diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,33633 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.2278725729175424,
+  "eval_steps": 500,
+  "global_step": 4800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007263977481669807,
+      "grad_norm": 63.2759549108439,
+      "learning_rate": 3.333333333333334e-08,
+      "loss": 7.3622,
+      "step": 1
+    },
+    {
+      "epoch": 0.0014527954963339613,
+      "grad_norm": 63.41948831357221,
+      "learning_rate": 6.666666666666668e-08,
+      "loss": 7.3563,
+      "step": 2
+    },
+    {
+      "epoch": 0.002179193244500942,
+      "grad_norm": 63.20335169729121,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 7.3613,
+      "step": 3
+    },
+    {
+      "epoch": 0.0029055909926679227,
+      "grad_norm": 63.088082938073725,
+      "learning_rate": 1.3333333333333336e-07,
+      "loss": 7.3498,
+      "step": 4
+    },
+    {
+      "epoch": 0.0036319887408349035,
+      "grad_norm": 66.62993398865497,
+      "learning_rate": 1.6666666666666668e-07,
+      "loss": 7.4933,
+      "step": 5
+    },
+    {
+      "epoch": 0.004358386489001884,
+      "grad_norm": 64.28207217419981,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 7.3739,
+      "step": 6
+    },
+    {
+      "epoch": 0.005084784237168865,
+      "grad_norm": 64.86547058325266,
+      "learning_rate": 2.3333333333333336e-07,
+      "loss": 7.4356,
+      "step": 7
+    },
+    {
+      "epoch": 0.005811181985335845,
+      "grad_norm": 62.520037864532036,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 7.3069,
+      "step": 8
+    },
+    {
+      "epoch": 0.006537579733502826,
+      "grad_norm": 64.86566194865063,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 7.4228,
+      "step": 9
+    },
+    {
+      "epoch": 0.007263977481669807,
+      "grad_norm": 62.970846908905436,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 7.3558,
+      "step": 10
+    },
+    {
+      "epoch": 0.007990375229836788,
+      "grad_norm": 63.631610429531904,
+      "learning_rate": 3.666666666666667e-07,
+      "loss": 7.3535,
+      "step": 11
+    },
+    {
+      "epoch": 0.008716772978003768,
+      "grad_norm": 63.25811371761601,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 7.3819,
+      "step": 12
+    },
+    {
+      "epoch": 0.00944317072617075,
+      "grad_norm": 63.66259077849509,
+      "learning_rate": 4.333333333333334e-07,
+      "loss": 7.3795,
+      "step": 13
+    },
+    {
+      "epoch": 0.01016956847433773,
+      "grad_norm": 64.15289940053654,
+      "learning_rate": 4.666666666666667e-07,
+      "loss": 7.3683,
+      "step": 14
+    },
+    {
+      "epoch": 0.01089596622250471,
+      "grad_norm": 64.59626603451477,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 7.3799,
+      "step": 15
+    },
+    {
+      "epoch": 0.01162236397067169,
+      "grad_norm": 63.26977989702329,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 7.3623,
+      "step": 16
+    },
+    {
+      "epoch": 0.012348761718838672,
+      "grad_norm": 60.067823922331,
+      "learning_rate": 5.666666666666667e-07,
+      "loss": 7.1842,
+      "step": 17
+    },
+    {
+      "epoch": 0.013075159467005652,
+      "grad_norm": 63.99871264593463,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 7.236,
+      "step": 18
+    },
+    {
+      "epoch": 0.013801557215172632,
+      "grad_norm": 59.629646935909086,
+      "learning_rate": 6.333333333333334e-07,
+      "loss": 7.1603,
+      "step": 19
+    },
+    {
+      "epoch": 0.014527954963339614,
+      "grad_norm": 62.1060954445667,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 7.283,
+      "step": 20
+    },
+    {
+      "epoch": 0.015254352711506594,
+      "grad_norm": 60.05206589269812,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 7.186,
+      "step": 21
+    },
+    {
+      "epoch": 0.015980750459673575,
+      "grad_norm": 59.645318014317375,
+      "learning_rate": 7.333333333333334e-07,
+      "loss": 7.1921,
+      "step": 22
+    },
+    {
+      "epoch": 0.016707148207840555,
+      "grad_norm": 57.48650185086973,
+      "learning_rate": 7.666666666666667e-07,
+      "loss": 6.9704,
+      "step": 23
+    },
+    {
+      "epoch": 0.017433545956007535,
+      "grad_norm": 53.42583321263901,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 6.8416,
+      "step": 24
+    },
+    {
+      "epoch": 0.01815994370417452,
+      "grad_norm": 53.633682206139234,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 6.8657,
+      "step": 25
+    },
+    {
+      "epoch": 0.0188863414523415,
+      "grad_norm": 53.611957520088204,
+      "learning_rate": 8.666666666666668e-07,
+      "loss": 6.8391,
+      "step": 26
+    },
+    {
+      "epoch": 0.01961273920050848,
+      "grad_norm": 53.90927959387481,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 6.8457,
+      "step": 27
+    },
+    {
+      "epoch": 0.02033913694867546,
+      "grad_norm": 52.34299208362781,
+      "learning_rate": 9.333333333333334e-07,
+      "loss": 6.7246,
+      "step": 28
+    },
+    {
+      "epoch": 0.021065534696842438,
+      "grad_norm": 52.44583643778421,
+      "learning_rate": 9.666666666666668e-07,
+      "loss": 6.7693,
+      "step": 29
+    },
+    {
+      "epoch": 0.02179193244500942,
+      "grad_norm": 50.904404917626735,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 6.6598,
+      "step": 30
+    },
+    {
+      "epoch": 0.0225183301931764,
+      "grad_norm": 51.76963606843732,
+      "learning_rate": 1.0333333333333333e-06,
+      "loss": 6.5898,
+      "step": 31
+    },
+    {
+      "epoch": 0.02324472794134338,
+      "grad_norm": 51.843062752094205,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 6.4639,
+      "step": 32
+    },
+    {
+      "epoch": 0.02397112568951036,
+      "grad_norm": 48.38768669965513,
+      "learning_rate": 1.1e-06,
+      "loss": 6.1057,
+      "step": 33
+    },
+    {
+      "epoch": 0.024697523437677345,
+      "grad_norm": 47.75728492688787,
+      "learning_rate": 1.1333333333333334e-06,
+      "loss": 6.0672,
+      "step": 34
+    },
+    {
+      "epoch": 0.025423921185844325,
+      "grad_norm": 47.1983740094125,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 5.9977,
+      "step": 35
+    },
+    {
+      "epoch": 0.026150318934011305,
+      "grad_norm": 47.80807555427671,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 6.0389,
+      "step": 36
+    },
+    {
+      "epoch": 0.026876716682178284,
+      "grad_norm": 46.78515259105002,
+      "learning_rate": 1.2333333333333335e-06,
+      "loss": 5.9493,
+      "step": 37
+    },
+    {
+      "epoch": 0.027603114430345264,
+      "grad_norm": 46.44941548286642,
+      "learning_rate": 1.2666666666666669e-06,
+      "loss": 5.9096,
+      "step": 38
+    },
+    {
+      "epoch": 0.028329512178512248,
+      "grad_norm": 47.87743744982369,
+      "learning_rate": 1.3e-06,
+      "loss": 5.957,
+      "step": 39
+    },
+    {
+      "epoch": 0.029055909926679228,
+      "grad_norm": 47.69020729829381,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 5.9052,
+      "step": 40
+    },
+    {
+      "epoch": 0.029782307674846208,
+      "grad_norm": 45.65840416434142,
+      "learning_rate": 1.3666666666666668e-06,
+      "loss": 5.7094,
+      "step": 41
+    },
+    {
+      "epoch": 0.030508705423013188,
+      "grad_norm": 46.52695760405757,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 5.6376,
+      "step": 42
+    },
+    {
+      "epoch": 0.03123510317118017,
+      "grad_norm": 47.27139026128274,
+      "learning_rate": 1.4333333333333335e-06,
+      "loss": 5.5706,
+      "step": 43
+    },
+    {
+      "epoch": 0.03196150091934715,
+      "grad_norm": 46.74277727155727,
+      "learning_rate": 1.4666666666666669e-06,
+      "loss": 5.4399,
+      "step": 44
+    },
+    {
+      "epoch": 0.03268789866751413,
+      "grad_norm": 46.31302781518561,
+      "learning_rate": 1.5e-06,
+      "loss": 5.305,
+      "step": 45
+    },
+    {
+      "epoch": 0.03341429641568111,
+      "grad_norm": 47.29508573577366,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 5.2508,
+      "step": 46
+    },
+    {
+      "epoch": 0.034140694163848094,
+      "grad_norm": 46.5497242137799,
+      "learning_rate": 1.566666666666667e-06,
+      "loss": 5.1042,
+      "step": 47
+    },
+    {
+      "epoch": 0.03486709191201507,
+      "grad_norm": 47.310997403673,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 5.0298,
+      "step": 48
+    },
+    {
+      "epoch": 0.035593489660182054,
+      "grad_norm": 47.85952489756574,
+      "learning_rate": 1.6333333333333335e-06,
+      "loss": 4.9936,
+      "step": 49
+    },
+    {
+      "epoch": 0.03631988740834904,
+      "grad_norm": 48.28390256638122,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 4.9282,
+      "step": 50
+    },
+    {
+      "epoch": 0.037046285156516014,
+      "grad_norm": 47.36438783340036,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 4.7947,
+      "step": 51
+    },
+    {
+      "epoch": 0.037772682904683,
+      "grad_norm": 47.97444500743773,
+      "learning_rate": 1.7333333333333336e-06,
+      "loss": 4.7598,
+      "step": 52
+    },
+    {
+      "epoch": 0.038499080652849973,
+      "grad_norm": 48.032965049328915,
+      "learning_rate": 1.7666666666666668e-06,
+      "loss": 4.7174,
+      "step": 53
+    },
+    {
+      "epoch": 0.03922547840101696,
+      "grad_norm": 48.54449670206776,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 4.6641,
+      "step": 54
+    },
+    {
+      "epoch": 0.03995187614918394,
+      "grad_norm": 48.098423617340245,
+      "learning_rate": 1.8333333333333333e-06,
+      "loss": 4.5518,
+      "step": 55
+    },
+    {
+      "epoch": 0.04067827389735092,
+      "grad_norm": 48.398816364766674,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 4.4794,
+      "step": 56
+    },
+    {
+      "epoch": 0.0414046716455179,
+      "grad_norm": 49.27527728561581,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 4.4303,
+      "step": 57
+    },
+    {
+      "epoch": 0.042131069393684877,
+      "grad_norm": 48.371954013728626,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 4.3185,
+      "step": 58
+    },
+    {
+      "epoch": 0.04285746714185186,
+      "grad_norm": 48.5608446730172,
+      "learning_rate": 1.9666666666666668e-06,
+      "loss": 4.2362,
+      "step": 59
+    },
+    {
+      "epoch": 0.04358386489001884,
+      "grad_norm": 48.368964459448684,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 4.1383,
+      "step": 60
+    },
+    {
+      "epoch": 0.04431026263818582,
+      "grad_norm": 48.57285815251858,
+      "learning_rate": 2.0333333333333335e-06,
+      "loss": 4.0723,
+      "step": 61
+    },
+    {
+      "epoch": 0.0450366603863528,
+      "grad_norm": 48.76734862112584,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 4.0063,
+      "step": 62
+    },
+    {
+      "epoch": 0.04576305813451978,
+      "grad_norm": 49.26145502427768,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 3.957,
+      "step": 63
+    },
+    {
+      "epoch": 0.04648945588268676,
+      "grad_norm": 49.50243947973019,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 3.8785,
+      "step": 64
+    },
+    {
+      "epoch": 0.047215853630853746,
+      "grad_norm": 49.828107092810185,
+      "learning_rate": 2.166666666666667e-06,
+      "loss": 3.8176,
+      "step": 65
+    },
+    {
+      "epoch": 0.04794225137902072,
+      "grad_norm": 49.26466685923202,
+      "learning_rate": 2.2e-06,
+      "loss": 3.7495,
+      "step": 66
+    },
+    {
+      "epoch": 0.048668649127187706,
+      "grad_norm": 48.62847142711153,
+      "learning_rate": 2.2333333333333333e-06,
+      "loss": 3.6826,
+      "step": 67
+    },
+    {
+      "epoch": 0.04939504687535469,
+      "grad_norm": 49.05074040352798,
+      "learning_rate": 2.266666666666667e-06,
+      "loss": 3.6296,
+      "step": 68
+    },
+    {
+      "epoch": 0.050121444623521666,
+      "grad_norm": 49.862009737363984,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 3.5896,
+      "step": 69
+    },
+    {
+      "epoch": 0.05084784237168865,
+      "grad_norm": 50.69198753633318,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 3.5724,
+      "step": 70
+    },
+    {
+      "epoch": 0.051574240119855626,
+      "grad_norm": 50.03195157652246,
+      "learning_rate": 2.3666666666666667e-06,
+      "loss": 3.5113,
+      "step": 71
+    },
+    {
+      "epoch": 0.05230063786802261,
+      "grad_norm": 49.23258661177582,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 3.4314,
+      "step": 72
+    },
+    {
+      "epoch": 0.05302703561618959,
+      "grad_norm": 50.580583101959995,
+      "learning_rate": 2.4333333333333335e-06,
+      "loss": 3.411,
+      "step": 73
+    },
+    {
+      "epoch": 0.05375343336435657,
+      "grad_norm": 50.72488608293446,
+      "learning_rate": 2.466666666666667e-06,
+      "loss": 3.3533,
+      "step": 74
+    },
+    {
+      "epoch": 0.05447983111252355,
+      "grad_norm": 50.6273218587908,
+      "learning_rate": 2.5e-06,
+      "loss": 3.3145,
+      "step": 75
+    },
+    {
+      "epoch": 0.05520622886069053,
+      "grad_norm": 50.73713708166495,
+      "learning_rate": 2.5333333333333338e-06,
+      "loss": 3.2674,
+      "step": 76
+    },
+    {
+      "epoch": 0.05593262660885751,
+      "grad_norm": 50.698736069610405,
+      "learning_rate": 2.566666666666667e-06,
+      "loss": 3.2224,
+      "step": 77
+    },
+    {
+      "epoch": 0.056659024357024496,
+      "grad_norm": 49.95915915848869,
+      "learning_rate": 2.6e-06,
+      "loss": 3.1458,
+      "step": 78
+    },
+    {
+      "epoch": 0.05738542210519147,
+      "grad_norm": 50.41144470850246,
+      "learning_rate": 2.6333333333333332e-06,
+      "loss": 3.1436,
+      "step": 79
+    },
+    {
+      "epoch": 0.058111819853358455,
+      "grad_norm": 50.69282224043935,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 3.0908,
+      "step": 80
+    },
+    {
+      "epoch": 0.05883821760152543,
+      "grad_norm": 49.800949375863084,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 3.0296,
+      "step": 81
+    },
+    {
+      "epoch": 0.059564615349692415,
+      "grad_norm": 51.039912830570024,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 2.9943,
+      "step": 82
+    },
+    {
+      "epoch": 0.0602910130978594,
+      "grad_norm": 49.70036616878881,
+      "learning_rate": 2.766666666666667e-06,
+      "loss": 2.9383,
+      "step": 83
+    },
+    {
+      "epoch": 0.061017410846026375,
+      "grad_norm": 50.93003291238077,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.9363,
+      "step": 84
+    },
+    {
+      "epoch": 0.06174380859419336,
+      "grad_norm": 50.34453682192858,
+      "learning_rate": 2.8333333333333335e-06,
+      "loss": 2.8795,
+      "step": 85
+    },
+    {
+      "epoch": 0.06247020634236034,
+      "grad_norm": 50.87658557775924,
+      "learning_rate": 2.866666666666667e-06,
+      "loss": 2.8499,
+      "step": 86
+    },
+    {
+      "epoch": 0.06319660409052733,
+      "grad_norm": 49.93087461716983,
+      "learning_rate": 2.9e-06,
+      "loss": 2.8086,
+      "step": 87
+    },
+    {
+      "epoch": 0.0639230018386943,
+      "grad_norm": 50.674279843425246,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 2.7729,
+      "step": 88
+    },
+    {
+      "epoch": 0.06464939958686128,
+      "grad_norm": 50.663286782284274,
+      "learning_rate": 2.9666666666666673e-06,
+      "loss": 2.7499,
+      "step": 89
+    },
+    {
+      "epoch": 0.06537579733502825,
+      "grad_norm": 50.35183232947615,
+      "learning_rate": 3e-06,
+      "loss": 2.7004,
+      "step": 90
+    },
+    {
+      "epoch": 0.06610219508319524,
+      "grad_norm": 50.447339204931495,
+      "learning_rate": 3.0333333333333337e-06,
+      "loss": 2.6782,
+      "step": 91
+    },
+    {
+      "epoch": 0.06682859283136222,
+      "grad_norm": 50.08671877556182,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 2.6386,
+      "step": 92
+    },
+    {
+      "epoch": 0.0675549905795292,
+      "grad_norm": 50.85016696537039,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 2.6105,
+      "step": 93
+    },
+    {
+      "epoch": 0.06828138832769619,
+      "grad_norm": 50.432484997573745,
+      "learning_rate": 3.133333333333334e-06,
+      "loss": 2.579,
+      "step": 94
+    },
+    {
+      "epoch": 0.06900778607586316,
+      "grad_norm": 49.822703178081824,
+      "learning_rate": 3.1666666666666667e-06,
+      "loss": 2.5516,
+      "step": 95
+    },
+    {
+      "epoch": 0.06973418382403014,
+      "grad_norm": 49.52642010085431,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.4905,
+      "step": 96
+    },
+    {
+      "epoch": 0.07046058157219713,
+      "grad_norm": 49.393946360680815,
+      "learning_rate": 3.2333333333333334e-06,
+      "loss": 2.4803,
+      "step": 97
+    },
+    {
+      "epoch": 0.07118697932036411,
+      "grad_norm": 48.897469338046086,
+      "learning_rate": 3.266666666666667e-06,
+      "loss": 2.4483,
+      "step": 98
+    },
+    {
+      "epoch": 0.07191337706853108,
+      "grad_norm": 49.378935763876314,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 2.4292,
+      "step": 99
+    },
+    {
+      "epoch": 0.07263977481669807,
+      "grad_norm": 50.41695005032472,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.4022,
+      "step": 100
+    },
+    {
+      "epoch": 0.07336617256486505,
+      "grad_norm": 50.2049353156204,
+      "learning_rate": 3.366666666666667e-06,
+      "loss": 2.3749,
+      "step": 101
+    },
+    {
+      "epoch": 0.07409257031303203,
+      "grad_norm": 49.693566201239186,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 2.3459,
+      "step": 102
+    },
+    {
+      "epoch": 0.074818968061199,
+      "grad_norm": 48.84685424655744,
+      "learning_rate": 3.4333333333333336e-06,
+      "loss": 2.334,
+      "step": 103
+    },
+    {
+      "epoch": 0.075545365809366,
+      "grad_norm": 47.62798075643798,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 2.2763,
+      "step": 104
+    },
+    {
+      "epoch": 0.07627176355753297,
+      "grad_norm": 48.19573541336029,
+      "learning_rate": 3.5e-06,
+      "loss": 2.265,
+      "step": 105
+    },
+    {
+      "epoch": 0.07699816130569995,
+      "grad_norm": 49.45197671503261,
+      "learning_rate": 3.5333333333333335e-06,
+      "loss": 2.2395,
+      "step": 106
+    },
+    {
+      "epoch": 0.07772455905386694,
+      "grad_norm": 47.40823487468658,
+      "learning_rate": 3.566666666666667e-06,
+      "loss": 2.2103,
+      "step": 107
+    },
+    {
+      "epoch": 0.07845095680203391,
+      "grad_norm": 47.671328668080086,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.1863,
+      "step": 108
+    },
+    {
+      "epoch": 0.07917735455020089,
+      "grad_norm": 47.279964152668605,
+      "learning_rate": 3.633333333333334e-06,
+      "loss": 2.1527,
+      "step": 109
+    },
+    {
+      "epoch": 0.07990375229836788,
+      "grad_norm": 46.06802031513997,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 2.1353,
+      "step": 110
+    },
+    {
+      "epoch": 0.08063015004653486,
+      "grad_norm": 46.73470181620622,
+      "learning_rate": 3.7e-06,
+      "loss": 2.1074,
+      "step": 111
+    },
+    {
+      "epoch": 0.08135654779470183,
+      "grad_norm": 47.34253122512511,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 2.0864,
+      "step": 112
+    },
+    {
+      "epoch": 0.08208294554286882,
+      "grad_norm": 46.64992121387836,
+      "learning_rate": 3.766666666666667e-06,
+      "loss": 2.072,
+      "step": 113
+    },
+    {
+      "epoch": 0.0828093432910358,
+      "grad_norm": 46.30918717601416,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 2.0391,
+      "step": 114
+    },
+    {
+      "epoch": 0.08353574103920278,
+      "grad_norm": 46.61354559014084,
+      "learning_rate": 3.833333333333334e-06,
+      "loss": 2.014,
+      "step": 115
+    },
+    {
+      "epoch": 0.08426213878736975,
+      "grad_norm": 45.95154255629814,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 2.0111,
+      "step": 116
+    },
+    {
+      "epoch": 0.08498853653553674,
+      "grad_norm": 45.17776725607547,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.9856,
+      "step": 117
+    },
+    {
+      "epoch": 0.08571493428370372,
+      "grad_norm": 45.131837668001495,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 1.9698,
+      "step": 118
+    },
+    {
+      "epoch": 0.0864413320318707,
+      "grad_norm": 44.877128843538,
+      "learning_rate": 3.966666666666667e-06,
+      "loss": 1.9418,
+      "step": 119
+    },
+    {
+      "epoch": 0.08716772978003769,
+      "grad_norm": 43.42147842225612,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.9271,
+      "step": 120
+    },
+    {
+      "epoch": 0.08789412752820466,
+      "grad_norm": 43.9251239135038,
+      "learning_rate": 4.033333333333333e-06,
+      "loss": 1.9074,
+      "step": 121
+    },
+    {
+      "epoch": 0.08862052527637164,
+      "grad_norm": 45.52313461713586,
+      "learning_rate": 4.066666666666667e-06,
+      "loss": 1.8972,
+      "step": 122
+    },
+    {
+      "epoch": 0.08934692302453863,
+      "grad_norm": 43.96000925438749,
+      "learning_rate": 4.1e-06,
+      "loss": 1.867,
+      "step": 123
+    },
+    {
+      "epoch": 0.0900733207727056,
+      "grad_norm": 43.27066380698389,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 1.8483,
+      "step": 124
+    },
+    {
+      "epoch": 0.09079971852087258,
+      "grad_norm": 42.70094664578436,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.8224,
+      "step": 125
+    },
+    {
+      "epoch": 0.09152611626903956,
+      "grad_norm": 41.5753156714769,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.8262,
+      "step": 126
+    },
+    {
+      "epoch": 0.09225251401720655,
+      "grad_norm": 41.898261769934656,
+      "learning_rate": 4.233333333333334e-06,
+      "loss": 1.787,
+      "step": 127
+    },
+    {
+      "epoch": 0.09297891176537353,
+      "grad_norm": 42.0872491446625,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.7713,
+      "step": 128
+    },
+    {
+      "epoch": 0.0937053095135405,
+      "grad_norm": 41.19380423975876,
+      "learning_rate": 4.3e-06,
+      "loss": 1.7568,
+      "step": 129
+    },
+    {
+      "epoch": 0.09443170726170749,
+      "grad_norm": 40.680968528405074,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.7495,
+      "step": 130
+    },
+    {
+      "epoch": 0.09515810500987447,
+      "grad_norm": 40.16939074899239,
+      "learning_rate": 4.366666666666667e-06,
+      "loss": 1.7404,
+      "step": 131
+    },
+    {
+      "epoch": 0.09588450275804145,
+      "grad_norm": 39.72263478772674,
+      "learning_rate": 4.4e-06,
+      "loss": 1.7212,
+      "step": 132
+    },
+    {
+      "epoch": 0.09661090050620844,
+      "grad_norm": 39.49181775364284,
+      "learning_rate": 4.433333333333334e-06,
+      "loss": 1.6969,
+      "step": 133
+    },
+    {
+      "epoch": 0.09733729825437541,
+      "grad_norm": 39.04439034347531,
+      "learning_rate": 4.4666666666666665e-06,
+      "loss": 1.6875,
+      "step": 134
+    },
+    {
+      "epoch": 0.09806369600254239,
+      "grad_norm": 38.443498008849005,
+      "learning_rate": 4.5e-06,
+      "loss": 1.693,
+      "step": 135
+    },
+    {
+      "epoch": 0.09879009375070938,
+      "grad_norm": 38.36227491759409,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.659,
+      "step": 136
+    },
+    {
+      "epoch": 0.09951649149887636,
+      "grad_norm": 37.3629590490201,
+      "learning_rate": 4.566666666666667e-06,
+      "loss": 1.6459,
+      "step": 137
+    },
+    {
+      "epoch": 0.10024288924704333,
+      "grad_norm": 37.54793761307316,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.6365,
+      "step": 138
+    },
+    {
+      "epoch": 0.10096928699521031,
+      "grad_norm": 36.18933616176955,
+      "learning_rate": 4.633333333333334e-06,
+      "loss": 1.631,
+      "step": 139
+    },
+    {
+      "epoch": 0.1016956847433773,
+      "grad_norm": 36.69815961845787,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.6211,
+      "step": 140
+    },
+    {
+      "epoch": 0.10242208249154428,
+      "grad_norm": 36.12103329793505,
+      "learning_rate": 4.7e-06,
+      "loss": 1.6008,
+      "step": 141
+    },
+    {
+      "epoch": 0.10314848023971125,
+      "grad_norm": 34.75085594638176,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 1.5941,
+      "step": 142
+    },
+    {
+      "epoch": 0.10387487798787824,
+      "grad_norm": 35.264396011673654,
+      "learning_rate": 4.766666666666667e-06,
+      "loss": 1.5712,
+      "step": 143
+    },
+    {
+      "epoch": 0.10460127573604522,
+      "grad_norm": 35.15237015020466,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.5638,
+      "step": 144
+    },
+    {
+      "epoch": 0.1053276734842122,
+      "grad_norm": 34.263602593586796,
+      "learning_rate": 4.833333333333333e-06,
+      "loss": 1.554,
+      "step": 145
+    },
+    {
+      "epoch": 0.10605407123237919,
+      "grad_norm": 33.753618127877616,
+      "learning_rate": 4.866666666666667e-06,
+      "loss": 1.5283,
+      "step": 146
+    },
+    {
+      "epoch": 0.10678046898054616,
+      "grad_norm": 33.839298156814806,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.5213,
+      "step": 147
+    },
+    {
+      "epoch": 0.10750686672871314,
+      "grad_norm": 33.52850494542813,
+      "learning_rate": 4.933333333333334e-06,
+      "loss": 1.5176,
+      "step": 148
+    },
+    {
+      "epoch": 0.10823326447688013,
+      "grad_norm": 32.69335919755698,
+      "learning_rate": 4.966666666666667e-06,
+      "loss": 1.5137,
+      "step": 149
+    },
+    {
+      "epoch": 0.1089596622250471,
+      "grad_norm": 32.83997712477471,
+      "learning_rate": 5e-06,
+      "loss": 1.4922,
+      "step": 150
+    },
+    {
+      "epoch": 0.10968605997321408,
+      "grad_norm": 31.70298573457259,
+      "learning_rate": 5.033333333333333e-06,
+      "loss": 1.4865,
+      "step": 151
+    },
+    {
+      "epoch": 0.11041245772138106,
+      "grad_norm": 31.601502772523528,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.4676,
+      "step": 152
+    },
+    {
+      "epoch": 0.11113885546954805,
+      "grad_norm": 31.921384648076824,
+      "learning_rate": 5.1e-06,
+      "loss": 1.4523,
+      "step": 153
+    },
+    {
+      "epoch": 0.11186525321771502,
+      "grad_norm": 31.225511458265583,
+      "learning_rate": 5.133333333333334e-06,
+      "loss": 1.4455,
+      "step": 154
+    },
+    {
+      "epoch": 0.112591650965882,
+      "grad_norm": 30.32788022406298,
+      "learning_rate": 5.1666666666666675e-06,
+      "loss": 1.4522,
+      "step": 155
+    },
+    {
+      "epoch": 0.11331804871404899,
+      "grad_norm": 30.167607122123012,
+      "learning_rate": 5.2e-06,
+      "loss": 1.4346,
+      "step": 156
+    },
+    {
+      "epoch": 0.11404444646221597,
+      "grad_norm": 29.780741495847135,
+      "learning_rate": 5.233333333333334e-06,
+      "loss": 1.4242,
+      "step": 157
+    },
+    {
+      "epoch": 0.11477084421038294,
+      "grad_norm": 29.145633047690872,
+      "learning_rate": 5.2666666666666665e-06,
+      "loss": 1.4239,
+      "step": 158
+    },
+    {
+      "epoch": 0.11549724195854993,
+      "grad_norm": 28.905395314413873,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 1.4028,
+      "step": 159
+    },
+    {
+      "epoch": 0.11622363970671691,
+      "grad_norm": 27.95492252225459,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.4024,
+      "step": 160
+    },
+    {
+      "epoch": 0.11695003745488389,
+      "grad_norm": 28.364454949347795,
+      "learning_rate": 5.366666666666666e-06,
+      "loss": 1.3966,
+      "step": 161
+    },
+    {
+      "epoch": 0.11767643520305086,
+      "grad_norm": 27.595047874488912,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.3804,
+      "step": 162
+    },
+    {
+      "epoch": 0.11840283295121785,
+      "grad_norm": 27.224371964119193,
+      "learning_rate": 5.4333333333333335e-06,
+      "loss": 1.374,
+      "step": 163
+    },
+    {
+      "epoch": 0.11912923069938483,
+      "grad_norm": 26.950172876003883,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 1.3664,
+      "step": 164
+    },
+    {
+      "epoch": 0.1198556284475518,
+      "grad_norm": 26.807043896561844,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.3463,
+      "step": 165
+    },
+    {
+      "epoch": 0.1205820261957188,
+      "grad_norm": 25.770582681589264,
+      "learning_rate": 5.533333333333334e-06,
+      "loss": 1.3464,
+      "step": 166
+    },
+    {
+      "epoch": 0.12130842394388577,
+      "grad_norm": 26.002994801093113,
+      "learning_rate": 5.566666666666667e-06,
+      "loss": 1.3223,
+      "step": 167
+    },
+    {
+      "epoch": 0.12203482169205275,
+      "grad_norm": 25.487072733988494,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.3284,
+      "step": 168
+    },
+    {
+      "epoch": 0.12276121944021974,
+      "grad_norm": 24.657995803131406,
+      "learning_rate": 5.633333333333334e-06,
+      "loss": 1.3275,
+      "step": 169
+    },
+    {
+      "epoch": 0.12348761718838672,
+      "grad_norm": 24.472525282080856,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.3198,
+      "step": 170
+    },
+    {
+      "epoch": 0.1242140149365537,
+      "grad_norm": 24.396214212939338,
+      "learning_rate": 5.7e-06,
+      "loss": 1.3064,
+      "step": 171
+    },
+    {
+      "epoch": 0.12494041268472068,
+      "grad_norm": 23.954881409211772,
+      "learning_rate": 5.733333333333334e-06,
+      "loss": 1.302,
+      "step": 172
+    },
+    {
+      "epoch": 0.12566681043288766,
+      "grad_norm": 23.260344993149822,
+      "learning_rate": 5.766666666666667e-06,
+      "loss": 1.2848,
+      "step": 173
+    },
+    {
+      "epoch": 0.12639320818105465,
+      "grad_norm": 23.023966618398276,
+      "learning_rate": 5.8e-06,
+      "loss": 1.284,
+      "step": 174
+    },
+    {
+      "epoch": 0.1271196059292216,
+      "grad_norm": 22.329009297749824,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.2843,
+      "step": 175
+    },
+    {
+      "epoch": 0.1278460036773886,
+      "grad_norm": 22.5351171358773,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 1.2682,
+      "step": 176
+    },
+    {
+      "epoch": 0.1285724014255556,
+      "grad_norm": 22.009496715465836,
+      "learning_rate": 5.9e-06,
+      "loss": 1.2625,
+      "step": 177
+    },
+    {
+      "epoch": 0.12929879917372256,
+      "grad_norm": 21.965579290780184,
+      "learning_rate": 5.933333333333335e-06,
+      "loss": 1.2579,
+      "step": 178
+    },
+    {
+      "epoch": 0.13002519692188955,
+      "grad_norm": 21.682618987574696,
+      "learning_rate": 5.966666666666667e-06,
+      "loss": 1.2549,
+      "step": 179
+    },
+    {
+      "epoch": 0.1307515946700565,
+      "grad_norm": 21.080469805135827,
+      "learning_rate": 6e-06,
+      "loss": 1.2388,
+      "step": 180
+    },
+    {
+      "epoch": 0.1314779924182235,
+      "grad_norm": 20.967610245749125,
+      "learning_rate": 6.033333333333335e-06,
+      "loss": 1.2265,
+      "step": 181
+    },
+    {
+      "epoch": 0.1322043901663905,
+      "grad_norm": 20.1492358534322,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 1.2454,
+      "step": 182
+    },
+    {
+      "epoch": 0.13293078791455745,
+      "grad_norm": 19.607412502409314,
+      "learning_rate": 6.1e-06,
+      "loss": 1.2614,
+      "step": 183
+    },
+    {
+      "epoch": 0.13365718566272444,
+      "grad_norm": 19.346758211452215,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.2399,
+      "step": 184
+    },
+    {
+      "epoch": 0.13438358341089143,
+      "grad_norm": 19.257583489746565,
+      "learning_rate": 6.166666666666667e-06,
+      "loss": 1.232,
+      "step": 185
+    },
+    {
+      "epoch": 0.1351099811590584,
+      "grad_norm": 18.86454673675082,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.219,
+      "step": 186
+    },
+    {
+      "epoch": 0.13583637890722539,
+      "grad_norm": 18.46080931559179,
+      "learning_rate": 6.2333333333333335e-06,
+      "loss": 1.2205,
+      "step": 187
+    },
+    {
+      "epoch": 0.13656277665539238,
+      "grad_norm": 18.28704328355964,
+      "learning_rate": 6.266666666666668e-06,
+      "loss": 1.2031,
+      "step": 188
+    },
+    {
+      "epoch": 0.13728917440355934,
+      "grad_norm": 18.229753602356926,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 1.2197,
+      "step": 189
+    },
+    {
+      "epoch": 0.13801557215172633,
+      "grad_norm": 17.836618035828575,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 1.1986,
+      "step": 190
+    },
+    {
+      "epoch": 0.13874196989989332,
+      "grad_norm": 17.524100548365034,
+      "learning_rate": 6.366666666666668e-06,
+      "loss": 1.2009,
+      "step": 191
+    },
+    {
+      "epoch": 0.13946836764806028,
+      "grad_norm": 17.365264025546907,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.1699,
+      "step": 192
+    },
+    {
+      "epoch": 0.14019476539622727,
+      "grad_norm": 17.146603082680905,
+      "learning_rate": 6.433333333333333e-06,
+      "loss": 1.1738,
+      "step": 193
+    },
+    {
+      "epoch": 0.14092116314439426,
+      "grad_norm": 16.859962202084272,
+      "learning_rate": 6.466666666666667e-06,
+      "loss": 1.1783,
+      "step": 194
+    },
+    {
+      "epoch": 0.14164756089256123,
+      "grad_norm": 16.420121457022596,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.1685,
+      "step": 195
+    },
+    {
+      "epoch": 0.14237395864072822,
+      "grad_norm": 16.275858682293666,
+      "learning_rate": 6.533333333333334e-06,
+      "loss": 1.1689,
+      "step": 196
+    },
+    {
+      "epoch": 0.1431003563888952,
+      "grad_norm": 15.72759862599527,
+      "learning_rate": 6.566666666666667e-06,
+      "loss": 1.1618,
+      "step": 197
+    },
+    {
+      "epoch": 0.14382675413706217,
+      "grad_norm": 15.213709034432256,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.1834,
+      "step": 198
+    },
+    {
+      "epoch": 0.14455315188522916,
+      "grad_norm": 15.44383632151497,
+      "learning_rate": 6.633333333333334e-06,
+      "loss": 1.1606,
+      "step": 199
+    },
+    {
+      "epoch": 0.14527954963339615,
+      "grad_norm": 15.0297402495118,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1643,
+      "step": 200
+    },
+    {
+      "epoch": 0.1460059473815631,
+      "grad_norm": 14.616753595774137,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 1.1572,
+      "step": 201
+    },
+    {
+      "epoch": 0.1467323451297301,
+      "grad_norm": 14.988102097685823,
+      "learning_rate": 6.733333333333334e-06,
+      "loss": 1.1223,
+      "step": 202
+    },
+    {
+      "epoch": 0.1474587428778971,
+      "grad_norm": 14.154002560143741,
+      "learning_rate": 6.7666666666666665e-06,
+      "loss": 1.1482,
+      "step": 203
+    },
+    {
+      "epoch": 0.14818514062606405,
+      "grad_norm": 14.070198384045403,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.137,
+      "step": 204
+    },
+    {
+      "epoch": 0.14891153837423105,
+      "grad_norm": 14.00683253820219,
+      "learning_rate": 6.833333333333334e-06,
+      "loss": 1.1349,
+      "step": 205
+    },
+    {
+      "epoch": 0.149637936122398,
+      "grad_norm": 13.651960917625363,
+      "learning_rate": 6.866666666666667e-06,
+      "loss": 1.1313,
+      "step": 206
+    },
+    {
+      "epoch": 0.150364333870565,
+      "grad_norm": 13.649166736813594,
+      "learning_rate": 6.9e-06,
+      "loss": 1.1372,
+      "step": 207
+    },
+    {
+      "epoch": 0.151090731618732,
+      "grad_norm": 13.321459953919788,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 1.121,
+      "step": 208
+    },
+    {
+      "epoch": 0.15181712936689895,
+      "grad_norm": 13.18972666324515,
+      "learning_rate": 6.966666666666667e-06,
+      "loss": 1.1196,
+      "step": 209
+    },
+    {
+      "epoch": 0.15254352711506594,
+      "grad_norm": 13.023718213326685,
+      "learning_rate": 7e-06,
+      "loss": 1.1161,
+      "step": 210
+    },
+    {
+      "epoch": 0.15326992486323293,
+      "grad_norm": 12.778850791700851,
+      "learning_rate": 7.033333333333334e-06,
+      "loss": 1.1333,
+      "step": 211
+    },
+    {
+      "epoch": 0.1539963226113999,
+      "grad_norm": 12.636516960484471,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 1.1124,
+      "step": 212
+    },
+    {
+      "epoch": 0.15472272035956688,
+      "grad_norm": 12.497406739744827,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 1.1297,
+      "step": 213
+    },
+    {
+      "epoch": 0.15544911810773387,
+      "grad_norm": 12.253086279848734,
+      "learning_rate": 7.133333333333334e-06,
+      "loss": 1.108,
+      "step": 214
+    },
+    {
+      "epoch": 0.15617551585590084,
+      "grad_norm": 12.301343331028344,
+      "learning_rate": 7.166666666666667e-06,
+      "loss": 1.0704,
+      "step": 215
+    },
+    {
+      "epoch": 0.15690191360406783,
+      "grad_norm": 11.856234190055382,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.106,
+      "step": 216
+    },
+    {
+      "epoch": 0.15762831135223482,
+      "grad_norm": 11.58489034219279,
+      "learning_rate": 7.233333333333334e-06,
+      "loss": 1.1039,
+      "step": 217
+    },
+    {
+      "epoch": 0.15835470910040178,
+      "grad_norm": 11.40452942412662,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 1.107,
+      "step": 218
+    },
+    {
+      "epoch": 0.15908110684856877,
+      "grad_norm": 11.492550763214666,
+      "learning_rate": 7.3e-06,
+      "loss": 1.0881,
+      "step": 219
+    },
+    {
+      "epoch": 0.15980750459673576,
+      "grad_norm": 11.269901967107916,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.0904,
+      "step": 220
+    },
+    {
+      "epoch": 0.16053390234490272,
+      "grad_norm": 10.96158775655295,
+      "learning_rate": 7.3666666666666676e-06,
+      "loss": 1.0952,
+      "step": 221
+    },
+    {
+      "epoch": 0.16126030009306971,
+      "grad_norm": 11.13901163124956,
+      "learning_rate": 7.4e-06,
+      "loss": 1.074,
+      "step": 222
+    },
+    {
+      "epoch": 0.1619866978412367,
+      "grad_norm": 10.84361490879539,
+      "learning_rate": 7.433333333333334e-06,
+      "loss": 1.0775,
+      "step": 223
+    },
+    {
+      "epoch": 0.16271309558940367,
+      "grad_norm": 10.809242398633671,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 1.0599,
+      "step": 224
+    },
+    {
+      "epoch": 0.16343949333757066,
+      "grad_norm": 10.357828962160399,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.0719,
+      "step": 225
+    },
+    {
+      "epoch": 0.16416589108573765,
+      "grad_norm": 10.16294538034498,
+      "learning_rate": 7.533333333333334e-06,
+      "loss": 1.0737,
+      "step": 226
+    },
+    {
+      "epoch": 0.1648922888339046,
+      "grad_norm": 10.39061500039731,
+      "learning_rate": 7.566666666666667e-06,
+      "loss": 1.0556,
+      "step": 227
+    },
+    {
+      "epoch": 0.1656186865820716,
+      "grad_norm": 10.221493241719058,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.0476,
+      "step": 228
+    },
+    {
+      "epoch": 0.16634508433023856,
+      "grad_norm": 9.907815020465414,
+      "learning_rate": 7.633333333333334e-06,
+      "loss": 1.0553,
+      "step": 229
+    },
+    {
+      "epoch": 0.16707148207840555,
+      "grad_norm": 9.581041133597106,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.0775,
+      "step": 230
+    },
+    {
+      "epoch": 0.16779787982657254,
+      "grad_norm": 9.71168279973024,
+      "learning_rate": 7.7e-06,
+      "loss": 1.0639,
+      "step": 231
+    },
+    {
+      "epoch": 0.1685242775747395,
+      "grad_norm": 9.551789851749682,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 1.0503,
+      "step": 232
+    },
+    {
+      "epoch": 0.1692506753229065,
+      "grad_norm": 9.376979470265642,
+      "learning_rate": 7.766666666666666e-06,
+      "loss": 1.0605,
+      "step": 233
+    },
+    {
+      "epoch": 0.1699770730710735,
+      "grad_norm": 9.215487743984946,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.0529,
+      "step": 234
+    },
+    {
+      "epoch": 0.17070347081924045,
+      "grad_norm": 9.023321477961552,
+      "learning_rate": 7.833333333333333e-06,
+      "loss": 1.06,
+      "step": 235
+    },
+    {
+      "epoch": 0.17142986856740744,
+      "grad_norm": 8.978074772661731,
+      "learning_rate": 7.866666666666667e-06,
+      "loss": 1.0296,
+      "step": 236
+    },
+    {
+      "epoch": 0.17215626631557443,
+      "grad_norm": 8.837716165817618,
+      "learning_rate": 7.9e-06,
+      "loss": 1.0255,
+      "step": 237
+    },
+    {
+      "epoch": 0.1728826640637414,
+      "grad_norm": 8.634089579695464,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 1.0436,
+      "step": 238
+    },
+    {
+      "epoch": 0.17360906181190838,
+      "grad_norm": 8.239794943770667,
+      "learning_rate": 7.966666666666668e-06,
+      "loss": 1.0565,
+      "step": 239
+    },
+    {
+      "epoch": 0.17433545956007537,
+      "grad_norm": 8.517474915317166,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0147,
+      "step": 240
+    },
+    {
+      "epoch": 0.17506185730824234,
+      "grad_norm": 8.270883045328144,
+      "learning_rate": 8.033333333333335e-06,
+      "loss": 1.0425,
+      "step": 241
+    },
+    {
+      "epoch": 0.17578825505640933,
+      "grad_norm": 8.19007825276564,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 1.0255,
+      "step": 242
+    },
+    {
+      "epoch": 0.17651465280457632,
+      "grad_norm": 8.020665101381212,
+      "learning_rate": 8.1e-06,
+      "loss": 1.0372,
+      "step": 243
+    },
+    {
+      "epoch": 0.17724105055274328,
+      "grad_norm": 8.078219951557776,
+      "learning_rate": 8.133333333333334e-06,
+      "loss": 1.023,
+      "step": 244
+    },
+    {
+      "epoch": 0.17796744830091027,
+      "grad_norm": 7.830702554013111,
+      "learning_rate": 8.166666666666668e-06,
+      "loss": 1.0366,
+      "step": 245
+    },
+    {
+      "epoch": 0.17869384604907726,
+      "grad_norm": 7.820391288845364,
+      "learning_rate": 8.2e-06,
+      "loss": 1.0234,
+      "step": 246
+    },
+    {
+      "epoch": 0.17942024379724422,
+      "grad_norm": 7.607710068511707,
+      "learning_rate": 8.233333333333335e-06,
+      "loss": 1.0226,
+      "step": 247
+    },
+    {
+      "epoch": 0.1801466415454112,
+      "grad_norm": 7.541222252515193,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 1.0205,
+      "step": 248
+    },
+    {
+      "epoch": 0.1808730392935782,
+      "grad_norm": 7.602914271454926,
+      "learning_rate": 8.3e-06,
+      "loss": 1.0057,
+      "step": 249
+    },
+    {
+      "epoch": 0.18159943704174517,
+      "grad_norm": 7.41561445398044,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.0091,
+      "step": 250
+    },
+    {
+      "epoch": 0.18232583478991216,
+      "grad_norm": 7.428385752483024,
+      "learning_rate": 8.366666666666667e-06,
+      "loss": 1.0091,
+      "step": 251
+    },
+    {
+      "epoch": 0.18305223253807912,
+      "grad_norm": 7.215990147777718,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.0165,
+      "step": 252
+    },
+    {
+      "epoch": 0.1837786302862461,
+      "grad_norm": 7.137875961690411,
+      "learning_rate": 8.433333333333334e-06,
+      "loss": 0.9942,
+      "step": 253
+    },
+    {
+      "epoch": 0.1845050280344131,
+      "grad_norm": 7.178095525054059,
+      "learning_rate": 8.466666666666668e-06,
+      "loss": 0.9913,
+      "step": 254
+    },
+    {
+      "epoch": 0.18523142578258006,
+      "grad_norm": 6.8632846366421285,
+      "learning_rate": 8.5e-06,
+      "loss": 1.0196,
+      "step": 255
+    },
+    {
+      "epoch": 0.18595782353074705,
+      "grad_norm": 6.91059218104985,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 1.0035,
+      "step": 256
+    },
+    {
+      "epoch": 0.18668422127891404,
+      "grad_norm": 6.756354737848418,
+      "learning_rate": 8.566666666666667e-06,
+      "loss": 1.0071,
+      "step": 257
+    },
+    {
+      "epoch": 0.187410619027081,
+      "grad_norm": 6.672245371900697,
+      "learning_rate": 8.6e-06,
+      "loss": 1.0146,
+      "step": 258
+    },
+    {
+      "epoch": 0.188137016775248,
+      "grad_norm": 6.664492580555999,
+      "learning_rate": 8.633333333333334e-06,
+      "loss": 1.0041,
+      "step": 259
+    },
+    {
+      "epoch": 0.18886341452341499,
+      "grad_norm": 6.630116579252964,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.9909,
+      "step": 260
+    },
+    {
+      "epoch": 0.18958981227158195,
+      "grad_norm": 6.540452718326813,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 0.9957,
+      "step": 261
+    },
+    {
+      "epoch": 0.19031621001974894,
+      "grad_norm": 6.341367627655762,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 1.0009,
+      "step": 262
+    },
+    {
+      "epoch": 0.19104260776791593,
+      "grad_norm": 6.234409930165991,
+      "learning_rate": 8.766666666666669e-06,
+      "loss": 1.001,
+      "step": 263
+    },
+    {
+      "epoch": 0.1917690055160829,
+      "grad_norm": 6.305502441912889,
+      "learning_rate": 8.8e-06,
+      "loss": 0.9935,
+      "step": 264
+    },
+    {
+      "epoch": 0.19249540326424988,
+      "grad_norm": 6.225158071475369,
+      "learning_rate": 8.833333333333334e-06,
+      "loss": 1.0081,
+      "step": 265
+    },
+    {
+      "epoch": 0.19322180101241687,
+      "grad_norm": 6.058144207484882,
+      "learning_rate": 8.866666666666668e-06,
+      "loss": 0.9841,
+      "step": 266
+    },
+    {
+      "epoch": 0.19394819876058383,
+      "grad_norm": 5.98628355833425,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 0.9876,
+      "step": 267
+    },
+    {
+      "epoch": 0.19467459650875082,
+      "grad_norm": 5.943991083458489,
+      "learning_rate": 8.933333333333333e-06,
+      "loss": 0.997,
+      "step": 268
+    },
+    {
+      "epoch": 0.19540099425691781,
+      "grad_norm": 5.873982099655574,
+      "learning_rate": 8.966666666666667e-06,
+      "loss": 0.978,
+      "step": 269
+    },
+    {
+      "epoch": 0.19612739200508478,
+      "grad_norm": 5.895132355616301,
+      "learning_rate": 9e-06,
+      "loss": 0.9741,
+      "step": 270
+    },
+    {
+      "epoch": 0.19685378975325177,
+      "grad_norm": 5.769369464159913,
+      "learning_rate": 9.033333333333334e-06,
+      "loss": 0.9732,
+      "step": 271
+    },
+    {
+      "epoch": 0.19758018750141876,
+      "grad_norm": 5.576420099083153,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 0.9892,
+      "step": 272
+    },
+    {
+      "epoch": 0.19830658524958572,
+      "grad_norm": 5.674605536835416,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 0.9647,
+      "step": 273
+    },
+    {
+      "epoch": 0.1990329829977527,
+      "grad_norm": 5.689962851483039,
+      "learning_rate": 9.133333333333335e-06,
+      "loss": 0.9561,
+      "step": 274
+    },
+    {
+      "epoch": 0.1997593807459197,
+      "grad_norm": 5.451787863746172,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.9693,
+      "step": 275
+    },
+    {
+      "epoch": 0.20048577849408666,
+      "grad_norm": 5.508273579548314,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.9476,
+      "step": 276
+    },
+    {
+      "epoch": 0.20121217624225365,
+      "grad_norm": 5.346002804833987,
+      "learning_rate": 9.233333333333334e-06,
+      "loss": 0.9492,
+      "step": 277
+    },
+    {
+      "epoch": 0.20193857399042062,
+      "grad_norm": 5.315336149698185,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 0.9608,
+      "step": 278
+    },
+    {
+      "epoch": 0.2026649717385876,
+      "grad_norm": 5.215861439520431,
+      "learning_rate": 9.3e-06,
+      "loss": 0.9854,
+      "step": 279
+    },
+    {
+      "epoch": 0.2033913694867546,
+      "grad_norm": 5.193177826479856,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.9604,
+      "step": 280
+    },
+    {
+      "epoch": 0.20411776723492156,
+      "grad_norm": 5.123020615984874,
+      "learning_rate": 9.366666666666668e-06,
+      "loss": 0.9585,
+      "step": 281
+    },
+    {
+      "epoch": 0.20484416498308855,
+      "grad_norm": 5.029202178322807,
+      "learning_rate": 9.4e-06,
+      "loss": 0.9733,
+      "step": 282
+    },
+    {
+      "epoch": 0.20557056273125554,
+      "grad_norm": 5.013483166254676,
+      "learning_rate": 9.433333333333335e-06,
+      "loss": 0.9684,
+      "step": 283
+    },
+    {
+      "epoch": 0.2062969604794225,
+      "grad_norm": 4.887824924314841,
+      "learning_rate": 9.466666666666667e-06,
+      "loss": 0.9734,
+      "step": 284
+    },
+    {
+      "epoch": 0.2070233582275895,
+      "grad_norm": 4.975153535316991,
+      "learning_rate": 9.5e-06,
+      "loss": 0.9635,
+      "step": 285
+    },
+    {
+      "epoch": 0.20774975597575648,
+      "grad_norm": 4.856621018396769,
+      "learning_rate": 9.533333333333334e-06,
+      "loss": 0.9523,
+      "step": 286
+    },
+    {
+      "epoch": 0.20847615372392345,
+      "grad_norm": 4.921703565848544,
+      "learning_rate": 9.566666666666668e-06,
+      "loss": 0.939,
+      "step": 287
+    },
+    {
+      "epoch": 0.20920255147209044,
+      "grad_norm": 4.759014226910551,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.9514,
+      "step": 288
+    },
+    {
+      "epoch": 0.20992894922025743,
+      "grad_norm": 4.850649945410091,
+      "learning_rate": 9.633333333333335e-06,
+      "loss": 0.9423,
+      "step": 289
+    },
+    {
+      "epoch": 0.2106553469684244,
+      "grad_norm": 4.7510300916535835,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.9375,
+      "step": 290
+    },
+    {
+      "epoch": 0.21138174471659138,
+      "grad_norm": 4.631844761995069,
+      "learning_rate": 9.7e-06,
+      "loss": 0.9498,
+      "step": 291
+    },
+    {
+      "epoch": 0.21210814246475837,
+      "grad_norm": 4.6840553260562094,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 0.9318,
+      "step": 292
+    },
+    {
+      "epoch": 0.21283454021292533,
+      "grad_norm": 4.64051963705808,
+      "learning_rate": 9.766666666666667e-06,
