+      "learning_rate": 1.0130753523309473e-06,
+      "loss": 0.4946,
+      "step": 21186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7225707550242724,
+      "learning_rate": 1.0130258015105874e-06,
+      "loss": 0.6124,
+      "step": 21187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.453183918641332,
+      "learning_rate": 1.0129763446222453e-06,
+      "loss": 0.3935,
+      "step": 21188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9900048070088523,
+      "learning_rate": 1.0129269816669563e-06,
+      "loss": 0.5304,
+      "step": 21189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4162891890103717,
+      "learning_rate": 1.0128777126457541e-06,
+      "loss": 0.4913,
+      "step": 21190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5816560247934377,
+      "learning_rate": 1.01282853755967e-06,
+      "loss": 0.4704,
+      "step": 21191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6631289110687837,
+      "learning_rate": 1.0127794564097336e-06,
+      "loss": 0.4104,
+      "step": 21192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7057015232632775,
+      "learning_rate": 1.012730469196972e-06,
+      "loss": 0.5281,
+      "step": 21193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0110608839667514,
+      "learning_rate": 1.0126815759224109e-06,
+      "loss": 0.6127,
+      "step": 21194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6447913148945124,
+      "learning_rate": 1.0126327765870738e-06,
+      "loss": 0.3771,
+      "step": 21195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1740153441606194,
+      "learning_rate": 1.0125840711919821e-06,
+      "loss": 0.6342,
+      "step": 21196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6624732897513508,
+      "learning_rate": 1.0125354597381554e-06,
+      "loss": 0.4867,
+      "step": 21197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8718287096322077,
+      "learning_rate": 1.0124869422266121e-06,
+      "loss": 0.437,
+      "step": 21198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8950652497695766,
+      "learning_rate": 1.0124385186583674e-06,
+      "loss": 0.4639,
+      "step": 21199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.712344262792496,
+      "learning_rate": 1.0123901890344346e-06,
+      "loss": 0.4255,
+      "step": 21200
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.4799458086490631,
+      "eval_runtime": 67.9925,
+      "eval_samples_per_second": 50.153,
+      "eval_steps_per_second": 1.574,
+      "step": 21200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4266585195229797,
+      "learning_rate": 1.0123419533558261e-06,
+      "loss": 0.3071,
+      "step": 21201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7433179162407888,
+      "learning_rate": 1.0122938116235515e-06,
+      "loss": 0.4077,
+      "step": 21202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6402691382924184,
+      "learning_rate": 1.0122457638386182e-06,
+      "loss": 0.4062,
+      "step": 21203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6944326380962185,
+      "learning_rate": 1.0121978100020328e-06,
+      "loss": 0.457,
+      "step": 21204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1533097796630383,
+      "learning_rate": 1.0121499501147983e-06,
+      "loss": 0.6714,
+      "step": 21205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.833918733166987,
+      "learning_rate": 1.0121021841779175e-06,
+      "loss": 0.4117,
+      "step": 21206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8973484954669428,
+      "learning_rate": 1.0120545121923897e-06,
+      "loss": 0.4582,
+      "step": 21207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.704598730683971,
+      "learning_rate": 1.0120069341592131e-06,
+      "loss": 0.4088,
+      "step": 21208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8651254579733778,
+      "learning_rate": 1.011959450079384e-06,
+      "loss": 0.4585,
+      "step": 21209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.696386383162375,
+      "learning_rate": 1.0119120599538961e-06,
+      "loss": 0.3904,
+      "step": 21210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1497547833605064,
+      "learning_rate": 1.0118647637837414e-06,
+      "loss": 0.4691,
+      "step": 21211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6645265377340166,
+      "learning_rate": 1.0118175615699103e-06,
+      "loss": 0.4428,
+      "step": 21212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5250802994907369,
+      "learning_rate": 1.0117704533133906e-06,
+      "loss": 0.3755,
+      "step": 21213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4799305398097384,
+      "learning_rate": 1.011723439015169e-06,
+      "loss": 0.4313,
+      "step": 21214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5656017227879635,
+      "learning_rate": 1.0116765186762295e-06,
+      "loss": 0.4114,
+      "step": 21215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.826974379148373,
+      "learning_rate": 1.0116296922975543e-06,
+      "loss": 0.4969,
+      "step": 21216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.67296299260725,
+      "learning_rate": 1.0115829598801237e-06,
+      "loss": 0.3947,
+      "step": 21217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7101029490543844,
+      "learning_rate": 1.0115363214249157e-06,
+      "loss": 0.421,
+      "step": 21218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.683329722730147,
+      "learning_rate": 1.0114897769329068e-06,
+      "loss": 0.4654,
+      "step": 21219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9724132896491542,
+      "learning_rate": 1.0114433264050715e-06,
+      "loss": 0.4361,
+      "step": 21220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3792620129013908,
+      "learning_rate": 1.011396969842383e-06,
+      "loss": 0.3444,
+      "step": 21221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5516891667231707,
+      "learning_rate": 1.01135070724581e-06,
+      "loss": 0.4638,
+      "step": 21222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7934382123833876,
+      "learning_rate": 1.0113045386163225e-06,
+      "loss": 0.5074,
+      "step": 21223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7774341545561054,
+      "learning_rate": 1.0112584639548864e-06,
+      "loss": 0.3524,
+      "step": 21224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9874229826532688,
+      "learning_rate": 1.0112124832624663e-06,
+      "loss": 0.5372,
+      "step": 21225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5987588823563195,
+      "learning_rate": 1.0111665965400245e-06,
+      "loss": 0.4374,
+      "step": 21226
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9838226059971444,
+      "learning_rate": 1.0111208037885224e-06,
+      "loss": 0.4839,
+      "step": 21227
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5309173229652953,
+      "learning_rate": 1.0110751050089173e-06,
+      "loss": 0.3763,
+      "step": 21228
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4618528342559676,
+      "learning_rate": 1.0110295002021673e-06,
+      "loss": 0.3029,
+      "step": 21229
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4840834810766728,
+      "learning_rate": 1.0109839893692267e-06,
+      "loss": 0.3575,
+      "step": 21230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9152206064063104,
+      "learning_rate": 1.0109385725110475e-06,
+      "loss": 0.4081,
+      "step": 21231
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.173757399476767,
+      "learning_rate": 1.0108932496285814e-06,
+      "loss": 0.518,
+      "step": 21232
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9186229183825274,
+      "learning_rate": 1.010848020722776e-06,
+      "loss": 0.4157,
+      "step": 21233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.053831386654381,
+      "learning_rate": 1.0108028857945801e-06,
+      "loss": 0.5847,
+      "step": 21234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.323697522067825,
+      "learning_rate": 1.0107578448449369e-06,
+      "loss": 0.5591,
+      "step": 21235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0886568470899833,
+      "learning_rate": 1.01071289787479e-06,
+      "loss": 0.5287,
+      "step": 21236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9155010795446443,
+      "learning_rate": 1.0106680448850802e-06,
+      "loss": 0.5173,
+      "step": 21237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4188575810199087,
+      "learning_rate": 1.0106232858767464e-06,
+      "loss": 0.3471,
+      "step": 21238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5764688468744688,
+      "learning_rate": 1.0105786208507257e-06,
+      "loss": 0.4325,
+      "step": 21239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.529579077885247,
+      "learning_rate": 1.0105340498079532e-06,
+      "loss": 0.3544,
+      "step": 21240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7492243978431103,
+      "learning_rate": 1.0104895727493616e-06,
+      "loss": 0.4898,
+      "step": 21241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7278517904157145,
+      "learning_rate": 1.0104451896758821e-06,
+      "loss": 0.5156,
+      "step": 21242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3765126554320193,
+      "learning_rate": 1.0104009005884446e-06,
+      "loss": 0.5718,
+      "step": 21243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6265594174773308,
+      "learning_rate": 1.010356705487975e-06,
+      "loss": 0.3796,
+      "step": 21244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.53637984870758,
+      "learning_rate": 1.0103126043753996e-06,
+      "loss": 0.4167,
+      "step": 21245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5068131976000576,
+      "learning_rate": 1.010268597251641e-06,
+      "loss": 0.4091,
+      "step": 21246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6825141766766332,
+      "learning_rate": 1.0102246841176203e-06,
+      "loss": 0.3917,
+      "step": 21247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.96478318236341,
+      "learning_rate": 1.0101808649742576e-06,
+      "loss": 0.3773,
+      "step": 21248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5039248031237205,
+      "learning_rate": 1.0101371398224699e-06,
+      "loss": 0.3312,
+      "step": 21249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2123002683970165,
+      "learning_rate": 1.0100935086631718e-06,
+      "loss": 0.4672,
+      "step": 21250
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.4800041913986206,
+      "eval_runtime": 67.9957,
+      "eval_samples_per_second": 50.15,
+      "eval_steps_per_second": 1.574,
+      "step": 21250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.313424135294873,
+      "learning_rate": 1.0100499714972778e-06,
+      "loss": 0.3712,
+      "step": 21251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.821366794527838,
+      "learning_rate": 1.0100065283256988e-06,
+      "loss": 0.4446,
+      "step": 21252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8510748962352535,
+      "learning_rate": 1.0099631791493442e-06,
+      "loss": 0.4092,
+      "step": 21253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9545819564909015,
+      "learning_rate": 1.0099199239691215e-06,
+      "loss": 0.4284,
+      "step": 21254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5900796865428153,
+      "learning_rate": 1.0098767627859365e-06,
+      "loss": 0.3099,
+      "step": 21255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8445713893679885,
+      "learning_rate": 1.0098336956006925e-06,
+      "loss": 0.4658,
+      "step": 21256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4971059600848013,
+      "learning_rate": 1.0097907224142913e-06,
+      "loss": 0.3922,
+      "step": 21257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9089833491116364,
+      "learning_rate": 1.009747843227632e-06,
+      "loss": 0.5065,
+      "step": 21258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.028867608284145,
+      "learning_rate": 1.0097050580416128e-06,
+      "loss": 0.5867,
+      "step": 21259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5089277304863464,
+      "learning_rate": 1.0096623668571293e-06,
+      "loss": 0.3097,
+      "step": 21260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.819289031681443,
+      "learning_rate": 1.0096197696750745e-06,
+      "loss": 0.438,
+      "step": 21261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6073633732409007,
+      "learning_rate": 1.0095772664963416e-06,
+      "loss": 0.3695,
+      "step": 21262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7536377427429393,
+      "learning_rate": 1.009534857321819e-06,
+      "loss": 0.3416,
+      "step": 21263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9652749841073072,
+      "learning_rate": 1.0094925421523949e-06,
+      "loss": 0.5781,