+      "loss": 0.935,
+      "step": 293
+    },
+    {
+      "epoch": 0.21356093796109232,
+      "grad_norm": 4.510960059913492,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.9431,
+      "step": 294
+    },
+    {
+      "epoch": 0.2142873357092593,
+      "grad_norm": 4.407943102592227,
+      "learning_rate": 9.833333333333333e-06,
+      "loss": 0.9538,
+      "step": 295
+    },
+    {
+      "epoch": 0.21501373345742628,
+      "grad_norm": 4.477665809742677,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 0.9414,
+      "step": 296
+    },
+    {
+      "epoch": 0.21574013120559327,
+      "grad_norm": 4.3658879124763885,
+      "learning_rate": 9.9e-06,
+      "loss": 0.9475,
+      "step": 297
+    },
+    {
+      "epoch": 0.21646652895376026,
+      "grad_norm": 4.3068037062213875,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 0.9423,
+      "step": 298
+    },
+    {
+      "epoch": 0.21719292670192722,
+      "grad_norm": 4.285180819974754,
+      "learning_rate": 9.966666666666667e-06,
+      "loss": 0.9399,
+      "step": 299
+    },
+    {
+      "epoch": 0.2179193244500942,
+      "grad_norm": 4.229079706272251,
+      "learning_rate": 1e-05,
+      "loss": 0.9482,
+      "step": 300
+    },
+    {
+      "epoch": 0.21864572219826117,
+      "grad_norm": 4.281554504080381,
+      "learning_rate": 9.99999943011404e-06,
+      "loss": 0.9175,
+      "step": 301
+    },
+    {
+      "epoch": 0.21937211994642816,
+      "grad_norm": 4.169777969455729,
+      "learning_rate": 9.999997720456285e-06,
+      "loss": 0.9479,
+      "step": 302
+    },
+    {
+      "epoch": 0.22009851769459515,
+      "grad_norm": 4.1699923475025855,
+      "learning_rate": 9.999994871027128e-06,
+      "loss": 0.925,
+      "step": 303
+    },
+    {
+      "epoch": 0.22082491544276212,
+      "grad_norm": 4.125376173668559,
+      "learning_rate": 9.999990881827218e-06,
+      "loss": 0.9263,
+      "step": 304
+    },
+    {
+      "epoch": 0.2215513131909291,
+      "grad_norm": 4.075622019823068,
+      "learning_rate": 9.999985752857464e-06,
+      "loss": 0.9415,
+      "step": 305
+    },
+    {
+      "epoch": 0.2222777109390961,
+      "grad_norm": 4.087232981051823,
+      "learning_rate": 9.999979484119035e-06,
+      "loss": 0.9023,
+      "step": 306
+    },
+    {
+      "epoch": 0.22300410868726306,
+      "grad_norm": 3.9626016761472567,
+      "learning_rate": 9.99997207561336e-06,
+      "loss": 0.9274,
+      "step": 307
+    },
+    {
+      "epoch": 0.22373050643543005,
+      "grad_norm": 3.9837800051798036,
+      "learning_rate": 9.999963527342128e-06,
+      "loss": 0.9256,
+      "step": 308
+    },
+    {
+      "epoch": 0.22445690418359704,
+      "grad_norm": 3.916777154550035,
+      "learning_rate": 9.999953839307287e-06,
+      "loss": 0.9217,
+      "step": 309
+    },
+    {
+      "epoch": 0.225183301931764,
+      "grad_norm": 3.8728958068319006,
+      "learning_rate": 9.999943011511045e-06,
+      "loss": 0.9275,
+      "step": 310
+    },
+    {
+      "epoch": 0.225909699679931,
+      "grad_norm": 3.8132974533118,
+      "learning_rate": 9.999931043955876e-06,
+      "loss": 0.9207,
+      "step": 311
+    },
+    {
+      "epoch": 0.22663609742809798,
+      "grad_norm": 3.841173198735151,
+      "learning_rate": 9.999917936644498e-06,
+      "loss": 0.9086,
+      "step": 312
+    },
+    {
+      "epoch": 0.22736249517626494,
+      "grad_norm": 3.8065529809421452,
+      "learning_rate": 9.999903689579906e-06,
+      "loss": 0.9273,
+      "step": 313
+    },
+    {
+      "epoch": 0.22808889292443194,
+      "grad_norm": 3.7651904589221785,
+      "learning_rate": 9.999888302765347e-06,
+      "loss": 0.9189,
+      "step": 314
+    },
+    {
+      "epoch": 0.22881529067259893,
+      "grad_norm": 3.7601408395621334,
+      "learning_rate": 9.999871776204325e-06,
+      "loss": 0.9134,
+      "step": 315
+    },
+    {
+      "epoch": 0.2295416884207659,
+      "grad_norm": 3.670695667395263,
+      "learning_rate": 9.99985410990061e-06,
+      "loss": 0.916,
+      "step": 316
+    },
+    {
+      "epoch": 0.23026808616893288,
+      "grad_norm": 3.6252352338521305,
+      "learning_rate": 9.999835303858231e-06,
+      "loss": 0.9134,
+      "step": 317
+    },
+    {
+      "epoch": 0.23099448391709987,
+      "grad_norm": 3.6643475103199106,
+      "learning_rate": 9.99981535808147e-06,
+      "loss": 0.9056,
+      "step": 318
+    },
+    {
+      "epoch": 0.23172088166526683,
+      "grad_norm": 3.6490134106546113,
+      "learning_rate": 9.99979427257488e-06,
+      "loss": 0.908,
+      "step": 319
+    },
+    {
+      "epoch": 0.23244727941343382,
+      "grad_norm": 3.595270690577472,
+      "learning_rate": 9.999772047343259e-06,
+      "loss": 0.9089,
+      "step": 320
+    },
+    {
+      "epoch": 0.2331736771616008,
+      "grad_norm": 3.5270211803260803,
+      "learning_rate": 9.999748682391682e-06,
+      "loss": 0.9042,
+      "step": 321
+    },
+    {
+      "epoch": 0.23390007490976777,
+      "grad_norm": 3.526672305447795,
+      "learning_rate": 9.99972417772547e-06,
+      "loss": 0.9027,
+      "step": 322
+    },
+    {
+      "epoch": 0.23462647265793476,
+      "grad_norm": 3.4884435048204785,
+      "learning_rate": 9.99969853335021e-06,
+      "loss": 0.9147,
+      "step": 323
+    },
+    {
+      "epoch": 0.23535287040610173,
+      "grad_norm": 3.466461119074874,
+      "learning_rate": 9.99967174927175e-06,
+      "loss": 0.8831,
+      "step": 324
+    },
+    {
+      "epoch": 0.23607926815426872,
+      "grad_norm": 3.383397947245138,
+      "learning_rate": 9.99964382549619e-06,
+      "loss": 0.9125,
+      "step": 325
+    },
+    {
+      "epoch": 0.2368056659024357,
+      "grad_norm": 3.442604009677502,
+      "learning_rate": 9.999614762029901e-06,
+      "loss": 0.8934,
+      "step": 326
+    },
+    {
+      "epoch": 0.23753206365060267,
+      "grad_norm": 3.355588421676241,
+      "learning_rate": 9.999584558879507e-06,
+      "loss": 0.9061,
+      "step": 327
+    },
+    {
+      "epoch": 0.23825846139876966,
+      "grad_norm": 3.266053823556341,
+      "learning_rate": 9.999553216051892e-06,
+      "loss": 0.9182,
+      "step": 328
+    },
+    {
+      "epoch": 0.23898485914693665,
+      "grad_norm": 3.3662912679017603,
+      "learning_rate": 9.9995207335542e-06,
+      "loss": 0.9014,
+      "step": 329
+    },
+    {
+      "epoch": 0.2397112568951036,
+      "grad_norm": 3.286115029330833,
+      "learning_rate": 9.999487111393836e-06,
+      "loss": 0.9102,
+      "step": 330
+    },
+    {
+      "epoch": 0.2404376546432706,
+      "grad_norm": 3.337503631773769,
+      "learning_rate": 9.999452349578465e-06,
+      "loss": 0.9128,
+      "step": 331
+    },
+    {
+      "epoch": 0.2411640523914376,
+      "grad_norm": 3.1928364364917896,
+      "learning_rate": 9.999416448116011e-06,
+      "loss": 0.9107,
+      "step": 332
+    },
+    {
+      "epoch": 0.24189045013960456,
+      "grad_norm": 3.1773858850837833,
+      "learning_rate": 9.999379407014656e-06,
+      "loss": 0.9102,
+      "step": 333
+    },
+    {
+      "epoch": 0.24261684788777155,
+      "grad_norm": 3.1494952919561117,
+      "learning_rate": 9.999341226282848e-06,
+      "loss": 0.9053,
+      "step": 334
+    },
+    {
+      "epoch": 0.24334324563593854,
+      "grad_norm": 3.1585806465130704,
+      "learning_rate": 9.999301905929286e-06,
+      "loss": 0.9027,
+      "step": 335
+    },
+    {
+      "epoch": 0.2440696433841055,
+      "grad_norm": 3.153378709522304,
+      "learning_rate": 9.999261445962936e-06,
+      "loss": 0.8965,
+      "step": 336
+    },
+    {
+      "epoch": 0.2447960411322725,
+      "grad_norm": 3.1271021116928215,
+      "learning_rate": 9.999219846393018e-06,
+      "loss": 0.9022,
+      "step": 337
+    },
+    {
+      "epoch": 0.24552243888043948,
+      "grad_norm": 3.1009031288222775,
+      "learning_rate": 9.999177107229019e-06,
+      "loss": 0.9013,
+      "step": 338
+    },
+    {
+      "epoch": 0.24624883662860644,
+      "grad_norm": 3.1087383815587244,
+      "learning_rate": 9.999133228480679e-06,
+      "loss": 0.9037,
+      "step": 339
+    },
+    {
+      "epoch": 0.24697523437677343,
+      "grad_norm": 3.080267158750676,
+      "learning_rate": 9.999088210158001e-06,
+      "loss": 0.8849,
+      "step": 340
+    },
+    {
+      "epoch": 0.24770163212494042,
+      "grad_norm": 2.999486292773668,
+      "learning_rate": 9.999042052271247e-06,
+      "loss": 0.8937,
+      "step": 341
+    },
+    {
+      "epoch": 0.2484280298731074,
+      "grad_norm": 3.013895982835644,
+      "learning_rate": 9.99899475483094e-06,
+      "loss": 0.8888,
+      "step": 342
+    },
+    {
+      "epoch": 0.24915442762127438,
+      "grad_norm": 3.073758208806406,
+      "learning_rate": 9.998946317847857e-06,
+      "loss": 0.8602,
+      "step": 343
+    },
+    {
+      "epoch": 0.24988082536944137,
+      "grad_norm": 2.9286726313576112,
+      "learning_rate": 9.998896741333047e-06,
+      "loss": 0.9008,
+      "step": 344
+    },
+    {
+      "epoch": 0.25060722311760836,
+      "grad_norm": 2.910650348601522,
+      "learning_rate": 9.998846025297804e-06,
+      "loss": 0.895,
+      "step": 345
+    },
+    {
+      "epoch": 0.2513336208657753,
+      "grad_norm": 2.8993230794505838,
+      "learning_rate": 9.998794169753694e-06,
+      "loss": 0.8841,
+      "step": 346
+    },
+    {
+      "epoch": 0.2520600186139423,
+      "grad_norm": 2.9168647806374617,
+      "learning_rate": 9.998741174712534e-06,
+      "loss": 0.8989,
+      "step": 347
+    },
+    {
+      "epoch": 0.2527864163621093,
+      "grad_norm": 2.9351727444896283,
+      "learning_rate": 9.998687040186407e-06,
+      "loss": 0.8851,
+      "step": 348
+    },
+    {
+      "epoch": 0.25351281411027626,
+      "grad_norm": 2.823829131544092,
+      "learning_rate": 9.998631766187651e-06,
+      "loss": 0.8873,
+      "step": 349
+    },
+    {
+      "epoch": 0.2542392118584432,
+      "grad_norm": 2.8207151518990448,
+      "learning_rate": 9.998575352728869e-06,
+      "loss": 0.8808,
+      "step": 350
+    },
+    {
+      "epoch": 0.25496560960661024,
+      "grad_norm": 2.881827280089076,
+      "learning_rate": 9.99851779982292e-06,
+      "loss": 0.8792,
+      "step": 351
+    },
+    {
+      "epoch": 0.2556920073547772,
+      "grad_norm": 2.839308859851771,
+      "learning_rate": 9.998459107482922e-06,
+      "loss": 0.8716,
+      "step": 352
+    },
+    {
+      "epoch": 0.25641840510294417,
+      "grad_norm": 2.7473965822142743,
+      "learning_rate": 9.998399275722253e-06,
+      "loss": 0.8842,
+      "step": 353
+    },
+    {
+      "epoch": 0.2571448028511112,
+      "grad_norm": 2.762628380371553,
+      "learning_rate": 9.998338304554553e-06,
+      "loss": 0.8847,
+      "step": 354
+    },
+    {
+      "epoch": 0.25787120059927815,
+      "grad_norm": 2.7322944562152616,
+      "learning_rate": 9.998276193993723e-06,
+      "loss": 0.8605,
+      "step": 355
+    },
+    {
+      "epoch": 0.2585975983474451,
+      "grad_norm": 2.666398101224024,
+      "learning_rate": 9.99821294405392e-06,
+      "loss": 0.8867,
+      "step": 356
+    },
+    {
+      "epoch": 0.25932399609561213,
+      "grad_norm": 2.710835958236394,
+      "learning_rate": 9.998148554749561e-06,
+      "loss": 0.8634,
+      "step": 357
+    },
+    {
+      "epoch": 0.2600503938437791,
+      "grad_norm": 2.6670423652778346,
+      "learning_rate": 9.998083026095323e-06,
+      "loss": 0.8878,
+      "step": 358
+    },
+    {
+      "epoch": 0.26077679159194606,
+      "grad_norm": 2.681536519147625,
+      "learning_rate": 9.998016358106147e-06,
+      "loss": 0.8686,
+      "step": 359
+    },
+    {
+      "epoch": 0.261503189340113,
+      "grad_norm": 2.597410074827163,
+      "learning_rate": 9.997948550797227e-06,
+      "loss": 0.8868,
+      "step": 360
+    },
+    {
+      "epoch": 0.26222958708828004,
+      "grad_norm": 2.621123678822491,
+      "learning_rate": 9.997879604184023e-06,
+      "loss": 0.887,
+      "step": 361
+    },
+    {
+      "epoch": 0.262955984836447,
+      "grad_norm": 2.604762010879466,
+      "learning_rate": 9.99780951828225e-06,
+      "loss": 0.8587,
+      "step": 362
+    },
+    {
+      "epoch": 0.26368238258461396,
+      "grad_norm": 2.653532685112411,
+      "learning_rate": 9.997738293107882e-06,
+      "loss": 0.8694,
+      "step": 363
+    },
+    {
+      "epoch": 0.264408780332781,
+      "grad_norm": 2.5929563189339575,
+      "learning_rate": 9.997665928677159e-06,
+      "loss": 0.8663,
+      "step": 364
+    },
+    {
+      "epoch": 0.26513517808094794,
+      "grad_norm": 2.511532491788452,
+      "learning_rate": 9.997592425006574e-06,
+      "loss": 0.8923,
+      "step": 365
+    },
+    {
+      "epoch": 0.2658615758291149,
+      "grad_norm": 2.5809881467050313,
+      "learning_rate": 9.997517782112885e-06,
+      "loss": 0.8544,
+      "step": 366
+    },
+    {
+      "epoch": 0.2665879735772819,
+      "grad_norm": 2.5451328703553964,
+      "learning_rate": 9.997442000013105e-06,
+      "loss": 0.869,
+      "step": 367
+    },
+    {
+      "epoch": 0.2673143713254489,
+      "grad_norm": 2.5718644330923253,
+      "learning_rate": 9.997365078724512e-06,
+      "loss": 0.8576,
+      "step": 368
+    },
+    {
+      "epoch": 0.26804076907361585,
+      "grad_norm": 2.5201758471763127,
+      "learning_rate": 9.997287018264637e-06,
+      "loss": 0.8471,
+      "step": 369
+    },
+    {
+      "epoch": 0.26876716682178287,
+      "grad_norm": 2.469545330155226,
+      "learning_rate": 9.997207818651273e-06,
+      "loss": 0.8716,
+      "step": 370
+    },
+    {
+      "epoch": 0.26949356456994983,
+      "grad_norm": 2.4963246591865556,
+      "learning_rate": 9.99712747990248e-06,
+      "loss": 0.8598,
+      "step": 371
+    },
+    {
+      "epoch": 0.2702199623181168,
+      "grad_norm": 2.4316216175071825,
+      "learning_rate": 9.997046002036566e-06,
+      "loss": 0.8687,
+      "step": 372
+    },
+    {
+      "epoch": 0.2709463600662838,
+      "grad_norm": 2.4339436736619677,
+      "learning_rate": 9.996963385072108e-06,
+      "loss": 0.8536,
+      "step": 373
+    },
+    {
+      "epoch": 0.27167275781445077,
+      "grad_norm": 2.447464955174159,
+      "learning_rate": 9.996879629027934e-06,
+      "loss": 0.8666,
+      "step": 374
+    },
+    {
+      "epoch": 0.27239915556261773,
+      "grad_norm": 2.408860632016767,
+      "learning_rate": 9.996794733923141e-06,
+      "loss": 0.858,
+      "step": 375
+    },
+    {
+      "epoch": 0.27312555331078475,
+      "grad_norm": 2.381748758234186,
+      "learning_rate": 9.996708699777081e-06,
+      "loss": 0.8671,
+      "step": 376
+    },
+    {
+      "epoch": 0.2738519510589517,
+      "grad_norm": 2.547484262667178,
+      "learning_rate": 9.996621526609364e-06,
+      "loss": 0.8385,
+      "step": 377
+    },
+    {
+      "epoch": 0.2745783488071187,
+      "grad_norm": 2.4047390006326905,
+      "learning_rate": 9.996533214439864e-06,
+      "loss": 0.8594,
+      "step": 378
+    },
+    {
+      "epoch": 0.2753047465552857,
+      "grad_norm": 2.392166393236436,
+      "learning_rate": 9.996443763288708e-06,
+      "loss": 0.855,
+      "step": 379
+    },
+    {
+      "epoch": 0.27603114430345266,
+      "grad_norm": 2.314177983313085,
+      "learning_rate": 9.99635317317629e-06,
+      "loss": 0.8613,
+      "step": 380
+    },
+    {
+      "epoch": 0.2767575420516196,
+      "grad_norm": 2.2953329879725706,
+      "learning_rate": 9.99626144412326e-06,
+      "loss": 0.8508,
+      "step": 381
+    },
+    {
+      "epoch": 0.27748393979978664,
+      "grad_norm": 2.373682548895982,
+      "learning_rate": 9.996168576150527e-06,
+      "loss": 0.8529,
+      "step": 382
+    },
+    {
+      "epoch": 0.2782103375479536,
+      "grad_norm": 2.272694418338483,
+      "learning_rate": 9.996074569279261e-06,
+      "loss": 0.8499,
+      "step": 383
+    },
+    {
+      "epoch": 0.27893673529612056,
+      "grad_norm": 2.27252146099723,
+      "learning_rate": 9.995979423530893e-06,
+      "loss": 0.8597,
+      "step": 384
+    },
+    {
+      "epoch": 0.2796631330442876,
+      "grad_norm": 2.2642442307360624,
+      "learning_rate": 9.995883138927109e-06,
+      "loss": 0.8642,
+      "step": 385
+    },
+    {
+      "epoch": 0.28038953079245454,
+      "grad_norm": 2.2782516681714635,
+      "learning_rate": 9.995785715489861e-06,
+      "loss": 0.8433,
+      "step": 386
+    },
+    {
+      "epoch": 0.2811159285406215,
+      "grad_norm": 2.2336277825210216,
+      "learning_rate": 9.995687153241353e-06,
+      "loss": 0.856,
+      "step": 387
+    },
+    {
+      "epoch": 0.2818423262887885,
+      "grad_norm": 2.195604639250854,
+      "learning_rate": 9.995587452204056e-06,
+      "loss": 0.8668,
+      "step": 388
+    },
+    {
+      "epoch": 0.2825687240369555,
+      "grad_norm": 2.1848982031790976,
+      "learning_rate": 9.995486612400695e-06,
+      "loss": 0.8638,
+      "step": 389
+    },
+    {
+      "epoch": 0.28329512178512245,
+      "grad_norm": 2.2079370575686292,
+      "learning_rate": 9.995384633854259e-06,
+      "loss": 0.8459,
+      "step": 390
+    },
+    {
+      "epoch": 0.28402151953328947,
+      "grad_norm": 2.4403396183778168,
+      "learning_rate": 9.995281516587992e-06,
+      "loss": 0.8442,
+      "step": 391
+    },
+    {
+      "epoch": 0.28474791728145643,
+      "grad_norm": 2.190782039428668,
+      "learning_rate": 9.995177260625401e-06,
+      "loss": 0.8571,
+      "step": 392
+    },
+    {
+      "epoch": 0.2854743150296234,
+      "grad_norm": 2.1783620914890447,
+      "learning_rate": 9.995071865990255e-06,
+      "loss": 0.8421,
+      "step": 393
+    },
+    {
+      "epoch": 0.2862007127777904,
+      "grad_norm": 2.160114786487807,
+      "learning_rate": 9.994965332706574e-06,
+      "loss": 0.8464,
+      "step": 394
+    },
+    {
+      "epoch": 0.2869271105259574,
+      "grad_norm": 2.1077114811406363,
+      "learning_rate": 9.994857660798644e-06,
+      "loss": 0.851,
+      "step": 395
+    },
+    {
+      "epoch": 0.28765350827412434,
+      "grad_norm": 2.130514190474774,
+      "learning_rate": 9.994748850291013e-06,
+      "loss": 0.8391,
+      "step": 396
+    },
+    {
+      "epoch": 0.28837990602229135,
+      "grad_norm": 2.305097773774775,
+      "learning_rate": 9.99463890120848e-06,
+      "loss": 0.8484,
+      "step": 397
+    },
+    {
+      "epoch": 0.2891063037704583,
+      "grad_norm": 2.3190227202687557,
+      "learning_rate": 9.994527813576111e-06,
+      "loss": 0.8687,
+      "step": 398
+    },
+    {
+      "epoch": 0.2898327015186253,
+      "grad_norm": 2.0697354319941184,
+      "learning_rate": 9.994415587419229e-06,
+      "loss": 0.871,
+      "step": 399
+    },
+    {
+      "epoch": 0.2905590992667923,
+      "grad_norm": 2.1179466292046802,
+      "learning_rate": 9.994302222763415e-06,
+      "loss": 0.8427,
+      "step": 400
+    },
+    {
+      "epoch": 0.29128549701495926,
+      "grad_norm": 2.0548380672609254,
+      "learning_rate": 9.994187719634512e-06,
+      "loss": 0.851,
+      "step": 401
+    },
+    {
+      "epoch": 0.2920118947631262,
+      "grad_norm": 2.004703885482171,
+      "learning_rate": 9.994072078058621e-06,
+      "loss": 0.8572,
+      "step": 402
+    },
+    {
+      "epoch": 0.29273829251129324,
+      "grad_norm": 2.040493240651109,
+      "learning_rate": 9.993955298062103e-06,
+      "loss": 0.8497,
+      "step": 403
+    },
+    {
+      "epoch": 0.2934646902594602,
+      "grad_norm": 2.0653918004675615,
+      "learning_rate": 9.99383737967158e-06,
+      "loss": 0.8361,
+      "step": 404
+    },
+    {
+      "epoch": 0.29419108800762717,
+      "grad_norm": 2.0162210298258088,
+      "learning_rate": 9.99371832291393e-06,
+      "loss": 0.8407,
+      "step": 405
+    },
+    {
+      "epoch": 0.2949174857557942,
+      "grad_norm": 2.039149148468403,
+      "learning_rate": 9.993598127816292e-06,
+      "loss": 0.8252,
+      "step": 406
+    },
+    {
+      "epoch": 0.29564388350396115,
+      "grad_norm": 1.9921985897639096,
+      "learning_rate": 9.993476794406068e-06,
+      "loss": 0.8467,
+      "step": 407
+    },
+    {
+      "epoch": 0.2963702812521281,
+      "grad_norm": 1.9853296697955654,
+      "learning_rate": 9.993354322710914e-06,
+      "loss": 0.8447,
+      "step": 408
+    },
+    {
+      "epoch": 0.29709667900029507,
+      "grad_norm": 1.933388097861427,
+      "learning_rate": 9.993230712758748e-06,
+      "loss": 0.8443,
+      "step": 409
+    },
+    {
+      "epoch": 0.2978230767484621,
+      "grad_norm": 2.015700970579226,
+      "learning_rate": 9.99310596457775e-06,
+      "loss": 0.8273,
+      "step": 410
+    },
+    {
+      "epoch": 0.29854947449662905,
+      "grad_norm": 1.9666969573202417,
+      "learning_rate": 9.992980078196355e-06,
+      "loss": 0.8348,
+      "step": 411
+    },
+    {
+      "epoch": 0.299275872244796,
+      "grad_norm": 1.9594216972253362,
+      "learning_rate": 9.992853053643257e-06,
+      "loss": 0.832,
+      "step": 412
+    },
+    {
+      "epoch": 0.30000226999296303,
+      "grad_norm": 1.9446630419498114,
+      "learning_rate": 9.992724890947417e-06,
+      "loss": 0.8267,
+      "step": 413
+    },
+    {
+      "epoch": 0.30072866774113,
+      "grad_norm": 1.9688606092767582,
+      "learning_rate": 9.992595590138045e-06,
+      "loss": 0.8368,
+      "step": 414
+    },
+    {
+      "epoch": 0.30145506548929696,
+      "grad_norm": 1.9423435194261474,
+      "learning_rate": 9.99246515124462e-06,
+      "loss": 0.8242,
+      "step": 415
+    },
+    {
+      "epoch": 0.302181463237464,
+      "grad_norm": 1.9125718876356885,
+      "learning_rate": 9.992333574296872e-06,
+      "loss": 0.8276,
+      "step": 416
+    },
+    {
+      "epoch": 0.30290786098563094,
+      "grad_norm": 1.883649957859309,
+      "learning_rate": 9.992200859324799e-06,
+      "loss": 0.8298,
+      "step": 417
+    },
+    {
+      "epoch": 0.3036342587337979,
+      "grad_norm": 1.879036782487961,
+      "learning_rate": 9.992067006358651e-06,
+      "loss": 0.8353,
+      "step": 418
+    },
+    {
+      "epoch": 0.3043606564819649,
+      "grad_norm": 1.8719975868291152,
+      "learning_rate": 9.991932015428941e-06,
+      "loss": 0.8443,
+      "step": 419
+    },
+    {
+      "epoch": 0.3050870542301319,
+      "grad_norm": 1.874484854453761,
+      "learning_rate": 9.991795886566443e-06,
+      "loss": 0.8292,
+      "step": 420
+    },
+    {
+      "epoch": 0.30581345197829884,
+      "grad_norm": 1.8775152039378569,
+      "learning_rate": 9.991658619802183e-06,
+      "loss": 0.8252,
+      "step": 421
+    },
+    {
+      "epoch": 0.30653984972646586,
+      "grad_norm": 1.8441233018367729,
+      "learning_rate": 9.991520215167455e-06,
+      "loss": 0.8272,
+      "step": 422
+    },
+    {
+      "epoch": 0.3072662474746328,
+      "grad_norm": 1.8281278187371393,
+      "learning_rate": 9.99138067269381e-06,
+      "loss": 0.8333,
+      "step": 423
+    },
+    {
+      "epoch": 0.3079926452227998,
+      "grad_norm": 1.9152044763911662,
+      "learning_rate": 9.991239992413056e-06,
+      "loss": 0.8289,
+      "step": 424
+    },
+    {
+      "epoch": 0.3087190429709668,
+      "grad_norm": 1.853675901429331,
+      "learning_rate": 9.99109817435726e-06,
+      "loss": 0.8206,
+      "step": 425
+    },
+    {
+      "epoch": 0.30944544071913377,
+      "grad_norm": 1.828312370971533,
+      "learning_rate": 9.990955218558751e-06,
+      "loss": 0.8017,
+      "step": 426
+    },
+    {
+      "epoch": 0.31017183846730073,
+      "grad_norm": 1.828150007798247,
+      "learning_rate": 9.99081112505012e-06,
+      "loss": 0.8268,
+      "step": 427
+    },
+    {
+      "epoch": 0.31089823621546775,
+      "grad_norm": 2.086629705375154,
+      "learning_rate": 9.990665893864208e-06,
+      "loss": 0.8187,
+      "step": 428
+    },
+    {
+      "epoch": 0.3116246339636347,
+      "grad_norm": 1.7980079821122612,
+      "learning_rate": 9.990519525034125e-06,
+      "loss": 0.8181,
+      "step": 429
+    },
+    {
+      "epoch": 0.3123510317118017,
+      "grad_norm": 1.8022071581091539,
+      "learning_rate": 9.990372018593236e-06,
+      "loss": 0.8107,
+      "step": 430
+    },
+    {
+      "epoch": 0.3130774294599687,
+      "grad_norm": 1.7817457263756484,
+      "learning_rate": 9.990223374575165e-06,
+      "loss": 0.8223,
+      "step": 431
+    },
+    {
+      "epoch": 0.31380382720813566,
+      "grad_norm": 1.8186788986310414,
+      "learning_rate": 9.990073593013794e-06,
+      "loss": 0.8306,
+      "step": 432
+    },
+    {
+      "epoch": 0.3145302249563026,
+      "grad_norm": 1.78999729473262,
+      "learning_rate": 9.989922673943271e-06,
+      "loss": 0.8218,
+      "step": 433
+    },
+    {
+      "epoch": 0.31525662270446964,
+      "grad_norm": 1.7922498131927898,
+      "learning_rate": 9.989770617397994e-06,
+      "loss": 0.8177,
+      "step": 434
+    },
+    {
+      "epoch": 0.3159830204526366,
+      "grad_norm": 1.7407032579649269,
+      "learning_rate": 9.989617423412627e-06,
+      "loss": 0.8172,
+      "step": 435
+    },
+    {
+      "epoch": 0.31670941820080356,
+      "grad_norm": 1.7377083293660502,
+      "learning_rate": 9.989463092022093e-06,
+      "loss": 0.8097,
+      "step": 436
+    },
+    {
+      "epoch": 0.3174358159489706,
+      "grad_norm": 1.8106149891609342,
+      "learning_rate": 9.989307623261571e-06,
+      "loss": 0.8324,
+      "step": 437
+    },
+    {
+      "epoch": 0.31816221369713754,
+      "grad_norm": 2.0548792095096537,
+      "learning_rate": 9.989151017166498e-06,
+      "loss": 0.8281,
+      "step": 438
+    },
+    {
+      "epoch": 0.3188886114453045,
+      "grad_norm": 1.7202909871253904,
+      "learning_rate": 9.988993273772578e-06,
+      "loss": 0.8152,
+      "step": 439
+    },
+    {
+      "epoch": 0.3196150091934715,
+      "grad_norm": 1.9416483022054316,
+      "learning_rate": 9.988834393115768e-06,
+      "loss": 0.8116,
+      "step": 440
+    },
+    {
+      "epoch": 0.3203414069416385,
+      "grad_norm": 1.700305712466284,
+      "learning_rate": 9.98867437523228e-06,
+      "loss": 0.8192,
+      "step": 441
+    },
+    {
+      "epoch": 0.32106780468980545,
+      "grad_norm": 1.6986113649343069,
+      "learning_rate": 9.9885132201586e-06,
+      "loss": 0.8206,
+      "step": 442
+    },
+    {
+      "epoch": 0.32179420243797247,
+      "grad_norm": 1.7568165040916024,
+      "learning_rate": 9.988350927931455e-06,
+      "loss": 0.8159,
+      "step": 443
+    },
+    {
+      "epoch": 0.32252060018613943,
+      "grad_norm": 1.6638610925315405,
+      "learning_rate": 9.98818749858785e-06,
+      "loss": 0.8261,
+      "step": 444
+    },
+    {
+      "epoch": 0.3232469979343064,
+      "grad_norm": 1.6400851065971984,
+      "learning_rate": 9.988022932165029e-06,
+      "loss": 0.822,
+      "step": 445
+    },
+    {
+      "epoch": 0.3239733956824734,
+      "grad_norm": 1.6637785482086707,
+      "learning_rate": 9.987857228700512e-06,
+      "loss": 0.7977,
+      "step": 446
+    },
+    {
+      "epoch": 0.32469979343064037,
+      "grad_norm": 1.6527505249510377,
+      "learning_rate": 9.987690388232071e-06,
+      "loss": 0.8109,
+      "step": 447
+    },
+    {
+      "epoch": 0.32542619117880733,
+      "grad_norm": 1.6589931492071577,
+      "learning_rate": 9.987522410797736e-06,
+      "loss": 0.8122,
+      "step": 448
+    },
+    {
+      "epoch": 0.32615258892697435,
+      "grad_norm": 1.6112609442519945,
+      "learning_rate": 9.987353296435802e-06,
+      "loss": 0.8223,
+      "step": 449
+    },
+    {
+      "epoch": 0.3268789866751413,
+      "grad_norm": 1.6299975467243404,
+      "learning_rate": 9.987183045184815e-06,
+      "loss": 0.8148,
+      "step": 450
+    },
+    {
+      "epoch": 0.3276053844233083,
+      "grad_norm": 1.6911871741513431,
+      "learning_rate": 9.987011657083587e-06,
+      "loss": 0.812,
+      "step": 451
+    },
+    {
+      "epoch": 0.3283317821714753,
+      "grad_norm": 1.6301988035526893,
+      "learning_rate": 9.986839132171186e-06,
+      "loss": 0.7894,
+      "step": 452
+    },
+    {
+      "epoch": 0.32905817991964226,
+      "grad_norm": 1.6050588368442675,
+      "learning_rate": 9.98666547048694e-06,
+      "loss": 0.7994,
+      "step": 453
+    },
+    {
+      "epoch": 0.3297845776678092,
+      "grad_norm": 1.5935740513279069,
+      "learning_rate": 9.986490672070438e-06,
+      "loss": 0.8133,
+      "step": 454
+    },
+    {
+      "epoch": 0.33051097541597624,
+      "grad_norm": 1.5800491187854209,
+      "learning_rate": 9.986314736961522e-06,
+      "loss": 0.8057,
+      "step": 455
+    },
+    {
+      "epoch": 0.3312373731641432,
+      "grad_norm": 1.5998802898110331,
+      "learning_rate": 9.986137665200298e-06,
+      "loss": 0.7788,
+      "step": 456
+    },
+    {
+      "epoch": 0.33196377091231016,
+      "grad_norm": 1.5993757442652543,
+      "learning_rate": 9.985959456827132e-06,
+      "loss": 0.7947,
+      "step": 457
+    },
+    {
+      "epoch": 0.3326901686604771,
+      "grad_norm": 1.599986123833006,
+      "learning_rate": 9.985780111882648e-06,
+      "loss": 0.8124,
+      "step": 458
+    },
+    {
+      "epoch": 0.33341656640864414,
+      "grad_norm": 1.567732652499849,
+      "learning_rate": 9.985599630407727e-06,
+      "loss": 0.7983,
+      "step": 459
+    },
+    {
+      "epoch": 0.3341429641568111,
+      "grad_norm": 1.568753470069939,
+      "learning_rate": 9.98541801244351e-06,
+      "loss": 0.7993,
+      "step": 460
+    },
+    {
+      "epoch": 0.33486936190497807,
+      "grad_norm": 1.53872532512622,
+      "learning_rate": 9.9852352580314e-06,
+      "loss": 0.7954,
+      "step": 461
+    },
+    {
+      "epoch": 0.3355957596531451,
+      "grad_norm": 1.573735628732098,
+      "learning_rate": 9.985051367213055e-06,
+      "loss": 0.8028,
+      "step": 462
+    },
+    {
+      "epoch": 0.33632215740131205,
+      "grad_norm": 1.5562781601601243,
+      "learning_rate": 9.984866340030393e-06,
+      "loss": 0.7991,
+      "step": 463
+    },
+    {
+      "epoch": 0.337048555149479,
+      "grad_norm": 1.5286823486792622,
+      "learning_rate": 9.984680176525591e-06,
+      "loss": 0.8086,
+      "step": 464
+    },
+    {
+      "epoch": 0.33777495289764603,
+      "grad_norm": 1.5206427472940696,
+      "learning_rate": 9.98449287674109e-06,
+      "loss": 0.7931,
+      "step": 465
+    },
+    {
+      "epoch": 0.338501350645813,
+      "grad_norm": 1.529018033590309,
+      "learning_rate": 9.984304440719582e-06,
+      "loss": 0.7976,
+      "step": 466
+    },
+    {
+      "epoch": 0.33922774839397996,
+      "grad_norm": 1.4941622511227373,
+      "learning_rate": 9.984114868504023e-06,
+      "loss": 0.8079,
+      "step": 467
+    },
+    {
+      "epoch": 0.339954146142147,
+      "grad_norm": 1.4781959326957563,
+      "learning_rate": 9.983924160137627e-06,
+      "loss": 0.8022,
+      "step": 468
+    },
+    {
+      "epoch": 0.34068054389031394,
+      "grad_norm": 1.4674192733847702,
+      "learning_rate": 9.983732315663865e-06,
+      "loss": 0.8091,
+      "step": 469
+    },
+    {
+      "epoch": 0.3414069416384809,
+      "grad_norm": 1.4885758046839543,
+      "learning_rate": 9.983539335126473e-06,
+      "loss": 0.8014,
+      "step": 470
+    },
+    {
+      "epoch": 0.3421333393866479,
+      "grad_norm": 1.4639151595858675,
+      "learning_rate": 9.983345218569438e-06,
+      "loss": 0.8076,
+      "step": 471
+    },
+    {
+      "epoch": 0.3428597371348149,
+      "grad_norm": 1.4519543436282374,
+      "learning_rate": 9.98314996603701e-06,
+      "loss": 0.8147,
+      "step": 472
+    },
+    {
+      "epoch": 0.34358613488298184,
+      "grad_norm": 1.4770141552096316,
+      "learning_rate": 9.9829535775737e-06,
+      "loss": 0.7926,
+      "step": 473
+    },
+    {
+      "epoch": 0.34431253263114886,
+      "grad_norm": 1.461845780579588,
+      "learning_rate": 9.982756053224274e-06,
+      "loss": 0.8123,
+      "step": 474
+    },
+    {
+      "epoch": 0.3450389303793158,
+      "grad_norm": 1.503728720014433,
+      "learning_rate": 9.982557393033758e-06,
+      "loss": 0.779,
+      "step": 475
+    },
+    {
+      "epoch": 0.3457653281274828,
+      "grad_norm": 1.4849653618968792,
+      "learning_rate": 9.982357597047438e-06,
+      "loss": 0.7866,
+      "step": 476
+    },
+    {
+      "epoch": 0.3464917258756498,
+      "grad_norm": 1.4519069940854028,
+      "learning_rate": 9.98215666531086e-06,
+      "loss": 0.7858,
+      "step": 477
+    },
+    {
+      "epoch": 0.34721812362381677,
+      "grad_norm": 1.441735470802499,
+      "learning_rate": 9.981954597869825e-06,
+      "loss": 0.8038,
+      "step": 478
+    },
+    {
+      "epoch": 0.34794452137198373,
+      "grad_norm": 1.4516397081642147,
+      "learning_rate": 9.981751394770395e-06,
+      "loss": 0.8039,
+      "step": 479
+    },
+    {
+      "epoch": 0.34867091912015075,
+      "grad_norm": 1.4238296243544586,
+      "learning_rate": 9.981547056058893e-06,
+      "loss": 0.8057,
+      "step": 480
+    },
+    {
+      "epoch": 0.3493973168683177,
+      "grad_norm": 1.4190132849441395,
+      "learning_rate": 9.981341581781899e-06,
+      "loss": 0.7899,
+      "step": 481
+    },
+    {
+      "epoch": 0.35012371461648467,
+      "grad_norm": 1.4326395273254544,
+      "learning_rate": 9.98113497198625e-06,
+      "loss": 0.7895,
+      "step": 482
+    },
+    {
+      "epoch": 0.3508501123646517,
+      "grad_norm": 1.399840454673465,
+      "learning_rate": 9.980927226719044e-06,
+      "loss": 0.7965,
+      "step": 483
+    },
+    {
+      "epoch": 0.35157651011281865,
+      "grad_norm": 1.4298889459244744,
+      "learning_rate": 9.980718346027636e-06,
+      "loss": 0.7716,
+      "step": 484
+    },
+    {
+      "epoch": 0.3523029078609856,
+      "grad_norm": 1.40248883066053,
+      "learning_rate": 9.980508329959646e-06,
+      "loss": 0.7886,
+      "step": 485
+    },
+    {
+      "epoch": 0.35302930560915263,
+      "grad_norm": 1.387451411929723,
+      "learning_rate": 9.980297178562943e-06,
+      "loss": 0.7865,
+      "step": 486
+    },
+    {
+      "epoch": 0.3537557033573196,
+      "grad_norm": 1.3729511131953516,
+      "learning_rate": 9.980084891885662e-06,
+      "loss": 0.7872,
+      "step": 487
+    },
+    {
+      "epoch": 0.35448210110548656,
+      "grad_norm": 1.3898531180126903,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.7692,
+      "step": 488
+    },
+    {
+      "epoch": 0.3552084988536536,
+      "grad_norm": 1.3687738845362856,
+      "learning_rate": 9.979656912883193e-06,
+      "loss": 0.7845,
+      "step": 489
+    },
+    {
+      "epoch": 0.35593489660182054,
+      "grad_norm": 1.381027925131146,
+      "learning_rate": 9.979441220655564e-06,
+      "loss": 0.7895,
+      "step": 490
+    },
+    {
+      "epoch": 0.3566612943499875,
+      "grad_norm": 1.346328819485624,
+      "learning_rate": 9.979224393342477e-06,
+      "loss": 0.8029,
+      "step": 491
+    },
+    {
+      "epoch": 0.3573876920981545,
+      "grad_norm": 1.384237345196671,
+      "learning_rate": 9.979006430993357e-06,
+      "loss": 0.8098,
+      "step": 492
+    },
+    {
+      "epoch": 0.3581140898463215,
+      "grad_norm": 1.3699677137536757,
+      "learning_rate": 9.97878733365789e-06,
+      "loss": 0.7804,
+      "step": 493
+    },
+    {
+      "epoch": 0.35884048759448844,
+      "grad_norm": 1.336258195408756,
+      "learning_rate": 9.978567101386021e-06,
+      "loss": 0.7782,
+      "step": 494
+    },
+    {
+      "epoch": 0.35956688534265546,
+      "grad_norm": 1.3486480096565705,
+      "learning_rate": 9.978345734227953e-06,
+      "loss": 0.791,
+      "step": 495
+    },
+    {
+      "epoch": 0.3602932830908224,
+      "grad_norm": 1.3402139669890587,
+      "learning_rate": 9.978123232234147e-06,
+      "loss": 0.7805,
+      "step": 496
+    },
+    {
+      "epoch": 0.3610196808389894,
+      "grad_norm": 1.3203806678784924,
+      "learning_rate": 9.977899595455324e-06,
+      "loss": 0.8059,
+      "step": 497
+    },
+    {
+      "epoch": 0.3617460785871564,
+      "grad_norm": 1.3240137088291317,
+      "learning_rate": 9.977674823942463e-06,
+      "loss": 0.7769,
+      "step": 498
+    },
+    {
+      "epoch": 0.36247247633532337,
+      "grad_norm": 1.3371647877624209,
+      "learning_rate": 9.9774489177468e-06,
+      "loss": 0.7828,
+      "step": 499
+    },
+    {
+      "epoch": 0.36319887408349033,
+      "grad_norm": 1.3435475467562,
+      "learning_rate": 9.977221876919833e-06,
+      "loss": 0.7942,
+      "step": 500
+    },
+    {
+      "epoch": 0.36392527183165735,
+      "grad_norm": 1.3153448965184218,
+      "learning_rate": 9.976993701513317e-06,
+      "loss": 0.7721,
+      "step": 501
+    },
+    {
+      "epoch": 0.3646516695798243,
+      "grad_norm": 1.314625926636882,
+      "learning_rate": 9.976764391579266e-06,
+      "loss": 0.7625,
+      "step": 502
+    },
+    {
+      "epoch": 0.3653780673279913,
+      "grad_norm": 1.2981750511740526,
+      "learning_rate": 9.976533947169948e-06,
+      "loss": 0.7834,
+      "step": 503
+    },
+    {
+      "epoch": 0.36610446507615824,
+      "grad_norm": 1.3305161678384152,
+      "learning_rate": 9.9763023683379e-06,
+      "loss": 0.7874,
+      "step": 504
+    },
+    {
+      "epoch": 0.36683086282432525,
+      "grad_norm": 1.3051284028811305,
+      "learning_rate": 9.976069655135906e-06,
+      "loss": 0.7657,
+      "step": 505
+    },
+    {
+      "epoch": 0.3675572605724922,
+      "grad_norm": 1.279969631814808,
+      "learning_rate": 9.975835807617019e-06,
+      "loss": 0.7822,
+      "step": 506
+    },
+    {
+      "epoch": 0.3682836583206592,
+      "grad_norm": 1.2960567138751116,
+      "learning_rate": 9.975600825834542e-06,
+      "loss": 0.7694,
+      "step": 507
+    },
+    {
+      "epoch": 0.3690100560688262,
+      "grad_norm": 1.306849416485967,
+      "learning_rate": 9.975364709842042e-06,
+      "loss": 0.786,
+      "step": 508
+    },
+    {
+      "epoch": 0.36973645381699316,
+      "grad_norm": 1.271523847952412,
+      "learning_rate": 9.97512745969334e-06,
+      "loss": 0.7761,
+      "step": 509
+    },
+    {
+      "epoch": 0.3704628515651601,
+      "grad_norm": 1.2914497316352134,
+      "learning_rate": 9.97488907544252e-06,
+      "loss": 0.7623,
+      "step": 510
+    },
+    {
+      "epoch": 0.37118924931332714,
+      "grad_norm": 1.309823954459763,
+      "learning_rate": 9.974649557143926e-06,
+      "loss": 0.777,
+      "step": 511
+    },
+    {
+      "epoch": 0.3719156470614941,
+      "grad_norm": 1.2827788113935663,
+      "learning_rate": 9.97440890485215e-06,
+      "loss": 0.7787,
+      "step": 512
+    },
+    {
+      "epoch": 0.37264204480966107,
+      "grad_norm": 1.3873355303110653,
+      "learning_rate": 9.974167118622056e-06,
+      "loss": 0.784,
+      "step": 513
+    },
+    {
+      "epoch": 0.3733684425578281,
+      "grad_norm": 1.3254270203436433,
+      "learning_rate": 9.973924198508758e-06,
+      "loss": 0.7701,
+      "step": 514
+    },
+    {
+      "epoch": 0.37409484030599505,
+      "grad_norm": 1.2944959469357342,
+      "learning_rate": 9.97368014456763e-06,
+      "loss": 0.76,
+      "step": 515
+    },
+    {
+      "epoch": 0.374821238054162,
+      "grad_norm": 1.2474253013121739,
+      "learning_rate": 9.973434956854306e-06,
+      "loss": 0.7666,
+      "step": 516
+    },
+    {
+      "epoch": 0.375547635802329,
+      "grad_norm": 1.2656514925156077,
+      "learning_rate": 9.97318863542468e-06,
+      "loss": 0.7684,
+      "step": 517
+    },
+    {
+      "epoch": 0.376274033550496,
+      "grad_norm": 1.2260343270931553,
+      "learning_rate": 9.972941180334896e-06,
+      "loss": 0.7752,
+      "step": 518
+    },
+    {
+      "epoch": 0.37700043129866295,
+      "grad_norm": 1.2365101027609489,
+      "learning_rate": 9.972692591641367e-06,
+      "loss": 0.7659,
+      "step": 519
+    },
+    {
+      "epoch": 0.37772682904682997,
+      "grad_norm": 1.2107248676220899,
+      "learning_rate": 9.97244286940076e-06,
+      "loss": 0.7958,
+      "step": 520
+    },
+    {
+      "epoch": 0.37845322679499693,
+      "grad_norm": 1.305074081461733,
+      "learning_rate": 9.972192013669998e-06,
+      "loss": 0.7608,
+      "step": 521
+    },
+    {
+      "epoch": 0.3791796245431639,
+      "grad_norm": 1.2188322690177842,
+      "learning_rate": 9.971940024506265e-06,
+      "loss": 0.7864,
+      "step": 522
+    },
+    {
+      "epoch": 0.3799060222913309,
+      "grad_norm": 1.2134567730823578,
+      "learning_rate": 9.971686901967006e-06,
+      "loss": 0.7777,
+      "step": 523
+    },
+    {
+      "epoch": 0.3806324200394979,
+      "grad_norm": 1.214745326766063,
+      "learning_rate": 9.971432646109919e-06,
+      "loss": 0.7643,
+      "step": 524
+    },
+    {
+      "epoch": 0.38135881778766484,
+      "grad_norm": 3.637168647480618,
+      "learning_rate": 9.971177256992961e-06,
+      "loss": 0.7689,
+      "step": 525
+    },
+    {
+      "epoch": 0.38208521553583186,
+      "grad_norm": 1.195297556379348,
+      "learning_rate": 9.970920734674355e-06,
+      "loss": 0.7732,
+      "step": 526
+    },
+    {
+      "epoch": 0.3828116132839988,
+      "grad_norm": 1.2164501623111703,
+      "learning_rate": 9.970663079212568e-06,
+      "loss": 0.7609,
+      "step": 527
+    },
+    {
+      "epoch": 0.3835380110321658,
+      "grad_norm": 1.2178209396112816,
+      "learning_rate": 9.97040429066634e-06,
+      "loss": 0.7654,
+      "step": 528
+    },
+    {
+      "epoch": 0.3842644087803328,
+      "grad_norm": 1.196664997899127,
+      "learning_rate": 9.970144369094663e-06,
+      "loss": 0.7604,
+      "step": 529
+    },
+    {
+      "epoch": 0.38499080652849976,
+      "grad_norm": 1.205518971317643,
+      "learning_rate": 9.969883314556784e-06,
+      "loss": 0.7621,
+      "step": 530
+    },
+    {
+      "epoch": 0.3857172042766667,
+      "grad_norm": 1.1951038596907404,
+      "learning_rate": 9.969621127112211e-06,
+      "loss": 0.7821,
+      "step": 531
+    },
+    {
+      "epoch": 0.38644360202483374,
+      "grad_norm": 1.1948561987593798,
+      "learning_rate": 9.969357806820714e-06,
+      "loss": 0.7613,
+      "step": 532
+    },
+    {
+      "epoch": 0.3871699997730007,
+      "grad_norm": 1.2101490112809141,
+      "learning_rate": 9.969093353742318e-06,
+      "loss": 0.7612,
+      "step": 533
+    },
+    {
+      "epoch": 0.38789639752116767,
+      "grad_norm": 1.171209913607891,
+      "learning_rate": 9.968827767937303e-06,
+      "loss": 0.7801,
+      "step": 534
+    },
+    {
+      "epoch": 0.3886227952693347,
+      "grad_norm": 1.2245907679426409,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.7712,
+      "step": 535
+    },
+    {
+      "epoch": 0.38934919301750165,
+      "grad_norm": 1.2097923964427286,
+      "learning_rate": 9.96829319838985e-06,
+      "loss": 0.7504,
+      "step": 536
+    },
+    {
+      "epoch": 0.3900755907656686,
+      "grad_norm": 1.2500277611138257,
+      "learning_rate": 9.968024214769265e-06,
+      "loss": 0.7662,
+      "step": 537
+    },
+    {
+      "epoch": 0.39080198851383563,
+      "grad_norm": 1.1852000915918435,
+      "learning_rate": 9.967754098665778e-06,
+      "loss": 0.7614,
+      "step": 538
+    },
+    {
+      "epoch": 0.3915283862620026,
+      "grad_norm": 1.194954947602848,
+      "learning_rate": 9.967482850140965e-06,
+      "loss": 0.7758,
+      "step": 539
+    },
+    {
+      "epoch": 0.39225478401016955,
+      "grad_norm": 1.1815956688439189,
+      "learning_rate": 9.967210469256657e-06,
+      "loss": 0.7598,
+      "step": 540
+    },
+    {
+      "epoch": 0.3929811817583366,
+      "grad_norm": 1.1881258495264189,
+      "learning_rate": 9.96693695607494e-06,
+      "loss": 0.7642,
+      "step": 541
+    },
+    {
+      "epoch": 0.39370757950650354,
+      "grad_norm": 1.1408535421371642,
+      "learning_rate": 9.96666231065817e-06,
+      "loss": 0.7645,
+      "step": 542
+    },
+    {
+      "epoch": 0.3944339772546705,
+      "grad_norm": 1.2935039433273636,
+      "learning_rate": 9.966386533068949e-06,
+      "loss": 0.7826,
+      "step": 543
+    },
+    {
+      "epoch": 0.3951603750028375,
+      "grad_norm": 1.2067339559583887,
+      "learning_rate": 9.96610962337014e-06,
+      "loss": 0.7424,
+      "step": 544
+    },
+    {
+      "epoch": 0.3958867727510045,
+      "grad_norm": 1.1225159513779392,
+      "learning_rate": 9.965831581624872e-06,
+      "loss": 0.7654,
+      "step": 545
+    },
+    {
+      "epoch": 0.39661317049917144,
+      "grad_norm": 1.180080430089498,
+      "learning_rate": 9.965552407896519e-06,
+      "loss": 0.7655,
+      "step": 546
+    },
+    {
+      "epoch": 0.39733956824733846,
+      "grad_norm": 1.1256118419038732,
+      "learning_rate": 9.965272102248726e-06,
+      "loss": 0.7581,
+      "step": 547
+    },
+    {
+      "epoch": 0.3980659659955054,
+      "grad_norm": 1.1436250068062523,
+      "learning_rate": 9.964990664745387e-06,
+      "loss": 0.7493,
+      "step": 548
+    },
+    {
+      "epoch": 0.3987923637436724,
+      "grad_norm": 1.125774487206082,
+      "learning_rate": 9.964708095450655e-06,
+      "loss": 0.7563,
+      "step": 549
+    },
+    {
+      "epoch": 0.3995187614918394,
+      "grad_norm": 1.136420520925349,
+      "learning_rate": 9.964424394428943e-06,
+      "loss": 0.7648,
+      "step": 550
+    },
+    {
+      "epoch": 0.40024515924000637,
+      "grad_norm": 1.1580527989641898,
+      "learning_rate": 9.964139561744927e-06,
+      "loss": 0.7468,
+      "step": 551
+    },
+    {
+      "epoch": 0.4009715569881733,
+      "grad_norm": 1.1434802807692184,
+      "learning_rate": 9.963853597463533e-06,
+      "loss": 0.7502,
+      "step": 552
+    },
+    {
+      "epoch": 0.4016979547363403,
+      "grad_norm": 1.1142997290828731,
+      "learning_rate": 9.963566501649946e-06,
+      "loss": 0.7355,
+      "step": 553
+    },
+    {
+      "epoch": 0.4024243524845073,
+      "grad_norm": 1.0993449364698016,
+      "learning_rate": 9.963278274369613e-06,
+      "loss": 0.7438,
+      "step": 554
+    },
+    {
+      "epoch": 0.40315075023267427,
+      "grad_norm": 1.0645134866074886,
+      "learning_rate": 9.962988915688234e-06,
+      "loss": 0.7581,
+      "step": 555
+    },
+    {
+      "epoch": 0.40387714798084123,
+      "grad_norm": 1.0874163008360718,
+      "learning_rate": 9.962698425671773e-06,
+      "loss": 0.7559,
+      "step": 556
+    },
+    {
+      "epoch": 0.40460354572900825,
+      "grad_norm": 1.0893436970495776,
+      "learning_rate": 9.962406804386447e-06,
+      "loss": 0.7392,
+      "step": 557
+    },
+    {
+      "epoch": 0.4053299434771752,
+      "grad_norm": 1.1791025461727056,
+      "learning_rate": 9.962114051898731e-06,
+      "loss": 0.7468,
+      "step": 558
+    },
+    {
+      "epoch": 0.4060563412253422,
+      "grad_norm": 1.0617478387032344,
+      "learning_rate": 9.96182016827536e-06,
+      "loss": 0.7549,
+      "step": 559
+    },
+    {
+      "epoch": 0.4067827389735092,
+      "grad_norm": 1.0774497292109935,
+      "learning_rate": 9.961525153583327e-06,
+      "loss": 0.7415,
+      "step": 560
+    },
+    {
+      "epoch": 0.40750913672167616,
+      "grad_norm": 1.0821338781986698,
+      "learning_rate": 9.961229007889883e-06,
+      "loss": 0.7307,
+      "step": 561
+    },
+    {
+      "epoch": 0.4082355344698431,
+      "grad_norm": 1.0892656942041017,
+      "learning_rate": 9.960931731262533e-06,
+      "loss": 0.7427,
+      "step": 562
+    },
+    {
+      "epoch": 0.40896193221801014,
+      "grad_norm": 1.062469625082668,
+      "learning_rate": 9.960633323769043e-06,
+      "loss": 0.7498,
+      "step": 563
+    },
+    {
+      "epoch": 0.4096883299661771,
+      "grad_norm": 1.0668837175759642,
+      "learning_rate": 9.960333785477437e-06,
+      "loss": 0.7478,
+      "step": 564
+    },
+    {
+      "epoch": 0.41041472771434406,
+      "grad_norm": 1.0441183230317423,
+      "learning_rate": 9.960033116455997e-06,
+      "loss": 0.7559,
+      "step": 565
+    },
+    {
+      "epoch": 0.4111411254625111,
+      "grad_norm": 1.4159202842843437,
+      "learning_rate": 9.95973131677326e-06,
+      "loss": 0.7347,
+      "step": 566
+    },
+    {
+      "epoch": 0.41186752321067804,
+      "grad_norm": 1.2970154589976308,
+      "learning_rate": 9.959428386498023e-06,
+      "loss": 0.7428,
+      "step": 567
+    },
+    {
+      "epoch": 0.412593920958845,
+      "grad_norm": 1.0635260939557447,
+      "learning_rate": 9.95912432569934e-06,
+      "loss": 0.7684,
+      "step": 568
+    },
+    {
+      "epoch": 0.413320318707012,
+      "grad_norm": 1.065670306952603,
+      "learning_rate": 9.958819134446527e-06,
+      "loss": 0.7305,
+      "step": 569
+    },
+    {
+      "epoch": 0.414046716455179,
+      "grad_norm": 1.051820184891295,
+      "learning_rate": 9.958512812809148e-06,
+      "loss": 0.7417,
+      "step": 570
+    },
+    {
+      "epoch": 0.41477311420334595,
+      "grad_norm": 1.0932314529080025,
+      "learning_rate": 9.958205360857031e-06,
+      "loss": 0.757,
+      "step": 571
+    },
+    {
+      "epoch": 0.41549951195151297,
+      "grad_norm": 1.0904112546627889,
+      "learning_rate": 9.957896778660266e-06,
+      "loss": 0.7325,
+      "step": 572
+    },
+    {
+      "epoch": 0.41622590969967993,
+      "grad_norm": 1.0710868590287677,
+      "learning_rate": 9.957587066289189e-06,
+      "loss": 0.7532,
+      "step": 573
+    },
+    {
+      "epoch": 0.4169523074478469,
+      "grad_norm": 1.029792935403967,
+      "learning_rate": 9.957276223814405e-06,
+      "loss": 0.7487,
+      "step": 574
+    },
+    {
+      "epoch": 0.4176787051960139,
+      "grad_norm": 1.039615004460751,
+      "learning_rate": 9.95696425130677e-06,
+      "loss": 0.7484,
+      "step": 575
+    },
+    {
+      "epoch": 0.4184051029441809,
+      "grad_norm": 1.027265520457917,
+      "learning_rate": 9.956651148837402e-06,
+      "loss": 0.7431,
+      "step": 576
+    },
+    {
+      "epoch": 0.41913150069234784,
+      "grad_norm": 1.0445331233584851,
+      "learning_rate": 9.956336916477672e-06,
+      "loss": 0.7518,
+      "step": 577
+    },
+    {
+      "epoch": 0.41985789844051485,
+      "grad_norm": 1.0056629352503992,