+      "step": 21264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.73910287433801,
+      "learning_rate": 1.0094503209889557e-06,
+      "loss": 0.4424,
+      "step": 21265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.927128945950807,
+      "learning_rate": 1.0094081938323844e-06,
+      "loss": 0.3986,
+      "step": 21266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9327327985864913,
+      "learning_rate": 1.0093661606835637e-06,
+      "loss": 0.4904,
+      "step": 21267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7919900254637449,
+      "learning_rate": 1.009324221543373e-06,
+      "loss": 0.4477,
+      "step": 21268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.621016862334438,
+      "learning_rate": 1.0092823764126906e-06,
+      "loss": 0.4056,
+      "step": 21269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6876162933071868,
+      "learning_rate": 1.009240625292392e-06,
+      "loss": 0.4567,
+      "step": 21270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6716396934723223,
+      "learning_rate": 1.0091989681833524e-06,
+      "loss": 0.4247,
+      "step": 21271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5360225362924356,
+      "learning_rate": 1.0091574050864422e-06,
+      "loss": 0.4164,
+      "step": 21272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.170636899982011,
+      "learning_rate": 1.0091159360025328e-06,
+      "loss": 0.5925,
+      "step": 21273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9014917887651455,
+      "learning_rate": 1.0090745609324916e-06,
+      "loss": 0.3439,
+      "step": 21274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2465750156223714,
+      "learning_rate": 1.0090332798771857e-06,
+      "loss": 0.4259,
+      "step": 21275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8565805574308456,
+      "learning_rate": 1.0089920928374779e-06,
+      "loss": 0.5387,
+      "step": 21276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6275657231275071,
+      "learning_rate": 1.0089509998142318e-06,
+      "loss": 0.4374,
+      "step": 21277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.14059613969362,
+      "learning_rate": 1.0089100008083064e-06,
+      "loss": 0.6121,
+      "step": 21278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.662893506980631,
+      "learning_rate": 1.008869095820561e-06,
+      "loss": 0.422,
+      "step": 21279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.190765378341776,
+      "learning_rate": 1.0088282848518516e-06,
+      "loss": 0.5638,
+      "step": 21280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.131418660508466,
+      "learning_rate": 1.0087875679030319e-06,
+      "loss": 0.5197,
+      "step": 21281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1797760376584516,
+      "learning_rate": 1.0087469449749555e-06,
+      "loss": 0.2904,
+      "step": 21282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9083447319768103,
+      "learning_rate": 1.0087064160684722e-06,
+      "loss": 0.5265,
+      "step": 21283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.766643683828314,
+      "learning_rate": 1.0086659811844301e-06,
+      "loss": 0.5421,
+      "step": 21284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.731425684231102,
+      "learning_rate": 1.0086256403236763e-06,
+      "loss": 0.4071,
+      "step": 21285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8209471153264791,
+      "learning_rate": 1.0085853934870548e-06,
+      "loss": 0.4404,
+      "step": 21286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8316533183566774,
+      "learning_rate": 1.0085452406754084e-06,
+      "loss": 0.426,
+      "step": 21287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.848463727079909,
+      "learning_rate": 1.0085051818895777e-06,
+      "loss": 0.3581,
+      "step": 21288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.962139171306417,
+      "learning_rate": 1.0084652171304012e-06,
+      "loss": 0.389,
+      "step": 21289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9205776061199837,
+      "learning_rate": 1.0084253463987156e-06,
+      "loss": 0.3887,
+      "step": 21290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7444001379735437,
+      "learning_rate": 1.0083855696953556e-06,
+      "loss": 0.6359,
+      "step": 21291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5735729936492426,
+      "learning_rate": 1.0083458870211539e-06,
+      "loss": 0.4594,
+      "step": 21292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7832870513027057,
+      "learning_rate": 1.008306298376941e-06,
+      "loss": 0.348,
+      "step": 21293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7924333778722301,
+      "learning_rate": 1.008266803763546e-06,
+      "loss": 0.4089,
+      "step": 21294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6673270557121036,
+      "learning_rate": 1.0082274031817952e-06,
+      "loss": 0.4244,
+      "step": 21295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3348904688384011,
+      "learning_rate": 1.0081880966325142e-06,
+      "loss": 0.3047,
+      "step": 21296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.168641663789537,
+      "learning_rate": 1.0081488841165249e-06,
+      "loss": 0.5462,
+      "step": 21297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9193025010846674,
+      "learning_rate": 1.0081097656346493e-06,
+      "loss": 0.3712,
+      "step": 21298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4825680692882508,
+      "learning_rate": 1.0080707411877059e-06,
+      "loss": 0.4101,
+      "step": 21299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.11577200595126,
+      "learning_rate": 1.0080318107765106e-06,
+      "loss": 0.4825,
+      "step": 21300
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.47983652353286743,
+      "eval_runtime": 68.0218,
+      "eval_samples_per_second": 50.131,
+      "eval_steps_per_second": 1.573,
+      "step": 21300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7612604183559286,
+      "learning_rate": 1.00799297440188e-06,
+      "loss": 0.4429,
+      "step": 21301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6945954570449067,
+      "learning_rate": 1.007954232064626e-06,
+      "loss": 0.4894,
+      "step": 21302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5648176545607424,
+      "learning_rate": 1.0079155837655605e-06,
+      "loss": 0.5077,
+      "step": 21303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.103708085344678,
+      "learning_rate": 1.007877029505492e-06,
+      "loss": 0.5099,
+      "step": 21304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9150988914077156,
+      "learning_rate": 1.0078385692852275e-06,
+      "loss": 0.4144,
+      "step": 21305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7568235756187842,
+      "learning_rate": 1.0078002031055722e-06,
+      "loss": 0.4157,
+      "step": 21306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9703330873796134,
+      "learning_rate": 1.0077619309673301e-06,
+      "loss": 0.4153,
+      "step": 21307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6054484490025343,
+      "learning_rate": 1.007723752871301e-06,
+      "loss": 0.4802,
+      "step": 21308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.639817972550069,
+      "learning_rate": 1.0076856688182853e-06,
+      "loss": 0.3662,
+      "step": 21309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7214145432855326,
+      "learning_rate": 1.0076476788090798e-06,
+      "loss": 0.4975,
+      "step": 21310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1146700924251305,
+      "learning_rate": 1.00760978284448e-06,
+      "loss": 0.5262,
+      "step": 21311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8122848045289308,
+      "learning_rate": 1.0075719809252786e-06,
+      "loss": 0.4624,
+      "step": 21312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5824998047667513,
+      "learning_rate": 1.0075342730522675e-06,
+      "loss": 0.4395,
+      "step": 21313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4836934203354462,
+      "learning_rate": 1.0074966592262367e-06,
+      "loss": 0.4534,
+      "step": 21314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8821005320005157,
+      "learning_rate": 1.0074591394479724e-06,
+      "loss": 0.5468,
+      "step": 21315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.024405723234824,
+      "learning_rate": 1.0074217137182606e-06,
+      "loss": 0.519,
+      "step": 21316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.576722368738577,
+      "learning_rate": 1.0073843820378846e-06,
+      "loss": 0.301,
+      "step": 21317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9706025047468279,
+      "learning_rate": 1.0073471444076263e-06,
+      "loss": 0.5863,
+      "step": 21318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.612902572443404,
+      "learning_rate": 1.007310000828265e-06,
+      "loss": 0.4773,
+      "step": 21319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6916927434330031,
+      "learning_rate": 1.0072729513005786e-06,
+      "loss": 0.4898,
+      "step": 21320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0298463291872695,
+      "learning_rate": 1.0072359958253422e-06,
+      "loss": 0.61,
+      "step": 21321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0110800376657774,
+      "learning_rate": 1.0071991344033295e-06,
+      "loss": 0.4759,
+      "step": 21322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.507711952544873,
+      "learning_rate": 1.0071623670353128e-06,
+      "loss": 0.3958,
+      "step": 21323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4937157791972377,
+      "learning_rate": 1.007125693722061e-06,
+      "loss": 0.3513,
+      "step": 21324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5210028358395287,
+      "learning_rate": 1.0070891144643424e-06,
+      "loss": 0.4326,
+      "step": 21325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.920791534422306,
+      "learning_rate": 1.0070526292629223e-06,
+      "loss": 0.5678,
+      "step": 21326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.093057219709613,
+      "learning_rate": 1.0070162381185648e-06,
+      "loss": 0.5368,
+      "step": 21327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3886543137003964,
+      "learning_rate": 1.0069799410320314e-06,
+      "loss": 0.2501,
+      "step": 21328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7209365856189534,
+      "learning_rate": 1.006943738004083e-06,
+      "loss": 0.3751,
+      "step": 21329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7804308565245366,
+      "learning_rate": 1.0069076290354757e-06,
+      "loss": 0.5071,
+      "step": 21330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6440597570948154,
+      "learning_rate": 1.0068716141269673e-06,
+      "loss": 0.4994,
+      "step": 21331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1297134151740416,
+      "learning_rate": 1.0068356932793104e-06,
+      "loss": 0.4386,
+      "step": 21332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8214987970986973,
+      "learning_rate": 1.0067998664932575e-06,
+      "loss": 0.6214,
+      "step": 21333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9232930205216738,
+      "learning_rate": 1.0067641337695585e-06,
+      "loss": 0.7743,
+      "step": 21334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.587817508945849,
+      "learning_rate": 1.0067284951089614e-06,
+      "loss": 0.3343,
+      "step": 21335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.62327442962715,
+      "learning_rate": 1.006692950512213e-06,
+      "loss": 0.4722,
+      "step": 21336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.181219517237209,
+      "learning_rate": 1.0066574999800562e-06,
+      "loss": 0.5116,
+      "step": 21337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.622278847346557,
+      "learning_rate": 1.006622143513234e-06,
+      "loss": 0.4952,
+      "step": 21338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8889161727779151,
+      "learning_rate": 1.0065868811124862e-06,
+      "loss": 0.4886,
+      "step": 21339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8991605089991377,
+      "learning_rate": 1.0065517127785512e-06,
+      "loss": 0.5771,
+      "step": 21340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.159083447479284,
+      "learning_rate": 1.0065166385121648e-06,
+      "loss": 0.5463,
+      "step": 21341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.053847735871777,
+      "learning_rate": 1.0064816583140618e-06,
+      "loss": 0.4793,
+      "step": 21342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7678152415244093,
+      "learning_rate": 1.006446772184974e-06,
+      "loss": 0.4767,
+      "step": 21343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0406628422678854,
+      "learning_rate": 1.0064119801256322e-06,
+      "loss": 0.4687,