+      "learning_rate": 9.95602155429921e-06,
+      "loss": 0.7442,
+      "step": 578
+    },
+    {
+      "epoch": 0.4205842961886818,
+      "grad_norm": 1.0364566761082272,
+      "learning_rate": 9.955705062373904e-06,
+      "loss": 0.7563,
+      "step": 579
+    },
+    {
+      "epoch": 0.4213106939368488,
+      "grad_norm": 1.0210209438240678,
+      "learning_rate": 9.955387440773902e-06,
+      "loss": 0.7484,
+      "step": 580
+    },
+    {
+      "epoch": 0.4220370916850158,
+      "grad_norm": 1.0282324113074215,
+      "learning_rate": 9.955068689571606e-06,
+      "loss": 0.7282,
+      "step": 581
+    },
+    {
+      "epoch": 0.42276348943318276,
+      "grad_norm": 1.1729764498651893,
+      "learning_rate": 9.954748808839675e-06,
+      "loss": 0.7542,
+      "step": 582
+    },
+    {
+      "epoch": 0.4234898871813497,
+      "grad_norm": 1.040462706700545,
+      "learning_rate": 9.95442779865103e-06,
+      "loss": 0.7194,
+      "step": 583
+    },
+    {
+      "epoch": 0.42421628492951674,
+      "grad_norm": 1.0062430347983087,
+      "learning_rate": 9.954105659078846e-06,
+      "loss": 0.7431,
+      "step": 584
+    },
+    {
+      "epoch": 0.4249426826776837,
+      "grad_norm": 1.0639035558270775,
+      "learning_rate": 9.953782390196554e-06,
+      "loss": 0.7492,
+      "step": 585
+    },
+    {
+      "epoch": 0.42566908042585067,
+      "grad_norm": 1.0425379187859118,
+      "learning_rate": 9.953457992077847e-06,
+      "loss": 0.7339,
+      "step": 586
+    },
+    {
+      "epoch": 0.4263954781740177,
+      "grad_norm": 1.0031591436611542,
+      "learning_rate": 9.953132464796674e-06,
+      "loss": 0.7358,
+      "step": 587
+    },
+    {
+      "epoch": 0.42712187592218465,
+      "grad_norm": 1.0015864996993507,
+      "learning_rate": 9.952805808427235e-06,
+      "loss": 0.7305,
+      "step": 588
+    },
+    {
+      "epoch": 0.4278482736703516,
+      "grad_norm": 1.8101689727197796,
+      "learning_rate": 9.952478023043999e-06,
+      "loss": 0.7501,
+      "step": 589
+    },
+    {
+      "epoch": 0.4285746714185186,
+      "grad_norm": 1.0253116313842865,
+      "learning_rate": 9.952149108721683e-06,
+      "loss": 0.7311,
+      "step": 590
+    },
+    {
+      "epoch": 0.4293010691666856,
+      "grad_norm": 0.9992846826039635,
+      "learning_rate": 9.951819065535265e-06,
+      "loss": 0.7172,
+      "step": 591
+    },
+    {
+      "epoch": 0.43002746691485255,
+      "grad_norm": 0.9817655480972577,
+      "learning_rate": 9.95148789355998e-06,
+      "loss": 0.7259,
+      "step": 592
+    },
+    {
+      "epoch": 0.43075386466301957,
+      "grad_norm": 1.063898145073969,
+      "learning_rate": 9.951155592871317e-06,
+      "loss": 0.7283,
+      "step": 593
+    },
+    {
+      "epoch": 0.43148026241118653,
+      "grad_norm": 0.9758454373620455,
+      "learning_rate": 9.950822163545032e-06,
+      "loss": 0.7472,
+      "step": 594
+    },
+    {
+      "epoch": 0.4322066601593535,
+      "grad_norm": 1.000367186605782,
+      "learning_rate": 9.950487605657126e-06,
+      "loss": 0.7452,
+      "step": 595
+    },
+    {
+      "epoch": 0.4329330579075205,
+      "grad_norm": 0.9540605112818284,
+      "learning_rate": 9.950151919283866e-06,
+      "loss": 0.742,
+      "step": 596
+    },
+    {
+      "epoch": 0.4336594556556875,
+      "grad_norm": 1.0039954162122182,
+      "learning_rate": 9.94981510450177e-06,
+      "loss": 0.734,
+      "step": 597
+    },
+    {
+      "epoch": 0.43438585340385444,
+      "grad_norm": 0.9744628011735277,
+      "learning_rate": 9.949477161387619e-06,
+      "loss": 0.7395,
+      "step": 598
+    },
+    {
+      "epoch": 0.43511225115202146,
+      "grad_norm": 0.9773584355212287,
+      "learning_rate": 9.94913809001845e-06,
+      "loss": 0.7315,
+      "step": 599
+    },
+    {
+      "epoch": 0.4358386489001884,
+      "grad_norm": 0.9760546875526702,
+      "learning_rate": 9.948797890471552e-06,
+      "loss": 0.731,
+      "step": 600
+    },
+    {
+      "epoch": 0.4365650466483554,
+      "grad_norm": 0.9911326295240364,
+      "learning_rate": 9.948456562824478e-06,
+      "loss": 0.7209,
+      "step": 601
+    },
+    {
+      "epoch": 0.43729144439652234,
+      "grad_norm": 0.9630389794461729,
+      "learning_rate": 9.948114107155032e-06,
+      "loss": 0.7185,
+      "step": 602
+    },
+    {
+      "epoch": 0.43801784214468936,
+      "grad_norm": 0.9643561574474725,
+      "learning_rate": 9.947770523541283e-06,
+      "loss": 0.7312,
+      "step": 603
+    },
+    {
+      "epoch": 0.4387442398928563,
+      "grad_norm": 0.9457862591453264,
+      "learning_rate": 9.947425812061546e-06,
+      "loss": 0.7334,
+      "step": 604
+    },
+    {
+      "epoch": 0.4394706376410233,
+      "grad_norm": 0.9764292467518713,
+      "learning_rate": 9.947079972794405e-06,
+      "loss": 0.719,
+      "step": 605
+    },
+    {
+      "epoch": 0.4401970353891903,
+      "grad_norm": 0.9652820848231466,
+      "learning_rate": 9.946733005818694e-06,
+      "loss": 0.7614,
+      "step": 606
+    },
+    {
+      "epoch": 0.44092343313735727,
+      "grad_norm": 0.9789870034754022,
+      "learning_rate": 9.946384911213504e-06,
+      "loss": 0.7505,
+      "step": 607
+    },
+    {
+      "epoch": 0.44164983088552423,
+      "grad_norm": 0.97261896897398,
+      "learning_rate": 9.946035689058189e-06,
+      "loss": 0.7269,
+      "step": 608
+    },
+    {
+      "epoch": 0.44237622863369125,
+      "grad_norm": 0.9512662501729403,
+      "learning_rate": 9.945685339432349e-06,
+      "loss": 0.7275,
+      "step": 609
+    },
+    {
+      "epoch": 0.4431026263818582,
+      "grad_norm": 0.9193936544152037,
+      "learning_rate": 9.945333862415855e-06,
+      "loss": 0.734,
+      "step": 610
+    },
+    {
+      "epoch": 0.4438290241300252,
+      "grad_norm": 0.9437084533207984,
+      "learning_rate": 9.944981258088822e-06,
+      "loss": 0.7455,
+      "step": 611
+    },
+    {
+      "epoch": 0.4445554218781922,
+      "grad_norm": 0.9366898434836213,
+      "learning_rate": 9.94462752653163e-06,
+      "loss": 0.7141,
+      "step": 612
+    },
+    {
+      "epoch": 0.44528181962635915,
+      "grad_norm": 0.9649835362940479,
+      "learning_rate": 9.944272667824915e-06,
+      "loss": 0.7172,
+      "step": 613
+    },
+    {
+      "epoch": 0.4460082173745261,
+      "grad_norm": 0.9480294246476275,
+      "learning_rate": 9.943916682049564e-06,
+      "loss": 0.7282,
+      "step": 614
+    },
+    {
+      "epoch": 0.44673461512269314,
+      "grad_norm": 0.9575970196238945,
+      "learning_rate": 9.943559569286731e-06,
+      "loss": 0.731,
+      "step": 615
+    },
+    {
+      "epoch": 0.4474610128708601,
+      "grad_norm": 1.0501616602935506,
+      "learning_rate": 9.943201329617819e-06,
+      "loss": 0.7344,
+      "step": 616
+    },
+    {
+      "epoch": 0.44818741061902706,
+      "grad_norm": 0.9169524659520641,
+      "learning_rate": 9.94284196312449e-06,
+      "loss": 0.7241,
+      "step": 617
+    },
+    {
+      "epoch": 0.4489138083671941,
+      "grad_norm": 0.8988178774810287,
+      "learning_rate": 9.942481469888664e-06,
+      "loss": 0.7263,
+      "step": 618
+    },
+    {
+      "epoch": 0.44964020611536104,
+      "grad_norm": 0.9133573838776243,
+      "learning_rate": 9.942119849992515e-06,
+      "loss": 0.7176,
+      "step": 619
+    },
+    {
+      "epoch": 0.450366603863528,
+      "grad_norm": 0.9312725209626799,
+      "learning_rate": 9.94175710351848e-06,
+      "loss": 0.7199,
+      "step": 620
+    },
+    {
+      "epoch": 0.451093001611695,
+      "grad_norm": 0.9231425800341202,
+      "learning_rate": 9.941393230549245e-06,
+      "loss": 0.721,
+      "step": 621
+    },
+    {
+      "epoch": 0.451819399359862,
+      "grad_norm": 0.8913757165093452,
+      "learning_rate": 9.941028231167756e-06,
+      "loss": 0.736,
+      "step": 622
+    },
+    {
+      "epoch": 0.45254579710802895,
+      "grad_norm": 0.9192511627739857,
+      "learning_rate": 9.940662105457219e-06,
+      "loss": 0.7311,
+      "step": 623
+    },
+    {
+      "epoch": 0.45327219485619596,
+      "grad_norm": 1.0126121327276585,
+      "learning_rate": 9.940294853501093e-06,
+      "loss": 0.7328,
+      "step": 624
+    },
+    {
+      "epoch": 0.4539985926043629,
+      "grad_norm": 0.9086059945770585,
+      "learning_rate": 9.939926475383093e-06,
+      "loss": 0.7306,
+      "step": 625
+    },
+    {
+      "epoch": 0.4547249903525299,
+      "grad_norm": 0.9006835415527965,
+      "learning_rate": 9.939556971187195e-06,
+      "loss": 0.7139,
+      "step": 626
+    },
+    {
+      "epoch": 0.4554513881006969,
+      "grad_norm": 0.8776531408479233,
+      "learning_rate": 9.939186340997629e-06,
+      "loss": 0.7256,
+      "step": 627
+    },
+    {
+      "epoch": 0.45617778584886387,
+      "grad_norm": 0.9131179128239798,
+      "learning_rate": 9.938814584898878e-06,
+      "loss": 0.7139,
+      "step": 628
+    },
+    {
+      "epoch": 0.45690418359703083,
+      "grad_norm": 0.911827974566791,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.7057,
+      "step": 629
+    },
+    {
+      "epoch": 0.45763058134519785,
+      "grad_norm": 0.9156346537960734,
+      "learning_rate": 9.938067695313062e-06,
+      "loss": 0.7087,
+      "step": 630
+    },
+    {
+      "epoch": 0.4583569790933648,
+      "grad_norm": 0.8880023303410371,
+      "learning_rate": 9.937692561996253e-06,
+      "loss": 0.718,
+      "step": 631
+    },
+    {
+      "epoch": 0.4590833768415318,
+      "grad_norm": 0.9401020700448064,
+      "learning_rate": 9.937316303110773e-06,
+      "loss": 0.7213,
+      "step": 632
+    },
+    {
+      "epoch": 0.4598097745896988,
+      "grad_norm": 0.8760498121504141,
+      "learning_rate": 9.936938918742396e-06,
+      "loss": 0.7087,
+      "step": 633
+    },
+    {
+      "epoch": 0.46053617233786576,
+      "grad_norm": 0.8471932856764717,
+      "learning_rate": 9.936560408977144e-06,
+      "loss": 0.7294,
+      "step": 634
+    },
+    {
+      "epoch": 0.4612625700860327,
+      "grad_norm": 0.8875415965820969,
+      "learning_rate": 9.936180773901305e-06,
+      "loss": 0.7232,
+      "step": 635
+    },
+    {
+      "epoch": 0.46198896783419974,
+      "grad_norm": 0.8747330864112066,
+      "learning_rate": 9.935800013601415e-06,
+      "loss": 0.7288,
+      "step": 636
+    },
+    {
+      "epoch": 0.4627153655823667,
+      "grad_norm": 0.8939346610383381,
+      "learning_rate": 9.935418128164271e-06,
+      "loss": 0.7422,
+      "step": 637
+    },
+    {
+      "epoch": 0.46344176333053366,
+      "grad_norm": 0.8639659321367306,
+      "learning_rate": 9.935035117676925e-06,
+      "loss": 0.7111,
+      "step": 638
+    },
+    {
+      "epoch": 0.4641681610787007,
+      "grad_norm": 0.8809990893660246,
+      "learning_rate": 9.934650982226687e-06,
+      "loss": 0.706,
+      "step": 639
+    },
+    {
+      "epoch": 0.46489455882686764,
+      "grad_norm": 0.8495844086453649,
+      "learning_rate": 9.93426572190112e-06,
+      "loss": 0.7274,
+      "step": 640
+    },
+    {
+      "epoch": 0.4656209565750346,
+      "grad_norm": 0.8457492178190025,
+      "learning_rate": 9.933879336788048e-06,
+      "loss": 0.7159,
+      "step": 641
+    },
+    {
+      "epoch": 0.4663473543232016,
+      "grad_norm": 0.872251965734634,
+      "learning_rate": 9.93349182697555e-06,
+      "loss": 0.7174,
+      "step": 642
+    },
+    {
+      "epoch": 0.4670737520713686,
+      "grad_norm": 0.8564380863748268,
+      "learning_rate": 9.933103192551958e-06,
+      "loss": 0.6997,
+      "step": 643
+    },
+    {
+      "epoch": 0.46780014981953555,
+      "grad_norm": 0.8539062885057994,
+      "learning_rate": 9.932713433605862e-06,
+      "loss": 0.7169,
+      "step": 644
+    },
+    {
+      "epoch": 0.46852654756770257,
+      "grad_norm": 0.8602195511545719,
+      "learning_rate": 9.932322550226114e-06,
+      "loss": 0.7267,
+      "step": 645
+    },
+    {
+      "epoch": 0.46925294531586953,
+      "grad_norm": 0.8660259689062596,
+      "learning_rate": 9.931930542501813e-06,
+      "loss": 0.7058,
+      "step": 646
+    },
+    {
+      "epoch": 0.4699793430640365,
+      "grad_norm": 0.8412500932338813,
+      "learning_rate": 9.931537410522323e-06,
+      "loss": 0.7117,
+      "step": 647
+    },
+    {
+      "epoch": 0.47070574081220345,
+      "grad_norm": 0.847365844467039,
+      "learning_rate": 9.931143154377257e-06,
+      "loss": 0.7038,
+      "step": 648
+    },
+    {
+      "epoch": 0.4714321385603705,
+      "grad_norm": 0.863076023805393,
+      "learning_rate": 9.930747774156485e-06,
+      "loss": 0.7236,
+      "step": 649
+    },
+    {
+      "epoch": 0.47215853630853744,
+      "grad_norm": 0.8578990699179848,
+      "learning_rate": 9.930351269950144e-06,
+      "loss": 0.7249,
+      "step": 650
+    },
+    {
+      "epoch": 0.4728849340567044,
+      "grad_norm": 0.8468784620880256,
+      "learning_rate": 9.92995364184861e-06,
+      "loss": 0.7111,
+      "step": 651
+    },
+    {
+      "epoch": 0.4736113318048714,
+      "grad_norm": 0.8548540373716411,
+      "learning_rate": 9.92955488994253e-06,
+      "loss": 0.7054,
+      "step": 652
+    },
+    {
+      "epoch": 0.4743377295530384,
+      "grad_norm": 0.8263217534184496,
+      "learning_rate": 9.9291550143228e-06,
+      "loss": 0.7013,
+      "step": 653
+    },
+    {
+      "epoch": 0.47506412730120534,
+      "grad_norm": 0.8558868202871573,
+      "learning_rate": 9.928754015080571e-06,
+      "loss": 0.6998,
+      "step": 654
+    },
+    {
+      "epoch": 0.47579052504937236,
+      "grad_norm": 0.8527525987219964,
+      "learning_rate": 9.928351892307254e-06,
+      "loss": 0.706,
+      "step": 655
+    },
+    {
+      "epoch": 0.4765169227975393,
+      "grad_norm": 0.8096658723028995,
+      "learning_rate": 9.927948646094514e-06,
+      "loss": 0.7104,
+      "step": 656
+    },
+    {
+      "epoch": 0.4772433205457063,
+      "grad_norm": 0.825386383857699,
+      "learning_rate": 9.927544276534275e-06,
+      "loss": 0.7263,
+      "step": 657
+    },
+    {
+      "epoch": 0.4779697182938733,
+      "grad_norm": 0.8110031620274161,
+      "learning_rate": 9.927138783718713e-06,
+      "loss": 0.7327,
+      "step": 658
+    },
+    {
+      "epoch": 0.47869611604204027,
+      "grad_norm": 0.8725373219371803,
+      "learning_rate": 9.926732167740262e-06,
+      "loss": 0.7062,
+      "step": 659
+    },
+    {
+      "epoch": 0.4794225137902072,
+      "grad_norm": 0.84156185662018,
+      "learning_rate": 9.926324428691612e-06,
+      "loss": 0.7184,
+      "step": 660
+    },
+    {
+      "epoch": 0.48014891153837425,
+      "grad_norm": 0.8234352820652021,
+      "learning_rate": 9.92591556666571e-06,
+      "loss": 0.6966,
+      "step": 661
+    },
+    {
+      "epoch": 0.4808753092865412,
+      "grad_norm": 0.8392923258120455,
+      "learning_rate": 9.925505581755754e-06,
+      "loss": 0.7033,
+      "step": 662
+    },
+    {
+      "epoch": 0.48160170703470817,
+      "grad_norm": 0.8175945014983422,
+      "learning_rate": 9.925094474055207e-06,
+      "loss": 0.7089,
+      "step": 663
+    },
+    {
+      "epoch": 0.4823281047828752,
+      "grad_norm": 0.8911331780707599,
+      "learning_rate": 9.92468224365778e-06,
+      "loss": 0.6947,
+      "step": 664
+    },
+    {
+      "epoch": 0.48305450253104215,
+      "grad_norm": 0.8111671666726761,
+      "learning_rate": 9.924268890657443e-06,
+      "loss": 0.7013,
+      "step": 665
+    },
+    {
+      "epoch": 0.4837809002792091,
+      "grad_norm": 0.820218156032821,
+      "learning_rate": 9.923854415148423e-06,
+      "loss": 0.7069,
+      "step": 666
+    },
+    {
+      "epoch": 0.48450729802737613,
+      "grad_norm": 0.8368104072898442,
+      "learning_rate": 9.923438817225198e-06,
+      "loss": 0.7056,
+      "step": 667
+    },
+    {
+      "epoch": 0.4852336957755431,
+      "grad_norm": 0.8172858374676331,
+      "learning_rate": 9.92302209698251e-06,
+      "loss": 0.7113,
+      "step": 668
+    },
+    {
+      "epoch": 0.48596009352371006,
+      "grad_norm": 0.8111333167554251,
+      "learning_rate": 9.922604254515348e-06,
+      "loss": 0.6986,
+      "step": 669
+    },
+    {
+      "epoch": 0.4866864912718771,
+      "grad_norm": 0.7939369425854435,
+      "learning_rate": 9.922185289918965e-06,
+      "loss": 0.7197,
+      "step": 670
+    },
+    {
+      "epoch": 0.48741288902004404,
+      "grad_norm": 0.8073816864202008,
+      "learning_rate": 9.921765203288862e-06,
+      "loss": 0.6938,
+      "step": 671
+    },
+    {
+      "epoch": 0.488139286768211,
+      "grad_norm": 0.8146062858167089,
+      "learning_rate": 9.921343994720803e-06,
+      "loss": 0.6901,
+      "step": 672
+    },
+    {
+      "epoch": 0.488865684516378,
+      "grad_norm": 0.8073049273734886,
+      "learning_rate": 9.920921664310801e-06,
+      "loss": 0.689,
+      "step": 673
+    },
+    {
+      "epoch": 0.489592082264545,
+      "grad_norm": 0.8000480150131922,
+      "learning_rate": 9.920498212155132e-06,
+      "loss": 0.6939,
+      "step": 674
+    },
+    {
+      "epoch": 0.49031848001271194,
+      "grad_norm": 0.8068016192638968,
+      "learning_rate": 9.920073638350319e-06,
+      "loss": 0.7025,
+      "step": 675
+    },
+    {
+      "epoch": 0.49104487776087896,
+      "grad_norm": 0.8260934820137759,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.6908,
+      "step": 676
+    },
+    {
+      "epoch": 0.4917712755090459,
+      "grad_norm": 0.7997873997828585,
+      "learning_rate": 9.91922112618066e-06,
+      "loss": 0.7041,
+      "step": 677
+    },
+    {
+      "epoch": 0.4924976732572129,
+      "grad_norm": 2.174322927976191,
+      "learning_rate": 9.918793188010147e-06,
+      "loss": 0.7143,
+      "step": 678
+    },
+    {
+      "epoch": 0.4932240710053799,
+      "grad_norm": 0.8298841272183457,
+      "learning_rate": 9.91836412857916e-06,
+      "loss": 0.7002,
+      "step": 679
+    },
+    {
+      "epoch": 0.49395046875354687,
+      "grad_norm": 0.815955296740638,
+      "learning_rate": 9.917933947985508e-06,
+      "loss": 0.6999,
+      "step": 680
+    },
+    {
+      "epoch": 0.49467686650171383,
+      "grad_norm": 0.8141154640169115,
+      "learning_rate": 9.917502646327247e-06,
+      "loss": 0.7185,
+      "step": 681
+    },
+    {
+      "epoch": 0.49540326424988085,
+      "grad_norm": 0.8441227422366394,
+      "learning_rate": 9.917070223702697e-06,
+      "loss": 0.7182,
+      "step": 682
+    },
+    {
+      "epoch": 0.4961296619980478,
+      "grad_norm": 0.7881967510667903,
+      "learning_rate": 9.916636680210431e-06,
+      "loss": 0.7189,
+      "step": 683
+    },
+    {
+      "epoch": 0.4968560597462148,
+      "grad_norm": 0.787513266332936,
+      "learning_rate": 9.916202015949277e-06,
+      "loss": 0.6961,
+      "step": 684
+    },
+    {
+      "epoch": 0.4975824574943818,
+      "grad_norm": 0.8297981418322181,
+      "learning_rate": 9.915766231018317e-06,
+      "loss": 0.7032,
+      "step": 685
+    },
+    {
+      "epoch": 0.49830885524254875,
+      "grad_norm": 0.7882950365342019,
+      "learning_rate": 9.915329325516894e-06,
+      "loss": 0.6936,
+      "step": 686
+    },
+    {
+      "epoch": 0.4990352529907157,
+      "grad_norm": 0.8012742339373384,
+      "learning_rate": 9.914891299544599e-06,
+      "loss": 0.6999,
+      "step": 687
+    },
+    {
+      "epoch": 0.49976165073888273,
+      "grad_norm": 0.7831837037525278,
+      "learning_rate": 9.914452153201282e-06,
+      "loss": 0.7051,
+      "step": 688
+    },
+    {
+      "epoch": 0.5004880484870496,
+      "grad_norm": 0.7862014823443961,
+      "learning_rate": 9.91401188658705e-06,
+      "loss": 0.6818,
+      "step": 689
+    },
+    {
+      "epoch": 0.5012144462352167,
+      "grad_norm": 0.7812613534102408,
+      "learning_rate": 9.91357049980226e-06,
+      "loss": 0.7019,
+      "step": 690
+    },
+    {
+      "epoch": 0.5019408439833837,
+      "grad_norm": 0.7697961842553203,
+      "learning_rate": 9.913127992947534e-06,
+      "loss": 0.7026,
+      "step": 691
+    },
+    {
+      "epoch": 0.5026672417315506,
+      "grad_norm": 0.7706971322719415,
+      "learning_rate": 9.91268436612374e-06,
+      "loss": 0.6966,
+      "step": 692
+    },
+    {
+      "epoch": 0.5033936394797176,
+      "grad_norm": 0.8218604590916418,
+      "learning_rate": 9.912239619432004e-06,
+      "loss": 0.7078,
+      "step": 693
+    },
+    {
+      "epoch": 0.5041200372278846,
+      "grad_norm": 0.8656142603168385,
+      "learning_rate": 9.911793752973711e-06,
+      "loss": 0.6914,
+      "step": 694
+    },
+    {
+      "epoch": 0.5048464349760515,
+      "grad_norm": 0.7750652898480885,
+      "learning_rate": 9.911346766850493e-06,
+      "loss": 0.6906,
+      "step": 695
+    },
+    {
+      "epoch": 0.5055728327242186,
+      "grad_norm": 0.7414098590179129,
+      "learning_rate": 9.910898661164249e-06,
+      "loss": 0.7109,
+      "step": 696
+    },
+    {
+      "epoch": 0.5062992304723856,
+      "grad_norm": 0.7569073529543336,
+      "learning_rate": 9.910449436017123e-06,
+      "loss": 0.7044,
+      "step": 697
+    },
+    {
+      "epoch": 0.5070256282205525,
+      "grad_norm": 0.7519787240625135,
+      "learning_rate": 9.909999091511516e-06,
+      "loss": 0.6964,
+      "step": 698
+    },
+    {
+      "epoch": 0.5077520259687195,
+      "grad_norm": 0.7733018395208994,
+      "learning_rate": 9.909547627750089e-06,
+      "loss": 0.7088,
+      "step": 699
+    },
+    {
+      "epoch": 0.5084784237168865,
+      "grad_norm": 0.7721039060567157,
+      "learning_rate": 9.909095044835755e-06,
+      "loss": 0.6913,
+      "step": 700
+    },
+    {
+      "epoch": 0.5092048214650534,
+      "grad_norm": 0.7385882659818571,
+      "learning_rate": 9.90864134287168e-06,
+      "loss": 0.693,
+      "step": 701
+    },
+    {
+      "epoch": 0.5099312192132205,
+      "grad_norm": 0.7421782318904324,
+      "learning_rate": 9.90818652196129e-06,
+      "loss": 0.7196,
+      "step": 702
+    },
+    {
+      "epoch": 0.5106576169613875,
+      "grad_norm": 0.7481323889267485,
+      "learning_rate": 9.907730582208261e-06,
+      "loss": 0.6999,
+      "step": 703
+    },
+    {
+      "epoch": 0.5113840147095544,
+      "grad_norm": 0.7601344672286628,
+      "learning_rate": 9.90727352371653e-06,
+      "loss": 0.698,
+      "step": 704
+    },
+    {
+      "epoch": 0.5121104124577214,
+      "grad_norm": 0.7483342620276054,
+      "learning_rate": 9.906815346590282e-06,
+      "loss": 0.7051,
+      "step": 705
+    },
+    {
+      "epoch": 0.5128368102058883,
+      "grad_norm": 0.818152958994249,
+      "learning_rate": 9.906356050933962e-06,
+      "loss": 0.6999,
+      "step": 706
+    },
+    {
+      "epoch": 0.5135632079540553,
+      "grad_norm": 0.7478510985456449,
+      "learning_rate": 9.905895636852268e-06,
+      "loss": 0.7063,
+      "step": 707
+    },
+    {
+      "epoch": 0.5142896057022224,
+      "grad_norm": 0.8360311926939497,
+      "learning_rate": 9.905434104450153e-06,
+      "loss": 0.7075,
+      "step": 708
+    },
+    {
+      "epoch": 0.5150160034503893,
+      "grad_norm": 0.728081494414813,
+      "learning_rate": 9.904971453832828e-06,
+      "loss": 0.6928,
+      "step": 709
+    },
+    {
+      "epoch": 0.5157424011985563,
+      "grad_norm": 0.7732823717838929,
+      "learning_rate": 9.904507685105753e-06,
+      "loss": 0.6804,
+      "step": 710
+    },
+    {
+      "epoch": 0.5164687989467233,
+      "grad_norm": 0.732175100079485,
+      "learning_rate": 9.904042798374647e-06,
+      "loss": 0.7015,
+      "step": 711
+    },
+    {
+      "epoch": 0.5171951966948902,
+      "grad_norm": 0.7521814228879359,
+      "learning_rate": 9.903576793745482e-06,
+      "loss": 0.6938,
+      "step": 712
+    },
+    {
+      "epoch": 0.5179215944430572,
+      "grad_norm": 0.727963567904356,
+      "learning_rate": 9.903109671324488e-06,
+      "loss": 0.6927,
+      "step": 713
+    },
+    {
+      "epoch": 0.5186479921912243,
+      "grad_norm": 0.8051400934341161,
+      "learning_rate": 9.902641431218149e-06,
+      "loss": 0.6845,
+      "step": 714
+    },
+    {
+      "epoch": 0.5193743899393912,
+      "grad_norm": 0.7399599754417058,
+      "learning_rate": 9.902172073533198e-06,
+      "loss": 0.702,
+      "step": 715
+    },
+    {
+      "epoch": 0.5201007876875582,
+      "grad_norm": 0.7177232225200838,
+      "learning_rate": 9.90170159837663e-06,
+      "loss": 0.6854,
+      "step": 716
+    },
+    {
+      "epoch": 0.5208271854357251,
+      "grad_norm": 0.7250491270529293,
+      "learning_rate": 9.90123000585569e-06,
+      "loss": 0.6718,
+      "step": 717
+    },
+    {
+      "epoch": 0.5215535831838921,
+      "grad_norm": 0.8226795700631545,
+      "learning_rate": 9.90075729607788e-06,
+      "loss": 0.6718,
+      "step": 718
+    },
+    {
+      "epoch": 0.5222799809320591,
+      "grad_norm": 0.7404005782833571,
+      "learning_rate": 9.900283469150959e-06,
+      "loss": 0.6894,
+      "step": 719
+    },
+    {
+      "epoch": 0.523006378680226,
+      "grad_norm": 0.7201992579567925,
+      "learning_rate": 9.899808525182935e-06,
+      "loss": 0.6878,
+      "step": 720
+    },
+    {
+      "epoch": 0.5237327764283931,
+      "grad_norm": 0.7085334339463191,
+      "learning_rate": 9.899332464282075e-06,
+      "loss": 0.6953,
+      "step": 721
+    },
+    {
+      "epoch": 0.5244591741765601,
+      "grad_norm": 0.7360701211727764,
+      "learning_rate": 9.898855286556897e-06,
+      "loss": 0.7087,
+      "step": 722
+    },
+    {
+      "epoch": 0.525185571924727,
+      "grad_norm": 0.7083495495920359,
+      "learning_rate": 9.898376992116179e-06,
+      "loss": 0.6957,
+      "step": 723
+    },
+    {
+      "epoch": 0.525911969672894,
+      "grad_norm": 0.7240208436144087,
+      "learning_rate": 9.897897581068948e-06,
+      "loss": 0.6986,
+      "step": 724
+    },
+    {
+      "epoch": 0.526638367421061,
+      "grad_norm": 0.7409795690878006,
+      "learning_rate": 9.897417053524487e-06,
+      "loss": 0.6898,
+      "step": 725
+    },
+    {
+      "epoch": 0.5273647651692279,
+      "grad_norm": 0.7292619216347412,
+      "learning_rate": 9.896935409592339e-06,
+      "loss": 0.6915,
+      "step": 726
+    },
+    {
+      "epoch": 0.528091162917395,
+      "grad_norm": 0.825328679283231,
+      "learning_rate": 9.896452649382291e-06,
+      "loss": 0.6867,
+      "step": 727
+    },
+    {
+      "epoch": 0.528817560665562,
+      "grad_norm": 0.731154122216054,
+      "learning_rate": 9.895968773004394e-06,
+      "loss": 0.6741,
+      "step": 728
+    },
+    {
+      "epoch": 0.5295439584137289,
+      "grad_norm": 0.7268817197791653,
+      "learning_rate": 9.895483780568949e-06,
+      "loss": 0.6728,
+      "step": 729
+    },
+    {
+      "epoch": 0.5302703561618959,
+      "grad_norm": 0.8850345181917018,
+      "learning_rate": 9.894997672186513e-06,
+      "loss": 0.6674,
+      "step": 730
+    },
+    {
+      "epoch": 0.5309967539100628,
+      "grad_norm": 0.7129326314830353,
+      "learning_rate": 9.894510447967893e-06,
+      "loss": 0.6949,
+      "step": 731
+    },
+    {
+      "epoch": 0.5317231516582298,
+      "grad_norm": 0.7445881485337487,
+      "learning_rate": 9.894022108024157e-06,
+      "loss": 0.6958,
+      "step": 732
+    },
+    {
+      "epoch": 0.5324495494063969,
+      "grad_norm": 0.7087267238053241,
+      "learning_rate": 9.893532652466625e-06,
+      "loss": 0.6826,
+      "step": 733
+    },
+    {
+      "epoch": 0.5331759471545638,
+      "grad_norm": 0.7029975399745358,
+      "learning_rate": 9.893042081406868e-06,
+      "loss": 0.691,
+      "step": 734
+    },
+    {
+      "epoch": 0.5339023449027308,
+      "grad_norm": 0.7357703371007258,
+      "learning_rate": 9.892550394956715e-06,
+      "loss": 0.6657,
+      "step": 735
+    },
+    {
+      "epoch": 0.5346287426508978,
+      "grad_norm": 0.7042415378243042,
+      "learning_rate": 9.892057593228248e-06,
+      "loss": 0.6669,
+      "step": 736
+    },
+    {
+      "epoch": 0.5353551403990647,
+      "grad_norm": 0.7537027415403048,
+      "learning_rate": 9.891563676333805e-06,
+      "loss": 0.7107,
+      "step": 737
+    },
+    {
+      "epoch": 0.5360815381472317,
+      "grad_norm": 0.716171888686636,
+      "learning_rate": 9.891068644385973e-06,
+      "loss": 0.679,
+      "step": 738
+    },
+    {
+      "epoch": 0.5368079358953988,
+      "grad_norm": 0.7988047089210955,
+      "learning_rate": 9.890572497497598e-06,
+      "loss": 0.6931,
+      "step": 739
+    },
+    {
+      "epoch": 0.5375343336435657,
+      "grad_norm": 0.7574454818586034,
+      "learning_rate": 9.89007523578178e-06,
+      "loss": 0.6924,
+      "step": 740
+    },
+    {
+      "epoch": 0.5382607313917327,
+      "grad_norm": 0.6985625682424247,
+      "learning_rate": 9.889576859351873e-06,
+      "loss": 0.6986,
+      "step": 741
+    },
+    {
+      "epoch": 0.5389871291398997,
+      "grad_norm": 0.9035477970405281,
+      "learning_rate": 9.88907736832148e-06,
+      "loss": 0.6931,
+      "step": 742
+    },
+    {
+      "epoch": 0.5397135268880666,
+      "grad_norm": 0.6894579767672432,
+      "learning_rate": 9.888576762804465e-06,
+      "loss": 0.6933,
+      "step": 743
+    },
+    {
+      "epoch": 0.5404399246362336,
+      "grad_norm": 0.6913186724338037,
+      "learning_rate": 9.888075042914946e-06,
+      "loss": 0.7082,
+      "step": 744
+    },
+    {
+      "epoch": 0.5411663223844007,
+      "grad_norm": 0.6914903982190204,
+      "learning_rate": 9.887572208767287e-06,
+      "loss": 0.6932,
+      "step": 745
+    },
+    {
+      "epoch": 0.5418927201325676,
+      "grad_norm": 5.796056874784305,
+      "learning_rate": 9.887068260476112e-06,
+      "loss": 0.6791,
+      "step": 746
+    },
+    {
+      "epoch": 0.5426191178807346,
+      "grad_norm": 0.8081792780426444,
+      "learning_rate": 9.886563198156302e-06,
+      "loss": 0.69,
+      "step": 747
+    },
+    {
+      "epoch": 0.5433455156289015,
+      "grad_norm": 0.9851928484816941,
+      "learning_rate": 9.886057021922984e-06,
+      "loss": 0.6944,
+      "step": 748
+    },
+    {
+      "epoch": 0.5440719133770685,
+      "grad_norm": 1.0116190695537544,
+      "learning_rate": 9.885549731891547e-06,
+      "loss": 0.6921,
+      "step": 749
+    },
+    {
+      "epoch": 0.5447983111252355,
+      "grad_norm": 1.1344844671703262,
+      "learning_rate": 9.885041328177625e-06,
+      "loss": 0.7013,
+      "step": 750
+    },
+    {
+      "epoch": 0.5455247088734025,
+      "grad_norm": 0.8914720633431127,
+      "learning_rate": 9.884531810897115e-06,
+      "loss": 0.6885,
+      "step": 751
+    },
+    {
+      "epoch": 0.5462511066215695,
+      "grad_norm": 0.8559206929302882,
+      "learning_rate": 9.884021180166164e-06,
+      "loss": 0.683,
+      "step": 752
+    },
+    {
+      "epoch": 0.5469775043697365,
+      "grad_norm": 0.8223803501330558,
+      "learning_rate": 9.883509436101168e-06,
+      "loss": 0.6741,
+      "step": 753
+    },
+    {
+      "epoch": 0.5477039021179034,
+      "grad_norm": 0.8233029896285243,
+      "learning_rate": 9.882996578818786e-06,
+      "loss": 0.6931,
+      "step": 754
+    },
+    {
+      "epoch": 0.5484302998660704,
+      "grad_norm": 0.8167149294453134,
+      "learning_rate": 9.882482608435924e-06,
+      "loss": 0.6798,
+      "step": 755
+    },
+    {
+      "epoch": 0.5491566976142374,
+      "grad_norm": 0.8157887270833556,
+      "learning_rate": 9.881967525069745e-06,
+      "loss": 0.6746,
+      "step": 756
+    },
+    {
+      "epoch": 0.5498830953624044,
+      "grad_norm": 0.786640198971054,
+      "learning_rate": 9.881451328837663e-06,
+      "loss": 0.7005,
+      "step": 757
+    },
+    {
+      "epoch": 0.5506094931105714,
+      "grad_norm": 0.7642969715869354,
+      "learning_rate": 9.880934019857346e-06,
+      "loss": 0.6806,
+      "step": 758
+    },
+    {
+      "epoch": 0.5513358908587384,
+      "grad_norm": 0.7827436991540868,
+      "learning_rate": 9.880415598246722e-06,
+      "loss": 0.6789,
+      "step": 759
+    },
+    {
+      "epoch": 0.5520622886069053,
+      "grad_norm": 0.7550650650499403,
+      "learning_rate": 9.879896064123961e-06,
+      "loss": 0.6859,
+      "step": 760
+    },
+    {
+      "epoch": 0.5527886863550723,
+      "grad_norm": 0.7314415455404134,
+      "learning_rate": 9.879375417607498e-06,
+      "loss": 0.673,
+      "step": 761
+    },
+    {
+      "epoch": 0.5535150841032392,
+      "grad_norm": 0.725667106303544,
+      "learning_rate": 9.878853658816015e-06,
+      "loss": 0.6722,
+      "step": 762
+    },
+    {
+      "epoch": 0.5542414818514063,
+      "grad_norm": 0.7253088889076802,
+      "learning_rate": 9.878330787868448e-06,
+      "loss": 0.6831,
+      "step": 763
+    },
+    {
+      "epoch": 0.5549678795995733,
+      "grad_norm": 0.7118746003515296,
+      "learning_rate": 9.87780680488399e-06,
+      "loss": 0.6854,
+      "step": 764
+    },
+    {
+      "epoch": 0.5556942773477402,
+      "grad_norm": 0.7124344265859947,
+      "learning_rate": 9.877281709982082e-06,
+      "loss": 0.6668,
+      "step": 765
+    },
+    {
+      "epoch": 0.5564206750959072,
+      "grad_norm": 0.6870086540392398,
+      "learning_rate": 9.876755503282425e-06,
+      "loss": 0.669,
+      "step": 766
+    },
+    {
+      "epoch": 0.5571470728440742,
+      "grad_norm": 0.6921474230850522,
+      "learning_rate": 9.87622818490497e-06,
+      "loss": 0.6735,
+      "step": 767
+    },
+    {
+      "epoch": 0.5578734705922411,
+      "grad_norm": 0.6691283320223589,
+      "learning_rate": 9.875699754969919e-06,
+      "loss": 0.6788,
+      "step": 768
+    },
+    {
+      "epoch": 0.5585998683404081,
+      "grad_norm": 0.7111757456950618,
+      "learning_rate": 9.875170213597731e-06,
+      "loss": 0.68,
+      "step": 769
+    },
+    {
+      "epoch": 0.5593262660885752,
+      "grad_norm": 0.6912163957683205,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.6835,
+      "step": 770
+    },
+    {
+      "epoch": 0.5600526638367421,
+      "grad_norm": 0.7059079937216522,
+      "learning_rate": 9.874107797025045e-06,
+      "loss": 0.6836,
+      "step": 771
+    },
+    {
+      "epoch": 0.5607790615849091,
+      "grad_norm": 0.6943502146356935,
+      "learning_rate": 9.87357492206673e-06,
+      "loss": 0.688,
+      "step": 772
+    },
+    {
+      "epoch": 0.561505459333076,
+      "grad_norm": 0.663025793246858,
+      "learning_rate": 9.87304093615564e-06,
+      "loss": 0.6801,
+      "step": 773
+    },
+    {
+      "epoch": 0.562231857081243,
+      "grad_norm": 0.680073642969289,
+      "learning_rate": 9.872505839413504e-06,
+      "loss": 0.6752,
+      "step": 774
+    },
+    {
+      "epoch": 0.56295825482941,
+      "grad_norm": 0.8247131197237683,
+      "learning_rate": 9.8719696319623e-06,
+      "loss": 0.6897,
+      "step": 775
+    },
+    {
+      "epoch": 0.563684652577577,
+      "grad_norm": 0.6800749444531121,
+      "learning_rate": 9.871432313924255e-06,
+      "loss": 0.6944,
+      "step": 776
+    },
+    {
+      "epoch": 0.564411050325744,
+      "grad_norm": 0.6589468936996172,
+      "learning_rate": 9.870893885421856e-06,
+      "loss": 0.6891,
+      "step": 777
+    },
+    {
+      "epoch": 0.565137448073911,
+      "grad_norm": 0.6699789187345543,
+      "learning_rate": 9.870354346577839e-06,
+      "loss": 0.6689,
+      "step": 778
+    },
+    {
+      "epoch": 0.5658638458220779,
+      "grad_norm": 0.6407811046259845,
+      "learning_rate": 9.869813697515197e-06,
+      "loss": 0.6662,
+      "step": 779
+    },
+    {
+      "epoch": 0.5665902435702449,
+      "grad_norm": 0.6621715745362748,
+      "learning_rate": 9.869271938357168e-06,
+      "loss": 0.6789,
+      "step": 780
+    },
+    {
+      "epoch": 0.5673166413184119,
+      "grad_norm": 0.6559166860078044,
+      "learning_rate": 9.868729069227253e-06,
+      "loss": 0.6734,
+      "step": 781
+    },
+    {
+      "epoch": 0.5680430390665789,
+      "grad_norm": 0.6934565492823604,
+      "learning_rate": 9.868185090249198e-06,
+      "loss": 0.6946,
+      "step": 782
+    },
+    {
+      "epoch": 0.5687694368147459,
+      "grad_norm": 0.6727050469819293,
+      "learning_rate": 9.867640001547007e-06,
+      "loss": 0.6851,
+      "step": 783
+    },
+    {
+      "epoch": 0.5694958345629129,
+      "grad_norm": 0.6829495739508391,
+      "learning_rate": 9.867093803244935e-06,
+      "loss": 0.664,
+      "step": 784
+    },
+    {
+      "epoch": 0.5702222323110798,
+      "grad_norm": 0.6481246326312373,
+      "learning_rate": 9.866546495467493e-06,
+      "loss": 0.6779,
+      "step": 785
+    },
+    {
+      "epoch": 0.5709486300592468,
+      "grad_norm": 0.6439766189909788,
+      "learning_rate": 9.86599807833944e-06,
+      "loss": 0.6581,
+      "step": 786
+    },
+    {
+      "epoch": 0.5716750278074137,
+      "grad_norm": 0.687659375861452,
+      "learning_rate": 9.865448551985788e-06,
+      "loss": 0.6766,
+      "step": 787
+    },
+    {
+      "epoch": 0.5724014255555808,
+      "grad_norm": 0.6283754813822696,
+      "learning_rate": 9.864897916531807e-06,
+      "loss": 0.6718,
+      "step": 788
+    },
+    {
+      "epoch": 0.5731278233037478,
+      "grad_norm": 0.6560916613823546,
+      "learning_rate": 9.864346172103016e-06,
+      "loss": 0.6791,
+      "step": 789
+    },
+    {
+      "epoch": 0.5738542210519147,
+      "grad_norm": 0.6579158607817118,
+      "learning_rate": 9.863793318825186e-06,
+      "loss": 0.6719,
+      "step": 790
+    },
+    {
+      "epoch": 0.5745806188000817,
+      "grad_norm": 0.6209174204285736,
+      "learning_rate": 9.863239356824346e-06,
+      "loss": 0.6668,
+      "step": 791
+    },
+    {
+      "epoch": 0.5753070165482487,
+      "grad_norm": 0.6674035902894433,
+      "learning_rate": 9.862684286226769e-06,
+      "loss": 0.663,
+      "step": 792
+    },
+    {
+      "epoch": 0.5760334142964156,
+      "grad_norm": 0.6390144070498434,
+      "learning_rate": 9.86212810715899e-06,
+      "loss": 0.6734,
+      "step": 793
+    },
+    {
+      "epoch": 0.5767598120445827,
+      "grad_norm": 0.6627367196299603,
+      "learning_rate": 9.86157081974779e-06,
+      "loss": 0.6527,
+      "step": 794
+    },
+    {
+      "epoch": 0.5774862097927497,
+      "grad_norm": 0.6596657350941416,
+      "learning_rate": 9.861012424120207e-06,
+      "loss": 0.6535,
+      "step": 795
+    },
+    {
+      "epoch": 0.5782126075409166,
+      "grad_norm": 0.6778250210754974,
+      "learning_rate": 9.860452920403528e-06,
+      "loss": 0.6806,
+      "step": 796
+    },
+    {
+      "epoch": 0.5789390052890836,
+      "grad_norm": 0.7402223552090328,
+      "learning_rate": 9.859892308725296e-06,
+      "loss": 0.6659,
+      "step": 797
+    },
+    {
+      "epoch": 0.5796654030372506,
+      "grad_norm": 0.6521000937099793,
+      "learning_rate": 9.859330589213302e-06,
+      "loss": 0.6747,
+      "step": 798
+    },
+    {
+      "epoch": 0.5803918007854175,
+      "grad_norm": 0.6589709652600224,
+      "learning_rate": 9.858767761995597e-06,
+      "loss": 0.6516,
+      "step": 799
+    },
+    {
+      "epoch": 0.5811181985335846,
+      "grad_norm": 0.6221340208951421,
+      "learning_rate": 9.858203827200477e-06,
+      "loss": 0.6575,
+      "step": 800
+    },
+    {
+      "epoch": 0.5818445962817516,
+      "grad_norm": 1.4392475113310477,
+      "learning_rate": 9.857638784956492e-06,
+      "loss": 0.6735,
+      "step": 801
+    },
+    {
+      "epoch": 0.5825709940299185,
+      "grad_norm": 0.6345408144241541,
+      "learning_rate": 9.857072635392449e-06,
+      "loss": 0.6704,
+      "step": 802
+    },
+    {
+      "epoch": 0.5832973917780855,
+      "grad_norm": 0.7024596655943957,
+      "learning_rate": 9.856505378637402e-06,
+      "loss": 0.6829,
+      "step": 803
+    },
+    {
+      "epoch": 0.5840237895262524,
+      "grad_norm": 0.6292491947390643,
+      "learning_rate": 9.85593701482066e-06,
+      "loss": 0.6877,
+      "step": 804
+    },
+    {
+      "epoch": 0.5847501872744194,
+      "grad_norm": 0.6444928744841502,
+      "learning_rate": 9.855367544071785e-06,
+      "loss": 0.664,
+      "step": 805
+    },
+    {
+      "epoch": 0.5854765850225865,
+      "grad_norm": 0.6334451683246128,
+      "learning_rate": 9.85479696652059e-06,
+      "loss": 0.6696,
+      "step": 806
+    },
+    {
+      "epoch": 0.5862029827707534,
+      "grad_norm": 0.6809663655418011,
+      "learning_rate": 9.854225282297141e-06,
+      "loss": 0.6951,
+      "step": 807
+    },
+    {
+      "epoch": 0.5869293805189204,
+      "grad_norm": 0.6633437669132748,
+      "learning_rate": 9.853652491531756e-06,
+      "loss": 0.6822,
+      "step": 808
+    },
+    {
+      "epoch": 0.5876557782670874,
+      "grad_norm": 0.6756561406729703,
+      "learning_rate": 9.853078594355003e-06,
+      "loss": 0.6729,
+      "step": 809
+    },
+    {
+      "epoch": 0.5883821760152543,
+      "grad_norm": 0.6175772675068528,
+      "learning_rate": 9.852503590897708e-06,
+      "loss": 0.6654,
+      "step": 810
+    },
+    {
+      "epoch": 0.5891085737634213,
+      "grad_norm": 0.6342072664742784,
+      "learning_rate": 9.851927481290943e-06,
+      "loss": 0.6813,
+      "step": 811
+    },
+    {
+      "epoch": 0.5898349715115884,
+      "grad_norm": 0.615481235538996,
+      "learning_rate": 9.851350265666034e-06,
+      "loss": 0.6569,
+      "step": 812
+    },
+    {
+      "epoch": 0.5905613692597553,
+      "grad_norm": 0.623506210218369,
+      "learning_rate": 9.850771944154563e-06,
+      "loss": 0.664,
+      "step": 813
+    },
+    {
+      "epoch": 0.5912877670079223,
+      "grad_norm": 0.6268532226267906,
+      "learning_rate": 9.850192516888357e-06,
+      "loss": 0.662,
+      "step": 814
+    },
+    {
+      "epoch": 0.5920141647560893,
+      "grad_norm": 0.6137453262606134,
+      "learning_rate": 9.849611983999503e-06,
+      "loss": 0.6481,
+      "step": 815
+    },
+    {
+      "epoch": 0.5927405625042562,
+      "grad_norm": 0.593844174928628,
+      "learning_rate": 9.849030345620333e-06,
+      "loss": 0.6664,
+      "step": 816
+    },
+    {
+      "epoch": 0.5934669602524232,
+      "grad_norm": 0.6066439569327111,
+      "learning_rate": 9.848447601883436e-06,
+      "loss": 0.6625,
+      "step": 817
+    },
+    {
+      "epoch": 0.5941933580005901,
+      "grad_norm": 0.6131220553241871,
+      "learning_rate": 9.847863752921649e-06,
+      "loss": 0.6769,
+      "step": 818
+    },
+    {
+      "epoch": 0.5949197557487572,
+      "grad_norm": 0.6127604757106416,
+      "learning_rate": 9.847278798868065e-06,
+      "loss": 0.6475,
+      "step": 819
+    },
+    {
+      "epoch": 0.5956461534969242,
+      "grad_norm": 0.6099319552963688,
+      "learning_rate": 9.846692739856023e-06,
+      "loss": 0.6714,
+      "step": 820
+    },
+    {
+      "epoch": 0.5963725512450911,
+      "grad_norm": 0.6400652371963013,
+      "learning_rate": 9.846105576019124e-06,
+      "loss": 0.6549,
+      "step": 821
+    },
+    {
+      "epoch": 0.5970989489932581,
+      "grad_norm": 0.6455148000021561,
+      "learning_rate": 9.84551730749121e-06,
+      "loss": 0.671,
+      "step": 822
+    },
+    {
+      "epoch": 0.5978253467414251,
+      "grad_norm": 0.6703898500769329,
+      "learning_rate": 9.844927934406381e-06,
+      "loss": 0.6672,
+      "step": 823
+    },
+    {
+      "epoch": 0.598551744489592,
+      "grad_norm": 0.6818434803159491,
+      "learning_rate": 9.844337456898985e-06,
+      "loss": 0.6559,
+      "step": 824
+    },
+    {
+      "epoch": 0.5992781422377591,
+      "grad_norm": 0.6307691400318461,
+      "learning_rate": 9.843745875103628e-06,
+      "loss": 0.6497,
+      "step": 825
+    },
+    {
+      "epoch": 0.6000045399859261,
+      "grad_norm": 0.605302882547688,
+      "learning_rate": 9.84315318915516e-06,
+      "loss": 0.6723,
+      "step": 826
+    },
+    {
+      "epoch": 0.600730937734093,
+      "grad_norm": 0.6250992899123806,
+      "learning_rate": 9.842559399188687e-06,
+      "loss": 0.6481,
+      "step": 827
+    },
+    {
+      "epoch": 0.60145733548226,
+      "grad_norm": 0.6090951991297758,
+      "learning_rate": 9.841964505339568e-06,
+      "loss": 0.6534,
+      "step": 828
+    },
+    {
+      "epoch": 0.602183733230427,
+      "grad_norm": 0.6731271428571752,
+      "learning_rate": 9.84136850774341e-06,
+      "loss": 0.6479,
+      "step": 829
+    },
+    {
+      "epoch": 0.6029101309785939,
+      "grad_norm": 0.6039650209903596,
+      "learning_rate": 9.840771406536073e-06,
+      "loss": 0.6731,
+      "step": 830
+    },
+    {
+      "epoch": 0.603636528726761,
+      "grad_norm": 0.6011413219903619,
+      "learning_rate": 9.84017320185367e-06,
+      "loss": 0.6838,
+      "step": 831
+    },
+    {
+      "epoch": 0.604362926474928,
+      "grad_norm": 0.6247705736995238,
+      "learning_rate": 9.839573893832564e-06,
+      "loss": 0.6723,
+      "step": 832
+    },
+    {
+      "epoch": 0.6050893242230949,
+      "grad_norm": 0.6220282560246873,
+      "learning_rate": 9.83897348260937e-06,
+      "loss": 0.6511,
+      "step": 833
+    },
+    {
+      "epoch": 0.6058157219712619,
+      "grad_norm": 0.6196889968424526,
+      "learning_rate": 9.838371968320951e-06,
+      "loss": 0.6669,
+      "step": 834
+    },
+    {
+      "epoch": 0.6065421197194288,
+      "grad_norm": 0.6183641181183734,
+      "learning_rate": 9.837769351104431e-06,
+      "loss": 0.6492,
+      "step": 835
+    },
+    {
+      "epoch": 0.6072685174675958,
+      "grad_norm": 0.7105892308887408,
+      "learning_rate": 9.837165631097177e-06,
+      "loss": 0.66,
+      "step": 836
+    },
+    {
+      "epoch": 0.6079949152157629,
+      "grad_norm": 0.5994749592259667,
+      "learning_rate": 9.836560808436806e-06,
+      "loss": 0.6766,
+      "step": 837
+    },
+    {
+      "epoch": 0.6087213129639298,
+      "grad_norm": 0.6002717353440816,
+      "learning_rate": 9.835954883261195e-06,
+      "loss": 0.6705,
+      "step": 838
+    },
+    {
+      "epoch": 0.6094477107120968,
+      "grad_norm": 0.6091493878958004,
+      "learning_rate": 9.835347855708464e-06,
+      "loss": 0.6582,
+      "step": 839
+    },
+    {
+      "epoch": 0.6101741084602638,
+      "grad_norm": 0.5968783813759632,
+      "learning_rate": 9.834739725916988e-06,
+      "loss": 0.6562,
+      "step": 840
+    },
+    {
+      "epoch": 0.6109005062084307,
+      "grad_norm": 0.7226406343809145,
+      "learning_rate": 9.834130494025395e-06,
+      "loss": 0.6682,
+      "step": 841
+    },
+    {
+      "epoch": 0.6116269039565977,
+      "grad_norm": 0.6236245151795735,
+      "learning_rate": 9.833520160172561e-06,
+      "loss": 0.6716,
+      "step": 842
+    },
+    {
+      "epoch": 0.6123533017047648,
+      "grad_norm": 0.649954848171266,
+      "learning_rate": 9.832908724497613e-06,
+      "loss": 0.6793,
+      "step": 843
+    },
+    {
+      "epoch": 0.6130796994529317,
+      "grad_norm": 0.8026775812212916,
+      "learning_rate": 9.83229618713993e-06,
+      "loss": 0.6482,
+      "step": 844
+    },
+    {
+      "epoch": 0.6138060972010987,
+      "grad_norm": 0.6549633112318536,
+      "learning_rate": 9.831682548239145e-06,
+      "loss": 0.6554,
+      "step": 845
+    },
+    {
+      "epoch": 0.6145324949492657,
+      "grad_norm": 0.6713749440634522,
+      "learning_rate": 9.83106780793514e-06,
+      "loss": 0.6501,
+      "step": 846
+    },
+    {
+      "epoch": 0.6152588926974326,
+      "grad_norm": 0.5976092035267176,
+      "learning_rate": 9.830451966368044e-06,
+      "loss": 0.6531,
+      "step": 847
+    },
+    {
+      "epoch": 0.6159852904455996,
+      "grad_norm": 0.6049813779727117,
+      "learning_rate": 9.829835023678243e-06,
+      "loss": 0.6511,
+      "step": 848
+    },
+    {
+      "epoch": 0.6167116881937666,
+      "grad_norm": 0.6075110205247508,
+      "learning_rate": 9.829216980006372e-06,
+      "loss": 0.6521,
+      "step": 849
+    },
+    {
+      "epoch": 0.6174380859419336,
+      "grad_norm": 0.6252951495072839,
+      "learning_rate": 9.828597835493319e-06,
+      "loss": 0.6458,
+      "step": 850
+    },
+    {
+      "epoch": 0.6181644836901006,
+      "grad_norm": 0.6655929334398094,
+      "learning_rate": 9.827977590280217e-06,
+      "loss": 0.6442,
+      "step": 851
+    },
+    {
+      "epoch": 0.6188908814382675,
+      "grad_norm": 0.5849309071329679,
+      "learning_rate": 9.827356244508455e-06,
+      "loss": 0.654,
+      "step": 852
+    },
+    {
+      "epoch": 0.6196172791864345,
+      "grad_norm": 0.6329417045286834,
+      "learning_rate": 9.82673379831967e-06,
+      "loss": 0.6541,
+      "step": 853
+    },
+    {
+      "epoch": 0.6203436769346015,
+      "grad_norm": 0.6746285553018466,
+      "learning_rate": 9.826110251855755e-06,
+      "loss": 0.6468,
+      "step": 854
+    },
+    {
+      "epoch": 0.6210700746827685,
+      "grad_norm": 0.6625458089858239,
+      "learning_rate": 9.825485605258846e-06,
+      "loss": 0.6475,
+      "step": 855
+    },
+    {
+      "epoch": 0.6217964724309355,
+      "grad_norm": 0.5924986665162189,
+      "learning_rate": 9.824859858671337e-06,
+      "loss": 0.6533,
+      "step": 856
+    },
+    {
+      "epoch": 0.6225228701791025,
+      "grad_norm": 0.5925928031908367,
+      "learning_rate": 9.824233012235869e-06,
+      "loss": 0.6523,
+      "step": 857
+    },
+    {
+      "epoch": 0.6232492679272694,
+      "grad_norm": 0.6605467068380393,
+      "learning_rate": 9.823605066095333e-06,
+      "loss": 0.6655,
+      "step": 858
+    },
+    {
+      "epoch": 0.6239756656754364,
+      "grad_norm": 0.5758378226313285,
+      "learning_rate": 9.822976020392874e-06,
+      "loss": 0.6527,
+      "step": 859
+    },
+    {
+      "epoch": 0.6247020634236033,
+      "grad_norm": 0.8187112632079937,
+      "learning_rate": 9.822345875271884e-06,
+      "loss": 0.6708,
+      "step": 860
+    },
+    {
+      "epoch": 0.6254284611717704,
+      "grad_norm": 0.6091453080384412,
+      "learning_rate": 9.821714630876009e-06,
+      "loss": 0.6741,
+      "step": 861
+    },
+    {
+      "epoch": 0.6261548589199374,
+      "grad_norm": 0.6029200767372679,
+      "learning_rate": 9.821082287349142e-06,
+      "loss": 0.652,
+      "step": 862
+    },
+    {
+      "epoch": 0.6268812566681043,
+      "grad_norm": 0.6106188486403386,
+      "learning_rate": 9.820448844835428e-06,
+      "loss": 0.6535,
+      "step": 863
+    },
+    {
+      "epoch": 0.6276076544162713,
+      "grad_norm": 0.5798197726465589,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.6448,
+      "step": 864
+    },
+    {
+      "epoch": 0.6283340521644383,
+      "grad_norm": 0.5943359460871429,
+      "learning_rate": 9.819178663425302e-06,
+      "loss": 0.6593,
+      "step": 865
+    },
+    {
+      "epoch": 0.6290604499126052,
+      "grad_norm": 0.5711044824370614,
+      "learning_rate": 9.818541924818432e-06,
+      "loss": 0.6672,
+      "step": 866
+    },
+    {
+      "epoch": 0.6297868476607722,
+      "grad_norm": 0.6952982586972026,
+      "learning_rate": 9.817904087803802e-06,
+      "loss": 0.6609,
+      "step": 867
+    },
+    {
+      "epoch": 0.6305132454089393,
+      "grad_norm": 0.5666054412389501,
+      "learning_rate": 9.817265152526811e-06,
+      "loss": 0.6608,
+      "step": 868
+    },
+    {
+      "epoch": 0.6312396431571062,
+      "grad_norm": 0.607448748434137,
+      "learning_rate": 9.816625119133109e-06,
+      "loss": 0.6532,
+      "step": 869
+    },
+    {
+      "epoch": 0.6319660409052732,
+      "grad_norm": 0.6135073373387316,
+      "learning_rate": 9.81598398776859e-06,
+      "loss": 0.657,
+      "step": 870
+    },
+    {
+      "epoch": 0.6326924386534402,
+      "grad_norm": 0.5926396668336961,
+      "learning_rate": 9.815341758579407e-06,
+      "loss": 0.6576,
+      "step": 871
+    },
+    {
+      "epoch": 0.6334188364016071,
+      "grad_norm": 0.5597381238227889,
+      "learning_rate": 9.814698431711957e-06,
+      "loss": 0.6547,
+      "step": 872
+    },
+    {
+      "epoch": 0.6341452341497741,
+      "grad_norm": 0.5823788201141964,
+      "learning_rate": 9.814054007312888e-06,
+      "loss": 0.6552,
+      "step": 873
+    },
+    {
+      "epoch": 0.6348716318979412,
+      "grad_norm": 0.5756772937219624,
+      "learning_rate": 9.813408485529103e-06,
+      "loss": 0.664,
+      "step": 874
+    },
+    {
+      "epoch": 0.6355980296461081,
+      "grad_norm": 0.5809430277549485,
+      "learning_rate": 9.812761866507748e-06,
+      "loss": 0.665,
+      "step": 875
+    },
+    {
+      "epoch": 0.6363244273942751,
+      "grad_norm": 0.5656279015120884,
+      "learning_rate": 9.812114150396223e-06,
+      "loss": 0.6685,
+      "step": 876
+    },
+    {
+      "epoch": 0.637050825142442,
+      "grad_norm": 0.5645400936452617,
+      "learning_rate": 9.81146533734218e-06,
+      "loss": 0.6661,
+      "step": 877
+    },
+    {
+      "epoch": 0.637777222890609,
+      "grad_norm": 0.6038920581138288,
+      "learning_rate": 9.810815427493516e-06,
+      "loss": 0.6513,
+      "step": 878
+    },
+    {
+      "epoch": 0.638503620638776,
+      "grad_norm": 0.6287288278527075,
+      "learning_rate": 9.810164420998385e-06,
+      "loss": 0.6378,
+      "step": 879
+    },
+    {
+      "epoch": 0.639230018386943,
+      "grad_norm": 0.5533432856351014,
+      "learning_rate": 9.80951231800518e-06,
+      "loss": 0.6503,
+      "step": 880
+    },
+    {
+      "epoch": 0.63995641613511,
+      "grad_norm": 0.5470769024145264,
+      "learning_rate": 9.808859118662558e-06,
+      "loss": 0.6487,
+      "step": 881
+    },
+    {
+      "epoch": 0.640682813883277,
+      "grad_norm": 0.5735833277625101,
+      "learning_rate": 9.808204823119414e-06,
+      "loss": 0.6383,
+      "step": 882
+    },
+    {
+      "epoch": 0.6414092116314439,
+      "grad_norm": 0.556322686472196,
+      "learning_rate": 9.8075494315249e-06,
+      "loss": 0.6465,
+      "step": 883
+    },
+    {
+      "epoch": 0.6421356093796109,
+      "grad_norm": 0.5843117762599995,
+      "learning_rate": 9.806892944028414e-06,
+      "loss": 0.662,
+      "step": 884
+    },
+    {
+      "epoch": 0.6428620071277779,
+      "grad_norm": 0.5499746337779424,
+      "learning_rate": 9.806235360779605e-06,
+      "loss": 0.6469,
+      "step": 885
+    },
+    {
+      "epoch": 0.6435884048759449,
+      "grad_norm": 0.5773912530605604,
+      "learning_rate": 9.805576681928373e-06,
+      "loss": 0.6573,
+      "step": 886
+    },
+    {
+      "epoch": 0.6443148026241119,
+      "grad_norm": 0.5699879773421873,
+      "learning_rate": 9.804916907624865e-06,
+      "loss": 0.6674,
+      "step": 887
+    },
+    {
+      "epoch": 0.6450412003722789,
+      "grad_norm": 0.5789871539035835,
+      "learning_rate": 9.804256038019482e-06,
+      "loss": 0.6588,
+      "step": 888
+    },
+    {
+      "epoch": 0.6457675981204458,
+      "grad_norm": 0.5261604408640989,
+      "learning_rate": 9.80359407326287e-06,
+      "loss": 0.6562,
+      "step": 889
+    },
+    {
+      "epoch": 0.6464939958686128,
+      "grad_norm": 0.5557539019590454,
+      "learning_rate": 9.802931013505929e-06,
+      "loss": 0.6409,
+      "step": 890
+    },
+    {
+      "epoch": 0.6472203936167797,
+      "grad_norm": 0.5481978932889137,
+      "learning_rate": 9.802266858899802e-06,
+      "loss": 0.645,
+      "step": 891
+    },
+    {
+      "epoch": 0.6479467913649468,
+      "grad_norm": 0.5563785530002918,
+      "learning_rate": 9.801601609595892e-06,
+      "loss": 0.6453,
+      "step": 892
+    },
+    {
+      "epoch": 0.6486731891131138,
+      "grad_norm": 0.5694891433544842,
+      "learning_rate": 9.80093526574584e-06,
+      "loss": 0.6409,
+      "step": 893
+    },
+    {
+      "epoch": 0.6493995868612807,
+      "grad_norm": 0.6636903196693987,
+      "learning_rate": 9.800267827501548e-06,
+      "loss": 0.6245,
+      "step": 894
+    },
+    {
+      "epoch": 0.6501259846094477,
+      "grad_norm": 0.6284598297309115,
+      "learning_rate": 9.799599295015154e-06,
+      "loss": 0.6366,
+      "step": 895
+    },
+    {
+      "epoch": 0.6508523823576147,
+      "grad_norm": 0.5437492801963629,
+      "learning_rate": 9.798929668439059e-06,
+      "loss": 0.6618,
+      "step": 896
+    },
+    {
+      "epoch": 0.6515787801057816,
+      "grad_norm": 0.5833253680539345,
+      "learning_rate": 9.798258947925905e-06,
+      "loss": 0.6482,
+      "step": 897
+    },
+    {
+      "epoch": 0.6523051778539487,
+      "grad_norm": 0.639449411834674,
+      "learning_rate": 9.797587133628586e-06,
+      "loss": 0.6472,
+      "step": 898
+    },
+    {
+      "epoch": 0.6530315756021157,
+      "grad_norm": 0.535584341412608,
+      "learning_rate": 9.796914225700243e-06,
+      "loss": 0.6593,
+      "step": 899
+    },
+    {
+      "epoch": 0.6537579733502826,
+      "grad_norm": 0.547567869110903,
+      "learning_rate": 9.79624022429427e-06,
+      "loss": 0.6393,
+      "step": 900
+    },
+    {
+      "epoch": 0.6544843710984496,
+      "grad_norm": 0.5522922741935506,
+      "learning_rate": 9.79556512956431e-06,
+      "loss": 0.6572,
+      "step": 901
+    },
+    {
+      "epoch": 0.6552107688466166,
+      "grad_norm": 0.5893600739438992,
+      "learning_rate": 9.794888941664253e-06,
+      "loss": 0.6416,
+      "step": 902
+    },
+    {
+      "epoch": 0.6559371665947835,
+      "grad_norm": 0.5351461232055588,
+      "learning_rate": 9.794211660748239e-06,
+      "loss": 0.6421,
+      "step": 903
+    },
+    {
+      "epoch": 0.6566635643429506,
+      "grad_norm": 0.5945928584609042,
+      "learning_rate": 9.793533286970655e-06,
+      "loss": 0.6706,
+      "step": 904
+    },
+    {
+      "epoch": 0.6573899620911176,
+      "grad_norm": 0.5570387654513104,
+      "learning_rate": 9.79285382048614e-06,
+      "loss": 0.6413,
+      "step": 905
+    },
+    {
+      "epoch": 0.6581163598392845,
+      "grad_norm": 0.5846520230724995,
+      "learning_rate": 9.792173261449586e-06,
+      "loss": 0.653,
+      "step": 906
+    },
+    {
+      "epoch": 0.6588427575874515,
+      "grad_norm": 0.5526390975042158,
+      "learning_rate": 9.791491610016123e-06,
+      "loss": 0.65,
+      "step": 907
+    },
+    {
+      "epoch": 0.6595691553356184,
+      "grad_norm": 0.5574061987078701,
+      "learning_rate": 9.79080886634114e-06,
+      "loss": 0.6425,
+      "step": 908
+    },
+    {
+      "epoch": 0.6602955530837854,
+      "grad_norm": 0.5341499920110605,
+      "learning_rate": 9.79012503058027e-06,
+      "loss": 0.6521,
+      "step": 909
+    },
+    {
+      "epoch": 0.6610219508319525,
+      "grad_norm": 0.5791055056403942,
+      "learning_rate": 9.789440102889396e-06,
+      "loss": 0.658,
+      "step": 910
+    },
+    {
+      "epoch": 0.6617483485801194,
+      "grad_norm": 0.5260761897292887,
+      "learning_rate": 9.788754083424654e-06,
+      "loss": 0.6401,
+      "step": 911
+    },
+    {
+      "epoch": 0.6624747463282864,
+      "grad_norm": 0.5109150407815444,
+      "learning_rate": 9.788066972342421e-06,
+      "loss": 0.6506,
+      "step": 912
+    },
+    {
+      "epoch": 0.6632011440764534,
+      "grad_norm": 0.5319111825792613,
+      "learning_rate": 9.787378769799326e-06,
+      "loss": 0.652,
+      "step": 913
+    },
+    {
+      "epoch": 0.6639275418246203,
+      "grad_norm": 0.5240092882015561,
+      "learning_rate": 9.786689475952254e-06,
+      "loss": 0.6445,
+      "step": 914
+    },
+    {
+      "epoch": 0.6646539395727873,
+      "grad_norm": 0.5440903216518904,
+      "learning_rate": 9.785999090958326e-06,
+      "loss": 0.6308,
+      "step": 915
+    },
+    {
+      "epoch": 0.6653803373209543,
+      "grad_norm": 0.5182863646250728,
+      "learning_rate": 9.785307614974922e-06,
+      "loss": 0.6353,
+      "step": 916
+    },
+    {
+      "epoch": 0.6661067350691213,
+      "grad_norm": 0.5407517444632997,
+      "learning_rate": 9.784615048159665e-06,
+      "loss": 0.6567,
+      "step": 917
+    },
+    {
+      "epoch": 0.6668331328172883,
+      "grad_norm": 0.5595401196963885,
+      "learning_rate": 9.78392139067043e-06,
+      "loss": 0.6618,
+      "step": 918
+    },
+    {
+      "epoch": 0.6675595305654553,
+      "grad_norm": 0.5281983812603945,
+      "learning_rate": 9.783226642665338e-06,
+      "loss": 0.637,
+      "step": 919
+    },
+    {
+      "epoch": 0.6682859283136222,
+      "grad_norm": 0.588468629477807,
+      "learning_rate": 9.782530804302763e-06,
+      "loss": 0.6406,
+      "step": 920
+    },
+    {
+      "epoch": 0.6690123260617892,
+      "grad_norm": 0.5387289554693768,
+      "learning_rate": 9.78183387574132e-06,
+      "loss": 0.6327,
+      "step": 921
+    },
+    {
+      "epoch": 0.6697387238099561,
+      "grad_norm": 0.5521325969515342,
+      "learning_rate": 9.781135857139881e-06,
+      "loss": 0.6374,
+      "step": 922
+    },
+    {
+      "epoch": 0.6704651215581232,
+      "grad_norm": 0.5453121610390395,
+      "learning_rate": 9.780436748657559e-06,
+      "loss": 0.6742,
+      "step": 923
+    },
+    {
+      "epoch": 0.6711915193062902,
+      "grad_norm": 0.527286669061359,
+      "learning_rate": 9.77973655045372e-06,
+      "loss": 0.648,
+      "step": 924
+    },
+    {
+      "epoch": 0.6719179170544571,
+      "grad_norm": 0.5625960128189396,
+      "learning_rate": 9.779035262687976e-06,
+      "loss": 0.6288,
+      "step": 925
+    },
+    {
+      "epoch": 0.6726443148026241,
+      "grad_norm": 0.5362827537791287,
+      "learning_rate": 9.778332885520195e-06,
+      "loss": 0.6277,
+      "step": 926
+    },
+    {
+      "epoch": 0.6733707125507911,
+      "grad_norm": 0.745633309649563,
+      "learning_rate": 9.777629419110478e-06,
+      "loss": 0.636,
+      "step": 927
+    },
+    {
+      "epoch": 0.674097110298958,
+      "grad_norm": 0.5743650133346059,
+      "learning_rate": 9.776924863619187e-06,
+      "loss": 0.6313,
+      "step": 928
+    },
+    {
+      "epoch": 0.6748235080471251,
+      "grad_norm": 0.5339811952680448,
+      "learning_rate": 9.776219219206933e-06,
+      "loss": 0.6534,
+      "step": 929
+    },
+    {
+      "epoch": 0.6755499057952921,
+      "grad_norm": 0.5412542702478056,
+      "learning_rate": 9.775512486034564e-06,
+      "loss": 0.6439,
+      "step": 930
+    },
+    {
+      "epoch": 0.676276303543459,
+      "grad_norm": 0.5390435892309067,
+      "learning_rate": 9.774804664263184e-06,
+      "loss": 0.658,
+      "step": 931
+    },
+    {
+      "epoch": 0.677002701291626,
+      "grad_norm": 0.6095626038690631,
+      "learning_rate": 9.774095754054147e-06,
+      "loss": 0.6488,
+      "step": 932
+    },
+    {
+      "epoch": 0.677729099039793,
+      "grad_norm": 0.5497445658882337,
+      "learning_rate": 9.773385755569052e-06,
+      "loss": 0.6394,
+      "step": 933
+    },
+    {
+      "epoch": 0.6784554967879599,
+      "grad_norm": 0.5966473206171481,
+      "learning_rate": 9.772674668969743e-06,
+      "loss": 0.6429,
+      "step": 934
+    },
+    {
+      "epoch": 0.679181894536127,
+      "grad_norm": 0.5590698378286874,
+      "learning_rate": 9.771962494418319e-06,
+      "loss": 0.6359,
+      "step": 935
+    },
+    {
+      "epoch": 0.679908292284294,
+      "grad_norm": 0.5339631288047351,
+      "learning_rate": 9.77124923207712e-06,
+      "loss": 0.6255,
+      "step": 936
+    },
+    {
+      "epoch": 0.6806346900324609,
+      "grad_norm": 0.5693835323242853,
+      "learning_rate": 9.77053488210874e-06,
+      "loss": 0.6454,
+      "step": 937
+    },
+    {
+      "epoch": 0.6813610877806279,
+      "grad_norm": 0.5136483877400199,
+      "learning_rate": 9.769819444676017e-06,
+      "loss": 0.6423,
+      "step": 938
+    },
+    {
+      "epoch": 0.6820874855287948,
+      "grad_norm": 0.5644953061878822,
+      "learning_rate": 9.769102919942038e-06,
+      "loss": 0.6327,
+      "step": 939
+    },
+    {
+      "epoch": 0.6828138832769618,
+      "grad_norm": 0.5243587189424204,
+      "learning_rate": 9.768385308070139e-06,
+      "loss": 0.6523,
+      "step": 940
+    },
+    {
+      "epoch": 0.6835402810251289,
+      "grad_norm": 0.53271625284346,
+      "learning_rate": 9.767666609223902e-06,
+      "loss": 0.6441,
+      "step": 941
+    },
+    {
+      "epoch": 0.6842666787732958,
+      "grad_norm": 0.5814209737223036,
+      "learning_rate": 9.766946823567157e-06,
+      "loss": 0.619,
+      "step": 942
+    },
+    {
+      "epoch": 0.6849930765214628,
+      "grad_norm": 0.5309212150196624,
+      "learning_rate": 9.766225951263982e-06,
+      "loss": 0.6108,
+      "step": 943
+    },
+    {
+      "epoch": 0.6857194742696298,
+      "grad_norm": 0.5374951063410888,
+      "learning_rate": 9.765503992478704e-06,
+      "loss": 0.627,
+      "step": 944
+    },
+    {
+      "epoch": 0.6864458720177967,
+      "grad_norm": 0.5142193404257767,
+      "learning_rate": 9.764780947375898e-06,
+      "loss": 0.65,
+      "step": 945
+    },
+    {
+      "epoch": 0.6871722697659637,
+      "grad_norm": 0.5385889304668647,
+      "learning_rate": 9.764056816120383e-06,
+      "loss": 0.6447,
+      "step": 946
+    },
+    {
+      "epoch": 0.6878986675141308,
+      "grad_norm": 0.5079349675047481,
+      "learning_rate": 9.763331598877229e-06,
+      "loss": 0.6252,
+      "step": 947
+    },
+    {
+      "epoch": 0.6886250652622977,
+      "grad_norm": 0.5252424116585818,
+      "learning_rate": 9.762605295811751e-06,
+      "loss": 0.6378,
+      "step": 948
+    },
+    {
+      "epoch": 0.6893514630104647,
+      "grad_norm": 0.5351541129299291,
+      "learning_rate": 9.761877907089515e-06,
+      "loss": 0.6316,
+      "step": 949
+    },
+    {
+      "epoch": 0.6900778607586316,
+      "grad_norm": 0.5213029147033851,
+      "learning_rate": 9.761149432876331e-06,
+      "loss": 0.6434,
+      "step": 950
+    },
+    {
+      "epoch": 0.6908042585067986,
+      "grad_norm": 0.5287068047571596,
+      "learning_rate": 9.760419873338261e-06,
+      "loss": 0.6434,
+      "step": 951
+    },
+    {
+      "epoch": 0.6915306562549656,
+      "grad_norm": 0.5534499293711467,
+      "learning_rate": 9.759689228641605e-06,
+      "loss": 0.6395,
+      "step": 952
+    },
+    {
+      "epoch": 0.6922570540031326,
+      "grad_norm": 0.586784866270694,
+      "learning_rate": 9.758957498952923e-06,
+      "loss": 0.653,
+      "step": 953
+    },
+    {
+      "epoch": 0.6929834517512996,
+      "grad_norm": 0.5266464417997495,
+      "learning_rate": 9.758224684439013e-06,
+      "loss": 0.6266,
+      "step": 954
+    },
+    {
+      "epoch": 0.6937098494994666,
+      "grad_norm": 0.5122612880356957,
+      "learning_rate": 9.757490785266924e-06,
+      "loss": 0.6347,
+      "step": 955
+    },
+    {
+      "epoch": 0.6944362472476335,
+      "grad_norm": 0.6375844075472203,
+      "learning_rate": 9.75675580160395e-06,
+      "loss": 0.625,
+      "step": 956
+    },
+    {
+      "epoch": 0.6951626449958005,
+      "grad_norm": 0.5212655025584345,
+      "learning_rate": 9.756019733617634e-06,
+      "loss": 0.6396,
+      "step": 957
+    },
+    {
+      "epoch": 0.6958890427439675,
+      "grad_norm": 0.5086286936385832,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.6342,
+      "step": 958
+    },
+    {
+      "epoch": 0.6966154404921344,
+      "grad_norm": 0.5128731180861688,
+      "learning_rate": 9.754544345346388e-06,
+      "loss": 0.6292,
+      "step": 959
+    },
+    {
+      "epoch": 0.6973418382403015,
+      "grad_norm": 0.4999617398047923,
+      "learning_rate": 9.75380502539778e-06,
+      "loss": 0.6339,
+      "step": 960
+    },
+    {
+      "epoch": 0.6980682359884685,
+      "grad_norm": 0.5282274063606959,
+      "learning_rate": 9.753064621798472e-06,
+      "loss": 0.6432,
+      "step": 961
+    },
+    {
+      "epoch": 0.6987946337366354,
+      "grad_norm": 0.5252127904810339,
+      "learning_rate": 9.752323134717244e-06,
+      "loss": 0.6281,
+      "step": 962
+    },
+    {
+      "epoch": 0.6995210314848024,
+      "grad_norm": 0.5082613730934245,
+      "learning_rate": 9.75158056432312e-06,
+      "loss": 0.6338,
+      "step": 963
+    },
+    {
+      "epoch": 0.7002474292329693,
+      "grad_norm": 0.4882829401463821,
+      "learning_rate": 9.750836910785372e-06,
+      "loss": 0.6304,
+      "step": 964
+    },
+    {
+      "epoch": 0.7009738269811363,
+      "grad_norm": 0.5538873366273053,
+      "learning_rate": 9.75009217427352e-06,
+      "loss": 0.6355,
+      "step": 965
+    },
+    {
+      "epoch": 0.7017002247293034,
+      "grad_norm": 0.7956754652954774,
+      "learning_rate": 9.749346354957331e-06,
+      "loss": 0.6462,
+      "step": 966
+    },
+    {
+      "epoch": 0.7024266224774703,
+      "grad_norm": 0.5219908542511834,
+      "learning_rate": 9.748599453006818e-06,
+      "loss": 0.6265,
+      "step": 967
+    },
+    {
+      "epoch": 0.7031530202256373,
+      "grad_norm": 0.5210128918221577,
+      "learning_rate": 9.74785146859224e-06,
+      "loss": 0.6537,
+      "step": 968
+    },
+    {
+      "epoch": 0.7038794179738043,
+      "grad_norm": 0.5040833546403364,
+      "learning_rate": 9.7471024018841e-06,
+      "loss": 0.6385,
+      "step": 969
+    },
+    {
+      "epoch": 0.7046058157219712,
+      "grad_norm": 0.5001664206030246,
+      "learning_rate": 9.746352253053155e-06,
+      "loss": 0.6313,
+      "step": 970
+    },
+    {
+      "epoch": 0.7053322134701382,
+      "grad_norm": 0.5098747327571058,
+      "learning_rate": 9.745601022270403e-06,
+      "loss": 0.633,
+      "step": 971
+    },
+    {
+      "epoch": 0.7060586112183053,
+      "grad_norm": 0.8219230135157451,
+      "learning_rate": 9.74484870970709e-06,
+      "loss": 0.6325,
+      "step": 972
+    },
+    {
+      "epoch": 0.7067850089664722,
+      "grad_norm": 0.5041778633169351,
+      "learning_rate": 9.744095315534711e-06,
+      "loss": 0.6515,
+      "step": 973
+    },
+    {
+      "epoch": 0.7075114067146392,
+      "grad_norm": 0.48370176730119585,
+      "learning_rate": 9.743340839925006e-06,
+      "loss": 0.6188,
+      "step": 974
+    },
+    {
+      "epoch": 0.7082378044628062,
+      "grad_norm": 0.5875403728810346,
+      "learning_rate": 9.742585283049957e-06,
+      "loss": 0.6616,
+      "step": 975
+    },
+    {
+      "epoch": 0.7089642022109731,
+      "grad_norm": 0.6310835977211383,
+      "learning_rate": 9.741828645081799e-06,
+      "loss": 0.6453,
+      "step": 976
+    },
+    {
+      "epoch": 0.7096905999591401,
+      "grad_norm": 0.545599500733757,
+      "learning_rate": 9.741070926193013e-06,
+      "loss": 0.6322,
+      "step": 977
+    },
+    {
+      "epoch": 0.7104169977073072,
+      "grad_norm": 0.5017837062704927,
+      "learning_rate": 9.74031212655632e-06,
+      "loss": 0.6512,
+      "step": 978
+    },
+    {
+      "epoch": 0.7111433954554741,
+      "grad_norm": 0.503250513244053,
+      "learning_rate": 9.739552246344692e-06,
+      "loss": 0.6547,
+      "step": 979
+    },
+    {
+      "epoch": 0.7118697932036411,
+      "grad_norm": 0.5163358870991519,
+      "learning_rate": 9.738791285731353e-06,
+      "loss": 0.6483,
+      "step": 980
+    },
+    {
+      "epoch": 0.712596190951808,
+      "grad_norm": 0.49564125514372964,
+      "learning_rate": 9.73802924488976e-06,
+      "loss": 0.6462,
+      "step": 981
+    },
+    {
+      "epoch": 0.713322588699975,
+      "grad_norm": 0.4913113083401416,
+      "learning_rate": 9.737266123993627e-06,
+      "loss": 0.6428,
+      "step": 982
+    },
+    {
+      "epoch": 0.714048986448142,
+      "grad_norm": 0.5095254384295473,
+      "learning_rate": 9.736501923216912e-06,
+      "loss": 0.6188,
+      "step": 983
+    },
+    {
+      "epoch": 0.714775384196309,
+      "grad_norm": 0.5189549653207551,
+      "learning_rate": 9.735736642733816e-06,
+      "loss": 0.6122,
+      "step": 984
+    },
+    {
+      "epoch": 0.715501781944476,
+      "grad_norm": 0.5112989759132592,
+      "learning_rate": 9.734970282718788e-06,
+      "loss": 0.6334,
+      "step": 985
+    },
+    {
+      "epoch": 0.716228179692643,
+      "grad_norm": 0.5107580279579825,
+      "learning_rate": 9.734202843346522e-06,
+      "loss": 0.6252,
+      "step": 986
+    },
+    {
+      "epoch": 0.7169545774408099,
+      "grad_norm": 0.5167080885659902,
+      "learning_rate": 9.733434324791962e-06,
+      "loss": 0.6401,
+      "step": 987
+    },
+    {
+      "epoch": 0.7176809751889769,
+      "grad_norm": 0.5023084975379272,
+      "learning_rate": 9.732664727230293e-06,
+      "loss": 0.6281,
+      "step": 988
+    },
+    {
+      "epoch": 0.7184073729371439,
+      "grad_norm": 0.537562102338342,
+      "learning_rate": 9.731894050836952e-06,
+      "loss": 0.6246,
+      "step": 989
+    },
+    {
+      "epoch": 0.7191337706853109,
+      "grad_norm": 0.48088299234652243,
+      "learning_rate": 9.731122295787611e-06,
+      "loss": 0.6354,
+      "step": 990
+    },
+    {
+      "epoch": 0.7198601684334779,
+      "grad_norm": 0.563544077811699,
+      "learning_rate": 9.730349462258202e-06,
+      "loss": 0.6261,
+      "step": 991
+    },
+    {
+      "epoch": 0.7205865661816448,
+      "grad_norm": 0.49876787412524776,
+      "learning_rate": 9.729575550424892e-06,
+      "loss": 0.6181,
+      "step": 992
+    },
+    {
+      "epoch": 0.7213129639298118,
+      "grad_norm": 0.527873118148596,
+      "learning_rate": 9.728800560464097e-06,
+      "loss": 0.6362,
+      "step": 993
+    },
+    {
+      "epoch": 0.7220393616779788,
+      "grad_norm": 0.6547177893128939,
+      "learning_rate": 9.728024492552484e-06,
+      "loss": 0.6312,
+      "step": 994
+    },
+    {
+      "epoch": 0.7227657594261457,
+      "grad_norm": 0.4938014707944413,
+      "learning_rate": 9.727247346866955e-06,
+      "loss": 0.6386,
+      "step": 995
+    },
+    {
+      "epoch": 0.7234921571743128,
+      "grad_norm": 0.4969962276566467,
+      "learning_rate": 9.726469123584668e-06,
+      "loss": 0.6301,
+      "step": 996
+    },
+    {
+      "epoch": 0.7242185549224798,
+      "grad_norm": 0.5610061497738755,
+      "learning_rate": 9.72568982288302e-06,
+      "loss": 0.6261,
+      "step": 997
+    },
+    {
+      "epoch": 0.7249449526706467,
+      "grad_norm": 0.48231769992974666,
+      "learning_rate": 9.724909444939657e-06,
+      "loss": 0.6321,
+      "step": 998
+    },
+    {
+      "epoch": 0.7256713504188137,
+      "grad_norm": 0.5566561152996569,
+      "learning_rate": 9.72412798993247e-06,
+      "loss": 0.6301,
+      "step": 999
+    },
+    {
+      "epoch": 0.7263977481669807,
+      "grad_norm": 0.4740045617404099,
+      "learning_rate": 9.723345458039595e-06,
+      "loss": 0.6333,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7271241459151476,
+      "grad_norm": 0.4673201163985113,
+      "learning_rate": 9.722561849439414e-06,
+      "loss": 0.6273,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7278505436633147,
+      "grad_norm": 0.5108892842569576,
+      "learning_rate": 9.72177716431055e-06,
+      "loss": 0.6239,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7285769414114817,
+      "grad_norm": 0.5098263048545012,
+      "learning_rate": 9.720991402831883e-06,
+      "loss": 0.6012,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7293033391596486,
+      "grad_norm": 0.5815800988061166,
+      "learning_rate": 9.720204565182522e-06,
+      "loss": 0.6204,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7300297369078156,
+      "grad_norm": 0.47676217623350187,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.6105,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7307561346559825,
+      "grad_norm": 0.47753745865698777,
+      "learning_rate": 9.718627662089435e-06,
+      "loss": 0.6395,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7314825324041495,
+      "grad_norm": 0.4879838290382639,
+      "learning_rate": 9.717837597005169e-06,
+      "loss": 0.639,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7322089301523165,
+      "grad_norm": 0.5013716494090865,
+      "learning_rate": 9.717046456469135e-06,
+      "loss": 0.6325,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7329353279004835,
+      "grad_norm": 0.4751272371204352,
+      "learning_rate": 9.71625424066168e-06,
+      "loss": 0.6241,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7336617256486505,
+      "grad_norm": 0.4699052765127366,
+      "learning_rate": 9.715460949763393e-06,
+      "loss": 0.6285,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7343881233968175,
+      "grad_norm": 0.5191334092088646,
+      "learning_rate": 9.714666583955108e-06,
+      "loss": 0.6228,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7351145211449844,
+      "grad_norm": 0.49374498413992257,
+      "learning_rate": 9.713871143417902e-06,
+      "loss": 0.6441,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7358409188931514,
+      "grad_norm": 0.4863582873661774,
+      "learning_rate": 9.713074628333102e-06,
+      "loss": 0.6254,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7365673166413184,
+      "grad_norm": 0.4498108134181778,
+      "learning_rate": 9.712277038882274e-06,
+      "loss": 0.6301,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7372937143894854,
+      "grad_norm": 0.5126199564731452,
+      "learning_rate": 9.711478375247237e-06,
+      "loss": 0.6335,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7380201121376524,
+      "grad_norm": 0.49607691586084646,
+      "learning_rate": 9.710678637610045e-06,
+      "loss": 0.6268,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7387465098858194,
+      "grad_norm": 0.458163972923393,
+      "learning_rate": 9.709877826153003e-06,
+      "loss": 0.6478,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7394729076339863,
+      "grad_norm": 0.4874017595835269,
+      "learning_rate": 9.709075941058661e-06,
+      "loss": 0.6251,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7401993053821533,
+      "grad_norm": 0.47017693490570733,
+      "learning_rate": 9.70827298250981e-06,
+      "loss": 0.6363,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7409257031303202,
+      "grad_norm": 0.6968313114729715,
+      "learning_rate": 9.70746895068949e-06,
+      "loss": 0.6149,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7416521008784873,
+      "grad_norm": 0.4965074758910424,
+      "learning_rate": 9.706663845780984e-06,
+      "loss": 0.6154,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7423784986266543,
+      "grad_norm": 0.48551084123096894,
+      "learning_rate": 9.705857667967818e-06,
+      "loss": 0.6312,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7431048963748212,
+      "grad_norm": 0.4866557666908839,
+      "learning_rate": 9.705050417433762e-06,
+      "loss": 0.6192,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7438312941229882,
+      "grad_norm": 0.5025491441411499,
+      "learning_rate": 9.704242094362834e-06,
+      "loss": 0.6086,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7445576918711552,
+      "grad_norm": 0.4786396836500014,
+      "learning_rate": 9.703432698939294e-06,
+      "loss": 0.6355,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7452840896193221,
+      "grad_norm": 0.5195325928953927,
+      "learning_rate": 9.70262223134765e-06,
+      "loss": 0.6176,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7460104873674892,
+      "grad_norm": 0.47770348548058295,
+      "learning_rate": 9.70181069177265e-06,
+      "loss": 0.6208,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7467368851156562,
+      "grad_norm": 0.5669878887712634,
+      "learning_rate": 9.700998080399287e-06,
+      "loss": 0.6118,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7474632828638231,
+      "grad_norm": 0.4697665700113132,
+      "learning_rate": 9.700184397412799e-06,
+      "loss": 0.6287,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7481896806119901,
+      "grad_norm": 0.4642121247185043,
+      "learning_rate": 9.699369642998671e-06,
+      "loss": 0.627,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7489160783601571,
+      "grad_norm": 0.48927006449177757,
+      "learning_rate": 9.698553817342627e-06,
+      "loss": 0.6185,
+      "step": 1031
+    },
+    {
+      "epoch": 0.749642476108324,
+      "grad_norm": 0.5077132474206864,
+      "learning_rate": 9.697736920630641e-06,
+      "loss": 0.6382,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7503688738564911,
+      "grad_norm": 0.4849060148767091,
+      "learning_rate": 9.696918953048925e-06,
+      "loss": 0.6371,
+      "step": 1033
+    },
+    {
+      "epoch": 0.751095271604658,
+      "grad_norm": 0.4952415892198469,
+      "learning_rate": 9.69609991478394e-06,
+      "loss": 0.6219,
+      "step": 1034
+    },
+    {
+      "epoch": 0.751821669352825,
+      "grad_norm": 0.46870581026203334,
+      "learning_rate": 9.695279806022391e-06,
+      "loss": 0.6062,
+      "step": 1035
+    },
+    {
+      "epoch": 0.752548067100992,
+      "grad_norm": 0.44552062529184516,
+      "learning_rate": 9.694458626951223e-06,
+      "loss": 0.6336,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7532744648491589,
+      "grad_norm": 0.4701788285387683,
+      "learning_rate": 9.693636377757628e-06,
+      "loss": 0.6348,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7540008625973259,
+      "grad_norm": 0.4930902095299255,
+      "learning_rate": 9.692813058629041e-06,
+      "loss": 0.6309,
+      "step": 1038
+    },
+    {
+      "epoch": 0.754727260345493,
+      "grad_norm": 0.513865274475057,
+      "learning_rate": 9.691988669753141e-06,
+      "loss": 0.6313,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7554536580936599,
+      "grad_norm": 0.5004488299102964,
+      "learning_rate": 9.691163211317853e-06,
+      "loss": 0.6186,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7561800558418269,
+      "grad_norm": 0.4632958610592306,
+      "learning_rate": 9.69033668351134e-06,
+      "loss": 0.6335,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7569064535899939,
+      "grad_norm": 0.5029063747150795,
+      "learning_rate": 9.689509086522019e-06,
+      "loss": 0.6247,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7576328513381608,
+      "grad_norm": 0.4634748963399614,
+      "learning_rate": 9.688680420538539e-06,
+      "loss": 0.6309,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7583592490863278,
+      "grad_norm": 0.4718510576080903,
+      "learning_rate": 9.6878506857498e-06,
+      "loss": 0.6371,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7590856468344949,
+      "grad_norm": 0.5131883999829031,
+      "learning_rate": 9.687019882344942e-06,
+      "loss": 0.6227,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7598120445826618,
+      "grad_norm": 0.5286803442259869,
+      "learning_rate": 9.68618801051335e-06,
+      "loss": 0.6347,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7605384423308288,
+      "grad_norm": 0.4592287328761153,
+      "learning_rate": 9.685355070444658e-06,
+      "loss": 0.6289,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7612648400789958,
+      "grad_norm": 0.46453734575529443,
+      "learning_rate": 9.684521062328734e-06,
+      "loss": 0.6167,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7619912378271627,
+      "grad_norm": 0.554800979203622,
+      "learning_rate": 9.683685986355692e-06,
+      "loss": 0.6174,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7627176355753297,
+      "grad_norm": 0.47868566678845503,
+      "learning_rate": 9.682849842715895e-06,
+      "loss": 0.6169,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7634440333234968,
+      "grad_norm": 0.4752654997453638,
+      "learning_rate": 9.682012631599945e-06,
+      "loss": 0.6115,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7641704310716637,
+      "grad_norm": 0.4779615895205545,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.6145,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7648968288198307,
+      "grad_norm": 0.48803918870499163,
+      "learning_rate": 9.68033500770321e-06,
+      "loss": 0.6149,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7656232265679976,
+      "grad_norm": 0.4696941476074349,
+      "learning_rate": 9.679494595304848e-06,
+      "loss": 0.6327,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7663496243161646,
+      "grad_norm": 0.4844539384776647,
+      "learning_rate": 9.678653116195174e-06,
+      "loss": 0.6239,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7670760220643316,
+      "grad_norm": 0.6568435300988817,
+      "learning_rate": 9.677810570566011e-06,
+      "loss": 0.6181,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7678024198124985,
+      "grad_norm": 0.5597089201336259,
+      "learning_rate": 9.676966958609417e-06,
+      "loss": 0.6265,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7685288175606656,
+      "grad_norm": 0.5421661140844151,
+      "learning_rate": 9.676122280517699e-06,
+      "loss": 0.6243,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7692552153088326,
+      "grad_norm": 0.45296333286927,
+      "learning_rate": 9.675276536483404e-06,
+      "loss": 0.6265,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7699816130569995,
+      "grad_norm": 0.4950287756900467,
+      "learning_rate": 9.674429726699324e-06,
+      "loss": 0.6139,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7707080108051665,
+      "grad_norm": 0.4530383471143353,
+      "learning_rate": 9.673581851358493e-06,
+      "loss": 0.613,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7714344085533335,
+      "grad_norm": 0.46904027613626054,
+      "learning_rate": 9.672732910654187e-06,
+      "loss": 0.61,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7721608063015004,
+      "grad_norm": 0.4562341083900913,
+      "learning_rate": 9.671882904779927e-06,
+      "loss": 0.6019,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7728872040496675,
+      "grad_norm": 0.47972675828391576,
+      "learning_rate": 9.671031833929474e-06,
+      "loss": 0.6012,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7736136017978344,
+      "grad_norm": 0.4889221372366856,
+      "learning_rate": 9.670179698296835e-06,
+      "loss": 0.6057,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7743399995460014,
+      "grad_norm": 0.4998365163823747,
+      "learning_rate": 9.669326498076258e-06,
+      "loss": 0.6244,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7750663972941684,
+      "grad_norm": 0.4461796783543215,
+      "learning_rate": 9.668472233462232e-06,
+      "loss": 0.5957,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7757927950423353,
+      "grad_norm": 0.4653827741141469,
+      "learning_rate": 9.667616904649491e-06,
+      "loss": 0.6223,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7765191927905023,
+      "grad_norm": 0.5101486161304896,
+      "learning_rate": 9.666760511833014e-06,
+      "loss": 0.6206,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7772455905386694,
+      "grad_norm": 0.44938945973356537,
+      "learning_rate": 9.665903055208013e-06,
+      "loss": 0.614,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7779719882868363,
+      "grad_norm": 0.48210108522020845,
+      "learning_rate": 9.665044534969957e-06,
+      "loss": 0.6294,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7786983860350033,
+      "grad_norm": 0.536826645391065,
+      "learning_rate": 9.664184951314542e-06,
+      "loss": 0.6216,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7794247837831703,
+      "grad_norm": 0.46290611415210603,
+      "learning_rate": 9.663324304437717e-06,
+      "loss": 0.6162,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7801511815313372,
+      "grad_norm": 0.4389114689546384,
+      "learning_rate": 9.662462594535674e-06,
+      "loss": 0.6107,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7808775792795042,
+      "grad_norm": 0.4572319500457869,
+      "learning_rate": 9.661599821804836e-06,
+      "loss": 0.6241,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7816039770276713,
+      "grad_norm": 0.4827522638295839,
+      "learning_rate": 9.660735986441881e-06,
+      "loss": 0.6122,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7823303747758382,
+      "grad_norm": 0.4740115259109735,
+      "learning_rate": 9.659871088643724e-06,
+      "loss": 0.6115,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7830567725240052,
+      "grad_norm": 0.4590734004873104,
+      "learning_rate": 9.659005128607518e-06,
+      "loss": 0.6185,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7837831702721721,
+      "grad_norm": 0.8749921694047537,
+      "learning_rate": 9.65813810653067e-06,
+      "loss": 0.5945,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7845095680203391,
+      "grad_norm": 0.7035666241156626,
+      "learning_rate": 9.657270022610814e-06,
+      "loss": 0.6091,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7852359657685061,
+      "grad_norm": 0.5346422298106133,
+      "learning_rate": 9.656400877045836e-06,
+      "loss": 0.6152,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7859623635166731,
+      "grad_norm": 0.4782527712539082,
+      "learning_rate": 9.655530670033866e-06,
+      "loss": 0.6225,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7866887612648401,
+      "grad_norm": 0.47908779964086995,
+      "learning_rate": 9.654659401773264e-06,
+      "loss": 0.6161,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7874151590130071,
+      "grad_norm": 0.4391207715095105,
+      "learning_rate": 9.653787072462644e-06,
+      "loss": 0.6166,
+      "step": 1084
+    },
+    {
+      "epoch": 0.788141556761174,
+      "grad_norm": 0.4423155828257749,
+      "learning_rate": 9.652913682300856e-06,
+      "loss": 0.5925,
+      "step": 1085
+    },
+    {
+      "epoch": 0.788867954509341,
+      "grad_norm": 0.4425248921573,
+      "learning_rate": 9.652039231486993e-06,
+      "loss": 0.6095,
+      "step": 1086
+    },
+    {
+      "epoch": 0.789594352257508,
+      "grad_norm": 0.4700153894358856,
+      "learning_rate": 9.651163720220391e-06,
+      "loss": 0.6084,
+      "step": 1087
+    },
+    {
+      "epoch": 0.790320750005675,
+      "grad_norm": 0.5178001098713066,
+      "learning_rate": 9.650287148700626e-06,
+      "loss": 0.6016,
+      "step": 1088
+    },
+    {
+      "epoch": 0.791047147753842,
+      "grad_norm": 0.47127932138489764,
+      "learning_rate": 9.649409517127515e-06,
+      "loss": 0.632,
+      "step": 1089
+    },
+    {
+      "epoch": 0.791773545502009,
+      "grad_norm": 0.5332912769127331,
+      "learning_rate": 9.648530825701118e-06,
+      "loss": 0.6172,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7924999432501759,
+      "grad_norm": 0.5257897079993382,
+      "learning_rate": 9.647651074621741e-06,
+      "loss": 0.6212,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7932263409983429,
+      "grad_norm": 0.5086619699814738,
+      "learning_rate": 9.646770264089921e-06,
+      "loss": 0.6206,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7939527387465098,
+      "grad_norm": 0.47689364802107714,
+      "learning_rate": 9.645888394306448e-06,
+      "loss": 0.6112,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7946791364946769,
+      "grad_norm": 0.5028782234896939,
+      "learning_rate": 9.645005465472345e-06,
+      "loss": 0.6258,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7954055342428439,
+      "grad_norm": 0.48796076270457783,
+      "learning_rate": 9.644121477788879e-06,
+      "loss": 0.6159,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7961319319910108,
+      "grad_norm": 0.4538044670324512,
+      "learning_rate": 9.643236431457561e-06,
+      "loss": 0.608,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7968583297391778,
+      "grad_norm": 0.4616355979057923,
+      "learning_rate": 9.64235032668014e-06,
+      "loss": 0.6164,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7975847274873448,
+      "grad_norm": 0.5297894983069743,
+      "learning_rate": 9.641463163658606e-06,
+      "loss": 0.6352,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7983111252355117,
+      "grad_norm": 0.4591303780851485,
+      "learning_rate": 9.640574942595195e-06,
+      "loss": 0.5976,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7990375229836788,
+      "grad_norm": 0.465064202103376,
+      "learning_rate": 9.63968566369238e-06,
+      "loss": 0.614,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7997639207318458,
+      "grad_norm": 0.4665514161467082,
+      "learning_rate": 9.638795327152872e-06,
+      "loss": 0.5919,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8004903184800127,
+      "grad_norm": 0.44878678129345034,
+      "learning_rate": 9.637903933179633e-06,
+      "loss": 0.6059,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8012167162281797,
+      "grad_norm": 0.4741833500595366,
+      "learning_rate": 9.637011481975858e-06,
+      "loss": 0.6378,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8019431139763467,
+      "grad_norm": 0.45791949887831834,
+      "learning_rate": 9.636117973744982e-06,
+      "loss": 0.6179,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8026695117245136,
+      "grad_norm": 0.5404917727674573,
+      "learning_rate": 9.635223408690688e-06,
+      "loss": 0.6081,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8033959094726806,
+      "grad_norm": 0.4476699095118304,
+      "learning_rate": 9.634327787016894e-06,
+      "loss": 0.6212,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8041223072208477,
+      "grad_norm": 0.4645084972315594,
+      "learning_rate": 9.633431108927764e-06,
+      "loss": 0.6103,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8048487049690146,
+      "grad_norm": 0.4591418402004859,
+      "learning_rate": 9.632533374627695e-06,
+      "loss": 0.624,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8055751027171816,
+      "grad_norm": 0.4264466790448802,
+      "learning_rate": 9.631634584321334e-06,
+      "loss": 0.604,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8063015004653485,
+      "grad_norm": 0.462102047288957,
+      "learning_rate": 9.630734738213561e-06,
+      "loss": 0.6193,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8070278982135155,
+      "grad_norm": 0.44591136659649727,
+      "learning_rate": 9.6298338365095e-06,
+      "loss": 0.5942,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8077542959616825,
+      "grad_norm": 0.5135813110167959,
+      "learning_rate": 9.628931879414519e-06,
+      "loss": 0.6073,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8084806937098495,
+      "grad_norm": 0.4301105343160582,
+      "learning_rate": 9.628028867134216e-06,
+      "loss": 0.5967,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8092070914580165,
+      "grad_norm": 0.6578782656668021,
+      "learning_rate": 9.627124799874446e-06,
+      "loss": 0.6026,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8099334892061835,
+      "grad_norm": 0.4958492633099008,
+      "learning_rate": 9.626219677841287e-06,
+      "loss": 0.6152,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8106598869543504,
+      "grad_norm": 0.4756410241779951,
+      "learning_rate": 9.625313501241069e-06,
+      "loss": 0.5805,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8113862847025174,
+      "grad_norm": 0.5433562306080713,
+      "learning_rate": 9.62440627028036e-06,
+      "loss": 0.626,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8121126824506844,
+      "grad_norm": 0.48943135901184903,
+      "learning_rate": 9.623497985165964e-06,
+      "loss": 0.6052,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8128390801988514,
+      "grad_norm": 0.5114674351446247,
+      "learning_rate": 9.622588646104934e-06,
+      "loss": 0.6023,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8135654779470184,
+      "grad_norm": 0.5220309988554744,
+      "learning_rate": 9.62167825330455e-06,
+      "loss": 0.6,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8142918756951854,
+      "grad_norm": 0.4994686649781941,
+      "learning_rate": 9.620766806972348e-06,
+      "loss": 0.5928,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8150182734433523,
+      "grad_norm": 0.4514390687848239,
+      "learning_rate": 9.61985430731609e-06,
+      "loss": 0.6155,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8157446711915193,
+      "grad_norm": 0.4681036653080852,
+      "learning_rate": 9.618940754543788e-06,
+      "loss": 0.5874,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8164710689396862,
+      "grad_norm": 0.4395085861388702,
+      "learning_rate": 9.618026148863689e-06,
+      "loss": 0.5947,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8171974666878533,
+      "grad_norm": 0.5053784535742852,
+      "learning_rate": 9.617110490484281e-06,
+      "loss": 0.6001,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8179238644360203,
+      "grad_norm": 0.4960217010563304,
+      "learning_rate": 9.616193779614294e-06,
+      "loss": 0.6125,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8186502621841872,
+      "grad_norm": 0.4852233185765258,
+      "learning_rate": 9.615276016462694e-06,
+      "loss": 0.631,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8193766599323542,
+      "grad_norm": 0.4550745865680181,
+      "learning_rate": 9.61435720123869e-06,
+      "loss": 0.6036,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8201030576805212,
+      "grad_norm": 0.46274283671479666,
+      "learning_rate": 9.613437334151731e-06,
+      "loss": 0.6051,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8208294554286881,
+      "grad_norm": 0.4768928920065255,
+      "learning_rate": 9.612516415411505e-06,
+      "loss": 0.6172,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8215558531768552,
+      "grad_norm": 0.4382177071505045,
+      "learning_rate": 9.611594445227939e-06,
+      "loss": 0.6208,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8222822509250222,
+      "grad_norm": 0.4637384188259406,
+      "learning_rate": 9.610671423811197e-06,
+      "loss": 0.6223,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8230086486731891,
+      "grad_norm": 0.4480710315760765,
+      "learning_rate": 9.60974735137169e-06,
+      "loss": 0.6044,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8237350464213561,
+      "grad_norm": 0.7018777620026823,
+      "learning_rate": 9.608822228120063e-06,
+      "loss": 0.6102,
+      "step": 1134
+    },
+    {
+      "epoch": 0.824461444169523,
+      "grad_norm": 0.4466079797386302,
+      "learning_rate": 9.607896054267201e-06,
+      "loss": 0.6009,
+      "step": 1135
+    },
+    {
+      "epoch": 0.82518784191769,
+      "grad_norm": 0.5051574955921055,
+      "learning_rate": 9.60696883002423e-06,
+      "loss": 0.6031,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8259142396658571,
+      "grad_norm": 0.4809646087862845,
+      "learning_rate": 9.606040555602516e-06,
+      "loss": 0.5941,
+      "step": 1137
+    },
+    {
+      "epoch": 0.826640637414024,
+      "grad_norm": 0.44871620180448013,
+      "learning_rate": 9.60511123121366e-06,
+      "loss": 0.6103,
+      "step": 1138
+    },
+    {
+      "epoch": 0.827367035162191,
+      "grad_norm": 0.49068878161291896,
+      "learning_rate": 9.604180857069509e-06,
+      "loss": 0.6228,
+      "step": 1139
+    },
+    {
+      "epoch": 0.828093432910358,
+      "grad_norm": 0.4768595004531384,
+      "learning_rate": 9.603249433382145e-06,
+      "loss": 0.6194,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8288198306585249,
+      "grad_norm": 0.48342908704288834,
+      "learning_rate": 9.602316960363889e-06,
+      "loss": 0.6033,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8295462284066919,
+      "grad_norm": 0.6161075106997672,
+      "learning_rate": 9.601383438227303e-06,
+      "loss": 0.6043,
+      "step": 1142
+    },
+    {
+      "epoch": 0.830272626154859,
+      "grad_norm": 0.4365995879925137,
+      "learning_rate": 9.600448867185188e-06,
+      "loss": 0.5937,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8309990239030259,
+      "grad_norm": 0.43924899950044805,
+      "learning_rate": 9.599513247450581e-06,
+      "loss": 0.6126,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8317254216511929,
+      "grad_norm": 0.4606375387545376,
+      "learning_rate": 9.598576579236765e-06,
+      "loss": 0.6205,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8324518193993599,
+      "grad_norm": 0.4370080723956695,
+      "learning_rate": 9.597638862757255e-06,
+      "loss": 0.6222,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8331782171475268,
+      "grad_norm": 0.45381509810289633,
+      "learning_rate": 9.596700098225806e-06,
+      "loss": 0.5969,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8339046148956938,