+      "step": 21344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4939413174686893,
+      "learning_rate": 1.0063772821367645e-06,
+      "loss": 0.4514,
+      "step": 21345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8708617902290832,
+      "learning_rate": 1.0063426782190972e-06,
+      "loss": 0.468,
+      "step": 21346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.016052990306432,
+      "learning_rate": 1.0063081683733552e-06,
+      "loss": 0.4187,
+      "step": 21347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4642762094727144,
+      "learning_rate": 1.0062737526002604e-06,
+      "loss": 0.354,
+      "step": 21348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6558994245491827,
+      "learning_rate": 1.006239430900533e-06,
+      "loss": 0.5358,
+      "step": 21349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7828197868025644,
+      "learning_rate": 1.0062052032748924e-06,
+      "loss": 0.4732,
+      "step": 21350
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.479937344789505,
+      "eval_runtime": 68.0051,
+      "eval_samples_per_second": 50.143,
+      "eval_steps_per_second": 1.573,
+      "step": 21350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8281820367027333,
+      "learning_rate": 1.0061710697240543e-06,
+      "loss": 0.3918,
+      "step": 21351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5822219801918578,
+      "learning_rate": 1.006137030248734e-06,
+      "loss": 0.3868,
+      "step": 21352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8445339619241277,
+      "learning_rate": 1.0061030848496439e-06,
+      "loss": 0.4578,
+      "step": 21353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8167080681644425,
+      "learning_rate": 1.0060692335274939e-06,
+      "loss": 0.48,
+      "step": 21354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.137203067485738,
+      "learning_rate": 1.0060354762829932e-06,
+      "loss": 0.5298,
+      "step": 21355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8470125055835234,
+      "learning_rate": 1.0060018131168491e-06,
+      "loss": 0.4614,
+      "step": 21356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8232798715205045,
+      "learning_rate": 1.0059682440297651e-06,
+      "loss": 0.4826,
+      "step": 21357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.689535917904947,
+      "learning_rate": 1.0059347690224449e-06,
+      "loss": 0.4901,
+      "step": 21358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7749391667982595,
+      "learning_rate": 1.0059013880955889e-06,
+      "loss": 0.442,
+      "step": 21359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.852497746022414,
+      "learning_rate": 1.0058681012498959e-06,
+      "loss": 0.3936,
+      "step": 21360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9813245515135507,
+      "learning_rate": 1.005834908486063e-06,
+      "loss": 0.4405,
+      "step": 21361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8306282036085006,
+      "learning_rate": 1.0058018098047847e-06,
+      "loss": 0.5402,
+      "step": 21362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6675767478863006,
+      "learning_rate": 1.0057688052067542e-06,
+      "loss": 0.4967,
+      "step": 21363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8577221934390906,
+      "learning_rate": 1.0057358946926618e-06,
+      "loss": 0.6057,
+      "step": 21364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0458114720639644,
+      "learning_rate": 1.0057030782631974e-06,
+      "loss": 0.5567,
+      "step": 21365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1696234619911907,
+      "learning_rate": 1.0056703559190476e-06,
+      "loss": 0.5219,
+      "step": 21366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.155424899419877,
+      "learning_rate": 1.0056377276608972e-06,
+      "loss": 0.3631,
+      "step": 21367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8124395035714136,
+      "learning_rate": 1.0056051934894295e-06,
+      "loss": 0.4703,
+      "step": 21368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1160922140112226,
+      "learning_rate": 1.0055727534053255e-06,
+      "loss": 0.5782,
+      "step": 21369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6990010155842472,
+      "learning_rate": 1.0055404074092638e-06,
+      "loss": 0.5423,
+      "step": 21370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9865353905530065,
+      "learning_rate": 1.0055081555019225e-06,
+      "loss": 0.409,
+      "step": 21371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7755328212542827,
+      "learning_rate": 1.0054759976839764e-06,
+      "loss": 0.55,
+      "step": 21372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.824389465264661,
+      "learning_rate": 1.0054439339560988e-06,
+      "loss": 0.478,
+      "step": 21373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7296361888778946,
+      "learning_rate": 1.0054119643189606e-06,
+      "loss": 0.5187,
+      "step": 21374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.752257148688262,
+      "learning_rate": 1.0053800887732313e-06,
+      "loss": 0.4724,
+      "step": 21375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6968335401227717,
+      "learning_rate": 1.0053483073195778e-06,
+      "loss": 0.4508,
+      "step": 21376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6843935267513983,
+      "learning_rate": 1.0053166199586657e-06,
+      "loss": 0.4252,
+      "step": 21377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7241491234784725,
+      "learning_rate": 1.005285026691159e-06,
+      "loss": 0.3557,
+      "step": 21378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7923810357927306,
+      "learning_rate": 1.005253527517718e-06,
+      "loss": 0.482,
+      "step": 21379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.760504991050395,
+      "learning_rate": 1.005222122439003e-06,
+      "loss": 0.349,
+      "step": 21380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7688084368714494,
+      "learning_rate": 1.0051908114556708e-06,
+      "loss": 0.5162,
+      "step": 21381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3578795866630426,
+      "learning_rate": 1.0051595945683772e-06,
+      "loss": 0.5031,
+      "step": 21382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4652884952297747,
+      "learning_rate": 1.0051284717777755e-06,
+      "loss": 0.487,
+      "step": 21383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7982216201779868,