+      "grad_norm": 0.45557634817071224,
+      "learning_rate": 9.595760285856418e-06,
+      "loss": 0.5993,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8346310126438609,
+      "grad_norm": 0.4413227228028858,
+      "learning_rate": 9.59481942586332e-06,
+      "loss": 0.6113,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8353574103920278,
+      "grad_norm": 0.45948096434027247,
+      "learning_rate": 9.593877518460988e-06,
+      "loss": 0.6003,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8360838081401948,
+      "grad_norm": 0.4077180909410102,
+      "learning_rate": 9.592934563864136e-06,
+      "loss": 0.6172,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8368102058883617,
+      "grad_norm": 0.43642720072650515,
+      "learning_rate": 9.59199056228771e-06,
+      "loss": 0.5942,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8375366036365287,
+      "grad_norm": 0.4566716571217164,
+      "learning_rate": 9.591045513946904e-06,
+      "loss": 0.6288,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8382630013846957,
+      "grad_norm": 0.4256375577799388,
+      "learning_rate": 9.590099419057142e-06,
+      "loss": 0.5925,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8389893991328626,
+      "grad_norm": 0.43304644846968315,
+      "learning_rate": 9.589152277834093e-06,
+      "loss": 0.6201,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8397157968810297,
+      "grad_norm": 0.4100228725712963,
+      "learning_rate": 9.58820409049366e-06,
+      "loss": 0.5943,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8404421946291967,
+      "grad_norm": 0.4099528807583989,
+      "learning_rate": 9.587254857251988e-06,
+      "loss": 0.6109,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8411685923773636,
+      "grad_norm": 0.491774985036076,
+      "learning_rate": 9.586304578325457e-06,
+      "loss": 0.6096,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8418949901255306,
+      "grad_norm": 0.559738280848998,
+      "learning_rate": 9.585353253930692e-06,
+      "loss": 0.6069,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8426213878736976,
+      "grad_norm": 0.40139731209421825,
+      "learning_rate": 9.584400884284546e-06,
+      "loss": 0.6131,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8433477856218645,
+      "grad_norm": 0.4400340950055217,
+      "learning_rate": 9.58344746960412e-06,
+      "loss": 0.6129,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8440741833700316,
+      "grad_norm": 0.4166850878820713,
+      "learning_rate": 9.582493010106744e-06,
+      "loss": 0.6041,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8448005811181986,
+      "grad_norm": 0.5165919941194519,
+      "learning_rate": 9.581537506009996e-06,
+      "loss": 0.6115,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8455269788663655,
+      "grad_norm": 0.4151737437790665,
+      "learning_rate": 9.580580957531685e-06,
+      "loss": 0.5974,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8462533766145325,
+      "grad_norm": 0.4550231106805591,
+      "learning_rate": 9.579623364889863e-06,
+      "loss": 0.6238,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8469797743626994,
+      "grad_norm": 0.4506499145111714,
+      "learning_rate": 9.578664728302813e-06,
+      "loss": 0.608,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8477061721108664,
+      "grad_norm": 0.5499080210212972,
+      "learning_rate": 9.577705047989065e-06,
+      "loss": 0.6023,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8484325698590335,
+      "grad_norm": 0.625121033765801,
+      "learning_rate": 9.57674432416738e-06,
+      "loss": 0.5984,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8491589676072004,
+      "grad_norm": 1.1862443305257935,
+      "learning_rate": 9.575782557056759e-06,
+      "loss": 0.6052,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8498853653553674,
+      "grad_norm": 0.47769553696473244,
+      "learning_rate": 9.57481974687644e-06,
+      "loss": 0.5912,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8506117631035344,
+      "grad_norm": 0.4080077932005349,
+      "learning_rate": 9.573855893845905e-06,
+      "loss": 0.6235,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8513381608517013,
+      "grad_norm": 0.4761375945040686,
+      "learning_rate": 9.572890998184862e-06,
+      "loss": 0.606,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8520645585998683,
+      "grad_norm": 0.5766409060005017,
+      "learning_rate": 9.571925060113267e-06,
+      "loss": 0.5979,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8527909563480354,
+      "grad_norm": 0.42596241668640383,
+      "learning_rate": 9.57095807985131e-06,
+      "loss": 0.5915,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8535173540962023,
+      "grad_norm": 0.5056126132604277,
+      "learning_rate": 9.569990057619414e-06,
+      "loss": 0.5898,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8542437518443693,
+      "grad_norm": 0.43335266407709333,
+      "learning_rate": 9.569020993638249e-06,
+      "loss": 0.6022,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8549701495925363,
+      "grad_norm": 0.44207808281110866,
+      "learning_rate": 9.568050888128716e-06,
+      "loss": 0.6071,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8556965473407032,
+      "grad_norm": 0.4341399555907341,
+      "learning_rate": 9.567079741311956e-06,
+      "loss": 0.6043,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8564229450888702,
+      "grad_norm": 0.4369089309389326,
+      "learning_rate": 9.566107553409341e-06,
+      "loss": 0.587,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8571493428370373,
+      "grad_norm": 0.4599730151913197,
+      "learning_rate": 9.565134324642491e-06,
+      "loss": 0.6102,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8578757405852042,
+      "grad_norm": 0.43977979280579654,
+      "learning_rate": 9.564160055233256e-06,
+      "loss": 0.5987,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8586021383333712,
+      "grad_norm": 0.46783053277652853,
+      "learning_rate": 9.563184745403725e-06,
+      "loss": 0.6029,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8593285360815381,
+      "grad_norm": 0.44458871164794883,
+      "learning_rate": 9.562208395376223e-06,
+      "loss": 0.593,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8600549338297051,
+      "grad_norm": 0.5733471186258695,
+      "learning_rate": 9.561231005373315e-06,
+      "loss": 0.6064,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8607813315778721,
+      "grad_norm": 0.45341893514784476,
+      "learning_rate": 9.560252575617798e-06,
+      "loss": 0.5947,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8615077293260391,
+      "grad_norm": 0.5303361989038317,
+      "learning_rate": 9.559273106332716e-06,
+      "loss": 0.5921,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8622341270742061,
+      "grad_norm": 0.4370520610409579,
+      "learning_rate": 9.558292597741337e-06,
+      "loss": 0.607,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8629605248223731,
+      "grad_norm": 0.4175970592323233,
+      "learning_rate": 9.557311050067175e-06,
+      "loss": 0.6018,
+      "step": 1188
+    },
+    {
+      "epoch": 0.86368692257054,
+      "grad_norm": 0.46416643253026196,
+      "learning_rate": 9.556328463533976e-06,
+      "loss": 0.6089,
+      "step": 1189
+    },
+    {
+      "epoch": 0.864413320318707,
+      "grad_norm": 0.6168553625857448,
+      "learning_rate": 9.555344838365727e-06,
+      "loss": 0.5861,
+      "step": 1190
+    },
+    {
+      "epoch": 0.865139718066874,
+      "grad_norm": 0.48915934761576074,
+      "learning_rate": 9.554360174786651e-06,
+      "loss": 0.6088,
+      "step": 1191
+    },
+    {
+      "epoch": 0.865866115815041,
+      "grad_norm": 0.44675004959116177,
+      "learning_rate": 9.553374473021204e-06,
+      "loss": 0.5922,
+      "step": 1192
+    },
+    {
+      "epoch": 0.866592513563208,
+      "grad_norm": 0.4385531991454444,
+      "learning_rate": 9.552387733294081e-06,
+      "loss": 0.5968,
+      "step": 1193
+    },
+    {
+      "epoch": 0.867318911311375,
+      "grad_norm": 0.472160005452995,
+      "learning_rate": 9.551399955830215e-06,
+      "loss": 0.6009,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8680453090595419,
+      "grad_norm": 0.5114615003958931,
+      "learning_rate": 9.550411140854772e-06,
+      "loss": 0.5957,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8687717068077089,
+      "grad_norm": 0.46197727438219466,
+      "learning_rate": 9.549421288593157e-06,
+      "loss": 0.6035,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8694981045558758,
+      "grad_norm": 0.4505938549631761,
+      "learning_rate": 9.548430399271013e-06,
+      "loss": 0.6237,
+      "step": 1197
+    },
+    {
+      "epoch": 0.8702245023040429,
+      "grad_norm": 0.43654822032779134,
+      "learning_rate": 9.547438473114219e-06,
+      "loss": 0.6089,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8709509000522099,
+      "grad_norm": 0.46067909068019247,
+      "learning_rate": 9.546445510348884e-06,
+      "loss": 0.61,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8716772978003768,
+      "grad_norm": 0.41694081713666115,
+      "learning_rate": 9.545451511201365e-06,
+      "loss": 0.5918,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8724036955485438,
+      "grad_norm": 0.46578768324984093,
+      "learning_rate": 9.54445647589824e-06,
+      "loss": 0.6005,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8731300932967108,
+      "grad_norm": 0.4438616477116534,
+      "learning_rate": 9.543460404666338e-06,
+      "loss": 0.5999,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8738564910448777,
+      "grad_norm": 0.48358201793997363,
+      "learning_rate": 9.542463297732716e-06,
+      "loss": 0.6064,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8745828887930447,
+      "grad_norm": 0.501153042263637,
+      "learning_rate": 9.54146515532467e-06,
+      "loss": 0.6203,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8753092865412118,
+      "grad_norm": 0.4521383558348136,
+      "learning_rate": 9.540465977669728e-06,
+      "loss": 0.6022,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8760356842893787,
+      "grad_norm": 0.5112596720285417,
+      "learning_rate": 9.539465764995658e-06,
+      "loss": 0.5977,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8767620820375457,
+      "grad_norm": 0.4861584352029897,
+      "learning_rate": 9.538464517530466e-06,
+      "loss": 0.5856,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8774884797857126,
+      "grad_norm": 0.4200265080369869,
+      "learning_rate": 9.537462235502385e-06,
+      "loss": 0.5866,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8782148775338796,
+      "grad_norm": 0.4261094133432591,
+      "learning_rate": 9.536458919139894e-06,
+      "loss": 0.5933,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8789412752820466,
+      "grad_norm": 0.4003104970002111,
+      "learning_rate": 9.535454568671705e-06,
+      "loss": 0.5844,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8796676730302136,
+      "grad_norm": 0.42295910863999364,
+      "learning_rate": 9.53444918432676e-06,
+      "loss": 0.6024,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8803940707783806,
+      "grad_norm": 0.4149494814967926,
+      "learning_rate": 9.53344276633424e-06,
+      "loss": 0.5863,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8811204685265476,
+      "grad_norm": 0.5110127220773379,
+      "learning_rate": 9.532435314923566e-06,
+      "loss": 0.5767,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8818468662747145,
+      "grad_norm": 0.43349333837655685,
+      "learning_rate": 9.531426830324388e-06,
+      "loss": 0.5951,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8825732640228815,
+      "grad_norm": 0.44997031713934815,
+      "learning_rate": 9.530417312766598e-06,
+      "loss": 0.6187,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8832996617710485,
+      "grad_norm": 0.4446591721155272,
+      "learning_rate": 9.529406762480318e-06,
+      "loss": 0.5985,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8840260595192155,
+      "grad_norm": 0.44120700182390005,
+      "learning_rate": 9.528395179695907e-06,
+      "loss": 0.5813,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8847524572673825,
+      "grad_norm": 0.4295735205310877,
+      "learning_rate": 9.527382564643957e-06,
+      "loss": 0.5971,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8854788550155495,
+      "grad_norm": 0.43730360118736156,
+      "learning_rate": 9.526368917555304e-06,
+      "loss": 0.6023,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8862052527637164,
+      "grad_norm": 0.43250587556589976,
+      "learning_rate": 9.52535423866101e-06,
+      "loss": 0.5893,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8869316505118834,
+      "grad_norm": 0.47439969227732015,
+      "learning_rate": 9.524338528192375e-06,
+      "loss": 0.5983,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8876580482600503,
+      "grad_norm": 0.43225531857326166,
+      "learning_rate": 9.523321786380936e-06,
+      "loss": 0.5784,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8883844460082174,
+      "grad_norm": 0.5170344250080916,
+      "learning_rate": 9.522304013458464e-06,
+      "loss": 0.6216,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8891108437563844,
+      "grad_norm": 0.42919086642378623,
+      "learning_rate": 9.521285209656964e-06,
+      "loss": 0.5881,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8898372415045513,
+      "grad_norm": 0.47659107028807385,
+      "learning_rate": 9.520265375208675e-06,
+      "loss": 0.5952,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8905636392527183,
+      "grad_norm": 0.41359128320071464,
+      "learning_rate": 9.519244510346076e-06,
+      "loss": 0.5908,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8912900370008853,
+      "grad_norm": 0.5644872809341976,
+      "learning_rate": 9.518222615301875e-06,
+      "loss": 0.5875,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8920164347490522,
+      "grad_norm": 0.47490025096662025,
+      "learning_rate": 9.51719969030902e-06,
+      "loss": 0.5917,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8927428324972193,
+      "grad_norm": 0.569369819329793,
+      "learning_rate": 9.516175735600688e-06,
+      "loss": 0.6024,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8934692302453863,
+      "grad_norm": 0.43225621864223424,
+      "learning_rate": 9.515150751410298e-06,
+      "loss": 0.6077,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8941956279935532,
+      "grad_norm": 0.708224935453519,
+      "learning_rate": 9.514124737971495e-06,
+      "loss": 0.5964,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8949220257417202,
+      "grad_norm": 0.4739759856133637,
+      "learning_rate": 9.51309769551817e-06,
+      "loss": 0.5763,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8956484234898872,
+      "grad_norm": 0.7728743977820823,
+      "learning_rate": 9.512069624284433e-06,
+      "loss": 0.5842,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8963748212380541,
+      "grad_norm": 0.43514666056639595,
+      "learning_rate": 9.511040524504645e-06,
+      "loss": 0.5967,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8971012189862212,
+      "grad_norm": 0.45557762839508126,
+      "learning_rate": 9.510010396413389e-06,
+      "loss": 0.5897,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8978276167343882,
+      "grad_norm": 0.5039340185306304,
+      "learning_rate": 9.508979240245489e-06,
+      "loss": 0.5773,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8985540144825551,
+      "grad_norm": 0.46341563582692274,
+      "learning_rate": 9.507947056236003e-06,
+      "loss": 0.5814,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8992804122307221,
+      "grad_norm": 0.47824940112014946,
+      "learning_rate": 9.506913844620217e-06,
+      "loss": 0.6016,
+      "step": 1238
+    },
+    {
+      "epoch": 0.900006809978889,
+      "grad_norm": 0.46306899110926264,
+      "learning_rate": 9.505879605633664e-06,
+      "loss": 0.5969,
+      "step": 1239
+    },
+    {
+      "epoch": 0.900733207727056,
+      "grad_norm": 0.4785112448900572,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.6181,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9014596054752231,
+      "grad_norm": 0.44806415943157596,
+      "learning_rate": 9.503808046491511e-06,
+      "loss": 0.5909,
+      "step": 1241
+    },
+    {
+      "epoch": 0.90218600322339,
+      "grad_norm": 0.4039053308077116,
+      "learning_rate": 9.502770726808133e-06,
+      "loss": 0.5942,
+      "step": 1242
+    },
+    {
+      "epoch": 0.902912400971557,
+      "grad_norm": 0.4988364921150387,
+      "learning_rate": 9.501732380698426e-06,
+      "loss": 0.6076,
+      "step": 1243
+    },
+    {
+      "epoch": 0.903638798719724,
+      "grad_norm": 0.4467206176328996,
+      "learning_rate": 9.500693008399085e-06,
+      "loss": 0.6061,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9043651964678909,
+      "grad_norm": 0.40186754448407863,
+      "learning_rate": 9.49965261014704e-06,
+      "loss": 0.5745,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9050915942160579,
+      "grad_norm": 0.4649796467570825,
+      "learning_rate": 9.498611186179454e-06,
+      "loss": 0.5891,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9058179919642249,
+      "grad_norm": 0.4076705157572678,
+      "learning_rate": 9.497568736733722e-06,
+      "loss": 0.6031,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9065443897123919,
+      "grad_norm": 0.4512970825986025,
+      "learning_rate": 9.496525262047479e-06,
+      "loss": 0.589,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9072707874605589,
+      "grad_norm": 0.42034195469706914,
+      "learning_rate": 9.495480762358585e-06,
+      "loss": 0.5858,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9079971852087259,
+      "grad_norm": 0.42929899360092316,
+      "learning_rate": 9.494435237905142e-06,
+      "loss": 0.5743,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9087235829568928,
+      "grad_norm": 0.4118235244162934,
+      "learning_rate": 9.493388688925481e-06,
+      "loss": 0.5844,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9094499807050598,
+      "grad_norm": 0.41506895210123806,
+      "learning_rate": 9.492341115658167e-06,
+      "loss": 0.6049,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9101763784532267,
+      "grad_norm": 0.4165328260065925,
+      "learning_rate": 9.491292518341998e-06,
+      "loss": 0.6008,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9109027762013938,
+      "grad_norm": 0.38977236916049524,
+      "learning_rate": 9.490242897216008e-06,
+      "loss": 0.615,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9116291739495608,
+      "grad_norm": 0.46505897512184374,
+      "learning_rate": 9.489192252519462e-06,
+      "loss": 0.5814,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9123555716977277,
+      "grad_norm": 0.446928251555708,
+      "learning_rate": 9.488140584491859e-06,
+      "loss": 0.6047,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9130819694458947,
+      "grad_norm": 0.43706570616826285,
+      "learning_rate": 9.487087893372933e-06,
+      "loss": 0.5741,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9138083671940617,
+      "grad_norm": 0.4537167354168948,
+      "learning_rate": 9.486034179402645e-06,
+      "loss": 0.5896,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9145347649422286,
+      "grad_norm": 0.41506148446047436,
+      "learning_rate": 9.484979442821199e-06,
+      "loss": 0.5949,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9152611626903957,
+      "grad_norm": 0.4638563917811123,
+      "learning_rate": 9.483923683869025e-06,
+      "loss": 0.584,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9159875604385627,
+      "grad_norm": 0.41081717634677034,
+      "learning_rate": 9.482866902786784e-06,
+      "loss": 0.5958,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9167139581867296,
+      "grad_norm": 0.44940733112945636,
+      "learning_rate": 9.481809099815382e-06,
+      "loss": 0.5868,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9174403559348966,
+      "grad_norm": 0.5181433591757858,
+      "learning_rate": 9.480750275195942e-06,
+      "loss": 0.609,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9181667536830636,
+      "grad_norm": 0.4264257820314425,
+      "learning_rate": 9.479690429169832e-06,
+      "loss": 0.585,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9188931514312305,
+      "grad_norm": 0.4385872968678699,
+      "learning_rate": 9.478629561978648e-06,
+      "loss": 0.5789,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9196195491793976,
+      "grad_norm": 0.4339437079303383,
+      "learning_rate": 9.477567673864217e-06,
+      "loss": 0.5947,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9203459469275646,
+      "grad_norm": 0.4415046209883479,
+      "learning_rate": 9.476504765068604e-06,
+      "loss": 0.6006,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9210723446757315,
+      "grad_norm": 0.42129146459625944,
+      "learning_rate": 9.475440835834103e-06,
+      "loss": 0.582,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9217987424238985,
+      "grad_norm": 0.48232608637628704,
+      "learning_rate": 9.474375886403239e-06,
+      "loss": 0.5914,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9225251401720654,
+      "grad_norm": 0.4770042456279813,
+      "learning_rate": 9.473309917018774e-06,
+      "loss": 0.6019,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9232515379202324,
+      "grad_norm": 0.4536442707571275,
+      "learning_rate": 9.4722429279237e-06,
+      "loss": 0.5864,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9239779356683995,
+      "grad_norm": 0.42626513154423673,
+      "learning_rate": 9.471174919361244e-06,
+      "loss": 0.5763,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9247043334165664,
+      "grad_norm": 0.6205482644129555,
+      "learning_rate": 9.47010589157486e-06,
+      "loss": 0.5901,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9254307311647334,
+      "grad_norm": 0.43225121195112626,
+      "learning_rate": 9.469035844808238e-06,
+      "loss": 0.5956,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9261571289129004,
+      "grad_norm": 0.41283182941994373,
+      "learning_rate": 9.467964779305304e-06,
+      "loss": 0.5997,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9268835266610673,
+      "grad_norm": 0.41546967158855563,
+      "learning_rate": 9.466892695310204e-06,
+      "loss": 0.584,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9276099244092343,
+      "grad_norm": 0.4300218140453633,
+      "learning_rate": 9.465819593067332e-06,
+      "loss": 0.5849,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9283363221574014,
+      "grad_norm": 0.4456416918954236,
+      "learning_rate": 9.464745472821302e-06,
+      "loss": 0.5973,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9290627199055683,
+      "grad_norm": 0.4482037090374508,
+      "learning_rate": 9.463670334816967e-06,
+      "loss": 0.58,
+      "step": 1279
+    },
+    {
+      "epoch": 0.9297891176537353,
+      "grad_norm": 0.46007508129114516,
+      "learning_rate": 9.462594179299408e-06,
+      "loss": 0.6001,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9305155154019022,
+      "grad_norm": 0.5011958549498003,
+      "learning_rate": 9.461517006513938e-06,
+      "loss": 0.5845,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9312419131500692,
+      "grad_norm": 0.42431402288059444,
+      "learning_rate": 9.460438816706106e-06,
+      "loss": 0.5905,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9319683108982362,
+      "grad_norm": 0.4502683304139396,
+      "learning_rate": 9.459359610121688e-06,
+      "loss": 0.5704,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9326947086464032,
+      "grad_norm": 0.3796874806100935,
+      "learning_rate": 9.458279387006695e-06,
+      "loss": 0.5768,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9334211063945702,
+      "grad_norm": 0.5004568592535342,
+      "learning_rate": 9.45719814760737e-06,
+      "loss": 0.5774,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9341475041427372,
+      "grad_norm": 0.4340577622957361,
+      "learning_rate": 9.456115892170183e-06,
+      "loss": 0.5911,
+      "step": 1286
+    },
+    {
+      "epoch": 0.9348739018909041,
+      "grad_norm": 0.45670721572589634,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.599,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9356002996390711,
+      "grad_norm": 0.45572780467011514,
+      "learning_rate": 9.453948334169279e-06,
+      "loss": 0.5747,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9363266973872381,
+      "grad_norm": 0.38841356192268106,
+      "learning_rate": 9.452863032099666e-06,
+      "loss": 0.5971,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9370530951354051,
+      "grad_norm": 0.4298037377188982,
+      "learning_rate": 9.451776714980402e-06,
+      "loss": 0.5913,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9377794928835721,
+      "grad_norm": 0.41341305861326055,
+      "learning_rate": 9.450689383059118e-06,
+      "loss": 0.5796,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9385058906317391,
+      "grad_norm": 0.42660791554657346,
+      "learning_rate": 9.449601036583673e-06,
+      "loss": 0.5819,
+      "step": 1292
+    },
+    {
+      "epoch": 0.939232288379906,
+      "grad_norm": 0.5598103689052227,
+      "learning_rate": 9.448511675802162e-06,
+      "loss": 0.5838,
+      "step": 1293
+    },
+    {
+      "epoch": 0.939958686128073,
+      "grad_norm": 0.4586706687352272,
+      "learning_rate": 9.447421300962911e-06,
+      "loss": 0.5976,
+      "step": 1294
+    },
+    {
+      "epoch": 0.94068508387624,
+      "grad_norm": 0.43063693058417746,
+      "learning_rate": 9.446329912314475e-06,
+      "loss": 0.5797,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9414114816244069,
+      "grad_norm": 0.4404865199677851,
+      "learning_rate": 9.44523751010564e-06,
+      "loss": 0.5937,
+      "step": 1296
+    },
+    {
+      "epoch": 0.942137879372574,
+      "grad_norm": 0.4531349959997388,
+      "learning_rate": 9.444144094585425e-06,
+      "loss": 0.5941,
+      "step": 1297
+    },
+    {
+      "epoch": 0.942864277120741,
+      "grad_norm": 0.5032016176400471,
+      "learning_rate": 9.443049666003077e-06,
+      "loss": 0.5945,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9435906748689079,
+      "grad_norm": 0.44126419665834615,
+      "learning_rate": 9.441954224608079e-06,
+      "loss": 0.5949,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9443170726170749,
+      "grad_norm": 0.45391630167344893,
+      "learning_rate": 9.440857770650139e-06,
+      "loss": 0.5898,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9450434703652418,
+      "grad_norm": 0.3956164187382792,
+      "learning_rate": 9.439760304379197e-06,
+      "loss": 0.5874,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9457698681134088,
+      "grad_norm": 0.4597343602525186,
+      "learning_rate": 9.438661826045429e-06,
+      "loss": 0.5841,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9464962658615759,
+      "grad_norm": 0.4933253827016849,
+      "learning_rate": 9.437562335899237e-06,
+      "loss": 0.5949,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9472226636097428,
+      "grad_norm": 0.42059079303606517,
+      "learning_rate": 9.436461834191254e-06,
+      "loss": 0.5735,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9479490613579098,
+      "grad_norm": 0.6604789856347021,
+      "learning_rate": 9.435360321172344e-06,
+      "loss": 0.5951,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9486754591060768,
+      "grad_norm": 0.6174395501507661,
+      "learning_rate": 9.434257797093602e-06,
+      "loss": 0.5904,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9494018568542437,
+      "grad_norm": 0.5221870207891371,
+      "learning_rate": 9.433154262206353e-06,
+      "loss": 0.6028,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9501282546024107,
+      "grad_norm": 0.4748141147379064,
+      "learning_rate": 9.432049716762151e-06,
+      "loss": 0.5967,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9508546523505778,
+      "grad_norm": 0.4557962640119291,
+      "learning_rate": 9.430944161012784e-06,
+      "loss": 0.5936,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9515810500987447,
+      "grad_norm": 0.4067491075418985,
+      "learning_rate": 9.42983759521027e-06,
+      "loss": 0.6003,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9523074478469117,
+      "grad_norm": 0.44211224151004935,
+      "learning_rate": 9.428730019606853e-06,
+      "loss": 0.5775,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9530338455950786,
+      "grad_norm": 0.407063349141801,
+      "learning_rate": 9.42762143445501e-06,
+      "loss": 0.59,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9537602433432456,
+      "grad_norm": 0.4028594981568491,
+      "learning_rate": 9.426511840007448e-06,
+      "loss": 0.5801,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9544866410914126,
+      "grad_norm": 0.3919749106342764,
+      "learning_rate": 9.425401236517104e-06,
+      "loss": 0.5905,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9552130388395796,
+      "grad_norm": 0.4032757710830502,
+      "learning_rate": 9.424289624237143e-06,
+      "loss": 0.6031,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9559394365877466,
+      "grad_norm": 0.4293862484316795,
+      "learning_rate": 9.423177003420966e-06,
+      "loss": 0.5708,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9566658343359136,
+      "grad_norm": 0.45285430043737923,
+      "learning_rate": 9.422063374322197e-06,
+      "loss": 0.5883,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9573922320840805,
+      "grad_norm": 0.5523709468876832,
+      "learning_rate": 9.420948737194693e-06,
+      "loss": 0.5866,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9581186298322475,
+      "grad_norm": 0.4459756657537176,
+      "learning_rate": 9.419833092292542e-06,
+      "loss": 0.5641,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9588450275804145,
+      "grad_norm": 0.48301638137861325,
+      "learning_rate": 9.418716439870056e-06,
+      "loss": 0.5865,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9595714253285815,
+      "grad_norm": 0.4128881676008887,
+      "learning_rate": 9.417598780181787e-06,
+      "loss": 0.6071,
+      "step": 1321
+    },
+    {
+      "epoch": 0.9602978230767485,
+      "grad_norm": 0.44753948046420106,
+      "learning_rate": 9.416480113482505e-06,
+      "loss": 0.5822,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9610242208249155,
+      "grad_norm": 0.4279853381773522,
+      "learning_rate": 9.415360440027216e-06,
+      "loss": 0.5804,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9617506185730824,
+      "grad_norm": 0.4338581451320516,
+      "learning_rate": 9.414239760071158e-06,
+      "loss": 0.5887,
+      "step": 1324
+    },
+    {
+      "epoch": 0.9624770163212494,
+      "grad_norm": 0.42860628721189403,
+      "learning_rate": 9.413118073869791e-06,
+      "loss": 0.5882,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9632034140694163,
+      "grad_norm": 0.44956892815122457,
+      "learning_rate": 9.411995381678811e-06,
+      "loss": 0.5782,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9639298118175834,
+      "grad_norm": 0.4814860194215449,
+      "learning_rate": 9.410871683754139e-06,
+      "loss": 0.5702,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9646562095657504,
+      "grad_norm": 0.42581017230827334,
+      "learning_rate": 9.409746980351927e-06,
+      "loss": 0.5678,
+      "step": 1328
+    },
+    {
+      "epoch": 0.9653826073139173,
+      "grad_norm": 0.44826183504581213,
+      "learning_rate": 9.408621271728555e-06,
+      "loss": 0.5591,
+      "step": 1329
+    },
+    {
+      "epoch": 0.9661090050620843,
+      "grad_norm": 0.5714153732960652,
+      "learning_rate": 9.407494558140636e-06,
+      "loss": 0.6072,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9668354028102513,
+      "grad_norm": 0.6836517336640814,
+      "learning_rate": 9.406366839845004e-06,
+      "loss": 0.577,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9675618005584182,
+      "grad_norm": 0.41510009176677465,
+      "learning_rate": 9.405238117098736e-06,
+      "loss": 0.582,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9682881983065853,
+      "grad_norm": 0.42043939764561933,
+      "learning_rate": 9.40410839015912e-06,
+      "loss": 0.5638,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9690145960547523,
+      "grad_norm": 0.47474607904645816,
+      "learning_rate": 9.40297765928369e-06,
+      "loss": 0.5808,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9697409938029192,
+      "grad_norm": 0.4404458653043789,
+      "learning_rate": 9.401845924730197e-06,
+      "loss": 0.6007,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9704673915510862,
+      "grad_norm": 0.41381749150233,
+      "learning_rate": 9.400713186756625e-06,
+      "loss": 0.5756,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9711937892992532,
+      "grad_norm": 0.40002352121048046,
+      "learning_rate": 9.399579445621187e-06,
+      "loss": 0.5845,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9719201870474201,
+      "grad_norm": 0.3819643412059154,
+      "learning_rate": 9.398444701582325e-06,
+      "loss": 0.5871,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9726465847955872,
+      "grad_norm": 0.4096279784042784,
+      "learning_rate": 9.397308954898708e-06,
+      "loss": 0.5602,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9733729825437542,
+      "grad_norm": 0.41853453487079706,
+      "learning_rate": 9.396172205829235e-06,
+      "loss": 0.5776,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9740993802919211,
+      "grad_norm": 0.40740299719446227,
+      "learning_rate": 9.395034454633032e-06,
+      "loss": 0.5805,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9748257780400881,
+      "grad_norm": 0.5010130777863815,
+      "learning_rate": 9.393895701569455e-06,
+      "loss": 0.5824,
+      "step": 1342
+    },
+    {
+      "epoch": 0.975552175788255,
+      "grad_norm": 0.423630646161036,
+      "learning_rate": 9.392755946898087e-06,
+      "loss": 0.5933,
+      "step": 1343
+    },
+    {
+      "epoch": 0.976278573536422,
+      "grad_norm": 0.5460848532649165,
+      "learning_rate": 9.391615190878742e-06,
+      "loss": 0.5716,
+      "step": 1344
+    },
+    {
+      "epoch": 0.977004971284589,
+      "grad_norm": 0.4314740817665518,
+      "learning_rate": 9.39047343377146e-06,
+      "loss": 0.5858,
+      "step": 1345
+    },
+    {
+      "epoch": 0.977731369032756,
+      "grad_norm": 0.4172447068705145,
+      "learning_rate": 9.389330675836508e-06,
+      "loss": 0.5855,
+      "step": 1346
+    },
+    {
+      "epoch": 0.978457766780923,
+      "grad_norm": 0.40537926860470647,
+      "learning_rate": 9.388186917334382e-06,
+      "loss": 0.5806,
+      "step": 1347
+    },
+    {
+      "epoch": 0.97918416452909,
+      "grad_norm": 0.42272881390227746,
+      "learning_rate": 9.38704215852581e-06,
+      "loss": 0.5833,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9799105622772569,
+      "grad_norm": 0.4071530915037966,
+      "learning_rate": 9.38589639967174e-06,
+      "loss": 0.5868,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9806369600254239,
+      "grad_norm": 0.4174487525760431,
+      "learning_rate": 9.384749641033358e-06,
+      "loss": 0.592,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9813633577735908,
+      "grad_norm": 0.5016160558691094,
+      "learning_rate": 9.38360188287207e-06,
+      "loss": 0.5688,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9820897555217579,
+      "grad_norm": 0.3709738274098191,
+      "learning_rate": 9.382453125449513e-06,
+      "loss": 0.5797,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9828161532699249,
+      "grad_norm": 0.4792601165573076,
+      "learning_rate": 9.381303369027552e-06,
+      "loss": 0.5799,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9835425510180918,
+      "grad_norm": 0.3842929548512234,
+      "learning_rate": 9.380152613868276e-06,
+      "loss": 0.5867,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9842689487662588,
+      "grad_norm": 0.443529556661457,
+      "learning_rate": 9.37900086023401e-06,
+      "loss": 0.573,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9849953465144258,
+      "grad_norm": 0.40568726593321786,
+      "learning_rate": 9.377848108387295e-06,
+      "loss": 0.5779,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9857217442625927,
+      "grad_norm": 0.4435967514139609,
+      "learning_rate": 9.37669435859091e-06,
+      "loss": 0.5798,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9864481420107598,
+      "grad_norm": 0.4245008371251557,
+      "learning_rate": 9.375539611107856e-06,
+      "loss": 0.5931,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9871745397589268,
+      "grad_norm": 0.4348440208427361,
+      "learning_rate": 9.374383866201364e-06,
+      "loss": 0.5866,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9879009375070937,
+      "grad_norm": 0.5474657163639375,
+      "learning_rate": 9.373227124134888e-06,
+      "loss": 0.5747,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9886273352552607,
+      "grad_norm": 0.46556069978991227,
+      "learning_rate": 9.372069385172115e-06,
+      "loss": 0.5951,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9893537330034277,
+      "grad_norm": 0.4871044110501798,
+      "learning_rate": 9.37091064957696e-06,
+      "loss": 0.5887,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9900801307515946,
+      "grad_norm": 0.4259323411994358,
+      "learning_rate": 9.369750917613554e-06,
+      "loss": 0.5907,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9908065284997617,
+      "grad_norm": 0.4100962064140313,
+      "learning_rate": 9.368590189546268e-06,
+      "loss": 0.5905,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9915329262479287,
+      "grad_norm": 0.4598716942351165,
+      "learning_rate": 9.367428465639696e-06,
+      "loss": 0.5791,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9922593239960956,
+      "grad_norm": 0.49479968191021007,
+      "learning_rate": 9.366265746158653e-06,
+      "loss": 0.5491,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9929857217442626,
+      "grad_norm": 0.40911147247756996,
+      "learning_rate": 9.365102031368191e-06,
+      "loss": 0.5765,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9937121194924295,
+      "grad_norm": 0.40613400942325284,
+      "learning_rate": 9.363937321533583e-06,
+      "loss": 0.5781,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9944385172405965,
+      "grad_norm": 0.41346069340622715,
+      "learning_rate": 9.362771616920328e-06,
+      "loss": 0.594,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9951649149887636,
+      "grad_norm": 0.42248849941999883,
+      "learning_rate": 9.361604917794155e-06,
+      "loss": 0.5784,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9958913127369305,
+      "grad_norm": 0.4395488497003654,
+      "learning_rate": 9.360437224421017e-06,
+      "loss": 0.5663,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9966177104850975,
+      "grad_norm": 0.48165754989833665,
+      "learning_rate": 9.359268537067094e-06,
+      "loss": 0.6047,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9973441082332645,
+      "grad_norm": 0.8462980679724643,
+      "learning_rate": 9.358098855998797e-06,
+      "loss": 0.5906,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9980705059814314,
+      "grad_norm": 0.44867266240084885,
+      "learning_rate": 9.356928181482757e-06,
+      "loss": 0.5807,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9987969037295984,
+      "grad_norm": 0.3922906494420764,
+      "learning_rate": 9.355756513785834e-06,
+      "loss": 0.5738,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9995233014777655,
+      "grad_norm": 0.3852398607524496,
+      "learning_rate": 9.354583853175117e-06,
+      "loss": 0.591,
+      "step": 1376
+    },
+    {
+      "epoch": 1.0002496992259324,
+      "grad_norm": 0.44617123921430374,
+      "learning_rate": 9.353410199917916e-06,
+      "loss": 0.5784,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0009760969740993,
+      "grad_norm": 0.38482197267852375,
+      "learning_rate": 9.352235554281775e-06,
+      "loss": 0.5642,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0017024947222664,
+      "grad_norm": 0.4127525659520533,
+      "learning_rate": 9.351059916534456e-06,
+      "loss": 0.5659,
+      "step": 1379
+    },
+    {
+      "epoch": 1.0024288924704334,
+      "grad_norm": 0.41330752590052344,
+      "learning_rate": 9.349883286943951e-06,
+      "loss": 0.5983,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0031552902186003,
+      "grad_norm": 0.40680416072814213,
+      "learning_rate": 9.348705665778479e-06,
+      "loss": 0.5733,
+      "step": 1381
+    },
+    {
+      "epoch": 1.0038816879667674,
+      "grad_norm": 0.4128818362833564,
+      "learning_rate": 9.347527053306482e-06,
+      "loss": 0.5882,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0046080857149342,
+      "grad_norm": 0.4170411955512922,
+      "learning_rate": 9.346347449796634e-06,
+      "loss": 0.5874,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0053344834631013,
+      "grad_norm": 0.45301343977096553,
+      "learning_rate": 9.345166855517827e-06,
+      "loss": 0.5915,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0060608812112684,
+      "grad_norm": 0.4174938754667378,
+      "learning_rate": 9.343985270739184e-06,
+      "loss": 0.5699,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0067872789594352,
+      "grad_norm": 0.6243604998156592,
+      "learning_rate": 9.34280269573005e-06,
+      "loss": 0.561,
+      "step": 1386
+    },
+    {
+      "epoch": 1.0075136767076023,
+      "grad_norm": 0.4128286014435526,
+      "learning_rate": 9.341619130760004e-06,
+      "loss": 0.5883,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0082400744557691,
+      "grad_norm": 0.4168211775574965,
+      "learning_rate": 9.34043457609884e-06,
+      "loss": 0.5734,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0089664722039362,
+      "grad_norm": 0.418443742472078,
+      "learning_rate": 9.339249032016584e-06,
+      "loss": 0.549,
+      "step": 1389
+    },
+    {
+      "epoch": 1.009692869952103,
+      "grad_norm": 0.412506877347843,
+      "learning_rate": 9.338062498783485e-06,
+      "loss": 0.5777,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0104192677002701,
+      "grad_norm": 0.4288751680263293,
+      "learning_rate": 9.336874976670018e-06,
+      "loss": 0.5745,
+      "step": 1391
+    },
+    {
+      "epoch": 1.0111456654484372,
+      "grad_norm": 0.48767089366316047,
+      "learning_rate": 9.335686465946888e-06,
+      "loss": 0.5835,
+      "step": 1392
+    },
+    {
+      "epoch": 1.011872063196604,
+      "grad_norm": 0.400362746791716,
+      "learning_rate": 9.334496966885017e-06,
+      "loss": 0.5917,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0125984609447711,
+      "grad_norm": 0.37674500749905987,
+      "learning_rate": 9.333306479755557e-06,
+      "loss": 0.5714,
+      "step": 1394
+    },
+    {
+      "epoch": 1.013324858692938,
+      "grad_norm": 0.43401884599608276,
+      "learning_rate": 9.332115004829885e-06,
+      "loss": 0.5727,
+      "step": 1395
+    },
+    {
+      "epoch": 1.014051256441105,
+      "grad_norm": 0.4355295649860661,
+      "learning_rate": 9.330922542379605e-06,
+      "loss": 0.5588,
+      "step": 1396
+    },
+    {
+      "epoch": 1.014777654189272,
+      "grad_norm": 0.45556005093229185,
+      "learning_rate": 9.329729092676542e-06,
+      "loss": 0.571,
+      "step": 1397
+    },
+    {
+      "epoch": 1.015504051937439,
+      "grad_norm": 0.43547493346898447,
+      "learning_rate": 9.328534655992747e-06,
+      "loss": 0.577,
+      "step": 1398
+    },
+    {
+      "epoch": 1.016230449685606,
+      "grad_norm": 0.39133492272444154,
+      "learning_rate": 9.3273392326005e-06,
+      "loss": 0.5845,
+      "step": 1399
+    },
+    {
+      "epoch": 1.016956847433773,
+      "grad_norm": 0.38477807853847923,
+      "learning_rate": 9.326142822772301e-06,
+      "loss": 0.5687,
+      "step": 1400
+    },
+    {
+      "epoch": 1.01768324518194,
+      "grad_norm": 0.4027839692661464,
+      "learning_rate": 9.324945426780879e-06,
+      "loss": 0.5788,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0184096429301068,
+      "grad_norm": 0.4215575711137664,
+      "learning_rate": 9.323747044899184e-06,
+      "loss": 0.5787,
+      "step": 1402
+    },
+    {
+      "epoch": 1.019136040678274,
+      "grad_norm": 0.4829986330178923,
+      "learning_rate": 9.322547677400391e-06,
+      "loss": 0.5848,
+      "step": 1403
+    },
+    {
+      "epoch": 1.019862438426441,
+      "grad_norm": 0.39452309057088686,
+      "learning_rate": 9.321347324557904e-06,
+      "loss": 0.5944,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0205888361746078,
+      "grad_norm": 0.4509442010918262,
+      "learning_rate": 9.320145986645348e-06,
+      "loss": 0.5865,
+      "step": 1405
+    },
+    {
+      "epoch": 1.021315233922775,
+      "grad_norm": 0.4257629876201213,
+      "learning_rate": 9.31894366393657e-06,
+      "loss": 0.5713,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0220416316709418,
+      "grad_norm": 0.48116645828094656,
+      "learning_rate": 9.317740356705649e-06,
+      "loss": 0.6006,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0227680294191088,
+      "grad_norm": 0.5090569406001516,
+      "learning_rate": 9.316536065226883e-06,
+      "loss": 0.5891,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0234944271672757,
+      "grad_norm": 0.5625989892238936,
+      "learning_rate": 9.315330789774795e-06,
+      "loss": 0.5852,
+      "step": 1409
+    },
+    {
+      "epoch": 1.0242208249154428,
+      "grad_norm": 1.4737104492769104,
+      "learning_rate": 9.31412453062413e-06,
+      "loss": 0.5987,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0249472226636098,
+      "grad_norm": 0.4422415845798024,
+      "learning_rate": 9.312917288049868e-06,
+      "loss": 0.5858,
+      "step": 1411
+    },
+    {
+      "epoch": 1.0256736204117767,
+      "grad_norm": 0.3877553271798556,
+      "learning_rate": 9.311709062327197e-06,
+      "loss": 0.5753,
+      "step": 1412
+    },
+    {
+      "epoch": 1.0264000181599438,
+      "grad_norm": 0.5253155430426784,
+      "learning_rate": 9.31049985373154e-06,
+      "loss": 0.5866,
+      "step": 1413
+    },
+    {
+      "epoch": 1.0271264159081106,
+      "grad_norm": 0.3967005737441752,
+      "learning_rate": 9.309289662538543e-06,
+      "loss": 0.5605,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0278528136562777,
+      "grad_norm": 0.429486431143533,
+      "learning_rate": 9.308078489024072e-06,
+      "loss": 0.5609,
+      "step": 1415
+    },
+    {
+      "epoch": 1.0285792114044447,
+      "grad_norm": 0.4650393072669976,
+      "learning_rate": 9.306866333464223e-06,
+      "loss": 0.5597,
+      "step": 1416
+    },
+    {
+      "epoch": 1.0293056091526116,
+      "grad_norm": 0.41440859115144746,
+      "learning_rate": 9.305653196135308e-06,
+      "loss": 0.6004,
+      "step": 1417
+    },
+    {
+      "epoch": 1.0300320069007787,
+      "grad_norm": 0.41388563070702816,
+      "learning_rate": 9.30443907731387e-06,
+      "loss": 0.5823,
+      "step": 1418
+    },
+    {
+      "epoch": 1.0307584046489455,
+      "grad_norm": 0.43655904948926433,
+      "learning_rate": 9.303223977276669e-06,
+      "loss": 0.568,
+      "step": 1419
+    },
+    {
+      "epoch": 1.0314848023971126,
+      "grad_norm": 0.3926843050634251,
+      "learning_rate": 9.302007896300697e-06,
+      "loss": 0.5635,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0322112001452795,
+      "grad_norm": 0.42957369993443373,
+      "learning_rate": 9.300790834663163e-06,
+      "loss": 0.5746,
+      "step": 1421
+    },
+    {
+      "epoch": 1.0329375978934465,
+      "grad_norm": 0.4035745493692686,
+      "learning_rate": 9.2995727926415e-06,
+      "loss": 0.5862,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0336639956416136,
+      "grad_norm": 0.3920427008419327,
+      "learning_rate": 9.298353770513368e-06,
+      "loss": 0.5663,
+      "step": 1423
+    },
+    {
+      "epoch": 1.0343903933897804,
+      "grad_norm": 0.46057274244672136,
+      "learning_rate": 9.297133768556649e-06,
+      "loss": 0.5812,
+      "step": 1424
+    },
+    {
+      "epoch": 1.0351167911379475,
+      "grad_norm": 0.5253446644849149,
+      "learning_rate": 9.295912787049447e-06,
+      "loss": 0.5916,
+      "step": 1425
+    },
+    {
+      "epoch": 1.0358431888861144,
+      "grad_norm": 0.4397820455055968,
+      "learning_rate": 9.294690826270086e-06,
+      "loss": 0.5773,
+      "step": 1426
+    },
+    {
+      "epoch": 1.0365695866342814,
+      "grad_norm": 0.3794802433265421,
+      "learning_rate": 9.293467886497123e-06,
+      "loss": 0.5638,
+      "step": 1427
+    },
+    {
+      "epoch": 1.0372959843824485,
+      "grad_norm": 0.39267603669754747,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.5825,
+      "step": 1428
+    },
+    {
+      "epoch": 1.0380223821306154,
+      "grad_norm": 0.39438762623612705,
+      "learning_rate": 9.291019071085707e-06,
+      "loss": 0.5683,
+      "step": 1429
+    },
+    {
+      "epoch": 1.0387487798787824,
+      "grad_norm": 0.4124722400201054,
+      "learning_rate": 9.289793196005472e-06,
+      "loss": 0.5887,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0394751776269493,
+      "grad_norm": 0.3710351913452333,
+      "learning_rate": 9.288566343048069e-06,
+      "loss": 0.5617,
+      "step": 1431
+    },
+    {
+      "epoch": 1.0402015753751164,
+      "grad_norm": 0.3696840977628092,
+      "learning_rate": 9.287338512493165e-06,
+      "loss": 0.5713,
+      "step": 1432
+    },
+    {
+      "epoch": 1.0409279731232832,
+      "grad_norm": 0.4711524354304988,
+      "learning_rate": 9.286109704620648e-06,
+      "loss": 0.5703,
+      "step": 1433
+    },
+    {
+      "epoch": 1.0416543708714503,
+      "grad_norm": 0.40615010982081595,
+      "learning_rate": 9.284879919710631e-06,
+      "loss": 0.564,
+      "step": 1434
+    },
+    {
+      "epoch": 1.0423807686196174,
+      "grad_norm": 0.41109169173341603,
+      "learning_rate": 9.283649158043452e-06,
+      "loss": 0.5732,
+      "step": 1435
+    },
+    {
+      "epoch": 1.0431071663677842,
+      "grad_norm": 0.44915827468482616,
+      "learning_rate": 9.282417419899664e-06,
+      "loss": 0.592,
+      "step": 1436
+    },
+    {
+      "epoch": 1.0438335641159513,
+      "grad_norm": 0.36242059051621556,
+      "learning_rate": 9.281184705560049e-06,
+      "loss": 0.575,
+      "step": 1437
+    },
+    {
+      "epoch": 1.0445599618641181,
+      "grad_norm": 0.3951607612016366,
+      "learning_rate": 9.279951015305611e-06,
+      "loss": 0.5629,
+      "step": 1438
+    },
+    {
+      "epoch": 1.0452863596122852,
+      "grad_norm": 0.42649878376009126,
+      "learning_rate": 9.27871634941757e-06,
+      "loss": 0.5697,
+      "step": 1439
+    },
+    {
+      "epoch": 1.0460127573604523,
+      "grad_norm": 0.5203166547135417,
+      "learning_rate": 9.27748070817738e-06,
+      "loss": 0.5837,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0467391551086191,
+      "grad_norm": 0.3776646098453339,
+      "learning_rate": 9.276244091866706e-06,
+      "loss": 0.5588,
+      "step": 1441
+    },
+    {
+      "epoch": 1.0474655528567862,
+      "grad_norm": 0.44599908364573615,
+      "learning_rate": 9.275006500767444e-06,
+      "loss": 0.5693,
+      "step": 1442
+    },
+    {
+      "epoch": 1.048191950604953,
+      "grad_norm": 0.38020056144781483,
+      "learning_rate": 9.273767935161705e-06,
+      "loss": 0.5741,
+      "step": 1443
+    },
+    {
+      "epoch": 1.0489183483531201,
+      "grad_norm": 0.4268767302486434,
+      "learning_rate": 9.272528395331826e-06,
+      "loss": 0.5732,
+      "step": 1444
+    },
+    {
+      "epoch": 1.049644746101287,
+      "grad_norm": 0.40829365584244665,
+      "learning_rate": 9.271287881560368e-06,
+      "loss": 0.5584,
+      "step": 1445
+    },
+    {
+      "epoch": 1.050371143849454,
+      "grad_norm": 0.39136192636878153,
+      "learning_rate": 9.27004639413011e-06,
+      "loss": 0.5784,
+      "step": 1446
+    },
+    {
+      "epoch": 1.0510975415976211,
+      "grad_norm": 0.39740978287858214,
+      "learning_rate": 9.268803933324053e-06,
+      "loss": 0.5862,
+      "step": 1447
+    },
+    {
+      "epoch": 1.051823939345788,
+      "grad_norm": 0.42772270593881806,
+      "learning_rate": 9.267560499425425e-06,
+      "loss": 0.5712,
+      "step": 1448
+    },
+    {
+      "epoch": 1.052550337093955,
+      "grad_norm": 0.4015339188340092,
+      "learning_rate": 9.266316092717666e-06,
+      "loss": 0.5804,
+      "step": 1449
+    },
+    {
+      "epoch": 1.053276734842122,
+      "grad_norm": 0.4721614746960808,
+      "learning_rate": 9.265070713484452e-06,
+      "loss": 0.5752,
+      "step": 1450
+    },
+    {
+      "epoch": 1.054003132590289,
+      "grad_norm": 0.46272828478575495,
+      "learning_rate": 9.263824362009667e-06,
+      "loss": 0.5725,
+      "step": 1451
+    },
+    {
+      "epoch": 1.0547295303384558,
+      "grad_norm": 0.43524012594081773,
+      "learning_rate": 9.262577038577423e-06,
+      "loss": 0.5745,
+      "step": 1452
+    },
+    {
+      "epoch": 1.055455928086623,
+      "grad_norm": 0.3697781664579044,
+      "learning_rate": 9.261328743472055e-06,
+      "loss": 0.5661,
+      "step": 1453
+    },
+    {
+      "epoch": 1.05618232583479,
+      "grad_norm": 0.44617814494960756,
+      "learning_rate": 9.260079476978117e-06,
+      "loss": 0.5852,
+      "step": 1454
+    },
+    {
+      "epoch": 1.0569087235829568,
+      "grad_norm": 0.4019452947154167,
+      "learning_rate": 9.25882923938038e-06,
+      "loss": 0.5939,
+      "step": 1455
+    },
+    {
+      "epoch": 1.057635121331124,
+      "grad_norm": 0.4191763655875106,
+      "learning_rate": 9.25757803096385e-06,
+      "loss": 0.5748,
+      "step": 1456
+    },
+    {
+      "epoch": 1.0583615190792908,
+      "grad_norm": 0.4126101858730787,
+      "learning_rate": 9.256325852013736e-06,
+      "loss": 0.5835,
+      "step": 1457
+    },
+    {
+      "epoch": 1.0590879168274578,
+      "grad_norm": 0.4339550235322551,
+      "learning_rate": 9.255072702815487e-06,
+      "loss": 0.5872,
+      "step": 1458
+    },
+    {
+      "epoch": 1.059814314575625,
+      "grad_norm": 0.388417450620657,
+      "learning_rate": 9.253818583654754e-06,
+      "loss": 0.5659,
+      "step": 1459
+    },
+    {
+      "epoch": 1.0605407123237918,
+      "grad_norm": 0.4540043570449997,
+      "learning_rate": 9.252563494817426e-06,
+      "loss": 0.5756,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0612671100719588,
+      "grad_norm": 0.38753097423844807,
+      "learning_rate": 9.251307436589605e-06,
+      "loss": 0.5716,
+      "step": 1461
+    },
+    {
+      "epoch": 1.0619935078201257,
+      "grad_norm": 0.37777144670282375,
+      "learning_rate": 9.250050409257612e-06,
+      "loss": 0.5586,
+      "step": 1462
+    },
+    {
+      "epoch": 1.0627199055682928,
+      "grad_norm": 0.445355054898785,
+      "learning_rate": 9.248792413107994e-06,
+      "loss": 0.5707,
+      "step": 1463
+    },
+    {
+      "epoch": 1.0634463033164596,
+      "grad_norm": 0.43309925235084556,
+      "learning_rate": 9.247533448427519e-06,
+      "loss": 0.5857,
+      "step": 1464
+    },
+    {
+      "epoch": 1.0641727010646267,
+      "grad_norm": 0.4464319853779842,
+      "learning_rate": 9.246273515503169e-06,
+      "loss": 0.5802,
+      "step": 1465
+    },
+    {
+      "epoch": 1.0648990988127938,
+      "grad_norm": 0.3924050059125471,
+      "learning_rate": 9.245012614622153e-06,
+      "loss": 0.5706,
+      "step": 1466
+    },
+    {
+      "epoch": 1.0656254965609606,
+      "grad_norm": 0.40147856573132024,
+      "learning_rate": 9.2437507460719e-06,
+      "loss": 0.5706,
+      "step": 1467
+    },
+    {
+      "epoch": 1.0663518943091277,
+      "grad_norm": 0.4001433161155672,
+      "learning_rate": 9.242487910140057e-06,
+      "loss": 0.5548,
+      "step": 1468
+    },
+    {
+      "epoch": 1.0670782920572945,
+      "grad_norm": 0.4079874843159747,
+      "learning_rate": 9.241224107114495e-06,
+      "loss": 0.5635,
+      "step": 1469
+    },
+    {
+      "epoch": 1.0678046898054616,
+      "grad_norm": 0.4560119919834794,
+      "learning_rate": 9.239959337283301e-06,
+      "loss": 0.5723,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0685310875536285,
+      "grad_norm": 0.4407800440761631,
+      "learning_rate": 9.238693600934785e-06,
+      "loss": 0.5775,
+      "step": 1471
+    },
+    {
+      "epoch": 1.0692574853017955,
+      "grad_norm": 0.4840882494667328,
+      "learning_rate": 9.23742689835748e-06,
+      "loss": 0.5754,
+      "step": 1472
+    },
+    {
+      "epoch": 1.0699838830499626,
+      "grad_norm": 0.37810093243018317,
+      "learning_rate": 9.236159229840133e-06,
+      "loss": 0.5566,
+      "step": 1473
+    },
+    {
+      "epoch": 1.0707102807981295,
+      "grad_norm": 0.3863492962548962,
+      "learning_rate": 9.234890595671717e-06,
+      "loss": 0.565,
+      "step": 1474
+    },
+    {
+      "epoch": 1.0714366785462965,
+      "grad_norm": 0.7615309489873513,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.5598,
+      "step": 1475
+    },
+    {
+      "epoch": 1.0721630762944634,
+      "grad_norm": 0.3976142210474501,
+      "learning_rate": 9.232350431538656e-06,
+      "loss": 0.568,
+      "step": 1476
+    },
+    {
+      "epoch": 1.0728894740426305,
+      "grad_norm": 0.3850052436582411,
+      "learning_rate": 9.231078902153058e-06,
+      "loss": 0.5587,
+      "step": 1477
+    },
+    {
+      "epoch": 1.0736158717907975,
+      "grad_norm": 0.7063675394401735,
+      "learning_rate": 9.229806408274467e-06,
+      "loss": 0.5527,
+      "step": 1478
+    },
+    {
+      "epoch": 1.0743422695389644,
+      "grad_norm": 0.4258570084193788,
+      "learning_rate": 9.228532950192964e-06,
+      "loss": 0.5688,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0750686672871315,
+      "grad_norm": 0.41356428724432126,
+      "learning_rate": 9.227258528198832e-06,
+      "loss": 0.5764,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0757950650352983,
+      "grad_norm": 0.3686752782723383,
+      "learning_rate": 9.225983142582584e-06,
+      "loss": 0.5571,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0765214627834654,
+      "grad_norm": 0.38160042743844014,
+      "learning_rate": 9.224706793634951e-06,
+      "loss": 0.5604,
+      "step": 1482
+    },
+    {
+      "epoch": 1.0772478605316325,
+      "grad_norm": 0.3707774059356207,
+      "learning_rate": 9.223429481646881e-06,
+      "loss": 0.581,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0779742582797993,
+      "grad_norm": 0.3695515047035739,
+      "learning_rate": 9.222151206909541e-06,
+      "loss": 0.5498,
+      "step": 1484
+    },
+    {
+      "epoch": 1.0787006560279664,
+      "grad_norm": 0.3831609762559241,
+      "learning_rate": 9.220871969714325e-06,
+      "loss": 0.5715,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0794270537761332,
+      "grad_norm": 0.36103940460289247,
+      "learning_rate": 9.219591770352836e-06,
+      "loss": 0.5713,
+      "step": 1486
+    },
+    {
+      "epoch": 1.0801534515243003,
+      "grad_norm": 0.4427569819582544,
+      "learning_rate": 9.2183106091169e-06,
+      "loss": 0.5873,
+      "step": 1487
+    },
+    {
+      "epoch": 1.0808798492724672,
+      "grad_norm": 0.45995302001148347,
+      "learning_rate": 9.217028486298567e-06,
+      "loss": 0.5714,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0816062470206342,
+      "grad_norm": 0.4569865672570247,
+      "learning_rate": 9.2157454021901e-06,
+      "loss": 0.5734,
+      "step": 1489
+    },
+    {
+      "epoch": 1.0823326447688013,
+      "grad_norm": 0.4061969443898093,
+      "learning_rate": 9.214461357083986e-06,
+      "loss": 0.5611,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0830590425169682,
+      "grad_norm": 0.36787594074816127,
+      "learning_rate": 9.213176351272926e-06,
+      "loss": 0.5771,
+      "step": 1491
+    },
+    {
+      "epoch": 1.0837854402651352,
+      "grad_norm": 0.5034943262793765,
+      "learning_rate": 9.211890385049845e-06,
+      "loss": 0.5707,
+      "step": 1492
+    },
+    {
+      "epoch": 1.084511838013302,
+      "grad_norm": 0.38467957622459686,
+      "learning_rate": 9.210603458707883e-06,
+      "loss": 0.5697,
+      "step": 1493
+    },
+    {
+      "epoch": 1.0852382357614692,
+      "grad_norm": 0.42418759374955084,
+      "learning_rate": 9.209315572540402e-06,
+      "loss": 0.5625,
+      "step": 1494
+    },
+    {
+      "epoch": 1.085964633509636,
+      "grad_norm": 0.3998370227866447,
+      "learning_rate": 9.20802672684098e-06,
+      "loss": 0.5813,
+      "step": 1495
+    },
+    {
+      "epoch": 1.086691031257803,
+      "grad_norm": 0.42451732990267654,
+      "learning_rate": 9.206736921903414e-06,
+      "loss": 0.5759,
+      "step": 1496
+    },
+    {
+      "epoch": 1.0874174290059702,
+      "grad_norm": 0.5284676685659139,
+      "learning_rate": 9.205446158021725e-06,
+      "loss": 0.5638,
+      "step": 1497
+    },
+    {
+      "epoch": 1.088143826754137,
+      "grad_norm": 0.4213863049015563,
+      "learning_rate": 9.204154435490143e-06,
+      "loss": 0.5794,
+      "step": 1498
+    },
+    {
+      "epoch": 1.088870224502304,
+      "grad_norm": 0.4080295627251175,
+      "learning_rate": 9.202861754603126e-06,
+      "loss": 0.5751,
+      "step": 1499
+    },
+    {
+      "epoch": 1.089596622250471,
+      "grad_norm": 0.39291554245061283,
+      "learning_rate": 9.201568115655343e-06,
+      "loss": 0.556,
+      "step": 1500
+    },
+    {
+      "epoch": 1.090323019998638,
+      "grad_norm": 0.3740570835089451,
+      "learning_rate": 9.200273518941688e-06,
+      "loss": 0.5753,
+      "step": 1501
+    },
+    {
+      "epoch": 1.091049417746805,
+      "grad_norm": 0.407499502900159,
+      "learning_rate": 9.198977964757266e-06,
+      "loss": 0.5729,
+      "step": 1502
+    },
+    {
+      "epoch": 1.091775815494972,
+      "grad_norm": 0.3719259405007259,
+      "learning_rate": 9.197681453397407e-06,
+      "loss": 0.5627,
+      "step": 1503
+    },
+    {
+      "epoch": 1.092502213243139,
+      "grad_norm": 0.4068817249799283,
+      "learning_rate": 9.196383985157657e-06,
+      "loss": 0.5682,
+      "step": 1504
+    },
+    {
+      "epoch": 1.0932286109913059,
+      "grad_norm": 0.3822730777415677,
+      "learning_rate": 9.195085560333777e-06,
+      "loss": 0.5613,
+      "step": 1505
+    },
+    {
+      "epoch": 1.093955008739473,
+      "grad_norm": 0.509101461961365,
+      "learning_rate": 9.193786179221751e-06,
+      "loss": 0.5821,
+      "step": 1506
+    },
+    {
+      "epoch": 1.0946814064876398,
+      "grad_norm": 0.36195978677680385,
+      "learning_rate": 9.192485842117777e-06,
+      "loss": 0.5572,
+      "step": 1507
+    },
+    {
+      "epoch": 1.0954078042358069,
+      "grad_norm": 0.43123149253343235,
+      "learning_rate": 9.191184549318275e-06,
+      "loss": 0.5676,
+      "step": 1508
+    },
+    {
+      "epoch": 1.096134201983974,
+      "grad_norm": 0.39497568720609605,
+      "learning_rate": 9.189882301119878e-06,
+      "loss": 0.5662,
+      "step": 1509
+    },
+    {
+      "epoch": 1.0968605997321408,
+      "grad_norm": 0.392750640761948,
+      "learning_rate": 9.188579097819439e-06,
+      "loss": 0.5673,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0975869974803079,
+      "grad_norm": 0.39765453193046085,
+      "learning_rate": 9.18727493971403e-06,
+      "loss": 0.5867,
+      "step": 1511
+    },
+    {
+      "epoch": 1.0983133952284747,
+      "grad_norm": 0.4802452358360034,
+      "learning_rate": 9.18596982710094e-06,
+      "loss": 0.5591,
+      "step": 1512
+    },
+    {
+      "epoch": 1.0990397929766418,
+      "grad_norm": 0.3956359994045311,
+      "learning_rate": 9.184663760277674e-06,
+      "loss": 0.5708,
+      "step": 1513
+    },
+    {
+      "epoch": 1.0997661907248089,
+      "grad_norm": 0.4135810979273268,
+      "learning_rate": 9.183356739541958e-06,
+      "loss": 0.5706,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1004925884729757,
+      "grad_norm": 0.38969981003076315,
+      "learning_rate": 9.182048765191729e-06,
+      "loss": 0.5751,
+      "step": 1515
+    },
+    {
+      "epoch": 1.1012189862211428,
+      "grad_norm": 0.4182089910086,
+      "learning_rate": 9.180739837525148e-06,
+      "loss": 0.5545,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1019453839693096,
+      "grad_norm": 0.42615730462223866,
+      "learning_rate": 9.17942995684059e-06,
+      "loss": 0.5654,
+      "step": 1517
+    },
+    {
+      "epoch": 1.1026717817174767,
+      "grad_norm": 0.38455182758076384,
+      "learning_rate": 9.178119123436651e-06,
+      "loss": 0.576,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1033981794656436,
+      "grad_norm": 0.4081208032550898,
+      "learning_rate": 9.176807337612136e-06,
+      "loss": 0.5673,
+      "step": 1519
+    },
+    {
+      "epoch": 1.1041245772138106,
+      "grad_norm": 0.43376128159499927,
+      "learning_rate": 9.175494599666078e-06,
+      "loss": 0.5721,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1048509749619777,
+      "grad_norm": 0.40332872980030554,
+      "learning_rate": 9.174180909897715e-06,
+      "loss": 0.5687,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1055773727101446,
+      "grad_norm": 0.38223586760682,
+      "learning_rate": 9.172866268606514e-06,
+      "loss": 0.5702,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1063037704583116,
+      "grad_norm": 0.40924832268015005,
+      "learning_rate": 9.171550676092151e-06,
+      "loss": 0.5802,
+      "step": 1523
+    },
+    {
+      "epoch": 1.1070301682064785,
+      "grad_norm": 0.4228837477832897,
+      "learning_rate": 9.170234132654521e-06,
+      "loss": 0.5779,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1077565659546456,
+      "grad_norm": 0.43068444275154966,
+      "learning_rate": 9.168916638593736e-06,
+      "loss": 0.5733,
+      "step": 1525
+    },
+    {
+      "epoch": 1.1084829637028126,
+      "grad_norm": 0.4368955020773527,
+      "learning_rate": 9.167598194210124e-06,
+      "loss": 0.5637,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1092093614509795,
+      "grad_norm": 0.46106898550934694,
+      "learning_rate": 9.166278799804232e-06,
+      "loss": 0.5621,
+      "step": 1527
+    },
+    {
+      "epoch": 1.1099357591991466,
+      "grad_norm": 0.4998543301741803,
+      "learning_rate": 9.164958455676818e-06,
+      "loss": 0.5738,
+      "step": 1528
+    },
+    {
+      "epoch": 1.1106621569473134,
+      "grad_norm": 0.3770729245790208,
+      "learning_rate": 9.163637162128865e-06,
+      "loss": 0.5618,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1113885546954805,
+      "grad_norm": 0.49041521965602436,
+      "learning_rate": 9.162314919461566e-06,
+      "loss": 0.5533,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1121149524436473,
+      "grad_norm": 0.3900108576009805,
+      "learning_rate": 9.16099172797633e-06,
+      "loss": 0.5529,
+      "step": 1531
+    },
+    {
+      "epoch": 1.1128413501918144,
+      "grad_norm": 0.37032433949378185,
+      "learning_rate": 9.159667587974786e-06,
+      "loss": 0.5696,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1135677479399815,
+      "grad_norm": 0.42684873849663624,
+      "learning_rate": 9.158342499758777e-06,
+      "loss": 0.5434,
+      "step": 1533
+    },
+    {
+      "epoch": 1.1142941456881483,
+      "grad_norm": 0.43643707876206966,
+      "learning_rate": 9.157016463630363e-06,
+      "loss": 0.5706,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1150205434363154,
+      "grad_norm": 0.4454221836123075,
+      "learning_rate": 9.15568947989182e-06,
+      "loss": 0.5708,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1157469411844823,
+      "grad_norm": 0.38248652147035744,
+      "learning_rate": 9.154361548845639e-06,
+      "loss": 0.5571,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1164733389326493,
+      "grad_norm": 0.39988857054600924,
+      "learning_rate": 9.153032670794527e-06,
+      "loss": 0.5705,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1171997366808162,
+      "grad_norm": 0.4055927377484529,
+      "learning_rate": 9.15170284604141e-06,
+      "loss": 0.5589,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1179261344289833,
+      "grad_norm": 0.6645482334421663,
+      "learning_rate": 9.150372074889427e-06,
+      "loss": 0.5665,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1186525321771503,
+      "grad_norm": 0.37649582235674955,
+      "learning_rate": 9.14904035764193e-06,
+      "loss": 0.5655,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1193789299253172,
+      "grad_norm": 0.456135150501283,
+      "learning_rate": 9.147707694602492e-06,
+      "loss": 0.575,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1201053276734843,
+      "grad_norm": 0.4340819757352092,
+      "learning_rate": 9.1463740860749e-06,
+      "loss": 0.5541,
+      "step": 1542
+    },
+    {
+      "epoch": 1.120831725421651,
+      "grad_norm": 0.5154115608102522,
+      "learning_rate": 9.145039532363156e-06,
+      "loss": 0.5728,
+      "step": 1543
+    },
+    {
+      "epoch": 1.1215581231698182,
+      "grad_norm": 0.415248125939544,
+      "learning_rate": 9.143704033771476e-06,
+      "loss": 0.573,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1222845209179853,
+      "grad_norm": 0.47571569465415514,
+      "learning_rate": 9.142367590604294e-06,
+      "loss": 0.5724,
+      "step": 1545
+    },
+    {
+      "epoch": 1.123010918666152,
+      "grad_norm": 0.4328702166226174,
+      "learning_rate": 9.141030203166256e-06,
+      "loss": 0.5881,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1237373164143192,
+      "grad_norm": 0.4029536592492071,
+      "learning_rate": 9.139691871762229e-06,
+      "loss": 0.5569,
+      "step": 1547
+    },
+    {
+      "epoch": 1.124463714162486,
+      "grad_norm": 0.4467378583250728,
+      "learning_rate": 9.138352596697287e-06,
+      "loss": 0.5776,
+      "step": 1548
+    },
+    {
+      "epoch": 1.125190111910653,
+      "grad_norm": 0.41357552315372287,
+      "learning_rate": 9.137012378276729e-06,
+      "loss": 0.552,
+      "step": 1549
+    },
+    {
+      "epoch": 1.1259165096588202,
+      "grad_norm": 0.3961970064201352,
+      "learning_rate": 9.135671216806057e-06,
+      "loss": 0.542,
+      "step": 1550
+    },
+    {
+      "epoch": 1.126642907406987,
+      "grad_norm": 0.3799347314928259,
+      "learning_rate": 9.134329112591e-06,
+      "loss": 0.5783,
+      "step": 1551
+    },
+    {
+      "epoch": 1.127369305155154,
+      "grad_norm": 0.4409962185466233,
+      "learning_rate": 9.132986065937495e-06,
+      "loss": 0.5833,
+      "step": 1552
+    },
+    {
+      "epoch": 1.128095702903321,
+      "grad_norm": 0.4199431422304019,
+      "learning_rate": 9.131642077151695e-06,
+      "loss": 0.5544,
+      "step": 1553
+    },
+    {
+      "epoch": 1.128822100651488,
+      "grad_norm": 0.4359092717576693,
+      "learning_rate": 9.130297146539967e-06,
+      "loss": 0.5601,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1295484983996549,
+      "grad_norm": 0.4386918332276292,
+      "learning_rate": 9.128951274408898e-06,
+      "loss": 0.5779,
+      "step": 1555
+    },
+    {
+      "epoch": 1.130274896147822,
+      "grad_norm": 0.43388198361205227,
+      "learning_rate": 9.12760446106528e-06,
+      "loss": 0.5601,
+      "step": 1556
+    },
+    {
+      "epoch": 1.1310012938959888,
+      "grad_norm": 0.4020892074156709,
+      "learning_rate": 9.126256706816129e-06,
+      "loss": 0.558,
+      "step": 1557
+    },
+    {
+      "epoch": 1.1317276916441559,
+      "grad_norm": 0.3612712750947298,
+      "learning_rate": 9.124908011968667e-06,
+      "loss": 0.5615,
+      "step": 1558
+    },
+    {
+      "epoch": 1.132454089392323,
+      "grad_norm": 0.4384818833438229,
+      "learning_rate": 9.123558376830342e-06,
+      "loss": 0.5683,
+      "step": 1559
+    },
+    {
+      "epoch": 1.1331804871404898,
+      "grad_norm": 0.4018494187243776,
+      "learning_rate": 9.122207801708802e-06,
+      "loss": 0.5727,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1339068848886569,
+      "grad_norm": 0.40446557479401735,
+      "learning_rate": 9.120856286911922e-06,
+      "loss": 0.5686,
+      "step": 1561
+    },
+    {
+      "epoch": 1.1346332826368237,
+      "grad_norm": 0.7222820279604052,
+      "learning_rate": 9.119503832747782e-06,
+      "loss": 0.5595,
+      "step": 1562
+    },
+    {
+      "epoch": 1.1353596803849908,
+      "grad_norm": 0.4118173443641317,
+      "learning_rate": 9.118150439524682e-06,
+      "loss": 0.5591,
+      "step": 1563
+    },
+    {
+      "epoch": 1.1360860781331579,
+      "grad_norm": 0.39203921438913475,
+      "learning_rate": 9.116796107551134e-06,
+      "loss": 0.5664,
+      "step": 1564
+    },
+    {
+      "epoch": 1.1368124758813247,
+      "grad_norm": 0.44213420522319913,
+      "learning_rate": 9.115440837135862e-06,
+      "loss": 0.5497,
+      "step": 1565
+    },
+    {
+      "epoch": 1.1375388736294918,
+      "grad_norm": 0.3915843971752154,
+      "learning_rate": 9.114084628587806e-06,
+      "loss": 0.5743,
+      "step": 1566
+    },
+    {
+      "epoch": 1.1382652713776586,
+      "grad_norm": 0.37457500840606167,
+      "learning_rate": 9.112727482216123e-06,
+      "loss": 0.5645,
+      "step": 1567
+    },
+    {
+      "epoch": 1.1389916691258257,
+      "grad_norm": 0.39111251152334303,
+      "learning_rate": 9.111369398330177e-06,
+      "loss": 0.5793,
+      "step": 1568
+    },
+    {
+      "epoch": 1.1397180668739928,
+      "grad_norm": 0.5449219416546525,
+      "learning_rate": 9.110010377239552e-06,
+      "loss": 0.582,
+      "step": 1569
+    },
+    {
+      "epoch": 1.1404444646221596,
+      "grad_norm": 0.49691466965603404,
+      "learning_rate": 9.108650419254037e-06,
+      "loss": 0.5646,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1411708623703267,
+      "grad_norm": 0.40646649401243895,
+      "learning_rate": 9.107289524683648e-06,
+      "loss": 0.5645,
+      "step": 1571
+    },
+    {
+      "epoch": 1.1418972601184936,
+      "grad_norm": 0.39891945566604176,
+      "learning_rate": 9.105927693838601e-06,
+      "loss": 0.5642,
+      "step": 1572
+    },
+    {
+      "epoch": 1.1426236578666606,
+      "grad_norm": 0.4048822149119588,
+      "learning_rate": 9.104564927029337e-06,
+      "loss": 0.5652,
+      "step": 1573
+    },
+    {
+      "epoch": 1.1433500556148275,
+      "grad_norm": 0.45448169948196127,
+      "learning_rate": 9.103201224566499e-06,
+      "loss": 0.5704,
+      "step": 1574
+    },
+    {
+      "epoch": 1.1440764533629946,
+      "grad_norm": 0.3903073163448049,
+      "learning_rate": 9.101836586760951e-06,
+      "loss": 0.5751,
+      "step": 1575
+    },
+    {
+      "epoch": 1.1448028511111616,
+      "grad_norm": 0.5097478191145266,
+      "learning_rate": 9.10047101392377e-06,
+      "loss": 0.5623,
+      "step": 1576
+    },
+    {
+      "epoch": 1.1455292488593285,
+      "grad_norm": 0.5564016875040064,
+      "learning_rate": 9.099104506366242e-06,
+      "loss": 0.5735,
+      "step": 1577
+    },
+    {
+      "epoch": 1.1462556466074956,
+      "grad_norm": 0.45792127134281246,
+      "learning_rate": 9.09773706439987e-06,
+      "loss": 0.562,
+      "step": 1578
+    },
+    {
+      "epoch": 1.1469820443556624,
+      "grad_norm": 0.40573516591270037,
+      "learning_rate": 9.096368688336365e-06,
+      "loss": 0.5712,
+      "step": 1579
+    },
+    {
+      "epoch": 1.1477084421038295,
+      "grad_norm": 0.3746216254001378,
+      "learning_rate": 9.094999378487659e-06,
+      "loss": 0.565,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1484348398519963,
+      "grad_norm": 0.4132489165843512,
+      "learning_rate": 9.09362913516589e-06,
+      "loss": 0.5611,
+      "step": 1581
+    },
+    {
+      "epoch": 1.1491612376001634,
+      "grad_norm": 0.4327030952311928,
+      "learning_rate": 9.092257958683411e-06,
+      "loss": 0.572,
+      "step": 1582
+    },
+    {
+      "epoch": 1.1498876353483305,
+      "grad_norm": 0.5657636997480442,
+      "learning_rate": 9.090885849352788e-06,
+      "loss": 0.5725,
+      "step": 1583
+    },
+    {
+      "epoch": 1.1506140330964973,
+      "grad_norm": 0.43552805930495025,
+      "learning_rate": 9.0895128074868e-06,
+      "loss": 0.5722,
+      "step": 1584
+    },
+    {
+      "epoch": 1.1513404308446644,
+      "grad_norm": 0.4021861507833854,
+      "learning_rate": 9.088138833398435e-06,
+      "loss": 0.5773,
+      "step": 1585
+    },
+    {
+      "epoch": 1.1520668285928313,
+      "grad_norm": 0.4019858854361604,
+      "learning_rate": 9.086763927400898e-06,
+      "loss": 0.538,
+      "step": 1586
+    },
+    {
+      "epoch": 1.1527932263409983,
+      "grad_norm": 0.5661378384808389,
+      "learning_rate": 9.085388089807607e-06,
+      "loss": 0.5519,
+      "step": 1587
+    },
+    {
+      "epoch": 1.1535196240891654,
+      "grad_norm": 0.3744204610801191,
+      "learning_rate": 9.08401132093219e-06,
+      "loss": 0.5452,
+      "step": 1588
+    },
+    {
+      "epoch": 1.1542460218373323,
+      "grad_norm": 0.6743822486879415,
+      "learning_rate": 9.082633621088483e-06,
+      "loss": 0.5664,
+      "step": 1589
+    },
+    {
+      "epoch": 1.1549724195854993,
+      "grad_norm": 0.4620844862314671,
+      "learning_rate": 9.081254990590542e-06,
+      "loss": 0.5443,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1556988173336662,
+      "grad_norm": 0.35175578314298,
+      "learning_rate": 9.079875429752633e-06,
+      "loss": 0.5412,
+      "step": 1591
+    },
+    {
+      "epoch": 1.1564252150818333,
+      "grad_norm": 0.391580626589242,
+      "learning_rate": 9.07849493888923e-06,
+      "loss": 0.5537,
+      "step": 1592
+    },
+    {
+      "epoch": 1.1571516128300003,
+      "grad_norm": 0.4529476753979372,
+      "learning_rate": 9.077113518315024e-06,
+      "loss": 0.5604,
+      "step": 1593
+    },
+    {
+      "epoch": 1.1578780105781672,
+      "grad_norm": 0.4550412254149466,
+      "learning_rate": 9.075731168344917e-06,
+      "loss": 0.5618,
+      "step": 1594
+    },
+    {
+      "epoch": 1.1586044083263343,
+      "grad_norm": 0.3809100704198691,
+      "learning_rate": 9.074347889294017e-06,
+      "loss": 0.5896,
+      "step": 1595
+    },
+    {
+      "epoch": 1.1593308060745011,
+      "grad_norm": 0.3905562748653443,
+      "learning_rate": 9.072963681477654e-06,
+      "loss": 0.5726,
+      "step": 1596
+    },
+    {
+      "epoch": 1.1600572038226682,
+      "grad_norm": 0.4091043456983746,
+      "learning_rate": 9.071578545211362e-06,
+      "loss": 0.5696,
+      "step": 1597
+    },
+    {
+      "epoch": 1.160783601570835,
+      "grad_norm": 0.4839681586665417,
+      "learning_rate": 9.070192480810888e-06,
+      "loss": 0.5621,
+      "step": 1598
+    },
+    {
+      "epoch": 1.1615099993190021,
+      "grad_norm": 0.40196070880566054,
+      "learning_rate": 9.068805488592191e-06,
+      "loss": 0.5779,
+      "step": 1599
+    },
+    {
+      "epoch": 1.162236397067169,
+      "grad_norm": 0.3918863091270513,
+      "learning_rate": 9.067417568871444e-06,
+      "loss": 0.5785,
+      "step": 1600
+    },
+    {
+      "epoch": 1.162962794815336,
+      "grad_norm": 0.36932354291748165,
+      "learning_rate": 9.06602872196503e-06,
+      "loss": 0.5688,
+      "step": 1601
+    },
+    {
+      "epoch": 1.1636891925635031,
+      "grad_norm": 0.511494494122197,
+      "learning_rate": 9.064638948189539e-06,
+      "loss": 0.5554,
+      "step": 1602
+    },
+    {
+      "epoch": 1.16441559031167,
+      "grad_norm": 0.37895248932231274,
+      "learning_rate": 9.06324824786178e-06,
+      "loss": 0.5575,
+      "step": 1603
+    },
+    {
+      "epoch": 1.165141988059837,
+      "grad_norm": 0.3832823773236108,
+      "learning_rate": 9.061856621298767e-06,
+      "loss": 0.5654,
+      "step": 1604
+    },
+    {
+      "epoch": 1.165868385808004,
+      "grad_norm": 0.43465330879731395,
+      "learning_rate": 9.060464068817728e-06,
+      "loss": 0.5738,
+      "step": 1605
+    },
+    {
+      "epoch": 1.166594783556171,
+      "grad_norm": 0.5551169241788557,
+      "learning_rate": 9.059070590736101e-06,
+      "loss": 0.5613,
+      "step": 1606
+    },
+    {
+      "epoch": 1.167321181304338,
+      "grad_norm": 0.5780876697750443,
+      "learning_rate": 9.057676187371536e-06,
+      "loss": 0.5462,
+      "step": 1607
+    },
+    {
+      "epoch": 1.168047579052505,
+      "grad_norm": 0.38976775310867845,
+      "learning_rate": 9.056280859041893e-06,
+      "loss": 0.5602,
+      "step": 1608
+    },
+    {
+      "epoch": 1.168773976800672,
+      "grad_norm": 0.4137511184661419,
+      "learning_rate": 9.054884606065243e-06,
+      "loss": 0.5359,
+      "step": 1609
+    },
+    {
+      "epoch": 1.1695003745488388,
+      "grad_norm": 0.4545114979291193,
+      "learning_rate": 9.053487428759869e-06,
+      "loss": 0.5566,
+      "step": 1610
+    },
+    {
+      "epoch": 1.170226772297006,
+      "grad_norm": 0.40448803287087315,
+      "learning_rate": 9.052089327444263e-06,
+      "loss": 0.5613,
+      "step": 1611
+    },
+    {
+      "epoch": 1.170953170045173,
+      "grad_norm": 0.4204853059368482,
+      "learning_rate": 9.050690302437128e-06,
+      "loss": 0.5506,
+      "step": 1612
+    },
+    {
+      "epoch": 1.1716795677933398,
+      "grad_norm": 0.41937616591275495,
+      "learning_rate": 9.049290354057379e-06,
+      "loss": 0.5584,
+      "step": 1613
+    },
+    {
+      "epoch": 1.172405965541507,
+      "grad_norm": 0.3812136298560557,
+      "learning_rate": 9.047889482624139e-06,
+      "loss": 0.5502,
+      "step": 1614
+    },
+    {
+      "epoch": 1.1731323632896737,
+      "grad_norm": 0.42863197566192074,
+      "learning_rate": 9.046487688456745e-06,
+      "loss": 0.5717,
+      "step": 1615
+    },
+    {
+      "epoch": 1.1738587610378408,
+      "grad_norm": 0.4416572927164445,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.5629,
+      "step": 1616
+    },
+    {
+      "epoch": 1.1745851587860079,
+      "grad_norm": 0.4318909760541322,
+      "learning_rate": 9.043681333197878e-06,
+      "loss": 0.5616,
+      "step": 1617
+    },
+    {
+      "epoch": 1.1753115565341747,
+      "grad_norm": 0.4913993162820547,
+      "learning_rate": 9.042276772746127e-06,
+      "loss": 0.5555,
+      "step": 1618
+    },
+    {
+      "epoch": 1.1760379542823418,
+      "grad_norm": 0.38866225966344226,
+      "learning_rate": 9.040871290839663e-06,
+      "loss": 0.5804,
+      "step": 1619
+    },
+    {
+      "epoch": 1.1767643520305087,
+      "grad_norm": 0.4473064645504217,
+      "learning_rate": 9.03946488779887e-06,
+      "loss": 0.558,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1774907497786757,
+      "grad_norm": 0.4002867177138744,
+      "learning_rate": 9.038057563944346e-06,
+      "loss": 0.5563,
+      "step": 1621
+    },
+    {
+      "epoch": 1.1782171475268426,
+      "grad_norm": 0.3619082015242835,
+      "learning_rate": 9.036649319596895e-06,
+      "loss": 0.5704,
+      "step": 1622
+    },
+    {
+      "epoch": 1.1789435452750097,
+      "grad_norm": 0.42685374009998284,
+      "learning_rate": 9.035240155077532e-06,
+      "loss": 0.572,
+      "step": 1623
+    },
+    {
+      "epoch": 1.1796699430231765,
+      "grad_norm": 0.38283761766805846,
+      "learning_rate": 9.033830070707485e-06,
+      "loss": 0.5586,
+      "step": 1624
+    },
+    {
+      "epoch": 1.1803963407713436,
+      "grad_norm": 0.5072262281997689,
+      "learning_rate": 9.032419066808184e-06,
+      "loss": 0.5596,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1811227385195107,
+      "grad_norm": 0.47901385507935723,
+      "learning_rate": 9.03100714370128e-06,
+      "loss": 0.5742,
+      "step": 1626
+    },
+    {
+      "epoch": 1.1818491362676775,
+      "grad_norm": 0.4621432294903584,
+      "learning_rate": 9.029594301708622e-06,
+      "loss": 0.559,
+      "step": 1627
+    },
+    {
+      "epoch": 1.1825755340158446,
+      "grad_norm": 0.35661902505888965,
+      "learning_rate": 9.028180541152275e-06,
+      "loss": 0.5427,
+      "step": 1628
+    },
+    {
+      "epoch": 1.1833019317640114,
+      "grad_norm": 0.39979814096373295,
+      "learning_rate": 9.026765862354512e-06,
+      "loss": 0.5621,
+      "step": 1629
+    },
+    {
+      "epoch": 1.1840283295121785,
+      "grad_norm": 0.6290392164977799,
+      "learning_rate": 9.025350265637816e-06,
+      "loss": 0.5631,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1847547272603456,
+      "grad_norm": 0.41675307891525165,
+      "learning_rate": 9.023933751324876e-06,
+      "loss": 0.56,
+      "step": 1631
+    },
+    {
+      "epoch": 1.1854811250085124,
+      "grad_norm": 0.40915617471251176,
+      "learning_rate": 9.022516319738598e-06,
+      "loss": 0.568,
+      "step": 1632
+    },
+    {
+      "epoch": 1.1862075227566795,
+      "grad_norm": 0.36057322747830917,
+      "learning_rate": 9.021097971202085e-06,
+      "loss": 0.548,
+      "step": 1633
+    },
+    {
+      "epoch": 1.1869339205048464,
+      "grad_norm": 0.398305816253801,
+      "learning_rate": 9.01967870603866e-06,
+      "loss": 0.5648,
+      "step": 1634
+    },
+    {
+      "epoch": 1.1876603182530134,
+      "grad_norm": 0.40047400844909886,
+      "learning_rate": 9.018258524571848e-06,
+      "loss": 0.5648,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1883867160011805,
+      "grad_norm": 0.36446634860953764,
+      "learning_rate": 9.016837427125389e-06,
+      "loss": 0.5458,
+      "step": 1636
+    },
+    {
+      "epoch": 1.1891131137493474,
+      "grad_norm": 0.4437052109459863,
+      "learning_rate": 9.015415414023226e-06,
+      "loss": 0.5598,
+      "step": 1637
+    },
+    {
+      "epoch": 1.1898395114975144,
+      "grad_norm": 0.40903270592078117,
+      "learning_rate": 9.013992485589513e-06,
+      "loss": 0.5544,
+      "step": 1638
+    },
+    {
+      "epoch": 1.1905659092456813,
+      "grad_norm": 0.39745344686830697,
+      "learning_rate": 9.012568642148615e-06,
+      "loss": 0.5776,
+      "step": 1639
+    },
+    {
+      "epoch": 1.1912923069938484,
+      "grad_norm": 0.5021202842523842,
+      "learning_rate": 9.0111438840251e-06,
+      "loss": 0.5513,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1920187047420152,
+      "grad_norm": 0.5116684721496529,
+      "learning_rate": 9.009718211543752e-06,
+      "loss": 0.5615,
+      "step": 1641
+    },
+    {
+      "epoch": 1.1927451024901823,
+      "grad_norm": 0.3758199164186561,
+      "learning_rate": 9.008291625029556e-06,
+      "loss": 0.5675,
+      "step": 1642
+    },
+    {
+      "epoch": 1.1934715002383491,
+      "grad_norm": 0.3805578386122031,
+      "learning_rate": 9.00686412480771e-06,
+      "loss": 0.5511,
+      "step": 1643
+    },
+    {
+      "epoch": 1.1941978979865162,
+      "grad_norm": 0.44933161613407974,
+      "learning_rate": 9.005435711203619e-06,
+      "loss": 0.5492,
+      "step": 1644
+    },
+    {
+      "epoch": 1.1949242957346833,
+      "grad_norm": 0.43365968490721074,
+      "learning_rate": 9.004006384542894e-06,
+      "loss": 0.5541,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1956506934828501,
+      "grad_norm": 0.3871348134506192,
+      "learning_rate": 9.002576145151359e-06,
+      "loss": 0.5641,
+      "step": 1646
+    },
+    {
+      "epoch": 1.1963770912310172,
+      "grad_norm": 0.4064010639567433,
+      "learning_rate": 9.001144993355042e-06,
+      "loss": 0.5753,
+      "step": 1647
+    },
+    {
+      "epoch": 1.197103488979184,
+      "grad_norm": 0.46830624397535536,
+      "learning_rate": 8.99971292948018e-06,
+      "loss": 0.5656,
+      "step": 1648
+    },
+    {
+      "epoch": 1.1978298867273511,
+      "grad_norm": 0.38093273605602024,
+      "learning_rate": 8.99827995385322e-06,
+      "loss": 0.5537,
+      "step": 1649
+    },
+    {
+      "epoch": 1.1985562844755182,
+      "grad_norm": 0.4356470351028575,
+      "learning_rate": 8.996846066800815e-06,
+      "loss": 0.5609,
+      "step": 1650
+    },
+    {
+      "epoch": 1.199282682223685,
+      "grad_norm": 0.39422074961192954,
+      "learning_rate": 8.995411268649823e-06,
+      "loss": 0.5608,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2000090799718521,
+      "grad_norm": 0.375468512449991,
+      "learning_rate": 8.993975559727316e-06,
+      "loss": 0.5668,
+      "step": 1652
+    },
+    {
+      "epoch": 1.200735477720019,
+      "grad_norm": 0.40297827175793516,
+      "learning_rate": 8.992538940360568e-06,
+      "loss": 0.5538,
+      "step": 1653
+    },
+    {
+      "epoch": 1.201461875468186,
+      "grad_norm": 0.38635504994472913,
+      "learning_rate": 8.991101410877064e-06,
+      "loss": 0.5646,
+      "step": 1654
+    },
+    {
+      "epoch": 1.2021882732163531,
+      "grad_norm": 0.33581752318849784,
+      "learning_rate": 8.989662971604491e-06,
+      "loss": 0.5413,
+      "step": 1655
+    },
+    {
+      "epoch": 1.20291467096452,
+      "grad_norm": 0.3409030181703885,
+      "learning_rate": 8.988223622870754e-06,
+      "loss": 0.5442,
+      "step": 1656
+    },
+    {
+      "epoch": 1.203641068712687,
+      "grad_norm": 0.4306916478895894,
+      "learning_rate": 8.986783365003955e-06,
+      "loss": 0.5695,
+      "step": 1657
+    },
+    {
+      "epoch": 1.204367466460854,
+      "grad_norm": 0.43933251871690226,
+      "learning_rate": 8.985342198332407e-06,
+      "loss": 0.5614,
+      "step": 1658
+    },
+    {
+      "epoch": 1.205093864209021,
+      "grad_norm": 0.4168589094786127,
+      "learning_rate": 8.983900123184634e-06,
+      "loss": 0.5626,
+      "step": 1659
+    },
+    {
+      "epoch": 1.205820261957188,
+      "grad_norm": 0.39213159298507183,
+      "learning_rate": 8.982457139889358e-06,
+      "loss": 0.5508,
+      "step": 1660
+    },
+    {
+      "epoch": 1.206546659705355,
+      "grad_norm": 0.3930281890036753,
+      "learning_rate": 8.981013248775516e-06,
+      "loss": 0.5575,
+      "step": 1661
+    },
+    {
+      "epoch": 1.207273057453522,
+      "grad_norm": 0.4032943583882293,
+      "learning_rate": 8.979568450172248e-06,
+      "loss": 0.5513,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2079994552016888,
+      "grad_norm": 0.35483539042175233,
+      "learning_rate": 8.978122744408905e-06,
+      "loss": 0.5521,
+      "step": 1663
+    },
+    {
+      "epoch": 1.208725852949856,
+      "grad_norm": 0.604044247562166,
+      "learning_rate": 8.976676131815041e-06,
+      "loss": 0.5669,
+      "step": 1664
+    },
+    {
+      "epoch": 1.2094522506980228,
+      "grad_norm": 0.42184739662747306,
+      "learning_rate": 8.975228612720415e-06,
+      "loss": 0.5497,
+      "step": 1665
+    },
+    {
+      "epoch": 1.2101786484461898,
+      "grad_norm": 0.4423424779217794,
+      "learning_rate": 8.973780187454999e-06,
+      "loss": 0.5508,
+      "step": 1666
+    },
+    {
+      "epoch": 1.2109050461943567,
+      "grad_norm": 0.41083410822149735,
+      "learning_rate": 8.972330856348965e-06,
+      "loss": 0.5541,
+      "step": 1667
+    },
+    {
+      "epoch": 1.2116314439425238,
+      "grad_norm": 0.3870919292235364,
+      "learning_rate": 8.970880619732695e-06,
+      "loss": 0.5364,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2123578416906908,
+      "grad_norm": 0.41285288636232365,
+      "learning_rate": 8.96942947793678e-06,
+      "loss": 0.5753,
+      "step": 1669
+    },
+    {
+      "epoch": 1.2130842394388577,
+      "grad_norm": 0.39234857911866095,
+      "learning_rate": 8.967977431292009e-06,
+      "loss": 0.5522,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2138106371870248,
+      "grad_norm": 0.4284657100652839,
+      "learning_rate": 8.966524480129386e-06,
+      "loss": 0.558,
+      "step": 1671
+    },
+    {
+      "epoch": 1.2145370349351916,
+      "grad_norm": 0.42743619885583367,
+      "learning_rate": 8.965070624780117e-06,
+      "loss": 0.5732,
+      "step": 1672
+    },
+    {
+      "epoch": 1.2152634326833587,
+      "grad_norm": 0.38064277890598813,
+      "learning_rate": 8.963615865575613e-06,
+      "loss": 0.541,
+      "step": 1673
+    },
+    {
+      "epoch": 1.2159898304315258,
+      "grad_norm": 0.39160633024799835,
+      "learning_rate": 8.962160202847494e-06,
+      "loss": 0.5461,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2167162281796926,
+      "grad_norm": 0.45820430910342846,
+      "learning_rate": 8.960703636927585e-06,
+      "loss": 0.5398,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2174426259278597,
+      "grad_norm": 0.4594650336331551,
+      "learning_rate": 8.959246168147915e-06,
+      "loss": 0.5433,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2181690236760265,
+      "grad_norm": 0.43281785549799423,
+      "learning_rate": 8.957787796840724e-06,
+      "loss": 0.564,
+      "step": 1677
+    },
+    {
+      "epoch": 1.2188954214241936,
+      "grad_norm": 0.3444644303493312,
+      "learning_rate": 8.95632852333845e-06,
+      "loss": 0.5567,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2196218191723607,
+      "grad_norm": 0.3718226041891921,
+      "learning_rate": 8.954868347973742e-06,
+      "loss": 0.5462,
+      "step": 1679
+    },
+    {
+      "epoch": 1.2203482169205275,
+      "grad_norm": 0.408627736693502,
+      "learning_rate": 8.953407271079456e-06,
+      "loss": 0.5365,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2210746146686946,
+      "grad_norm": 0.37139547247679977,
+      "learning_rate": 8.951945292988647e-06,
+      "loss": 0.5471,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2218010124168615,
+      "grad_norm": 0.3724276111177844,
+      "learning_rate": 8.950482414034583e-06,
+      "loss": 0.5477,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2225274101650285,
+      "grad_norm": 0.39946058697205866,
+      "learning_rate": 8.94901863455073e-06,
+      "loss": 0.5371,
+      "step": 1683
+    },
+    {
+      "epoch": 1.2232538079131954,
+      "grad_norm": 0.36753474211608916,
+      "learning_rate": 8.947553954870765e-06,
+      "loss": 0.5574,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2239802056613625,
+      "grad_norm": 0.5392178107861165,
+      "learning_rate": 8.946088375328569e-06,
+      "loss": 0.5587,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2247066034095295,
+      "grad_norm": 0.3644470759919107,
+      "learning_rate": 8.944621896258226e-06,
+      "loss": 0.5621,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2254330011576964,
+      "grad_norm": 0.44219202544370106,
+      "learning_rate": 8.943154517994026e-06,
+      "loss": 0.5536,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2261593989058635,
+      "grad_norm": 0.3764996895947529,
+      "learning_rate": 8.941686240870464e-06,
+      "loss": 0.5558,
+      "step": 1688
+    },
+    {
+      "epoch": 1.2268857966540303,
+      "grad_norm": 0.3462514441906337,
+      "learning_rate": 8.940217065222241e-06,
+      "loss": 0.5434,
+      "step": 1689
+    },
+    {
+      "epoch": 1.2276121944021974,
+      "grad_norm": 0.37441059075120486,
+      "learning_rate": 8.938746991384264e-06,
+      "loss": 0.5576,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2283385921503642,
+      "grad_norm": 0.4985554959055178,
+      "learning_rate": 8.937276019691638e-06,
+      "loss": 0.5577,
+      "step": 1691
+    },
+    {
+      "epoch": 1.2290649898985313,
+      "grad_norm": 0.3864813217862178,
+      "learning_rate": 8.935804150479683e-06,
+      "loss": 0.5442,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2297913876466984,
+      "grad_norm": 0.3845478893478187,
+      "learning_rate": 8.934331384083914e-06,
+      "loss": 0.5554,
+      "step": 1693
+    },
+    {
+      "epoch": 1.2305177853948652,
+      "grad_norm": 1.6701398796419848,
+      "learning_rate": 8.932857720840056e-06,
+      "loss": 0.5666,
+      "step": 1694
+    },
+    {
+      "epoch": 1.2312441831430323,
+      "grad_norm": 0.4220956274028827,
+      "learning_rate": 8.931383161084037e-06,
+      "loss": 0.5488,
+      "step": 1695
+    },
+    {
+      "epoch": 1.2319705808911992,
+      "grad_norm": 0.38464465303061424,
+      "learning_rate": 8.92990770515199e-06,
+      "loss": 0.5392,
+      "step": 1696
+    },
+    {
+      "epoch": 1.2326969786393662,
+      "grad_norm": 0.38261868231706775,
+      "learning_rate": 8.928431353380249e-06,
+      "loss": 0.551,
+      "step": 1697
+    },
+    {
+      "epoch": 1.2334233763875333,
+      "grad_norm": 0.4963784011318068,
+      "learning_rate": 8.926954106105358e-06,
+      "loss": 0.5572,
+      "step": 1698
+    },
+    {
+      "epoch": 1.2341497741357002,
+      "grad_norm": 0.4774876780938401,
+      "learning_rate": 8.92547596366406e-06,
+      "loss": 0.5671,
+      "step": 1699
+    },
+    {
+      "epoch": 1.2348761718838672,
+      "grad_norm": 0.5664978767143851,
+      "learning_rate": 8.923996926393306e-06,
+      "loss": 0.5343,
+      "step": 1700
+    },
+    {
+      "epoch": 1.235602569632034,
+      "grad_norm": 0.4431798160139461,
+      "learning_rate": 8.922516994630246e-06,
+      "loss": 0.5484,
+      "step": 1701
+    },
+    {
+      "epoch": 1.2363289673802011,
+      "grad_norm": 0.39504209873632384,
+      "learning_rate": 8.921036168712241e-06,
+      "loss": 0.5636,
+      "step": 1702
+    },
+    {
+      "epoch": 1.2370553651283682,
+      "grad_norm": 1.8434104950338672,
+      "learning_rate": 8.919554448976848e-06,
+      "loss": 0.5624,
+      "step": 1703
+    },
+    {
+      "epoch": 1.237781762876535,
+      "grad_norm": 2.226707270583763,
+      "learning_rate": 8.918071835761833e-06,
+      "loss": 0.5408,
+      "step": 1704
+    },
+    {
+      "epoch": 1.2385081606247021,
+      "grad_norm": 0.46962685875100246,
+      "learning_rate": 8.916588329405164e-06,
+      "loss": 0.5627,
+      "step": 1705
+    },
+    {
+      "epoch": 1.239234558372869,
+      "grad_norm": 0.41155769724241303,
+      "learning_rate": 8.915103930245015e-06,
+      "loss": 0.56,
+      "step": 1706
+    },
+    {
+      "epoch": 1.239960956121036,
+      "grad_norm": 0.4924690840390668,
+      "learning_rate": 8.913618638619757e-06,
+      "loss": 0.5503,
+      "step": 1707
+    },
+    {
+      "epoch": 1.240687353869203,
+      "grad_norm": 0.44712240099824757,
+      "learning_rate": 8.912132454867972e-06,
+      "loss": 0.5658,
+      "step": 1708
+    },
+    {
+      "epoch": 1.24141375161737,
+      "grad_norm": 0.5552373112222078,
+      "learning_rate": 8.910645379328442e-06,
+      "loss": 0.5483,
+      "step": 1709
+    },
+    {
+      "epoch": 1.2421401493655368,
+      "grad_norm": 0.7778708121407731,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.5438,
+      "step": 1710
+    },
+    {
+      "epoch": 1.242866547113704,
+      "grad_norm": 0.48332018315159286,
+      "learning_rate": 8.907668554242286e-06,
+      "loss": 0.5506,
+      "step": 1711
+    },
+    {
+      "epoch": 1.243592944861871,
+      "grad_norm": 0.4372921808238204,
+      "learning_rate": 8.906178805374243e-06,
+      "loss": 0.5508,
+      "step": 1712
+    },
+    {
+      "epoch": 1.2443193426100378,
+      "grad_norm": 0.4716974668725763,
+      "learning_rate": 8.904688166075614e-06,
+      "loss": 0.5515,
+      "step": 1713
+    },
+    {
+      "epoch": 1.245045740358205,
+      "grad_norm": 0.5165417601435339,
+      "learning_rate": 8.903196636686198e-06,
+      "loss": 0.5343,
+      "step": 1714
+    },
+    {