+      "learning_rate": 1.0050974430845172e-06,
+      "loss": 0.4836,
+      "step": 21384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6993313520739126,
+      "learning_rate": 1.0050665084892523e-06,
+      "loss": 0.4068,
+      "step": 21385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8279802354764403,
+      "learning_rate": 1.0050356679926284e-06,
+      "loss": 0.46,
+      "step": 21386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.853396008442298,
+      "learning_rate": 1.0050049215952908e-06,
+      "loss": 0.552,
+      "step": 21387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6473834213418326,
+      "learning_rate": 1.004974269297883e-06,
+      "loss": 0.471,
+      "step": 21388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0012967614403196,
+      "learning_rate": 1.004943711101047e-06,
+      "loss": 0.5846,
+      "step": 21389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6213837822678483,
+      "learning_rate": 1.0049132470054223e-06,
+      "loss": 0.4226,
+      "step": 21390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7685000494867653,
+      "learning_rate": 1.004882877011647e-06,
+      "loss": 0.4657,
+      "step": 21391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2436501313718553,
+      "learning_rate": 1.0048526011203566e-06,
+      "loss": 0.4944,
+      "step": 21392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.761787157610436,
+      "learning_rate": 1.0048224193321843e-06,
+      "loss": 0.386,
+      "step": 21393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7058879218208962,
+      "learning_rate": 1.0047923316477635e-06,
+      "loss": 0.4862,
+      "step": 21394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7742441123188317,
+      "learning_rate": 1.004762338067723e-06,
+      "loss": 0.467,
+      "step": 21395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1274011669165342,
+      "learning_rate": 1.0047324385926904e-06,
+      "loss": 0.3761,
+      "step": 21396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.035121779975127,
+      "learning_rate": 1.0047026332232923e-06,
+      "loss": 0.5124,
+      "step": 21397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2613421912417797,
+      "learning_rate": 1.0046729219601525e-06,
+      "loss": 0.4802,
+      "step": 21398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.652872126968749,
+      "learning_rate": 1.0046433048038932e-06,
+      "loss": 0.4737,
+      "step": 21399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7337973901948744,
+      "learning_rate": 1.0046137817551335e-06,
+      "loss": 0.4587,
+      "step": 21400
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.48005878925323486,
+      "eval_runtime": 67.9942,
+      "eval_samples_per_second": 50.151,
+      "eval_steps_per_second": 1.574,
+      "step": 21400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8581658995121337,
+      "learning_rate": 1.0045843528144926e-06,
+      "loss": 0.4753,
+      "step": 21401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6089772430996674,
+      "learning_rate": 1.004555017982586e-06,
+      "loss": 0.3729,
+      "step": 21402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6800121174723517,
+      "learning_rate": 1.0045257772600278e-06,
+      "loss": 0.4695,
+      "step": 21403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6563805600787675,
+      "learning_rate": 1.0044966306474302e-06,
+      "loss": 0.3846,
+      "step": 21404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4138948143544876,
+      "learning_rate": 1.004467578145403e-06,
+      "loss": 0.3954,
+      "step": 21405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4993975082204525,
+      "learning_rate": 1.0044386197545552e-06,
+      "loss": 0.3003,
+      "step": 21406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9508611240592637,
+      "learning_rate": 1.004409755475493e-06,
+      "loss": 0.5788,
+      "step": 21407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5037742521933886,
+      "learning_rate": 1.0043809853088193e-06,
+      "loss": 0.2799,
+      "step": 21408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6368237640708305,
+      "learning_rate": 1.0043523092551384e-06,
+      "loss": 0.4684,
+      "step": 21409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.765722725005604,
+      "learning_rate": 1.004323727315049e-06,
+      "loss": 0.5808,
+      "step": 21410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7369838464453626,
+      "learning_rate": 1.0042952394891497e-06,
+      "loss": 0.3824,
+      "step": 21411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6156985720999775,
+      "learning_rate": 1.0042668457780374e-06,
+      "loss": 0.4137,
+      "step": 21412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.267268231756679,
+      "learning_rate": 1.0042385461823068e-06,
+      "loss": 0.2764,
+      "step": 21413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7841512444328642,
+      "learning_rate": 1.0042103407025493e-06,
+      "loss": 0.4704,
+      "step": 21414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6145249121891336,
+      "learning_rate": 1.0041822293393556e-06,
+      "loss": 0.4646,
+      "step": 21415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7973655625698344,
+      "learning_rate": 1.004154212093315e-06,
+      "loss": 0.4977,
+      "step": 21416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9394830033102195,
+      "learning_rate": 1.0041262889650135e-06,
+      "loss": 0.5106,
+      "step": 21417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8913946363445842,
+      "learning_rate": 1.0040984599550354e-06,
+      "loss": 0.4423,
+      "step": 21418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1068341399244295,
+      "learning_rate": 1.0040707250639635e-06,
+      "loss": 0.5695,
+      "step": 21419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5779889854942672,
+      "learning_rate": 1.0040430842923787e-06,
+      "loss": 0.3018,
+      "step": 21420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3700659985684596,
+      "learning_rate": 1.0040155376408594e-06,
+      "loss": 0.3534,
+      "step": 21421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8136523223992018,
+      "learning_rate": 1.0039880851099818e-06,
+      "loss": 0.4834,
+      "step": 21422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2018267960318116,