+      "epoch": 1.2457721381063718,
+      "grad_norm": 0.47588955550056855,
+      "learning_rate": 8.901704217545995e-06,
+      "loss": 0.5535,
+      "step": 1715
+    },
+    {
+      "epoch": 1.2464985358545388,
+      "grad_norm": 0.41375809514814743,
+      "learning_rate": 8.900210908995207e-06,
+      "loss": 0.5495,
+      "step": 1716
+    },
+    {
+      "epoch": 1.247224933602706,
+      "grad_norm": 0.3667236648873448,
+      "learning_rate": 8.898716711374243e-06,
+      "loss": 0.5657,
+      "step": 1717
+    },
+    {
+      "epoch": 1.2479513313508728,
+      "grad_norm": 0.472340981602009,
+      "learning_rate": 8.89722162502371e-06,
+      "loss": 0.541,
+      "step": 1718
+    },
+    {
+      "epoch": 1.2486777290990398,
+      "grad_norm": 0.4053169462679586,
+      "learning_rate": 8.89572565028442e-06,
+      "loss": 0.5594,
+      "step": 1719
+    },
+    {
+      "epoch": 1.2494041268472067,
+      "grad_norm": 0.5282769709840852,
+      "learning_rate": 8.894228787497389e-06,
+      "loss": 0.5687,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2501305245953738,
+      "grad_norm": 0.4227663685562685,
+      "learning_rate": 8.89273103700383e-06,
+      "loss": 0.5313,
+      "step": 1721
+    },
+    {
+      "epoch": 1.2508569223435408,
+      "grad_norm": 0.48655409953672724,
+      "learning_rate": 8.891232399145164e-06,
+      "loss": 0.5461,
+      "step": 1722
+    },
+    {
+      "epoch": 1.2515833200917077,
+      "grad_norm": 0.4797567850949162,
+      "learning_rate": 8.88973287426301e-06,
+      "loss": 0.5577,
+      "step": 1723
+    },
+    {
+      "epoch": 1.2523097178398748,
+      "grad_norm": 0.37881110627643433,
+      "learning_rate": 8.888232462699196e-06,
+      "loss": 0.5517,
+      "step": 1724
+    },
+    {
+      "epoch": 1.2530361155880416,
+      "grad_norm": 0.4458797743651538,
+      "learning_rate": 8.88673116479574e-06,
+      "loss": 0.5386,
+      "step": 1725
+    },
+    {
+      "epoch": 1.2537625133362087,
+      "grad_norm": 0.47035221294764884,
+      "learning_rate": 8.885228980894877e-06,
+      "loss": 0.5675,
+      "step": 1726
+    },
+    {
+      "epoch": 1.2544889110843758,
+      "grad_norm": 0.4037941136698805,
+      "learning_rate": 8.883725911339032e-06,
+      "loss": 0.5492,
+      "step": 1727
+    },
+    {
+      "epoch": 1.2552153088325426,
+      "grad_norm": 0.43376466055770363,
+      "learning_rate": 8.882221956470838e-06,
+      "loss": 0.5576,
+      "step": 1728
+    },
+    {
+      "epoch": 1.2559417065807095,
+      "grad_norm": 0.4404900043276078,
+      "learning_rate": 8.880717116633126e-06,
+      "loss": 0.5497,
+      "step": 1729
+    },
+    {
+      "epoch": 1.2566681043288765,
+      "grad_norm": 0.4159398594778771,
+      "learning_rate": 8.87921139216893e-06,
+      "loss": 0.5403,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2573945020770436,
+      "grad_norm": 0.40016155319282093,
+      "learning_rate": 8.877704783421492e-06,
+      "loss": 0.5596,
+      "step": 1731
+    },
+    {
+      "epoch": 1.2581208998252105,
+      "grad_norm": 0.6201551996430268,
+      "learning_rate": 8.876197290734247e-06,
+      "loss": 0.5486,
+      "step": 1732
+    },
+    {
+      "epoch": 1.2588472975733775,
+      "grad_norm": 0.423114797642255,
+      "learning_rate": 8.874688914450833e-06,
+      "loss": 0.5619,
+      "step": 1733
+    },
+    {
+      "epoch": 1.2595736953215444,
+      "grad_norm": 0.4015212537751768,
+      "learning_rate": 8.873179654915093e-06,
+      "loss": 0.5506,
+      "step": 1734
+    },
+    {
+      "epoch": 1.2603000930697115,
+      "grad_norm": 0.4138072597511688,
+      "learning_rate": 8.871669512471068e-06,
+      "loss": 0.5414,
+      "step": 1735
+    },
+    {
+      "epoch": 1.2610264908178785,
+      "grad_norm": 0.37740103115812385,
+      "learning_rate": 8.870158487463003e-06,
+      "loss": 0.5652,
+      "step": 1736
+    },
+    {
+      "epoch": 1.2617528885660454,
+      "grad_norm": 0.4672154959611961,
+      "learning_rate": 8.86864658023534e-06,
+      "loss": 0.5509,
+      "step": 1737
+    },
+    {
+      "epoch": 1.2624792863142125,
+      "grad_norm": 0.3794877865595524,
+      "learning_rate": 8.867133791132729e-06,
+      "loss": 0.5378,
+      "step": 1738
+    },
+    {
+      "epoch": 1.2632056840623793,
+      "grad_norm": 0.3941484564763056,
+      "learning_rate": 8.865620120500017e-06,
+      "loss": 0.5313,
+      "step": 1739
+    },
+    {
+      "epoch": 1.2639320818105464,
+      "grad_norm": 0.433625775037335,
+      "learning_rate": 8.864105568682245e-06,
+      "loss": 0.5477,
+      "step": 1740
+    },
+    {
+      "epoch": 1.2646584795587135,
+      "grad_norm": 0.4256197777407076,
+      "learning_rate": 8.862590136024668e-06,
+      "loss": 0.547,
+      "step": 1741
+    },
+    {
+      "epoch": 1.2653848773068803,
+      "grad_norm": 0.36925502674557587,
+      "learning_rate": 8.861073822872735e-06,
+      "loss": 0.5558,
+      "step": 1742
+    },
+    {
+      "epoch": 1.2661112750550474,
+      "grad_norm": 0.4404635405799064,
+      "learning_rate": 8.859556629572095e-06,
+      "loss": 0.5528,
+      "step": 1743
+    },
+    {
+      "epoch": 1.2668376728032142,
+      "grad_norm": 0.4142530403240027,
+      "learning_rate": 8.858038556468598e-06,
+      "loss": 0.5577,
+      "step": 1744
+    },
+    {
+      "epoch": 1.2675640705513813,
+      "grad_norm": 0.418909820251429,
+      "learning_rate": 8.856519603908295e-06,
+      "loss": 0.5482,
+      "step": 1745
+    },
+    {
+      "epoch": 1.2682904682995484,
+      "grad_norm": 0.47128421655077196,
+      "learning_rate": 8.854999772237442e-06,
+      "loss": 0.5437,
+      "step": 1746
+    },
+    {
+      "epoch": 1.2690168660477152,
+      "grad_norm": 0.5111805391386319,
+      "learning_rate": 8.853479061802489e-06,
+      "loss": 0.5317,
+      "step": 1747
+    },
+    {
+      "epoch": 1.2697432637958823,
+      "grad_norm": 0.3996020838695995,
+      "learning_rate": 8.851957472950086e-06,
+      "loss": 0.5593,
+      "step": 1748
+    },
+    {
+      "epoch": 1.2704696615440492,
+      "grad_norm": 0.4130100221068787,
+      "learning_rate": 8.85043500602709e-06,
+      "loss": 0.5549,
+      "step": 1749
+    },
+    {
+      "epoch": 1.2711960592922162,
+      "grad_norm": 0.3513680463359487,
+      "learning_rate": 8.84891166138055e-06,
+      "loss": 0.5423,
+      "step": 1750
+    },
+    {
+      "epoch": 1.271922457040383,
+      "grad_norm": 0.4395399002655679,
+      "learning_rate": 8.847387439357725e-06,
+      "loss": 0.5557,
+      "step": 1751
+    },
+    {
+      "epoch": 1.2726488547885502,
+      "grad_norm": 0.38446859995160715,
+      "learning_rate": 8.84586234030606e-06,
+      "loss": 0.5463,
+      "step": 1752
+    },
+    {
+      "epoch": 1.273375252536717,
+      "grad_norm": 0.3957873922173363,
+      "learning_rate": 8.844336364573214e-06,
+      "loss": 0.5602,
+      "step": 1753
+    },
+    {
+      "epoch": 1.274101650284884,
+      "grad_norm": 0.38643960679786543,
+      "learning_rate": 8.842809512507038e-06,
+      "loss": 0.5458,
+      "step": 1754
+    },
+    {
+      "epoch": 1.2748280480330512,
+      "grad_norm": 0.3706602353251092,
+      "learning_rate": 8.841281784455586e-06,
+      "loss": 0.5602,
+      "step": 1755
+    },
+    {
+      "epoch": 1.275554445781218,
+      "grad_norm": 0.3796213188989138,
+      "learning_rate": 8.839753180767108e-06,
+      "loss": 0.5635,
+      "step": 1756
+    },
+    {
+      "epoch": 1.276280843529385,
+      "grad_norm": 0.45928464006098907,
+      "learning_rate": 8.838223701790057e-06,
+      "loss": 0.5426,
+      "step": 1757
+    },
+    {
+      "epoch": 1.277007241277552,
+      "grad_norm": 0.3818371835551699,
+      "learning_rate": 8.836693347873084e-06,
+      "loss": 0.5463,
+      "step": 1758
+    },
+    {
+      "epoch": 1.277733639025719,
+      "grad_norm": 0.36166963772964084,
+      "learning_rate": 8.835162119365042e-06,
+      "loss": 0.5264,
+      "step": 1759
+    },
+    {
+      "epoch": 1.278460036773886,
+      "grad_norm": 0.385670474786461,
+      "learning_rate": 8.833630016614976e-06,
+      "loss": 0.543,
+      "step": 1760
+    },
+    {
+      "epoch": 1.279186434522053,
+      "grad_norm": 0.3990063492555785,
+      "learning_rate": 8.832097039972144e-06,
+      "loss": 0.5338,
+      "step": 1761
+    },
+    {
+      "epoch": 1.27991283227022,
+      "grad_norm": 0.6653963038347863,
+      "learning_rate": 8.830563189785986e-06,
+      "loss": 0.5375,
+      "step": 1762
+    },
+    {
+      "epoch": 1.2806392300183869,
+      "grad_norm": 0.39286885480174005,
+      "learning_rate": 8.829028466406156e-06,
+      "loss": 0.5575,
+      "step": 1763
+    },
+    {
+      "epoch": 1.281365627766554,
+      "grad_norm": 0.39103188453512855,
+      "learning_rate": 8.827492870182496e-06,
+      "loss": 0.5467,
+      "step": 1764
+    },
+    {
+      "epoch": 1.282092025514721,
+      "grad_norm": 0.3530567148397571,
+      "learning_rate": 8.825956401465056e-06,
+      "loss": 0.5501,
+      "step": 1765
+    },
+    {
+      "epoch": 1.2828184232628879,
+      "grad_norm": 0.3962145821734583,
+      "learning_rate": 8.824419060604081e-06,
+      "loss": 0.5512,
+      "step": 1766
+    },
+    {
+      "epoch": 1.283544821011055,
+      "grad_norm": 0.4155954746083735,
+      "learning_rate": 8.822880847950011e-06,
+      "loss": 0.543,
+      "step": 1767
+    },
+    {
+      "epoch": 1.2842712187592218,
+      "grad_norm": 0.36229776590475643,
+      "learning_rate": 8.82134176385349e-06,
+      "loss": 0.5494,
+      "step": 1768
+    },
+    {
+      "epoch": 1.2849976165073889,
+      "grad_norm": 0.3892187196516395,
+      "learning_rate": 8.819801808665361e-06,
+      "loss": 0.5606,
+      "step": 1769
+    },
+    {
+      "epoch": 1.285724014255556,
+      "grad_norm": 0.6272602907160172,
+      "learning_rate": 8.818260982736662e-06,
+      "loss": 0.5466,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2864504120037228,
+      "grad_norm": 0.3920063283771201,
+      "learning_rate": 8.81671928641863e-06,
+      "loss": 0.5346,
+      "step": 1771
+    },
+    {
+      "epoch": 1.2871768097518896,
+      "grad_norm": 0.34977300906231246,
+      "learning_rate": 8.815176720062701e-06,
+      "loss": 0.5343,
+      "step": 1772
+    },
+    {
+      "epoch": 1.2879032075000567,
+      "grad_norm": 0.4164422641969452,
+      "learning_rate": 8.813633284020512e-06,
+      "loss": 0.5552,
+      "step": 1773
+    },
+    {
+      "epoch": 1.2886296052482238,
+      "grad_norm": 0.389797438250589,
+      "learning_rate": 8.812088978643894e-06,
+      "loss": 0.5468,
+      "step": 1774
+    },
+    {
+      "epoch": 1.2893560029963906,
+      "grad_norm": 0.8495184137168471,
+      "learning_rate": 8.810543804284879e-06,
+      "loss": 0.5501,
+      "step": 1775
+    },
+    {
+      "epoch": 1.2900824007445577,
+      "grad_norm": 0.46604450197114805,
+      "learning_rate": 8.808997761295698e-06,
+      "loss": 0.5455,
+      "step": 1776
+    },
+    {
+      "epoch": 1.2908087984927246,
+      "grad_norm": 0.39491912481678554,
+      "learning_rate": 8.807450850028776e-06,
+      "loss": 0.5362,
+      "step": 1777
+    },
+    {
+      "epoch": 1.2915351962408916,
+      "grad_norm": 0.49234274419716917,
+      "learning_rate": 8.805903070836738e-06,
+      "loss": 0.5406,
+      "step": 1778
+    },
+    {
+      "epoch": 1.2922615939890587,
+      "grad_norm": 0.34933402697149396,
+      "learning_rate": 8.80435442407241e-06,
+      "loss": 0.5448,
+      "step": 1779
+    },
+    {
+      "epoch": 1.2929879917372256,
+      "grad_norm": 0.4826316850229999,
+      "learning_rate": 8.80280491008881e-06,
+      "loss": 0.5451,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2937143894853926,
+      "grad_norm": 0.39767972884114156,
+      "learning_rate": 8.801254529239156e-06,
+      "loss": 0.5499,
+      "step": 1781
+    },
+    {
+      "epoch": 1.2944407872335595,
+      "grad_norm": 0.4612738879037992,
+      "learning_rate": 8.799703281876866e-06,
+      "loss": 0.5473,
+      "step": 1782
+    },
+    {
+      "epoch": 1.2951671849817266,
+      "grad_norm": 0.4211579062948049,
+      "learning_rate": 8.798151168355555e-06,
+      "loss": 0.5347,
+      "step": 1783
+    },
+    {
+      "epoch": 1.2958935827298936,
+      "grad_norm": 0.4422061037768507,
+      "learning_rate": 8.79659818902903e-06,
+      "loss": 0.5535,
+      "step": 1784
+    },
+    {
+      "epoch": 1.2966199804780605,
+      "grad_norm": 0.5858687372390492,
+      "learning_rate": 8.795044344251302e-06,
+      "loss": 0.547,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2973463782262276,
+      "grad_norm": 0.39423622712419204,
+      "learning_rate": 8.793489634376576e-06,
+      "loss": 0.5372,
+      "step": 1786
+    },
+    {
+      "epoch": 1.2980727759743944,
+      "grad_norm": 0.4078027764596252,
+      "learning_rate": 8.791934059759256e-06,
+      "loss": 0.53,
+      "step": 1787
+    },
+    {
+      "epoch": 1.2987991737225615,
+      "grad_norm": 0.4120785404484633,
+      "learning_rate": 8.79037762075394e-06,
+      "loss": 0.5541,
+      "step": 1788
+    },
+    {
+      "epoch": 1.2995255714707286,
+      "grad_norm": 0.36186240974562306,
+      "learning_rate": 8.788820317715427e-06,
+      "loss": 0.5424,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3002519692188954,
+      "grad_norm": 0.4251288194120092,
+      "learning_rate": 8.787262150998713e-06,
+      "loss": 0.5549,
+      "step": 1790
+    },
+    {
+      "epoch": 1.3009783669670625,
+      "grad_norm": 0.4915263914925816,
+      "learning_rate": 8.785703120958984e-06,
+      "loss": 0.5326,
+      "step": 1791
+    },
+    {
+      "epoch": 1.3017047647152293,
+      "grad_norm": 0.3983668140161739,
+      "learning_rate": 8.784143227951628e-06,
+      "loss": 0.5303,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3024311624633964,
+      "grad_norm": 0.3754965619413052,
+      "learning_rate": 8.782582472332236e-06,
+      "loss": 0.5393,
+      "step": 1793
+    },
+    {
+      "epoch": 1.3031575602115635,
+      "grad_norm": 0.3773856242093586,
+      "learning_rate": 8.781020854456582e-06,
+      "loss": 0.535,
+      "step": 1794
+    },
+    {
+      "epoch": 1.3038839579597303,
+      "grad_norm": 0.4490866762349985,
+      "learning_rate": 8.779458374680646e-06,
+      "loss": 0.5308,
+      "step": 1795
+    },
+    {
+      "epoch": 1.3046103557078972,
+      "grad_norm": 0.40733809453212066,
+      "learning_rate": 8.777895033360603e-06,
+      "loss": 0.5478,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3053367534560643,
+      "grad_norm": 0.42410169866054975,
+      "learning_rate": 8.776330830852825e-06,
+      "loss": 0.5534,
+      "step": 1797
+    },
+    {
+      "epoch": 1.3060631512042313,
+      "grad_norm": 0.34776512661909453,
+      "learning_rate": 8.774765767513876e-06,
+      "loss": 0.5411,
+      "step": 1798
+    },
+    {
+      "epoch": 1.3067895489523982,
+      "grad_norm": 0.3769902398807461,
+      "learning_rate": 8.773199843700518e-06,
+      "loss": 0.5309,
+      "step": 1799
+    },
+    {
+      "epoch": 1.3075159467005653,
+      "grad_norm": 0.37330832388241414,
+      "learning_rate": 8.771633059769712e-06,
+      "loss": 0.549,
+      "step": 1800
+    },
+    {
+      "epoch": 1.308242344448732,
+      "grad_norm": 0.39550029782820983,
+      "learning_rate": 8.770065416078615e-06,
+      "loss": 0.5539,
+      "step": 1801
+    },
+    {
+      "epoch": 1.3089687421968992,
+      "grad_norm": 0.3819609212474707,
+      "learning_rate": 8.768496912984574e-06,
+      "loss": 0.5346,
+      "step": 1802
+    },
+    {
+      "epoch": 1.3096951399450663,
+      "grad_norm": 0.4118791910236496,
+      "learning_rate": 8.76692755084514e-06,
+      "loss": 0.5563,
+      "step": 1803
+    },
+    {
+      "epoch": 1.310421537693233,
+      "grad_norm": 0.44571217160818744,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.5438,
+      "step": 1804
+    },
+    {
+      "epoch": 1.3111479354414002,
+      "grad_norm": 0.4121837514951978,
+      "learning_rate": 8.763786250861258e-06,
+      "loss": 0.5497,
+      "step": 1805
+    },
+    {
+      "epoch": 1.311874333189567,
+      "grad_norm": 0.4083163261875877,
+      "learning_rate": 8.762214313732881e-06,
+      "loss": 0.5552,
+      "step": 1806
+    },
+    {
+      "epoch": 1.312600730937734,
+      "grad_norm": 0.4181152080467721,
+      "learning_rate": 8.760641518991257e-06,
+      "loss": 0.564,
+      "step": 1807
+    },
+    {
+      "epoch": 1.3133271286859012,
+      "grad_norm": 0.38364370277204013,
+      "learning_rate": 8.75906786699491e-06,
+      "loss": 0.5378,
+      "step": 1808
+    },
+    {
+      "epoch": 1.314053526434068,
+      "grad_norm": 0.41869608928831936,
+      "learning_rate": 8.75749335810256e-06,
+      "loss": 0.5464,
+      "step": 1809
+    },
+    {
+      "epoch": 1.314779924182235,
+      "grad_norm": 0.42528426296113725,
+      "learning_rate": 8.755917992673126e-06,
+      "loss": 0.5463,
+      "step": 1810
+    },
+    {
+      "epoch": 1.315506321930402,
+      "grad_norm": 0.38483565529483227,
+      "learning_rate": 8.754341771065716e-06,
+      "loss": 0.5398,
+      "step": 1811
+    },
+    {
+      "epoch": 1.316232719678569,
+      "grad_norm": 0.46536447305861245,
+      "learning_rate": 8.75276469363964e-06,
+      "loss": 0.5464,
+      "step": 1812
+    },
+    {
+      "epoch": 1.316959117426736,
+      "grad_norm": 0.41489340357941334,
+      "learning_rate": 8.751186760754397e-06,
+      "loss": 0.5319,
+      "step": 1813
+    },
+    {
+      "epoch": 1.317685515174903,
+      "grad_norm": 0.4578205762462449,
+      "learning_rate": 8.749607972769685e-06,
+      "loss": 0.5439,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3184119129230698,
+      "grad_norm": 0.49155744319398526,
+      "learning_rate": 8.748028330045395e-06,
+      "loss": 0.5508,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3191383106712369,
+      "grad_norm": 0.37190008962866994,
+      "learning_rate": 8.746447832941614e-06,
+      "loss": 0.5534,
+      "step": 1816
+    },
+    {
+      "epoch": 1.319864708419404,
+      "grad_norm": 0.3769835428333359,
+      "learning_rate": 8.744866481818624e-06,
+      "loss": 0.5389,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3205911061675708,
+      "grad_norm": 0.41417801427669604,
+      "learning_rate": 8.743284277036899e-06,
+      "loss": 0.5271,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3213175039157379,
+      "grad_norm": 0.3655224383838069,
+      "learning_rate": 8.74170121895711e-06,
+      "loss": 0.5684,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3220439016639047,
+      "grad_norm": 0.5029173229142005,
+      "learning_rate": 8.740117307940123e-06,
+      "loss": 0.5374,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3227702994120718,
+      "grad_norm": 0.4438745767992502,
+      "learning_rate": 8.738532544346998e-06,
+      "loss": 0.544,
+      "step": 1821
+    },
+    {
+      "epoch": 1.3234966971602389,
+      "grad_norm": 0.3818937011598134,
+      "learning_rate": 8.736946928538988e-06,
+      "loss": 0.5456,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3242230949084057,
+      "grad_norm": 0.4314164584511377,
+      "learning_rate": 8.73536046087754e-06,
+      "loss": 0.5336,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3249494926565728,
+      "grad_norm": 0.39232944801618264,
+      "learning_rate": 8.733773141724298e-06,
+      "loss": 0.5363,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3256758904047397,
+      "grad_norm": 0.505176054594776,
+      "learning_rate": 8.732184971441098e-06,
+      "loss": 0.5478,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3264022881529067,
+      "grad_norm": 0.4141339957842407,
+      "learning_rate": 8.730595950389968e-06,
+      "loss": 0.5758,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3271286859010738,
+      "grad_norm": 0.3771489964312615,
+      "learning_rate": 8.729006078933136e-06,
+      "loss": 0.5457,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3278550836492407,
+      "grad_norm": 0.3418085527128151,
+      "learning_rate": 8.727415357433018e-06,
+      "loss": 0.541,
+      "step": 1828
+    },
+    {
+      "epoch": 1.3285814813974077,
+      "grad_norm": 0.36072786868922635,
+      "learning_rate": 8.725823786252226e-06,
+      "loss": 0.55,
+      "step": 1829
+    },
+    {
+      "epoch": 1.3293078791455746,
+      "grad_norm": 0.35881277088044755,
+      "learning_rate": 8.724231365753567e-06,
+      "loss": 0.5497,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3300342768937417,
+      "grad_norm": 0.4034179783998635,
+      "learning_rate": 8.722638096300037e-06,
+      "loss": 0.5403,
+      "step": 1831
+    },
+    {
+      "epoch": 1.3307606746419087,
+      "grad_norm": 0.38808308265799407,
+      "learning_rate": 8.721043978254833e-06,
+      "loss": 0.5521,
+      "step": 1832
+    },
+    {
+      "epoch": 1.3314870723900756,
+      "grad_norm": 0.37141607365637724,
+      "learning_rate": 8.71944901198134e-06,
+      "loss": 0.5288,
+      "step": 1833
+    },
+    {
+      "epoch": 1.3322134701382427,
+      "grad_norm": 0.4153027779362984,
+      "learning_rate": 8.717853197843134e-06,
+      "loss": 0.563,
+      "step": 1834
+    },
+    {
+      "epoch": 1.3329398678864095,
+      "grad_norm": 0.4063731663632702,
+      "learning_rate": 8.716256536203992e-06,
+      "loss": 0.5381,
+      "step": 1835
+    },
+    {
+      "epoch": 1.3336662656345766,
+      "grad_norm": 0.3720949582825728,
+      "learning_rate": 8.714659027427878e-06,
+      "loss": 0.5378,
+      "step": 1836
+    },
+    {
+      "epoch": 1.3343926633827436,
+      "grad_norm": 0.38950561831594427,
+      "learning_rate": 8.713060671878953e-06,
+      "loss": 0.5408,
+      "step": 1837
+    },
+    {
+      "epoch": 1.3351190611309105,
+      "grad_norm": 0.3921429083344398,
+      "learning_rate": 8.711461469921568e-06,
+      "loss": 0.5313,
+      "step": 1838
+    },
+    {
+      "epoch": 1.3358454588790774,
+      "grad_norm": 0.3842738383447121,
+      "learning_rate": 8.709861421920267e-06,
+      "loss": 0.5579,
+      "step": 1839
+    },
+    {
+      "epoch": 1.3365718566272444,
+      "grad_norm": 0.4466227404643763,
+      "learning_rate": 8.708260528239788e-06,
+      "loss": 0.5391,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3372982543754115,
+      "grad_norm": 0.40000502784137687,
+      "learning_rate": 8.706658789245064e-06,
+      "loss": 0.5329,
+      "step": 1841
+    },
+    {
+      "epoch": 1.3380246521235784,
+      "grad_norm": 0.3681548592380934,
+      "learning_rate": 8.705056205301217e-06,
+      "loss": 0.543,
+      "step": 1842
+    },
+    {
+      "epoch": 1.3387510498717454,
+      "grad_norm": 0.5442305579733531,
+      "learning_rate": 8.703452776773563e-06,
+      "loss": 0.5572,
+      "step": 1843
+    },
+    {
+      "epoch": 1.3394774476199123,
+      "grad_norm": 0.41846847295976275,
+      "learning_rate": 8.70184850402761e-06,
+      "loss": 0.5336,
+      "step": 1844
+    },
+    {
+      "epoch": 1.3402038453680793,
+      "grad_norm": 0.39723343483159357,
+      "learning_rate": 8.700243387429061e-06,
+      "loss": 0.5432,
+      "step": 1845
+    },
+    {
+      "epoch": 1.3409302431162464,
+      "grad_norm": 0.38939768776621403,
+      "learning_rate": 8.698637427343809e-06,
+      "loss": 0.5229,
+      "step": 1846
+    },
+    {
+      "epoch": 1.3416566408644133,
+      "grad_norm": 0.4013658344287758,
+      "learning_rate": 8.697030624137937e-06,
+      "loss": 0.5599,
+      "step": 1847
+    },
+    {
+      "epoch": 1.3423830386125803,
+      "grad_norm": 0.38252834404139674,
+      "learning_rate": 8.695422978177724e-06,
+      "loss": 0.5364,
+      "step": 1848
+    },
+    {
+      "epoch": 1.3431094363607472,
+      "grad_norm": 0.37487413981443296,
+      "learning_rate": 8.693814489829643e-06,
+      "loss": 0.5305,
+      "step": 1849
+    },
+    {
+      "epoch": 1.3438358341089143,
+      "grad_norm": 0.5045145749139772,
+      "learning_rate": 8.692205159460349e-06,
+      "loss": 0.5324,
+      "step": 1850
+    },
+    {
+      "epoch": 1.3445622318570813,
+      "grad_norm": 0.3834848107935937,
+      "learning_rate": 8.690594987436705e-06,
+      "loss": 0.5373,
+      "step": 1851
+    },
+    {
+      "epoch": 1.3452886296052482,
+      "grad_norm": 0.402784897329814,
+      "learning_rate": 8.68898397412575e-06,
+      "loss": 0.5405,
+      "step": 1852
+    },
+    {
+      "epoch": 1.3460150273534153,
+      "grad_norm": 0.42427307386128277,
+      "learning_rate": 8.687372119894723e-06,
+      "loss": 0.5583,
+      "step": 1853
+    },
+    {
+      "epoch": 1.3467414251015821,
+      "grad_norm": 0.3911558225143398,
+      "learning_rate": 8.685759425111056e-06,
+      "loss": 0.5583,
+      "step": 1854
+    },
+    {
+      "epoch": 1.3474678228497492,
+      "grad_norm": 0.3543476050936793,
+      "learning_rate": 8.684145890142366e-06,
+      "loss": 0.5393,
+      "step": 1855
+    },
+    {
+      "epoch": 1.3481942205979163,
+      "grad_norm": 0.4137430614844222,
+      "learning_rate": 8.682531515356467e-06,
+      "loss": 0.5418,
+      "step": 1856
+    },
+    {
+      "epoch": 1.3489206183460831,
+      "grad_norm": 0.44802159447591655,
+      "learning_rate": 8.680916301121365e-06,
+      "loss": 0.5558,
+      "step": 1857
+    },
+    {
+      "epoch": 1.34964701609425,
+      "grad_norm": 0.39855739423499525,
+      "learning_rate": 8.679300247805252e-06,
+      "loss": 0.5488,
+      "step": 1858
+    },
+    {
+      "epoch": 1.350373413842417,
+      "grad_norm": 0.35349526794579944,
+      "learning_rate": 8.677683355776515e-06,
+      "loss": 0.5348,
+      "step": 1859
+    },
+    {
+      "epoch": 1.3510998115905841,
+      "grad_norm": 0.4079156229909949,
+      "learning_rate": 8.676065625403733e-06,
+      "loss": 0.5355,
+      "step": 1860
+    },
+    {
+      "epoch": 1.351826209338751,
+      "grad_norm": 0.4101135183395125,
+      "learning_rate": 8.674447057055673e-06,
+      "loss": 0.5592,
+      "step": 1861
+    },
+    {
+      "epoch": 1.352552607086918,
+      "grad_norm": 0.4373664713387905,
+      "learning_rate": 8.672827651101297e-06,
+      "loss": 0.5451,
+      "step": 1862
+    },
+    {
+      "epoch": 1.353279004835085,
+      "grad_norm": 0.3981073036267596,
+      "learning_rate": 8.671207407909754e-06,
+      "loss": 0.5478,
+      "step": 1863
+    },
+    {
+      "epoch": 1.354005402583252,
+      "grad_norm": 0.3773769438352129,
+      "learning_rate": 8.669586327850386e-06,
+      "loss": 0.5414,
+      "step": 1864
+    },
+    {
+      "epoch": 1.354731800331419,
+      "grad_norm": 0.4121506936367948,
+      "learning_rate": 8.667964411292725e-06,
+      "loss": 0.5418,
+      "step": 1865
+    },
+    {
+      "epoch": 1.355458198079586,
+      "grad_norm": 0.3721842200340298,
+      "learning_rate": 8.666341658606493e-06,
+      "loss": 0.533,
+      "step": 1866
+    },
+    {
+      "epoch": 1.356184595827753,
+      "grad_norm": 0.4303316522403638,
+      "learning_rate": 8.664718070161607e-06,
+      "loss": 0.5486,
+      "step": 1867
+    },
+    {
+      "epoch": 1.3569109935759198,
+      "grad_norm": 0.3803968632015748,
+      "learning_rate": 8.663093646328166e-06,
+      "loss": 0.5482,
+      "step": 1868
+    },
+    {
+      "epoch": 1.357637391324087,
+      "grad_norm": 0.4155775190064203,
+      "learning_rate": 8.661468387476471e-06,
+      "loss": 0.5249,
+      "step": 1869
+    },
+    {
+      "epoch": 1.358363789072254,
+      "grad_norm": 0.41649246283951763,
+      "learning_rate": 8.659842293977e-06,
+      "loss": 0.5353,
+      "step": 1870
+    },
+    {
+      "epoch": 1.3590901868204208,
+      "grad_norm": 0.4117337284313234,
+      "learning_rate": 8.658215366200433e-06,
+      "loss": 0.5451,
+      "step": 1871
+    },
+    {
+      "epoch": 1.359816584568588,
+      "grad_norm": 0.4392355311303106,
+      "learning_rate": 8.656587604517635e-06,
+      "loss": 0.5366,
+      "step": 1872
+    },
+    {
+      "epoch": 1.3605429823167547,
+      "grad_norm": 0.39648896071408246,
+      "learning_rate": 8.654959009299657e-06,
+      "loss": 0.5404,
+      "step": 1873
+    },
+    {
+      "epoch": 1.3612693800649218,
+      "grad_norm": 0.3872675445529076,
+      "learning_rate": 8.653329580917752e-06,
+      "loss": 0.5475,
+      "step": 1874
+    },
+    {
+      "epoch": 1.361995777813089,
+      "grad_norm": 0.4218469959253971,
+      "learning_rate": 8.651699319743348e-06,
+      "loss": 0.5293,
+      "step": 1875
+    },
+    {
+      "epoch": 1.3627221755612557,
+      "grad_norm": 0.41767884109068387,
+      "learning_rate": 8.650068226148074e-06,
+      "loss": 0.5317,
+      "step": 1876
+    },
+    {
+      "epoch": 1.3634485733094228,
+      "grad_norm": 0.3859802721551015,
+      "learning_rate": 8.648436300503742e-06,
+      "loss": 0.5293,
+      "step": 1877
+    },
+    {
+      "epoch": 1.3641749710575897,
+      "grad_norm": 0.42945118279457706,
+      "learning_rate": 8.646803543182361e-06,
+      "loss": 0.5376,
+      "step": 1878
+    },
+    {
+      "epoch": 1.3649013688057567,
+      "grad_norm": 0.37671501911345834,
+      "learning_rate": 8.645169954556123e-06,
+      "loss": 0.5395,
+      "step": 1879
+    },
+    {
+      "epoch": 1.3656277665539238,
+      "grad_norm": 0.369319222907703,
+      "learning_rate": 8.64353553499741e-06,
+      "loss": 0.5338,
+      "step": 1880
+    },
+    {
+      "epoch": 1.3663541643020907,
+      "grad_norm": 0.34095153700697944,
+      "learning_rate": 8.641900284878798e-06,
+      "loss": 0.5418,
+      "step": 1881
+    },
+    {
+      "epoch": 1.3670805620502575,
+      "grad_norm": 0.6919819010110789,
+      "learning_rate": 8.640264204573049e-06,
+      "loss": 0.5415,
+      "step": 1882
+    },
+    {
+      "epoch": 1.3678069597984246,
+      "grad_norm": 0.3531349893202495,
+      "learning_rate": 8.63862729445311e-06,
+      "loss": 0.5551,
+      "step": 1883
+    },
+    {
+      "epoch": 1.3685333575465917,
+      "grad_norm": 0.37080974188770977,
+      "learning_rate": 8.63698955489213e-06,
+      "loss": 0.5424,
+      "step": 1884
+    },
+    {
+      "epoch": 1.3692597552947585,
+      "grad_norm": 0.40867422693303573,
+      "learning_rate": 8.635350986263431e-06,
+      "loss": 0.5451,
+      "step": 1885
+    },
+    {
+      "epoch": 1.3699861530429256,
+      "grad_norm": 0.4680403849119622,
+      "learning_rate": 8.633711588940538e-06,
+      "loss": 0.5471,
+      "step": 1886
+    },
+    {
+      "epoch": 1.3707125507910924,
+      "grad_norm": 0.3835701618473411,
+      "learning_rate": 8.632071363297155e-06,
+      "loss": 0.5423,
+      "step": 1887
+    },
+    {
+      "epoch": 1.3714389485392595,
+      "grad_norm": 0.38961022680836604,
+      "learning_rate": 8.63043030970718e-06,
+      "loss": 0.5414,
+      "step": 1888
+    },
+    {
+      "epoch": 1.3721653462874266,
+      "grad_norm": 0.36131709249510413,
+      "learning_rate": 8.628788428544698e-06,
+      "loss": 0.5316,
+      "step": 1889
+    },
+    {
+      "epoch": 1.3728917440355934,
+      "grad_norm": 0.3730237230261701,
+      "learning_rate": 8.627145720183986e-06,
+      "loss": 0.5461,
+      "step": 1890
+    },
+    {
+      "epoch": 1.3736181417837605,
+      "grad_norm": 0.44393089036469924,
+      "learning_rate": 8.6255021849995e-06,
+      "loss": 0.5606,
+      "step": 1891
+    },
+    {
+      "epoch": 1.3743445395319274,
+      "grad_norm": 0.3878934576688868,
+      "learning_rate": 8.623857823365896e-06,
+      "loss": 0.5303,
+      "step": 1892
+    },
+    {
+      "epoch": 1.3750709372800944,
+      "grad_norm": 0.36740185024361566,
+      "learning_rate": 8.622212635658015e-06,
+      "loss": 0.5318,
+      "step": 1893
+    },
+    {
+      "epoch": 1.3757973350282615,
+      "grad_norm": 0.3873406422624257,
+      "learning_rate": 8.620566622250878e-06,
+      "loss": 0.5422,
+      "step": 1894
+    },
+    {
+      "epoch": 1.3765237327764284,
+      "grad_norm": 0.43456363913952073,
+      "learning_rate": 8.618919783519706e-06,
+      "loss": 0.5392,
+      "step": 1895
+    },
+    {
+      "epoch": 1.3772501305245954,
+      "grad_norm": 0.40024030236124175,
+      "learning_rate": 8.617272119839903e-06,
+      "loss": 0.525,
+      "step": 1896
+    },
+    {
+      "epoch": 1.3779765282727623,
+      "grad_norm": 0.3761797253882858,
+      "learning_rate": 8.61562363158706e-06,
+      "loss": 0.541,
+      "step": 1897
+    },
+    {
+      "epoch": 1.3787029260209294,
+      "grad_norm": 0.3707955979046983,
+      "learning_rate": 8.613974319136959e-06,
+      "loss": 0.5317,
+      "step": 1898
+    },
+    {
+      "epoch": 1.3794293237690964,
+      "grad_norm": 0.3760358723392047,
+      "learning_rate": 8.612324182865565e-06,
+      "loss": 0.5465,
+      "step": 1899
+    },
+    {
+      "epoch": 1.3801557215172633,
+      "grad_norm": 0.4256562839209359,
+      "learning_rate": 8.610673223149036e-06,
+      "loss": 0.5414,
+      "step": 1900
+    },
+    {
+      "epoch": 1.3808821192654301,
+      "grad_norm": 0.4279063471862655,
+      "learning_rate": 8.609021440363712e-06,
+      "loss": 0.5535,
+      "step": 1901
+    },
+    {
+      "epoch": 1.3816085170135972,
+      "grad_norm": 0.3509451349635699,
+      "learning_rate": 8.60736883488613e-06,
+      "loss": 0.5357,
+      "step": 1902
+    },
+    {
+      "epoch": 1.3823349147617643,
+      "grad_norm": 0.581332340740415,
+      "learning_rate": 8.605715407093005e-06,
+      "loss": 0.5311,
+      "step": 1903
+    },
+    {
+      "epoch": 1.3830613125099311,
+      "grad_norm": 0.4407596324959687,
+      "learning_rate": 8.604061157361243e-06,
+      "loss": 0.5467,
+      "step": 1904
+    },
+    {
+      "epoch": 1.3837877102580982,
+      "grad_norm": 0.4659283840524074,
+      "learning_rate": 8.602406086067938e-06,
+      "loss": 0.5415,
+      "step": 1905
+    },
+    {
+      "epoch": 1.384514108006265,
+      "grad_norm": 0.3802760596789848,
+      "learning_rate": 8.60075019359037e-06,
+      "loss": 0.5515,
+      "step": 1906
+    },
+    {
+      "epoch": 1.3852405057544321,
+      "grad_norm": 0.38316526842068516,
+      "learning_rate": 8.599093480306009e-06,
+      "loss": 0.5577,
+      "step": 1907
+    },
+    {
+      "epoch": 1.3859669035025992,
+      "grad_norm": 0.49345679046274693,
+      "learning_rate": 8.59743594659251e-06,
+      "loss": 0.5567,
+      "step": 1908
+    },
+    {
+      "epoch": 1.386693301250766,
+      "grad_norm": 0.4407146899861917,
+      "learning_rate": 8.595777592827713e-06,
+      "loss": 0.5374,
+      "step": 1909
+    },
+    {
+      "epoch": 1.3874196989989331,
+      "grad_norm": 0.3686260662439781,
+      "learning_rate": 8.594118419389648e-06,
+      "loss": 0.5358,
+      "step": 1910
+    },
+    {
+      "epoch": 1.3881460967471,
+      "grad_norm": 0.43813341796941524,
+      "learning_rate": 8.592458426656531e-06,
+      "loss": 0.5292,
+      "step": 1911
+    },
+    {
+      "epoch": 1.388872494495267,
+      "grad_norm": 0.5289201011280882,
+      "learning_rate": 8.590797615006763e-06,
+      "loss": 0.5314,
+      "step": 1912
+    },
+    {
+      "epoch": 1.3895988922434341,
+      "grad_norm": 0.3545435738165874,
+      "learning_rate": 8.589135984818936e-06,
+      "loss": 0.5331,
+      "step": 1913
+    },
+    {
+      "epoch": 1.390325289991601,
+      "grad_norm": 0.3880181286798106,
+      "learning_rate": 8.587473536471824e-06,
+      "loss": 0.5292,
+      "step": 1914
+    },
+    {
+      "epoch": 1.391051687739768,
+      "grad_norm": 0.3711953855294708,
+      "learning_rate": 8.585810270344391e-06,
+      "loss": 0.5398,
+      "step": 1915
+    },
+    {
+      "epoch": 1.391778085487935,
+      "grad_norm": 0.3606452388095397,
+      "learning_rate": 8.584146186815785e-06,
+      "loss": 0.5388,
+      "step": 1916
+    },
+    {
+      "epoch": 1.392504483236102,
+      "grad_norm": 0.49144573790253854,
+      "learning_rate": 8.582481286265341e-06,
+      "loss": 0.5331,
+      "step": 1917
+    },
+    {
+      "epoch": 1.393230880984269,
+      "grad_norm": 0.44018366120876584,
+      "learning_rate": 8.580815569072579e-06,
+      "loss": 0.553,
+      "step": 1918
+    },
+    {
+      "epoch": 1.393957278732436,
+      "grad_norm": 0.4114140131639351,
+      "learning_rate": 8.57914903561721e-06,
+      "loss": 0.5438,
+      "step": 1919
+    },
+    {
+      "epoch": 1.394683676480603,
+      "grad_norm": 0.3779639707831259,
+      "learning_rate": 8.577481686279123e-06,
+      "loss": 0.5315,
+      "step": 1920
+    },
+    {
+      "epoch": 1.3954100742287698,
+      "grad_norm": 0.5961234774932549,
+      "learning_rate": 8.575813521438401e-06,
+      "loss": 0.53,
+      "step": 1921
+    },
+    {
+      "epoch": 1.396136471976937,
+      "grad_norm": 0.5384898478460642,
+      "learning_rate": 8.574144541475309e-06,
+      "loss": 0.5325,
+      "step": 1922
+    },
+    {
+      "epoch": 1.396862869725104,
+      "grad_norm": 0.40857235280824467,
+      "learning_rate": 8.572474746770298e-06,
+      "loss": 0.5478,
+      "step": 1923
+    },
+    {
+      "epoch": 1.3975892674732708,
+      "grad_norm": 0.425386958242189,
+      "learning_rate": 8.570804137704005e-06,
+      "loss": 0.5312,
+      "step": 1924
+    },
+    {
+      "epoch": 1.3983156652214377,
+      "grad_norm": 0.7508299384321971,
+      "learning_rate": 8.56913271465725e-06,
+      "loss": 0.5324,
+      "step": 1925
+    },
+    {
+      "epoch": 1.3990420629696048,
+      "grad_norm": 0.36782172271514246,
+      "learning_rate": 8.567460478011048e-06,
+      "loss": 0.5435,
+      "step": 1926
+    },
+    {
+      "epoch": 1.3997684607177718,
+      "grad_norm": 0.38067589919289535,
+      "learning_rate": 8.565787428146586e-06,
+      "loss": 0.536,
+      "step": 1927
+    },
+    {
+      "epoch": 1.4004948584659387,
+      "grad_norm": 0.5934857184987411,
+      "learning_rate": 8.564113565445247e-06,
+      "loss": 0.5341,
+      "step": 1928
+    },
+    {
+      "epoch": 1.4012212562141058,
+      "grad_norm": 0.3596646337745538,
+      "learning_rate": 8.56243889028859e-06,
+      "loss": 0.5165,
+      "step": 1929
+    },
+    {
+      "epoch": 1.4019476539622726,
+      "grad_norm": 0.4280239182874118,
+      "learning_rate": 8.560763403058375e-06,
+      "loss": 0.5389,
+      "step": 1930
+    },
+    {
+      "epoch": 1.4026740517104397,
+      "grad_norm": 0.41277127732364816,
+      "learning_rate": 8.559087104136525e-06,
+      "loss": 0.5437,
+      "step": 1931
+    },
+    {
+      "epoch": 1.4034004494586068,
+      "grad_norm": 0.35672219835666513,
+      "learning_rate": 8.557409993905165e-06,
+      "loss": 0.5297,
+      "step": 1932
+    },
+    {
+      "epoch": 1.4041268472067736,
+      "grad_norm": 0.3761535339499059,
+      "learning_rate": 8.5557320727466e-06,
+      "loss": 0.5419,
+      "step": 1933
+    },
+    {
+      "epoch": 1.4048532449549407,
+      "grad_norm": 0.403660946737396,
+      "learning_rate": 8.55405334104332e-06,
+      "loss": 0.5283,
+      "step": 1934
+    },
+    {
+      "epoch": 1.4055796427031075,
+      "grad_norm": 0.387444269843542,
+      "learning_rate": 8.552373799177997e-06,
+      "loss": 0.5436,
+      "step": 1935
+    },
+    {
+      "epoch": 1.4063060404512746,
+      "grad_norm": 0.3691292665136819,
+      "learning_rate": 8.550693447533493e-06,
+      "loss": 0.5567,
+      "step": 1936
+    },
+    {
+      "epoch": 1.4070324381994417,
+      "grad_norm": 0.405303329778648,
+      "learning_rate": 8.549012286492848e-06,
+      "loss": 0.5315,
+      "step": 1937
+    },
+    {
+      "epoch": 1.4077588359476085,
+      "grad_norm": 0.6503597959594926,
+      "learning_rate": 8.54733031643929e-06,
+      "loss": 0.5369,
+      "step": 1938
+    },
+    {
+      "epoch": 1.4084852336957756,
+      "grad_norm": 0.39065256110950697,
+      "learning_rate": 8.545647537756236e-06,
+      "loss": 0.5332,
+      "step": 1939
+    },
+    {
+      "epoch": 1.4092116314439425,
+      "grad_norm": 0.4609381596916133,
+      "learning_rate": 8.543963950827279e-06,
+      "loss": 0.5442,
+      "step": 1940
+    },
+    {
+      "epoch": 1.4099380291921095,
+      "grad_norm": 0.3649780548895913,
+      "learning_rate": 8.5422795560362e-06,
+      "loss": 0.548,
+      "step": 1941
+    },
+    {
+      "epoch": 1.4106644269402766,
+      "grad_norm": 0.4168463039717824,
+      "learning_rate": 8.540594353766966e-06,
+      "loss": 0.5426,
+      "step": 1942
+    },
+    {
+      "epoch": 1.4113908246884435,
+      "grad_norm": 0.3605395594921,
+      "learning_rate": 8.538908344403724e-06,
+      "loss": 0.522,
+      "step": 1943
+    },
+    {
+      "epoch": 1.4121172224366103,
+      "grad_norm": 0.43891493266382947,
+      "learning_rate": 8.53722152833081e-06,
+      "loss": 0.5485,
+      "step": 1944
+    },
+    {
+      "epoch": 1.4128436201847774,
+      "grad_norm": 0.3616022146521361,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.5428,
+      "step": 1945
+    },
+    {
+      "epoch": 1.4135700179329445,
+      "grad_norm": 0.4830840461126677,
+      "learning_rate": 8.533845477594212e-06,
+      "loss": 0.5359,
+      "step": 1946
+    },
+    {
+      "epoch": 1.4142964156811113,
+      "grad_norm": 0.44953993531184466,
+      "learning_rate": 8.532156243700114e-06,
+      "loss": 0.5396,
+      "step": 1947
+    },
+    {
+      "epoch": 1.4150228134292784,
+      "grad_norm": 0.34692781192390154,
+      "learning_rate": 8.530466204635514e-06,
+      "loss": 0.5394,
+      "step": 1948
+    },
+    {
+      "epoch": 1.4157492111774452,
+      "grad_norm": 0.383099285241623,
+      "learning_rate": 8.528775360785665e-06,
+      "loss": 0.5512,
+      "step": 1949
+    },
+    {
+      "epoch": 1.4164756089256123,
+      "grad_norm": 0.38018322615703837,
+      "learning_rate": 8.527083712535998e-06,
+      "loss": 0.5334,
+      "step": 1950
+    },
+    {
+      "epoch": 1.4172020066737794,
+      "grad_norm": 0.4071584173973299,
+      "learning_rate": 8.525391260272134e-06,
+      "loss": 0.5376,
+      "step": 1951
+    },
+    {
+      "epoch": 1.4179284044219462,
+      "grad_norm": 0.33721951307742976,
+      "learning_rate": 8.523698004379878e-06,
+      "loss": 0.5279,
+      "step": 1952
+    },
+    {
+      "epoch": 1.4186548021701133,
+      "grad_norm": 0.8918147939342991,
+      "learning_rate": 8.52200394524521e-06,
+      "loss": 0.5394,
+      "step": 1953
+    },
+    {
+      "epoch": 1.4193811999182802,
+      "grad_norm": 0.38716601747441487,
+      "learning_rate": 8.520309083254301e-06,
+      "loss": 0.5348,
+      "step": 1954
+    },
+    {
+      "epoch": 1.4201075976664472,
+      "grad_norm": 0.4193252319693734,
+      "learning_rate": 8.518613418793502e-06,
+      "loss": 0.5232,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4208339954146143,
+      "grad_norm": 0.39452248506913185,
+      "learning_rate": 8.516916952249346e-06,
+      "loss": 0.5293,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4215603931627812,
+      "grad_norm": 0.47356394532545604,
+      "learning_rate": 8.51521968400855e-06,
+      "loss": 0.5335,
+      "step": 1957
+    },
+    {
+      "epoch": 1.4222867909109482,
+      "grad_norm": 0.4721127066478117,
+      "learning_rate": 8.513521614458015e-06,
+      "loss": 0.5395,
+      "step": 1958
+    },
+    {
+      "epoch": 1.423013188659115,
+      "grad_norm": 0.37657901193003795,
+      "learning_rate": 8.511822743984824e-06,
+      "loss": 0.5396,
+      "step": 1959
+    },
+    {
+      "epoch": 1.4237395864072822,
+      "grad_norm": 0.39741495285863576,
+      "learning_rate": 8.51012307297624e-06,
+      "loss": 0.5153,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4244659841554492,
+      "grad_norm": 0.41305756922410347,
+      "learning_rate": 8.508422601819713e-06,
+      "loss": 0.5241,
+      "step": 1961
+    },
+    {
+      "epoch": 1.425192381903616,
+      "grad_norm": 0.37996646400149187,
+      "learning_rate": 8.506721330902869e-06,
+      "loss": 0.5295,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4259187796517832,
+      "grad_norm": 0.3852696944338915,
+      "learning_rate": 8.505019260613523e-06,
+      "loss": 0.5305,
+      "step": 1963
+    },
+    {
+      "epoch": 1.42664517739995,
+      "grad_norm": 0.3958821782540526,
+      "learning_rate": 8.503316391339668e-06,
+      "loss": 0.5323,
+      "step": 1964
+    },
+    {
+      "epoch": 1.427371575148117,
+      "grad_norm": 0.37831995894612974,
+      "learning_rate": 8.501612723469483e-06,
+      "loss": 0.5181,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4280979728962842,
+      "grad_norm": 0.39926135943658303,
+      "learning_rate": 8.499908257391324e-06,
+      "loss": 0.5417,
+      "step": 1966
+    },
+    {
+      "epoch": 1.428824370644451,
+      "grad_norm": 0.358842675107982,
+      "learning_rate": 8.49820299349373e-06,
+      "loss": 0.5428,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4295507683926179,
+      "grad_norm": 0.40179609610989964,
+      "learning_rate": 8.49649693216543e-06,
+      "loss": 0.5369,
+      "step": 1968
+    },
+    {
+      "epoch": 1.430277166140785,
+      "grad_norm": 0.3789589574251804,
+      "learning_rate": 8.494790073795323e-06,
+      "loss": 0.5423,
+      "step": 1969
+    },
+    {
+      "epoch": 1.431003563888952,
+      "grad_norm": 0.3590287073648434,
+      "learning_rate": 8.493082418772494e-06,
+      "loss": 0.5353,
+      "step": 1970
+    },
+    {
+      "epoch": 1.4317299616371189,
+      "grad_norm": 0.6134094833899469,
+      "learning_rate": 8.491373967486212e-06,
+      "loss": 0.5349,
+      "step": 1971
+    },
+    {
+      "epoch": 1.432456359385286,
+      "grad_norm": 0.36762452749849744,
+      "learning_rate": 8.489664720325928e-06,
+      "loss": 0.5275,
+      "step": 1972
+    },
+    {
+      "epoch": 1.4331827571334528,
+      "grad_norm": 0.3692895289988314,
+      "learning_rate": 8.487954677681269e-06,
+      "loss": 0.54,
+      "step": 1973
+    },
+    {
+      "epoch": 1.4339091548816199,
+      "grad_norm": 0.3763720711023518,
+      "learning_rate": 8.486243839942048e-06,
+      "loss": 0.5268,
+      "step": 1974
+    },
+    {
+      "epoch": 1.434635552629787,
+      "grad_norm": 0.3806714456242494,
+      "learning_rate": 8.48453220749826e-06,
+      "loss": 0.5287,
+      "step": 1975
+    },
+    {
+      "epoch": 1.4353619503779538,
+      "grad_norm": 0.35038099734210776,
+      "learning_rate": 8.482819780740076e-06,
+      "loss": 0.5479,
+      "step": 1976
+    },
+    {
+      "epoch": 1.4360883481261209,
+      "grad_norm": 0.4317281985340517,
+      "learning_rate": 8.481106560057852e-06,
+      "loss": 0.5417,
+      "step": 1977
+    },
+    {
+      "epoch": 1.4368147458742877,
+      "grad_norm": 0.3701376689135592,
+      "learning_rate": 8.479392545842126e-06,
+      "loss": 0.5301,
+      "step": 1978
+    },
+    {
+      "epoch": 1.4375411436224548,
+      "grad_norm": 0.3774269963014525,
+      "learning_rate": 8.477677738483614e-06,
+      "loss": 0.5379,
+      "step": 1979
+    },
+    {
+      "epoch": 1.4382675413706218,
+      "grad_norm": 0.3857892291442041,
+      "learning_rate": 8.475962138373212e-06,
+      "loss": 0.5456,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4389939391187887,
+      "grad_norm": 0.3887872479758726,
+      "learning_rate": 8.474245745902002e-06,
+      "loss": 0.5261,
+      "step": 1981
+    },
+    {
+      "epoch": 1.4397203368669558,
+      "grad_norm": 0.3862173859493997,
+      "learning_rate": 8.47252856146124e-06,
+      "loss": 0.5383,
+      "step": 1982
+    },
+    {
+      "epoch": 1.4404467346151226,
+      "grad_norm": 0.4132140002809266,
+      "learning_rate": 8.470810585442367e-06,
+      "loss": 0.5508,
+      "step": 1983
+    },
+    {
+      "epoch": 1.4411731323632897,
+      "grad_norm": 0.44883189440473237,
+      "learning_rate": 8.469091818237004e-06,
+      "loss": 0.5507,
+      "step": 1984
+    },
+    {
+      "epoch": 1.4418995301114568,
+      "grad_norm": 0.9548703866806727,
+      "learning_rate": 8.467372260236951e-06,
+      "loss": 0.5275,
+      "step": 1985
+    },
+    {
+      "epoch": 1.4426259278596236,
+      "grad_norm": 0.3560892250131109,
+      "learning_rate": 8.465651911834187e-06,
+      "loss": 0.5245,
+      "step": 1986
+    },
+    {
+      "epoch": 1.4433523256077907,
+      "grad_norm": 0.3750927938822984,
+      "learning_rate": 8.463930773420874e-06,
+      "loss": 0.5278,
+      "step": 1987
+    },
+    {
+      "epoch": 1.4440787233559575,
+      "grad_norm": 0.3861372560323569,
+      "learning_rate": 8.462208845389356e-06,
+      "loss": 0.5357,
+      "step": 1988
+    },
+    {
+      "epoch": 1.4448051211041246,
+      "grad_norm": 0.4953428079781934,
+      "learning_rate": 8.460486128132151e-06,
+      "loss": 0.5322,
+      "step": 1989
+    },
+    {
+      "epoch": 1.4455315188522915,
+      "grad_norm": 0.7225665784307765,
+      "learning_rate": 8.458762622041959e-06,
+      "loss": 0.5328,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4462579166004585,
+      "grad_norm": 0.3635088658724107,
+      "learning_rate": 8.457038327511663e-06,
+      "loss": 0.5302,
+      "step": 1991
+    },
+    {
+      "epoch": 1.4469843143486254,
+      "grad_norm": 0.3951491124497268,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.5407,
+      "step": 1992
+    },
+    {
+      "epoch": 1.4477107120967925,
+      "grad_norm": 0.6229750799649701,
+      "learning_rate": 8.45358737470318e-06,
+      "loss": 0.5444,
+      "step": 1993
+    },
+    {
+      "epoch": 1.4484371098449595,
+      "grad_norm": 0.4284237318323039,
+      "learning_rate": 8.451860717211653e-06,
+      "loss": 0.5441,
+      "step": 1994
+    },
+    {
+      "epoch": 1.4491635075931264,
+      "grad_norm": 0.5036938209369309,
+      "learning_rate": 8.450133272853338e-06,
+      "loss": 0.5438,
+      "step": 1995
+    },
+    {
+      "epoch": 1.4498899053412935,
+      "grad_norm": 0.5598656231285419,
+      "learning_rate": 8.448405042022018e-06,
+      "loss": 0.5328,
+      "step": 1996
+    },
+    {
+      "epoch": 1.4506163030894603,
+      "grad_norm": 0.4352745677117185,
+      "learning_rate": 8.44667602511165e-06,
+      "loss": 0.5353,
+      "step": 1997
+    },
+    {
+      "epoch": 1.4513427008376274,