+      "learning_rate": 1.0039607267003214e-06,
+      "loss": 0.5546,
+      "step": 21423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.315678614134791,
+      "learning_rate": 1.0039334624124507e-06,
+      "loss": 0.7236,
+      "step": 21424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1399033595811403,
+      "learning_rate": 1.0039062922469402e-06,
+      "loss": 0.5032,
+      "step": 21425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8769385474500841,
+      "learning_rate": 1.0038792162043586e-06,
+      "loss": 0.5724,
+      "step": 21426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.828110715561251,
+      "learning_rate": 1.0038522342852737e-06,
+      "loss": 0.5007,
+      "step": 21427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7730359498546127,
+      "learning_rate": 1.0038253464902485e-06,
+      "loss": 0.4864,
+      "step": 21428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1897950690704664,
+      "learning_rate": 1.0037985528198477e-06,
+      "loss": 0.4355,
+      "step": 21429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.175053270320593,
+      "learning_rate": 1.0037718532746315e-06,
+      "loss": 0.4086,
+      "step": 21430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4368623273252934,
+      "learning_rate": 1.003745247855159e-06,
+      "loss": 0.4696,
+      "step": 21431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5802550501138448,
+      "learning_rate": 1.0037187365619869e-06,
+      "loss": 0.3871,
+      "step": 21432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.714312182821789,
+      "learning_rate": 1.0036923193956703e-06,
+      "loss": 0.4325,
+      "step": 21433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7411762815991292,
+      "learning_rate": 1.0036659963567624e-06,
+      "loss": 0.5381,
+      "step": 21434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.821503277835487,
+      "learning_rate": 1.0036397674458135e-06,
+      "loss": 0.4854,
+      "step": 21435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6332251147217582,
+      "learning_rate": 1.0036136326633735e-06,
+      "loss": 0.3887,
+      "step": 21436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.932967313003342,
+      "learning_rate": 1.0035875920099892e-06,
+      "loss": 0.3959,
+      "step": 21437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5736441665594632,
+      "learning_rate": 1.0035616454862063e-06,
+      "loss": 0.4487,
+      "step": 21438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5653345169887403,
+      "learning_rate": 1.0035357930925673e-06,
+      "loss": 0.3598,
+      "step": 21439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6579163859494763,
+      "learning_rate": 1.0035100348296135e-06,
+      "loss": 0.5845,
+      "step": 21440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.817250887765402,
+      "learning_rate": 1.003484370697884e-06,
+      "loss": 0.5109,
+      "step": 21441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5390678191699967,
+      "learning_rate": 1.0034588006979167e-06,
+      "loss": 0.3189,
+      "step": 21442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.089883565477259,
+      "learning_rate": 1.0034333248302468e-06,
+      "loss": 0.6042,
+      "step": 21443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9447742719425345,
+      "learning_rate": 1.003407943095407e-06,
+      "loss": 0.7023,
+      "step": 21444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5640713882492154,
+      "learning_rate": 1.0033826554939285e-06,
+      "loss": 0.3818,
+      "step": 21445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0383455138093054,
+      "learning_rate": 1.0033574620263415e-06,
+      "loss": 0.514,
+      "step": 21446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5791549618268312,
+      "learning_rate": 1.003332362693173e-06,
+      "loss": 0.4014,
+      "step": 21447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.877686339580635,
+      "learning_rate": 1.0033073574949492e-06,
+      "loss": 0.3873,
+      "step": 21448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.642321311586344,
+      "learning_rate": 1.0032824464321918e-06,
+      "loss": 0.5012,
+      "step": 21449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5240777769267795,
+      "learning_rate": 1.0032576295054238e-06,
+      "loss": 0.3865,
+      "step": 21450
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.47963061928749084,
+      "eval_runtime": 67.9988,
+      "eval_samples_per_second": 50.148,
+      "eval_steps_per_second": 1.574,
+      "step": 21450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.420362048798033,
+      "learning_rate": 1.0032329067151644e-06,
+      "loss": 0.5233,
+      "step": 21451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5051709974121208,
+      "learning_rate": 1.0032082780619312e-06,
+      "loss": 0.5117,
+      "step": 21452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8254572059079321,
+      "learning_rate": 1.0031837435462395e-06,
+      "loss": 0.4807,
+      "step": 21453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7687066163077765,
+      "learning_rate": 1.0031593031686028e-06,
+      "loss": 0.5011,
+      "step": 21454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4868193436532782,
+      "learning_rate": 1.0031349569295335e-06,
+      "loss": 0.3778,
+      "step": 21455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7323981455361084,
+      "learning_rate": 1.0031107048295403e-06,
+      "loss": 0.5533,
+      "step": 21456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9916022659024688,
+      "learning_rate": 1.0030865468691317e-06,
+      "loss": 0.4483,
+      "step": 21457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.96016326191401,
+      "learning_rate": 1.0030624830488128e-06,
+      "loss": 0.4885,
+      "step": 21458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9290998942653823,
+      "learning_rate": 1.003038513369088e-06,
+      "loss": 0.5031,
+      "step": 21459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.960343381503972,
+      "learning_rate": 1.0030146378304584e-06,
+      "loss": 0.6274,
+      "step": 21460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8952433860732176,
+      "learning_rate": 1.0029908564334245e-06,
+      "loss": 0.5681,
+      "step": 21461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8735329576334594,
+      "learning_rate": 1.002967169178484e-06,
+      "loss": 0.4356,
+      "step": 21462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3561120964747053,
+      "learning_rate": 1.0029435760661322e-06,
+      "loss": 0.281,
+      "step": 21463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5735943989231023,
+      "learning_rate": 1.0029200770968631e-06,
+      "loss": 0.3989,
+      "step": 21464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9964266182830135,
+      "learning_rate": 1.0028966722711697e-06,
+      "loss": 0.5538,
+      "step": 21465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4863458975917505,
+      "learning_rate": 1.0028733615895406e-06,
+      "loss": 0.3255,
+      "step": 21466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0035474694367954,
+      "learning_rate": 1.0028501450524649e-06,
+      "loss": 0.5922,
+      "step": 21467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7038413317851473,
+      "learning_rate": 1.0028270226604277e-06,
+      "loss": 0.4366,
+      "step": 21468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.587708387164177,
+      "learning_rate": 1.0028039944139137e-06,
+      "loss": 0.3979,
+      "step": 21469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8555810788872558,
+      "learning_rate": 1.0027810603134044e-06,
+      "loss": 0.5657,
+      "step": 21470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7178027956852566,
+      "learning_rate": 1.0027582203593808e-06,
+      "loss": 0.3246,
+      "step": 21471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4885959032207856,
+      "learning_rate": 1.0027354745523204e-06,
+      "loss": 0.3294,
+      "step": 21472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.960581831304703,
+      "learning_rate": 1.0027128228926995e-06,
+      "loss": 0.3617,
+      "step": 21473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.096640923303757,
+      "learning_rate": 1.002690265380992e-06,
+      "loss": 0.4893,
+      "step": 21474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.087607961212787,
+      "learning_rate": 1.002667802017671e-06,
+      "loss": 0.4245,
+      "step": 21475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.978125853669614,
+      "learning_rate": 1.0026454328032058e-06,
+      "loss": 0.6439,
+      "step": 21476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.547258076985003,
+      "learning_rate": 1.0026231577380649e-06,
+      "loss": 0.398,
+      "step": 21477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5293982781399056,
+      "learning_rate": 1.002600976822715e-06,
+      "loss": 0.3863,
+      "step": 21478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9179188952123543,
+      "learning_rate": 1.00257889005762e-06,
+      "loss": 0.5301,
+      "step": 21479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0810898413538617,
+      "learning_rate": 1.0025568974432424e-06,
+      "loss": 0.4554,
+      "step": 21480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6931712372764114,
+      "learning_rate": 1.0025349989800432e-06,
+      "loss": 0.4937,
+      "step": 21481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.520565597713062,
+      "learning_rate": 1.0025131946684798e-06,
+      "loss": 0.3269,
+      "step": 21482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8067130289042836,
+      "learning_rate": 1.0024914845090092e-06,
+      "loss": 0.5272,
+      "step": 21483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7983326999807752,
+      "learning_rate": 1.0024698685020863e-06,
+      "loss": 0.5644,
+      "step": 21484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425293052715267,
+      "learning_rate": 1.0024483466481627e-06,
+      "loss": 0.3024,
+      "step": 21485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6897161949971595,
+      "learning_rate": 1.0024269189476898e-06,
+      "loss": 0.4534,
+      "step": 21486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9280599599892982,
+      "learning_rate": 1.0024055854011156e-06,
+      "loss": 0.474,
+      "step": 21487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.789481380277184,
+      "learning_rate": 1.002384346008887e-06,
+      "loss": 0.4103,
+      "step": 21488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0152005275006717,
+      "learning_rate": 1.0023632007714484e-06,
+      "loss": 0.4581,
+      "step": 21489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4297729439599907,
+      "learning_rate": 1.002342149689243e-06,
+      "loss": 0.3455,
+      "step": 21490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.499826953008089,
+      "learning_rate": 1.002321192762711e-06,
+      "loss": 0.398,
+      "step": 21491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.449428514727956,
+      "learning_rate": 1.0023003299922906e-06,
+      "loss": 0.402,
+      "step": 21492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8059217448500469,
+      "learning_rate": 1.0022795613784194e-06,
+      "loss": 0.4854,
+      "step": 21493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8706487755751875,
+      "learning_rate": 1.0022588869215325e-06,
+      "loss": 0.6042,
+      "step": 21494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5532172778190347,
+      "learning_rate": 1.002238306622062e-06,
+      "loss": 0.4511,
+      "step": 21495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9768370362386767,
+      "learning_rate": 1.0022178204804383e-06,
+      "loss": 0.5226,
+      "step": 21496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8063563074824667,
+      "learning_rate": 1.0021974284970916e-06,
+      "loss": 0.577,
+      "step": 21497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6502778323267064,
+      "learning_rate": 1.0021771306724473e-06,
+      "loss": 0.5152,
+      "step": 21498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0525327688721955,
+      "learning_rate": 1.0021569270069317e-06,
+      "loss": 0.5226,
+      "step": 21499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8083285381367475,
+      "learning_rate": 1.0021368175009668e-06,
+      "loss": 0.4701,
+      "step": 21500
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.47988539934158325,
+      "eval_runtime": 68.0346,
+      "eval_samples_per_second": 50.122,
+      "eval_steps_per_second": 1.573,
+      "step": 21500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 21713,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 4501662597120000.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}