+      "grad_norm": 0.3727822813407849,
+      "learning_rate": 8.444946222516369e-06,
+      "loss": 0.5382,
+      "step": 1998
+    },
+    {
+      "epoch": 1.4520690985857945,
+      "grad_norm": 0.3673506875698348,
+      "learning_rate": 8.443215634630493e-06,
+      "loss": 0.5312,
+      "step": 1999
+    },
+    {
+      "epoch": 1.4527954963339613,
+      "grad_norm": 0.42066624613930265,
+      "learning_rate": 8.441484261848514e-06,
+      "loss": 0.5572,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4535218940821284,
+      "grad_norm": 0.4167581664824996,
+      "learning_rate": 8.43975210456511e-06,
+      "loss": 0.5439,
+      "step": 2001
+    },
+    {
+      "epoch": 1.4542482918302952,
+      "grad_norm": 0.4384370655418264,
+      "learning_rate": 8.438019163175132e-06,
+      "loss": 0.5454,
+      "step": 2002
+    },
+    {
+      "epoch": 1.4549746895784623,
+      "grad_norm": 0.41078736364713037,
+      "learning_rate": 8.436285438073612e-06,
+      "loss": 0.5278,
+      "step": 2003
+    },
+    {
+      "epoch": 1.4557010873266294,
+      "grad_norm": 0.5068341665843389,
+      "learning_rate": 8.43455092965576e-06,
+      "loss": 0.532,
+      "step": 2004
+    },
+    {
+      "epoch": 1.4564274850747962,
+      "grad_norm": 0.3361925311738195,
+      "learning_rate": 8.432815638316964e-06,
+      "loss": 0.5221,
+      "step": 2005
+    },
+    {
+      "epoch": 1.4571538828229633,
+      "grad_norm": 0.3907476605116234,
+      "learning_rate": 8.431079564452794e-06,
+      "loss": 0.5259,
+      "step": 2006
+    },
+    {
+      "epoch": 1.4578802805711302,
+      "grad_norm": 0.39243927084717206,
+      "learning_rate": 8.429342708458991e-06,
+      "loss": 0.5211,
+      "step": 2007
+    },
+    {
+      "epoch": 1.4586066783192972,
+      "grad_norm": 0.4033731808083159,
+      "learning_rate": 8.427605070731482e-06,
+      "loss": 0.5342,
+      "step": 2008
+    },
+    {
+      "epoch": 1.4593330760674643,
+      "grad_norm": 0.38186916648202734,
+      "learning_rate": 8.42586665166637e-06,
+      "loss": 0.5202,
+      "step": 2009
+    },
+    {
+      "epoch": 1.4600594738156312,
+      "grad_norm": 0.3752026451214445,
+      "learning_rate": 8.424127451659933e-06,
+      "loss": 0.5299,
+      "step": 2010
+    },
+    {
+      "epoch": 1.460785871563798,
+      "grad_norm": 0.40256241040537144,
+      "learning_rate": 8.422387471108631e-06,
+      "loss": 0.536,
+      "step": 2011
+    },
+    {
+      "epoch": 1.461512269311965,
+      "grad_norm": 0.38801726970593625,
+      "learning_rate": 8.420646710409099e-06,
+      "loss": 0.5298,
+      "step": 2012
+    },
+    {
+      "epoch": 1.4622386670601322,
+      "grad_norm": 0.3491406745354709,
+      "learning_rate": 8.418905169958152e-06,
+      "loss": 0.5358,
+      "step": 2013
+    },
+    {
+      "epoch": 1.462965064808299,
+      "grad_norm": 0.3743679163458471,
+      "learning_rate": 8.41716285015278e-06,
+      "loss": 0.5434,
+      "step": 2014
+    },
+    {
+      "epoch": 1.463691462556466,
+      "grad_norm": 0.37363032051278366,
+      "learning_rate": 8.415419751390155e-06,
+      "loss": 0.5306,
+      "step": 2015
+    },
+    {
+      "epoch": 1.464417860304633,
+      "grad_norm": 0.3793987257982108,
+      "learning_rate": 8.413675874067622e-06,
+      "loss": 0.5419,
+      "step": 2016
+    },
+    {
+      "epoch": 1.4651442580528,
+      "grad_norm": 0.4060396681017493,
+      "learning_rate": 8.411931218582706e-06,
+      "loss": 0.5306,
+      "step": 2017
+    },
+    {
+      "epoch": 1.465870655800967,
+      "grad_norm": 0.7718495729878065,
+      "learning_rate": 8.410185785333111e-06,
+      "loss": 0.5528,
+      "step": 2018
+    },
+    {
+      "epoch": 1.466597053549134,
+      "grad_norm": 0.3809931135117274,
+      "learning_rate": 8.408439574716712e-06,
+      "loss": 0.5151,
+      "step": 2019
+    },
+    {
+      "epoch": 1.467323451297301,
+      "grad_norm": 0.3464947862227488,
+      "learning_rate": 8.406692587131569e-06,
+      "loss": 0.5268,
+      "step": 2020
+    },
+    {
+      "epoch": 1.4680498490454679,
+      "grad_norm": 0.37560537023157703,
+      "learning_rate": 8.404944822975914e-06,
+      "loss": 0.5582,
+      "step": 2021
+    },
+    {
+      "epoch": 1.468776246793635,
+      "grad_norm": 0.3527908030684601,
+      "learning_rate": 8.403196282648156e-06,
+      "loss": 0.5464,
+      "step": 2022
+    },
+    {
+      "epoch": 1.469502644541802,
+      "grad_norm": 0.3744135464359761,
+      "learning_rate": 8.401446966546885e-06,
+      "loss": 0.5411,
+      "step": 2023
+    },
+    {
+      "epoch": 1.4702290422899689,
+      "grad_norm": 0.3898440069807425,
+      "learning_rate": 8.399696875070864e-06,
+      "loss": 0.5505,
+      "step": 2024
+    },
+    {
+      "epoch": 1.470955440038136,
+      "grad_norm": 0.44165838423074744,
+      "learning_rate": 8.397946008619035e-06,
+      "loss": 0.5151,
+      "step": 2025
+    },
+    {
+      "epoch": 1.4716818377863028,
+      "grad_norm": 0.4152275308276423,
+      "learning_rate": 8.396194367590515e-06,
+      "loss": 0.5274,
+      "step": 2026
+    },
+    {
+      "epoch": 1.4724082355344699,
+      "grad_norm": 0.3324160779039609,
+      "learning_rate": 8.394441952384597e-06,
+      "loss": 0.5406,
+      "step": 2027
+    },
+    {
+      "epoch": 1.473134633282637,
+      "grad_norm": 0.41106911809184316,
+      "learning_rate": 8.392688763400755e-06,
+      "loss": 0.532,
+      "step": 2028
+    },
+    {
+      "epoch": 1.4738610310308038,
+      "grad_norm": 0.3924212014605282,
+      "learning_rate": 8.390934801038632e-06,
+      "loss": 0.5233,
+      "step": 2029
+    },
+    {
+      "epoch": 1.4745874287789709,
+      "grad_norm": 0.414921189006647,
+      "learning_rate": 8.389180065698055e-06,
+      "loss": 0.5319,
+      "step": 2030
+    },
+    {
+      "epoch": 1.4753138265271377,
+      "grad_norm": 0.8311008021451181,
+      "learning_rate": 8.387424557779022e-06,
+      "loss": 0.551,
+      "step": 2031
+    },
+    {
+      "epoch": 1.4760402242753048,
+      "grad_norm": 0.361198159956959,
+      "learning_rate": 8.385668277681709e-06,
+      "loss": 0.5131,
+      "step": 2032
+    },
+    {
+      "epoch": 1.4767666220234719,
+      "grad_norm": 0.38745901241674185,
+      "learning_rate": 8.383911225806468e-06,
+      "loss": 0.5474,
+      "step": 2033
+    },
+    {
+      "epoch": 1.4774930197716387,
+      "grad_norm": 0.37053434565256693,
+      "learning_rate": 8.382153402553825e-06,
+      "loss": 0.5409,
+      "step": 2034
+    },
+    {
+      "epoch": 1.4782194175198056,
+      "grad_norm": 0.3666466711492958,
+      "learning_rate": 8.380394808324484e-06,
+      "loss": 0.5218,
+      "step": 2035
+    },
+    {
+      "epoch": 1.4789458152679726,
+      "grad_norm": 0.43161781550679396,
+      "learning_rate": 8.378635443519327e-06,
+      "loss": 0.5225,
+      "step": 2036
+    },
+    {
+      "epoch": 1.4796722130161397,
+      "grad_norm": 0.47557759375567127,
+      "learning_rate": 8.376875308539406e-06,
+      "loss": 0.5555,
+      "step": 2037
+    },
+    {
+      "epoch": 1.4803986107643066,
+      "grad_norm": 0.4422804846214722,
+      "learning_rate": 8.375114403785953e-06,
+      "loss": 0.5229,
+      "step": 2038
+    },
+    {
+      "epoch": 1.4811250085124736,
+      "grad_norm": 0.35316778327064985,
+      "learning_rate": 8.373352729660373e-06,
+      "loss": 0.532,
+      "step": 2039
+    },
+    {
+      "epoch": 1.4818514062606405,
+      "grad_norm": 0.43635553639514263,
+      "learning_rate": 8.371590286564247e-06,
+      "loss": 0.5267,
+      "step": 2040
+    },
+    {
+      "epoch": 1.4825778040088076,
+      "grad_norm": 0.3647953718925828,
+      "learning_rate": 8.369827074899333e-06,
+      "loss": 0.5172,
+      "step": 2041
+    },
+    {
+      "epoch": 1.4833042017569746,
+      "grad_norm": 0.43758131127371425,
+      "learning_rate": 8.368063095067564e-06,
+      "loss": 0.5299,
+      "step": 2042
+    },
+    {
+      "epoch": 1.4840305995051415,
+      "grad_norm": 0.3481965420729561,
+      "learning_rate": 8.366298347471043e-06,
+      "loss": 0.5529,
+      "step": 2043
+    },
+    {
+      "epoch": 1.4847569972533086,
+      "grad_norm": 0.3809271562611856,
+      "learning_rate": 8.364532832512055e-06,
+      "loss": 0.5298,
+      "step": 2044
+    },
+    {
+      "epoch": 1.4854833950014754,
+      "grad_norm": 0.39900438510168235,
+      "learning_rate": 8.362766550593055e-06,
+      "loss": 0.553,
+      "step": 2045
+    },
+    {
+      "epoch": 1.4862097927496425,
+      "grad_norm": 0.4024051219002026,
+      "learning_rate": 8.360999502116678e-06,
+      "loss": 0.5251,
+      "step": 2046
+    },
+    {
+      "epoch": 1.4869361904978096,
+      "grad_norm": 0.5664636582845303,
+      "learning_rate": 8.359231687485724e-06,
+      "loss": 0.5268,
+      "step": 2047
+    },
+    {
+      "epoch": 1.4876625882459764,
+      "grad_norm": 0.321380891619596,
+      "learning_rate": 8.357463107103182e-06,
+      "loss": 0.5262,
+      "step": 2048
+    },
+    {
+      "epoch": 1.4883889859941435,
+      "grad_norm": 0.3781206568900052,
+      "learning_rate": 8.355693761372203e-06,
+      "loss": 0.5265,
+      "step": 2049
+    },
+    {
+      "epoch": 1.4891153837423103,
+      "grad_norm": 0.4901907426957913,
+      "learning_rate": 8.353923650696119e-06,
+      "loss": 0.5352,
+      "step": 2050
+    },
+    {
+      "epoch": 1.4898417814904774,
+      "grad_norm": 0.4383754725203851,
+      "learning_rate": 8.352152775478431e-06,
+      "loss": 0.5351,
+      "step": 2051
+    },
+    {
+      "epoch": 1.4905681792386445,
+      "grad_norm": 0.3793071731555198,
+      "learning_rate": 8.350381136122823e-06,
+      "loss": 0.533,
+      "step": 2052
+    },
+    {
+      "epoch": 1.4912945769868113,
+      "grad_norm": 0.3936571281921193,
+      "learning_rate": 8.348608733033143e-06,
+      "loss": 0.5423,
+      "step": 2053
+    },
+    {
+      "epoch": 1.4920209747349782,
+      "grad_norm": 0.5586778199848161,
+      "learning_rate": 8.34683556661342e-06,
+      "loss": 0.542,
+      "step": 2054
+    },
+    {
+      "epoch": 1.4927473724831453,
+      "grad_norm": 0.4083828436352312,
+      "learning_rate": 8.345061637267858e-06,
+      "loss": 0.5437,
+      "step": 2055
+    },
+    {
+      "epoch": 1.4934737702313123,
+      "grad_norm": 0.39828639694777723,
+      "learning_rate": 8.343286945400827e-06,
+      "loss": 0.5404,
+      "step": 2056
+    },
+    {
+      "epoch": 1.4942001679794792,
+      "grad_norm": 0.4341716027611792,
+      "learning_rate": 8.341511491416877e-06,
+      "loss": 0.5419,
+      "step": 2057
+    },
+    {
+      "epoch": 1.4949265657276463,
+      "grad_norm": 0.3771833000560631,
+      "learning_rate": 8.339735275720735e-06,
+      "loss": 0.5272,
+      "step": 2058
+    },
+    {
+      "epoch": 1.4956529634758131,
+      "grad_norm": 0.4369176038701713,
+      "learning_rate": 8.337958298717293e-06,
+      "loss": 0.5344,
+      "step": 2059
+    },
+    {
+      "epoch": 1.4963793612239802,
+      "grad_norm": 0.3844043904597473,
+      "learning_rate": 8.336180560811619e-06,
+      "loss": 0.5323,
+      "step": 2060
+    },
+    {
+      "epoch": 1.4971057589721473,
+      "grad_norm": 0.3560037681982256,
+      "learning_rate": 8.334402062408962e-06,
+      "loss": 0.53,
+      "step": 2061
+    },
+    {
+      "epoch": 1.4978321567203141,
+      "grad_norm": 0.3683510079298262,
+      "learning_rate": 8.332622803914734e-06,
+      "loss": 0.5232,
+      "step": 2062
+    },
+    {
+      "epoch": 1.4985585544684812,
+      "grad_norm": 0.37159086861045465,
+      "learning_rate": 8.330842785734525e-06,
+      "loss": 0.5415,
+      "step": 2063
+    },
+    {
+      "epoch": 1.499284952216648,
+      "grad_norm": 0.6276047408660594,
+      "learning_rate": 8.3290620082741e-06,
+      "loss": 0.5279,
+      "step": 2064
+    },
+    {
+      "epoch": 1.500011349964815,
+      "grad_norm": 0.372171797078306,
+      "learning_rate": 8.327280471939392e-06,
+      "loss": 0.534,
+      "step": 2065
+    },
+    {
+      "epoch": 1.5007377477129822,
+      "grad_norm": 0.36192295787407436,
+      "learning_rate": 8.325498177136514e-06,
+      "loss": 0.5276,
+      "step": 2066
+    },
+    {
+      "epoch": 1.501464145461149,
+      "grad_norm": 0.4151167746324928,
+      "learning_rate": 8.323715124271745e-06,
+      "loss": 0.5465,
+      "step": 2067
+    },
+    {
+      "epoch": 1.5021905432093159,
+      "grad_norm": 0.5345148655537466,
+      "learning_rate": 8.32193131375154e-06,
+      "loss": 0.5248,
+      "step": 2068
+    },
+    {
+      "epoch": 1.502916940957483,
+      "grad_norm": 0.4010625713543608,
+      "learning_rate": 8.320146745982528e-06,
+      "loss": 0.5394,
+      "step": 2069
+    },
+    {
+      "epoch": 1.50364333870565,
+      "grad_norm": 0.4242426717100806,
+      "learning_rate": 8.318361421371507e-06,
+      "loss": 0.5432,
+      "step": 2070
+    },
+    {
+      "epoch": 1.504369736453817,
+      "grad_norm": 0.3769937648385446,
+      "learning_rate": 8.31657534032545e-06,
+      "loss": 0.5404,
+      "step": 2071
+    },
+    {
+      "epoch": 1.505096134201984,
+      "grad_norm": 0.4002762206548443,
+      "learning_rate": 8.314788503251506e-06,
+      "loss": 0.5268,
+      "step": 2072
+    },
+    {
+      "epoch": 1.5058225319501508,
+      "grad_norm": 0.49513683689221805,
+      "learning_rate": 8.313000910556986e-06,
+      "loss": 0.5325,
+      "step": 2073
+    },
+    {
+      "epoch": 1.5065489296983179,
+      "grad_norm": 0.37293996421107484,
+      "learning_rate": 8.311212562649383e-06,
+      "loss": 0.5347,
+      "step": 2074
+    },
+    {
+      "epoch": 1.507275327446485,
+      "grad_norm": 0.3735325900469624,
+      "learning_rate": 8.30942345993636e-06,
+      "loss": 0.541,
+      "step": 2075
+    },
+    {
+      "epoch": 1.508001725194652,
+      "grad_norm": 0.513815208479049,
+      "learning_rate": 8.307633602825746e-06,
+      "loss": 0.5335,
+      "step": 2076
+    },
+    {
+      "epoch": 1.5087281229428189,
+      "grad_norm": 0.42651791780626036,
+      "learning_rate": 8.305842991725552e-06,
+      "loss": 0.5359,
+      "step": 2077
+    },
+    {
+      "epoch": 1.5094545206909857,
+      "grad_norm": 0.7149161389443535,
+      "learning_rate": 8.304051627043952e-06,
+      "loss": 0.5337,
+      "step": 2078
+    },
+    {
+      "epoch": 1.5101809184391528,
+      "grad_norm": 0.5309869942657326,
+      "learning_rate": 8.302259509189299e-06,
+      "loss": 0.5139,
+      "step": 2079
+    },
+    {
+      "epoch": 1.5109073161873199,
+      "grad_norm": 0.44461112251803003,
+      "learning_rate": 8.30046663857011e-06,
+      "loss": 0.5289,
+      "step": 2080
+    },
+    {
+      "epoch": 1.511633713935487,
+      "grad_norm": 0.3676877067349289,
+      "learning_rate": 8.298673015595083e-06,
+      "loss": 0.5325,
+      "step": 2081
+    },
+    {
+      "epoch": 1.5123601116836538,
+      "grad_norm": 0.5061828033162785,
+      "learning_rate": 8.296878640673077e-06,
+      "loss": 0.5153,
+      "step": 2082
+    },
+    {
+      "epoch": 1.5130865094318207,
+      "grad_norm": 0.414329324752058,
+      "learning_rate": 8.29508351421313e-06,
+      "loss": 0.5184,
+      "step": 2083
+    },
+    {
+      "epoch": 1.5138129071799877,
+      "grad_norm": 0.44121955910229965,
+      "learning_rate": 8.293287636624447e-06,
+      "loss": 0.5379,
+      "step": 2084
+    },
+    {
+      "epoch": 1.5145393049281548,
+      "grad_norm": 0.4154035008830126,
+      "learning_rate": 8.291491008316409e-06,
+      "loss": 0.5287,
+      "step": 2085
+    },
+    {
+      "epoch": 1.5152657026763217,
+      "grad_norm": 0.3225024595739773,
+      "learning_rate": 8.289693629698564e-06,
+      "loss": 0.5381,
+      "step": 2086
+    },
+    {
+      "epoch": 1.5159921004244887,
+      "grad_norm": 0.5334881633732634,
+      "learning_rate": 8.287895501180632e-06,
+      "loss": 0.5442,
+      "step": 2087
+    },
+    {
+      "epoch": 1.5167184981726556,
+      "grad_norm": 0.44290513871966686,
+      "learning_rate": 8.286096623172506e-06,
+      "loss": 0.5247,
+      "step": 2088
+    },
+    {
+      "epoch": 1.5174448959208227,
+      "grad_norm": 0.4355594276107615,
+      "learning_rate": 8.284296996084244e-06,
+      "loss": 0.5331,
+      "step": 2089
+    },
+    {
+      "epoch": 1.5181712936689897,
+      "grad_norm": 0.47966811639145174,
+      "learning_rate": 8.282496620326085e-06,
+      "loss": 0.5411,
+      "step": 2090
+    },
+    {
+      "epoch": 1.5188976914171566,
+      "grad_norm": 0.43302335492933264,
+      "learning_rate": 8.280695496308428e-06,
+      "loss": 0.5407,
+      "step": 2091
+    },
+    {
+      "epoch": 1.5196240891653234,
+      "grad_norm": 0.3857374366033102,
+      "learning_rate": 8.278893624441849e-06,
+      "loss": 0.5361,
+      "step": 2092
+    },
+    {
+      "epoch": 1.5203504869134905,
+      "grad_norm": 0.375465296561444,
+      "learning_rate": 8.27709100513709e-06,
+      "loss": 0.5375,
+      "step": 2093
+    },
+    {
+      "epoch": 1.5210768846616576,
+      "grad_norm": 0.3682581338047892,
+      "learning_rate": 8.275287638805069e-06,
+      "loss": 0.5327,
+      "step": 2094
+    },
+    {
+      "epoch": 1.5218032824098247,
+      "grad_norm": 0.3535435043848988,
+      "learning_rate": 8.273483525856871e-06,
+      "loss": 0.5423,
+      "step": 2095
+    },
+    {
+      "epoch": 1.5225296801579915,
+      "grad_norm": 0.36136545072253023,
+      "learning_rate": 8.27167866670375e-06,
+      "loss": 0.5315,
+      "step": 2096
+    },
+    {
+      "epoch": 1.5232560779061584,
+      "grad_norm": 0.4133601831129233,
+      "learning_rate": 8.269873061757133e-06,
+      "loss": 0.5307,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5239824756543254,
+      "grad_norm": 0.5370894632116522,
+      "learning_rate": 8.268066711428614e-06,
+      "loss": 0.543,
+      "step": 2098
+    },
+    {
+      "epoch": 1.5247088734024925,
+      "grad_norm": 0.3613559837813718,
+      "learning_rate": 8.266259616129959e-06,
+      "loss": 0.5283,
+      "step": 2099
+    },
+    {
+      "epoch": 1.5254352711506596,
+      "grad_norm": 0.35068415153350707,
+      "learning_rate": 8.264451776273104e-06,
+      "loss": 0.5246,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5261616688988264,
+      "grad_norm": 0.4693794087880394,
+      "learning_rate": 8.262643192270152e-06,
+      "loss": 0.5271,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5268880666469933,
+      "grad_norm": 0.36685476737449635,
+      "learning_rate": 8.26083386453338e-06,
+      "loss": 0.5402,
+      "step": 2102
+    },
+    {
+      "epoch": 1.5276144643951604,
+      "grad_norm": 0.4240709701742305,
+      "learning_rate": 8.259023793475233e-06,
+      "loss": 0.531,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5283408621433274,
+      "grad_norm": 0.3992723948801737,
+      "learning_rate": 8.257212979508321e-06,
+      "loss": 0.5403,
+      "step": 2104
+    },
+    {
+      "epoch": 1.5290672598914945,
+      "grad_norm": 0.39217915177008994,
+      "learning_rate": 8.25540142304543e-06,
+      "loss": 0.5176,
+      "step": 2105
+    },
+    {
+      "epoch": 1.5297936576396614,
+      "grad_norm": 0.3691939997812454,
+      "learning_rate": 8.253589124499513e-06,
+      "loss": 0.5234,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5305200553878282,
+      "grad_norm": 0.34426922828140977,
+      "learning_rate": 8.251776084283687e-06,
+      "loss": 0.5255,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5312464531359953,
+      "grad_norm": 0.4228685867855801,
+      "learning_rate": 8.249962302811246e-06,
+      "loss": 0.5355,
+      "step": 2108
+    },
+    {
+      "epoch": 1.5319728508841624,
+      "grad_norm": 0.6702335615461068,
+      "learning_rate": 8.248147780495648e-06,
+      "loss": 0.53,
+      "step": 2109
+    },
+    {
+      "epoch": 1.5326992486323292,
+      "grad_norm": 0.3620865144375418,
+      "learning_rate": 8.246332517750524e-06,
+      "loss": 0.5256,
+      "step": 2110
+    },
+    {
+      "epoch": 1.533425646380496,
+      "grad_norm": 0.3508763292761052,
+      "learning_rate": 8.244516514989667e-06,
+      "loss": 0.5156,
+      "step": 2111
+    },
+    {
+      "epoch": 1.5341520441286631,
+      "grad_norm": 0.42627555830890596,
+      "learning_rate": 8.242699772627044e-06,
+      "loss": 0.5355,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5348784418768302,
+      "grad_norm": 0.3684308599352157,
+      "learning_rate": 8.240882291076794e-06,
+      "loss": 0.5182,
+      "step": 2113
+    },
+    {
+      "epoch": 1.5356048396249973,
+      "grad_norm": 0.43590036510920216,
+      "learning_rate": 8.239064070753213e-06,
+      "loss": 0.5165,
+      "step": 2114
+    },
+    {
+      "epoch": 1.5363312373731641,
+      "grad_norm": 0.3774981089091263,
+      "learning_rate": 8.237245112070775e-06,
+      "loss": 0.5433,
+      "step": 2115
+    },
+    {
+      "epoch": 1.537057635121331,
+      "grad_norm": 0.5796489601067345,
+      "learning_rate": 8.235425415444123e-06,
+      "loss": 0.5255,
+      "step": 2116
+    },
+    {
+      "epoch": 1.537784032869498,
+      "grad_norm": 0.4040368368012006,
+      "learning_rate": 8.233604981288059e-06,
+      "loss": 0.5207,
+      "step": 2117
+    },
+    {
+      "epoch": 1.5385104306176651,
+      "grad_norm": 0.4050182472662472,
+      "learning_rate": 8.231783810017562e-06,
+      "loss": 0.5281,
+      "step": 2118
+    },
+    {
+      "epoch": 1.5392368283658322,
+      "grad_norm": 0.43834797443504225,
+      "learning_rate": 8.229961902047777e-06,
+      "loss": 0.5269,
+      "step": 2119
+    },
+    {
+      "epoch": 1.539963226113999,
+      "grad_norm": 0.3867432482382013,
+      "learning_rate": 8.228139257794012e-06,
+      "loss": 0.5305,
+      "step": 2120
+    },
+    {
+      "epoch": 1.540689623862166,
+      "grad_norm": 0.3334578775651982,
+      "learning_rate": 8.22631587767175e-06,
+      "loss": 0.53,
+      "step": 2121
+    },
+    {
+      "epoch": 1.541416021610333,
+      "grad_norm": 0.3670398743223365,
+      "learning_rate": 8.22449176209664e-06,
+      "loss": 0.5321,
+      "step": 2122
+    },
+    {
+      "epoch": 1.5421424193585,
+      "grad_norm": 0.4172991422978223,
+      "learning_rate": 8.222666911484494e-06,
+      "loss": 0.5422,
+      "step": 2123
+    },
+    {
+      "epoch": 1.5428688171066671,
+      "grad_norm": 0.394926083100875,
+      "learning_rate": 8.220841326251297e-06,
+      "loss": 0.5323,
+      "step": 2124
+    },
+    {
+      "epoch": 1.543595214854834,
+      "grad_norm": 0.3701822483507409,
+      "learning_rate": 8.219015006813197e-06,
+      "loss": 0.5333,
+      "step": 2125
+    },
+    {
+      "epoch": 1.5443216126030008,
+      "grad_norm": 0.3830765074130067,
+      "learning_rate": 8.217187953586512e-06,
+      "loss": 0.533,
+      "step": 2126
+    },
+    {
+      "epoch": 1.545048010351168,
+      "grad_norm": 0.3495916719722912,
+      "learning_rate": 8.215360166987728e-06,
+      "loss": 0.529,
+      "step": 2127
+    },
+    {
+      "epoch": 1.545774408099335,
+      "grad_norm": 0.38420382108991424,
+      "learning_rate": 8.213531647433494e-06,
+      "loss": 0.5231,
+      "step": 2128
+    },
+    {
+      "epoch": 1.5465008058475018,
+      "grad_norm": 0.3664634654071994,
+      "learning_rate": 8.211702395340633e-06,
+      "loss": 0.5232,
+      "step": 2129
+    },
+    {
+      "epoch": 1.547227203595669,
+      "grad_norm": 0.39373304275204196,
+      "learning_rate": 8.209872411126127e-06,
+      "loss": 0.5404,
+      "step": 2130
+    },
+    {
+      "epoch": 1.5479536013438358,
+      "grad_norm": 0.5226087869025603,
+      "learning_rate": 8.208041695207134e-06,
+      "loss": 0.5271,
+      "step": 2131
+    },
+    {
+      "epoch": 1.5486799990920028,
+      "grad_norm": 0.37897080800573535,
+      "learning_rate": 8.20621024800097e-06,
+      "loss": 0.5257,
+      "step": 2132
+    },
+    {
+      "epoch": 1.54940639684017,
+      "grad_norm": 0.3607192702673823,
+      "learning_rate": 8.204378069925121e-06,
+      "loss": 0.5278,
+      "step": 2133
+    },
+    {
+      "epoch": 1.5501327945883367,
+      "grad_norm": 0.4491567190506299,
+      "learning_rate": 8.202545161397242e-06,
+      "loss": 0.5212,
+      "step": 2134
+    },
+    {
+      "epoch": 1.5508591923365036,
+      "grad_norm": 0.4648783169043336,
+      "learning_rate": 8.200711522835153e-06,
+      "loss": 0.5255,
+      "step": 2135
+    },
+    {
+      "epoch": 1.5515855900846707,
+      "grad_norm": 0.40825688941196847,
+      "learning_rate": 8.198877154656838e-06,
+      "loss": 0.5187,
+      "step": 2136
+    },
+    {
+      "epoch": 1.5523119878328377,
+      "grad_norm": 0.5055858786813612,
+      "learning_rate": 8.19704205728045e-06,
+      "loss": 0.5214,
+      "step": 2137
+    },
+    {
+      "epoch": 1.5530383855810048,
+      "grad_norm": 0.452686470398235,
+      "learning_rate": 8.195206231124309e-06,
+      "loss": 0.5272,
+      "step": 2138
+    },
+    {
+      "epoch": 1.5537647833291717,
+      "grad_norm": 0.37942785376928945,
+      "learning_rate": 8.193369676606896e-06,
+      "loss": 0.5348,
+      "step": 2139
+    },
+    {
+      "epoch": 1.5544911810773385,
+      "grad_norm": 0.4065149027015626,
+      "learning_rate": 8.191532394146865e-06,
+      "loss": 0.5292,
+      "step": 2140
+    },
+    {
+      "epoch": 1.5552175788255056,
+      "grad_norm": 0.5083598648566944,
+      "learning_rate": 8.189694384163032e-06,
+      "loss": 0.5176,
+      "step": 2141
+    },
+    {
+      "epoch": 1.5559439765736727,
+      "grad_norm": 0.3722742765668192,
+      "learning_rate": 8.187855647074376e-06,
+      "loss": 0.5262,
+      "step": 2142
+    },
+    {
+      "epoch": 1.5566703743218397,
+      "grad_norm": 0.40226452791696343,
+      "learning_rate": 8.186016183300052e-06,
+      "loss": 0.5231,
+      "step": 2143
+    },
+    {
+      "epoch": 1.5573967720700066,
+      "grad_norm": 0.43419760435625404,
+      "learning_rate": 8.184175993259367e-06,
+      "loss": 0.5258,
+      "step": 2144
+    },
+    {
+      "epoch": 1.5581231698181734,
+      "grad_norm": 0.4013434470364981,
+      "learning_rate": 8.182335077371803e-06,
+      "loss": 0.5259,
+      "step": 2145
+    },
+    {
+      "epoch": 1.5588495675663405,
+      "grad_norm": 0.5035945890529423,
+      "learning_rate": 8.180493436057008e-06,
+      "loss": 0.5227,
+      "step": 2146
+    },
+    {
+      "epoch": 1.5595759653145076,
+      "grad_norm": 0.37150966714124883,
+      "learning_rate": 8.178651069734787e-06,
+      "loss": 0.5336,
+      "step": 2147
+    },
+    {
+      "epoch": 1.5603023630626747,
+      "grad_norm": 0.4429586606794134,
+      "learning_rate": 8.17680797882512e-06,
+      "loss": 0.5453,
+      "step": 2148
+    },
+    {
+      "epoch": 1.5610287608108415,
+      "grad_norm": 0.37270998378690173,
+      "learning_rate": 8.174964163748142e-06,
+      "loss": 0.5269,
+      "step": 2149
+    },
+    {
+      "epoch": 1.5617551585590084,
+      "grad_norm": 0.4122813391997847,
+      "learning_rate": 8.173119624924164e-06,
+      "loss": 0.5204,
+      "step": 2150
+    },
+    {
+      "epoch": 1.5624815563071754,
+      "grad_norm": 0.37256623087813406,
+      "learning_rate": 8.171274362773657e-06,
+      "loss": 0.5284,
+      "step": 2151
+    },
+    {
+      "epoch": 1.5632079540553425,
+      "grad_norm": 0.366519633839237,
+      "learning_rate": 8.169428377717253e-06,
+      "loss": 0.5292,
+      "step": 2152
+    },
+    {
+      "epoch": 1.5639343518035094,
+      "grad_norm": 0.39571030872429297,
+      "learning_rate": 8.167581670175752e-06,
+      "loss": 0.5244,
+      "step": 2153
+    },
+    {
+      "epoch": 1.5646607495516764,
+      "grad_norm": 0.49398026531249895,
+      "learning_rate": 8.165734240570124e-06,
+      "loss": 0.5162,
+      "step": 2154
+    },
+    {
+      "epoch": 1.5653871472998433,
+      "grad_norm": 0.3702564028006646,
+      "learning_rate": 8.163886089321493e-06,
+      "loss": 0.5471,
+      "step": 2155
+    },
+    {
+      "epoch": 1.5661135450480104,
+      "grad_norm": 0.3958082047337379,
+      "learning_rate": 8.162037216851158e-06,
+      "loss": 0.5361,
+      "step": 2156
+    },
+    {
+      "epoch": 1.5668399427961774,
+      "grad_norm": 0.38027394888118066,
+      "learning_rate": 8.160187623580575e-06,
+      "loss": 0.5101,
+      "step": 2157
+    },
+    {
+      "epoch": 1.5675663405443443,
+      "grad_norm": 0.6188493033488481,
+      "learning_rate": 8.158337309931365e-06,
+      "loss": 0.5206,
+      "step": 2158
+    },
+    {
+      "epoch": 1.5682927382925111,
+      "grad_norm": 0.485014339706956,
+      "learning_rate": 8.15648627632532e-06,
+      "loss": 0.522,
+      "step": 2159
+    },
+    {
+      "epoch": 1.5690191360406782,
+      "grad_norm": 0.3372921872528581,
+      "learning_rate": 8.154634523184389e-06,
+      "loss": 0.5204,
+      "step": 2160
+    },
+    {
+      "epoch": 1.5697455337888453,
+      "grad_norm": 0.3775246242830134,
+      "learning_rate": 8.152782050930685e-06,
+      "loss": 0.5139,
+      "step": 2161
+    },
+    {
+      "epoch": 1.5704719315370124,
+      "grad_norm": 0.3441463438997507,
+      "learning_rate": 8.150928859986488e-06,
+      "loss": 0.5341,
+      "step": 2162
+    },
+    {
+      "epoch": 1.5711983292851792,
+      "grad_norm": 0.3958596216052891,
+      "learning_rate": 8.149074950774244e-06,
+      "loss": 0.5325,
+      "step": 2163
+    },
+    {
+      "epoch": 1.571924727033346,
+      "grad_norm": 0.3555124159077814,
+      "learning_rate": 8.14722032371656e-06,
+      "loss": 0.531,
+      "step": 2164
+    },
+    {
+      "epoch": 1.5726511247815131,
+      "grad_norm": 0.4197910562498298,
+      "learning_rate": 8.145364979236201e-06,
+      "loss": 0.5183,
+      "step": 2165
+    },
+    {
+      "epoch": 1.5733775225296802,
+      "grad_norm": 0.38377856071557065,
+      "learning_rate": 8.143508917756105e-06,
+      "loss": 0.539,
+      "step": 2166
+    },
+    {
+      "epoch": 1.5741039202778473,
+      "grad_norm": 0.3498504206619673,
+      "learning_rate": 8.14165213969937e-06,
+      "loss": 0.5318,
+      "step": 2167
+    },
+    {
+      "epoch": 1.5748303180260141,
+      "grad_norm": 0.35939784890857573,
+      "learning_rate": 8.139794645489252e-06,
+      "loss": 0.5244,
+      "step": 2168
+    },
+    {
+      "epoch": 1.575556715774181,
+      "grad_norm": 0.5191106846293712,
+      "learning_rate": 8.13793643554918e-06,
+      "loss": 0.5236,
+      "step": 2169
+    },
+    {
+      "epoch": 1.576283113522348,
+      "grad_norm": 0.3754443761805671,
+      "learning_rate": 8.13607751030274e-06,
+      "loss": 0.5218,
+      "step": 2170
+    },
+    {
+      "epoch": 1.5770095112705151,
+      "grad_norm": 0.3765275515814297,
+      "learning_rate": 8.13421787017368e-06,
+      "loss": 0.525,
+      "step": 2171
+    },
+    {
+      "epoch": 1.577735909018682,
+      "grad_norm": 0.3619776638617869,
+      "learning_rate": 8.132357515585913e-06,
+      "loss": 0.5387,
+      "step": 2172
+    },
+    {
+      "epoch": 1.578462306766849,
+      "grad_norm": 0.43674111006208605,
+      "learning_rate": 8.130496446963518e-06,
+      "loss": 0.528,
+      "step": 2173
+    },
+    {
+      "epoch": 1.579188704515016,
+      "grad_norm": 0.45352096537017633,
+      "learning_rate": 8.128634664730734e-06,
+      "loss": 0.5223,
+      "step": 2174
+    },
+    {
+      "epoch": 1.579915102263183,
+      "grad_norm": 0.3638288881531629,
+      "learning_rate": 8.126772169311959e-06,
+      "loss": 0.5272,
+      "step": 2175
+    },
+    {
+      "epoch": 1.58064150001135,
+      "grad_norm": 0.3678937439443033,
+      "learning_rate": 8.124908961131759e-06,
+      "loss": 0.5327,
+      "step": 2176
+    },
+    {
+      "epoch": 1.581367897759517,
+      "grad_norm": 0.42819149568379095,
+      "learning_rate": 8.123045040614859e-06,
+      "loss": 0.5273,
+      "step": 2177
+    },
+    {
+      "epoch": 1.5820942955076838,
+      "grad_norm": 0.451401867343068,
+      "learning_rate": 8.121180408186151e-06,
+      "loss": 0.52,
+      "step": 2178
+    },
+    {
+      "epoch": 1.5828206932558508,
+      "grad_norm": 0.35404518347187564,
+      "learning_rate": 8.119315064270683e-06,
+      "loss": 0.5418,
+      "step": 2179
+    },
+    {
+      "epoch": 1.583547091004018,
+      "grad_norm": 0.3874862248814074,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.515,
+      "step": 2180
+    },
+    {
+      "epoch": 1.584273488752185,
+      "grad_norm": 0.35996595384110947,
+      "learning_rate": 8.115582243680484e-06,
+      "loss": 0.5231,
+      "step": 2181
+    },
+    {
+      "epoch": 1.5849998865003518,
+      "grad_norm": 0.49032221582780916,
+      "learning_rate": 8.113714767856668e-06,
+      "loss": 0.533,
+      "step": 2182
+    },
+    {
+      "epoch": 1.5857262842485187,
+      "grad_norm": 0.3789203895555697,
+      "learning_rate": 8.111846582247917e-06,
+      "loss": 0.5204,
+      "step": 2183
+    },
+    {
+      "epoch": 1.5864526819966858,
+      "grad_norm": 0.44070282653119874,
+      "learning_rate": 8.109977687280095e-06,
+      "loss": 0.5182,
+      "step": 2184
+    },
+    {
+      "epoch": 1.5871790797448528,
+      "grad_norm": 0.3734368116904144,
+      "learning_rate": 8.108108083379224e-06,
+      "loss": 0.5239,
+      "step": 2185
+    },
+    {
+      "epoch": 1.58790547749302,
+      "grad_norm": 0.7794357692303779,
+      "learning_rate": 8.106237770971486e-06,
+      "loss": 0.5348,
+      "step": 2186
+    },
+    {
+      "epoch": 1.5886318752411868,
+      "grad_norm": 0.40048407359634847,
+      "learning_rate": 8.10436675048323e-06,
+      "loss": 0.5329,
+      "step": 2187
+    },
+    {
+      "epoch": 1.5893582729893536,
+      "grad_norm": 0.41996079281176507,
+      "learning_rate": 8.102495022340962e-06,
+      "loss": 0.5348,
+      "step": 2188
+    },
+    {
+      "epoch": 1.5900846707375207,
+      "grad_norm": 0.4141606229227644,
+      "learning_rate": 8.100622586971349e-06,
+      "loss": 0.5231,
+      "step": 2189
+    },
+    {
+      "epoch": 1.5908110684856878,
+      "grad_norm": 0.39613241249403197,
+      "learning_rate": 8.098749444801226e-06,
+      "loss": 0.5154,
+      "step": 2190
+    },
+    {
+      "epoch": 1.5915374662338548,
+      "grad_norm": 0.40699509976783327,
+      "learning_rate": 8.096875596257578e-06,
+      "loss": 0.5361,
+      "step": 2191
+    },
+    {
+      "epoch": 1.5922638639820217,
+      "grad_norm": 0.39720514745810015,
+      "learning_rate": 8.095001041767561e-06,
+      "loss": 0.5384,
+      "step": 2192
+    },
+    {
+      "epoch": 1.5929902617301885,
+      "grad_norm": 0.379944597113303,
+      "learning_rate": 8.093125781758485e-06,
+      "loss": 0.5191,
+      "step": 2193
+    },
+    {
+      "epoch": 1.5937166594783556,
+      "grad_norm": 0.40124115305103786,
+      "learning_rate": 8.091249816657826e-06,
+      "loss": 0.5319,
+      "step": 2194
+    },
+    {
+      "epoch": 1.5944430572265227,
+      "grad_norm": 0.36199001054735935,
+      "learning_rate": 8.089373146893216e-06,
+      "loss": 0.53,
+      "step": 2195
+    },
+    {
+      "epoch": 1.5951694549746895,
+      "grad_norm": 0.3839271041750099,
+      "learning_rate": 8.087495772892455e-06,
+      "loss": 0.5209,
+      "step": 2196
+    },
+    {
+      "epoch": 1.5958958527228566,
+      "grad_norm": 0.3633683466203577,
+      "learning_rate": 8.085617695083493e-06,
+      "loss": 0.5127,
+      "step": 2197
+    },
+    {
+      "epoch": 1.5966222504710235,
+      "grad_norm": 0.4480574212500137,
+      "learning_rate": 8.08373891389445e-06,
+      "loss": 0.5171,
+      "step": 2198
+    },
+    {
+      "epoch": 1.5973486482191905,
+      "grad_norm": 0.4441680664452495,
+      "learning_rate": 8.0818594297536e-06,
+      "loss": 0.5235,
+      "step": 2199
+    },
+    {
+      "epoch": 1.5980750459673576,
+      "grad_norm": 0.3848708364131823,
+      "learning_rate": 8.07997924308938e-06,
+      "loss": 0.5272,
+      "step": 2200
+    },
+    {
+      "epoch": 1.5988014437155245,
+      "grad_norm": 0.46643645595697314,
+      "learning_rate": 8.078098354330386e-06,
+      "loss": 0.5416,
+      "step": 2201
+    },
+    {
+      "epoch": 1.5995278414636913,
+      "grad_norm": 0.42482180204664727,
+      "learning_rate": 8.076216763905379e-06,
+      "loss": 0.5199,
+      "step": 2202
+    },
+    {
+      "epoch": 1.6002542392118584,
+      "grad_norm": 0.38351767411017756,
+      "learning_rate": 8.074334472243273e-06,
+      "loss": 0.5385,
+      "step": 2203
+    },
+    {
+      "epoch": 1.6009806369600255,
+      "grad_norm": 0.44335436503062614,
+      "learning_rate": 8.072451479773143e-06,
+      "loss": 0.5221,
+      "step": 2204
+    },
+    {
+      "epoch": 1.6017070347081925,
+      "grad_norm": 0.33046482661721827,
+      "learning_rate": 8.070567786924228e-06,
+      "loss": 0.5243,
+      "step": 2205
+    },
+    {
+      "epoch": 1.6024334324563594,
+      "grad_norm": 0.36748965271151585,
+      "learning_rate": 8.068683394125923e-06,
+      "loss": 0.531,
+      "step": 2206
+    },
+    {
+      "epoch": 1.6031598302045262,
+      "grad_norm": 0.3833859585923725,
+      "learning_rate": 8.066798301807782e-06,
+      "loss": 0.5254,
+      "step": 2207
+    },
+    {
+      "epoch": 1.6038862279526933,
+      "grad_norm": 0.3393603194185105,
+      "learning_rate": 8.064912510399524e-06,
+      "loss": 0.522,
+      "step": 2208
+    },
+    {
+      "epoch": 1.6046126257008604,
+      "grad_norm": 0.41474782851520825,
+      "learning_rate": 8.06302602033102e-06,
+      "loss": 0.5344,
+      "step": 2209
+    },
+    {
+      "epoch": 1.6053390234490275,
+      "grad_norm": 0.4576990738112103,
+      "learning_rate": 8.061138832032304e-06,
+      "loss": 0.5024,
+      "step": 2210
+    },
+    {
+      "epoch": 1.6060654211971943,
+      "grad_norm": 0.349589634316685,
+      "learning_rate": 8.05925094593357e-06,
+      "loss": 0.5231,
+      "step": 2211
+    },
+    {
+      "epoch": 1.6067918189453612,
+      "grad_norm": 0.3592881325763187,
+      "learning_rate": 8.05736236246517e-06,
+      "loss": 0.5243,
+      "step": 2212
+    },
+    {
+      "epoch": 1.6075182166935282,
+      "grad_norm": 0.342386535794566,
+      "learning_rate": 8.055473082057615e-06,
+      "loss": 0.5243,
+      "step": 2213
+    },
+    {
+      "epoch": 1.6082446144416953,
+      "grad_norm": 0.34698494413806036,
+      "learning_rate": 8.053583105141573e-06,
+      "loss": 0.5267,
+      "step": 2214
+    },
+    {
+      "epoch": 1.6089710121898622,
+      "grad_norm": 0.3444169193095043,
+      "learning_rate": 8.051692432147876e-06,
+      "loss": 0.5124,
+      "step": 2215
+    },
+    {
+      "epoch": 1.6096974099380292,
+      "grad_norm": 0.3523998685772173,
+      "learning_rate": 8.049801063507505e-06,
+      "loss": 0.5279,
+      "step": 2216
+    },
+    {
+      "epoch": 1.610423807686196,
+      "grad_norm": 0.3600581649454914,
+      "learning_rate": 8.047908999651613e-06,
+      "loss": 0.5243,
+      "step": 2217
+    },
+    {
+      "epoch": 1.6111502054343632,
+      "grad_norm": 0.3770696606265363,
+      "learning_rate": 8.0460162410115e-06,
+      "loss": 0.5323,
+      "step": 2218
+    },
+    {
+      "epoch": 1.6118766031825302,
+      "grad_norm": 0.3627755782168038,
+      "learning_rate": 8.04412278801863e-06,
+      "loss": 0.5322,
+      "step": 2219
+    },
+    {
+      "epoch": 1.612603000930697,
+      "grad_norm": 0.39535887557139776,
+      "learning_rate": 8.042228641104622e-06,
+      "loss": 0.5211,
+      "step": 2220
+    },
+    {
+      "epoch": 1.613329398678864,
+      "grad_norm": 0.5402785851323881,
+      "learning_rate": 8.040333800701258e-06,
+      "loss": 0.533,
+      "step": 2221
+    },
+    {
+      "epoch": 1.614055796427031,
+      "grad_norm": 0.3958450342645416,
+      "learning_rate": 8.038438267240473e-06,
+      "loss": 0.5338,
+      "step": 2222
+    },
+    {
+      "epoch": 1.614782194175198,
+      "grad_norm": 0.3819125676640373,
+      "learning_rate": 8.036542041154363e-06,
+      "loss": 0.5186,
+      "step": 2223
+    },
+    {
+      "epoch": 1.6155085919233652,
+      "grad_norm": 0.5474682629830405,
+      "learning_rate": 8.034645122875183e-06,
+      "loss": 0.5155,
+      "step": 2224
+    },
+    {
+      "epoch": 1.616234989671532,
+      "grad_norm": 0.36761066631477124,
+      "learning_rate": 8.032747512835338e-06,
+      "loss": 0.5128,
+      "step": 2225
+    },
+    {
+      "epoch": 1.6169613874196989,
+      "grad_norm": 0.3553974845209146,
+      "learning_rate": 8.030849211467401e-06,
+      "loss": 0.5339,
+      "step": 2226
+    },
+    {
+      "epoch": 1.617687785167866,
+      "grad_norm": 0.4054275992193891,
+      "learning_rate": 8.0289502192041e-06,
+      "loss": 0.5192,
+      "step": 2227
+    },
+    {
+      "epoch": 1.618414182916033,
+      "grad_norm": 0.5004251882280716,
+      "learning_rate": 8.027050536478315e-06,
+      "loss": 0.5182,
+      "step": 2228
+    },
+    {
+      "epoch": 1.6191405806642,
+      "grad_norm": 0.505586412419713,
+      "learning_rate": 8.025150163723087e-06,
+      "loss": 0.5328,
+      "step": 2229
+    },
+    {
+      "epoch": 1.619866978412367,
+      "grad_norm": 0.33396295149729055,
+      "learning_rate": 8.023249101371616e-06,
+      "loss": 0.5086,
+      "step": 2230
+    },
+    {
+      "epoch": 1.6205933761605338,
+      "grad_norm": 0.3865239183606997,
+      "learning_rate": 8.021347349857258e-06,
+      "loss": 0.5302,
+      "step": 2231
+    },
+    {
+      "epoch": 1.6213197739087009,
+      "grad_norm": 0.5417414484743928,
+      "learning_rate": 8.019444909613524e-06,
+      "loss": 0.5416,
+      "step": 2232
+    },
+    {
+      "epoch": 1.622046171656868,
+      "grad_norm": 0.381620451485267,
+      "learning_rate": 8.017541781074083e-06,
+      "loss": 0.5274,
+      "step": 2233
+    },
+    {
+      "epoch": 1.622772569405035,
+      "grad_norm": 0.4633290472028216,
+      "learning_rate": 8.015637964672764e-06,
+      "loss": 0.524,
+      "step": 2234
+    },
+    {
+      "epoch": 1.6234989671532019,
+      "grad_norm": 0.3435908519465226,
+      "learning_rate": 8.013733460843546e-06,
+      "loss": 0.5147,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6242253649013687,
+      "grad_norm": 0.3724847066613284,
+      "learning_rate": 8.011828270020575e-06,
+      "loss": 0.5137,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6249517626495358,
+      "grad_norm": 0.41822288543551656,
+      "learning_rate": 8.009922392638143e-06,
+      "loss": 0.5292,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6256781603977029,
+      "grad_norm": 0.36389885402184613,
+      "learning_rate": 8.008015829130704e-06,
+      "loss": 0.5559,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6264045581458697,
+      "grad_norm": 0.49820298992503304,
+      "learning_rate": 8.006108579932869e-06,
+      "loss": 0.5115,
+      "step": 2239
+    },
+    {
+      "epoch": 1.6271309558940368,
+      "grad_norm": 0.3858200969364259,
+      "learning_rate": 8.004200645479403e-06,
+      "loss": 0.5297,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6278573536422036,
+      "grad_norm": 0.36454219334100507,
+      "learning_rate": 8.002292026205229e-06,
+      "loss": 0.5304,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6285837513903707,
+      "grad_norm": 0.4297533946554223,
+      "learning_rate": 8.000382722545423e-06,
+      "loss": 0.5203,
+      "step": 2242
+    },
+    {
+      "epoch": 1.6293101491385378,
+      "grad_norm": 0.4333838989231198,
+      "learning_rate": 7.99847273493522e-06,
+      "loss": 0.5165,
+      "step": 2243
+    },
+    {
+      "epoch": 1.6300365468867046,
+      "grad_norm": 0.3353522195529111,
+      "learning_rate": 7.996562063810009e-06,
+      "loss": 0.521,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6307629446348715,
+      "grad_norm": 0.3506370164538408,
+      "learning_rate": 7.994650709605338e-06,
+      "loss": 0.5357,
+      "step": 2245
+    },
+    {
+      "epoch": 1.6314893423830386,
+      "grad_norm": 0.3793671010753837,
+      "learning_rate": 7.992738672756909e-06,
+      "loss": 0.5138,
+      "step": 2246
+    },
+    {
+      "epoch": 1.6322157401312056,
+      "grad_norm": 0.5060055389925995,
+      "learning_rate": 7.990825953700577e-06,
+      "loss": 0.5169,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6329421378793727,
+      "grad_norm": 0.3501754362435583,
+      "learning_rate": 7.988912552872354e-06,
+      "loss": 0.5225,
+      "step": 2248
+    },
+    {
+      "epoch": 1.6336685356275396,
+      "grad_norm": 0.42524904353834486,
+      "learning_rate": 7.98699847070841e-06,
+      "loss": 0.5216,
+      "step": 2249
+    },
+    {
+      "epoch": 1.6343949333757064,
+      "grad_norm": 0.38999245002865135,
+      "learning_rate": 7.985083707645071e-06,
+      "loss": 0.5222,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6351213311238735,
+      "grad_norm": 0.4078484544716897,
+      "learning_rate": 7.983168264118811e-06,
+      "loss": 0.5276,
+      "step": 2251
+    },
+    {
+      "epoch": 1.6358477288720406,
+      "grad_norm": 0.36978836915791335,
+      "learning_rate": 7.981252140566264e-06,
+      "loss": 0.5311,
+      "step": 2252
+    },
+    {
+      "epoch": 1.6365741266202076,
+      "grad_norm": 0.36068856400219423,
+      "learning_rate": 7.979335337424222e-06,
+      "loss": 0.5069,
+      "step": 2253
+    },
+    {
+      "epoch": 1.6373005243683745,
+      "grad_norm": 0.4777656356045701,
+      "learning_rate": 7.977417855129626e-06,
+      "loss": 0.5218,
+      "step": 2254
+    },
+    {
+      "epoch": 1.6380269221165413,
+      "grad_norm": 0.40092908368657715,
+      "learning_rate": 7.975499694119576e-06,
+      "loss": 0.5226,
+      "step": 2255
+    },
+    {
+      "epoch": 1.6387533198647084,
+      "grad_norm": 0.4112514216593581,
+      "learning_rate": 7.973580854831323e-06,
+      "loss": 0.5216,
+      "step": 2256
+    },
+    {
+      "epoch": 1.6394797176128755,
+      "grad_norm": 0.557912888011925,
+      "learning_rate": 7.971661337702278e-06,
+      "loss": 0.5367,
+      "step": 2257
+    },
+    {
+      "epoch": 1.6402061153610423,
+      "grad_norm": 0.36306247064629654,
+      "learning_rate": 7.969741143170003e-06,
+      "loss": 0.5131,
+      "step": 2258
+    },
+    {
+      "epoch": 1.6409325131092094,
+      "grad_norm": 0.4098049589502914,
+      "learning_rate": 7.967820271672211e-06,
+      "loss": 0.5294,
+      "step": 2259
+    },
+    {
+      "epoch": 1.6416589108573763,
+      "grad_norm": 0.42971189819772837,
+      "learning_rate": 7.965898723646777e-06,
+      "loss": 0.5162,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6423853086055433,
+      "grad_norm": 0.43974465607779384,
+      "learning_rate": 7.963976499531724e-06,
+      "loss": 0.5198,
+      "step": 2261
+    },
+    {
+      "epoch": 1.6431117063537104,
+      "grad_norm": 0.3782390833088789,
+      "learning_rate": 7.962053599765234e-06,
+      "loss": 0.5199,
+      "step": 2262
+    },
+    {
+      "epoch": 1.6438381041018773,
+      "grad_norm": 0.3642711346135813,
+      "learning_rate": 7.960130024785636e-06,
+      "loss": 0.5045,
+      "step": 2263
+    },
+    {
+      "epoch": 1.644564501850044,
+      "grad_norm": 0.4462608703081043,
+      "learning_rate": 7.958205775031423e-06,
+      "loss": 0.5165,
+      "step": 2264
+    },
+    {
+      "epoch": 1.6452908995982112,
+      "grad_norm": 0.3937762431924642,
+      "learning_rate": 7.95628085094123e-06,
+      "loss": 0.5187,
+      "step": 2265
+    },
+    {
+      "epoch": 1.6460172973463782,
+      "grad_norm": 0.3763979930994422,
+      "learning_rate": 7.954355252953859e-06,
+      "loss": 0.5141,
+      "step": 2266
+    },
+    {
+      "epoch": 1.6467436950945453,
+      "grad_norm": 0.7372133545800805,
+      "learning_rate": 7.952428981508254e-06,
+      "loss": 0.532,
+      "step": 2267
+    },
+    {
+      "epoch": 1.6474700928427122,
+      "grad_norm": 0.3950337311935571,
+      "learning_rate": 7.950502037043515e-06,
+      "loss": 0.5261,
+      "step": 2268
+    },
+    {
+      "epoch": 1.648196490590879,
+      "grad_norm": 0.48949465520274943,
+      "learning_rate": 7.948574419998903e-06,
+      "loss": 0.5136,
+      "step": 2269
+    },
+    {
+      "epoch": 1.648922888339046,
+      "grad_norm": 0.4338054113514785,
+      "learning_rate": 7.946646130813822e-06,
+      "loss": 0.5222,
+      "step": 2270
+    },
+    {
+      "epoch": 1.6496492860872132,
+      "grad_norm": 0.4816646656555247,
+      "learning_rate": 7.944717169927838e-06,
+      "loss": 0.5307,
+      "step": 2271
+    },
+    {
+      "epoch": 1.6503756838353802,
+      "grad_norm": 0.44931497631662126,
+      "learning_rate": 7.942787537780663e-06,
+      "loss": 0.518,
+      "step": 2272
+    },
+    {
+      "epoch": 1.651102081583547,
+      "grad_norm": 0.5414627251303079,
+      "learning_rate": 7.940857234812164e-06,
+      "loss": 0.5002,
+      "step": 2273
+    },
+    {
+      "epoch": 1.651828479331714,
+      "grad_norm": 0.39261828520787145,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.5141,
+      "step": 2274
+    },
+    {
+      "epoch": 1.652554877079881,
+      "grad_norm": 0.4903351786500484,
+      "learning_rate": 7.93699461817144e-06,
+      "loss": 0.5296,
+      "step": 2275
+    },
+    {
+      "epoch": 1.653281274828048,
+      "grad_norm": 0.47308665326073535,
+      "learning_rate": 7.935062305379715e-06,
+      "loss": 0.5141,
+      "step": 2276
+    },
+    {
+      "epoch": 1.6540076725762152,
+      "grad_norm": 0.4171785888876544,
+      "learning_rate": 7.933129323527668e-06,
+      "loss": 0.5163,
+      "step": 2277
+    },
+    {
+      "epoch": 1.654734070324382,
+      "grad_norm": 0.3629005440782251,
+      "learning_rate": 7.931195673055932e-06,
+      "loss": 0.5203,
+      "step": 2278
+    },
+    {
+      "epoch": 1.6554604680725489,
+      "grad_norm": 0.45630944853496325,
+      "learning_rate": 7.929261354405289e-06,
+      "loss": 0.5286,
+      "step": 2279
+    },
+    {
+      "epoch": 1.656186865820716,
+      "grad_norm": 0.39043909477075656,
+      "learning_rate": 7.927326368016677e-06,
+      "loss": 0.5315,
+      "step": 2280
+    },
+    {
+      "epoch": 1.656913263568883,
+      "grad_norm": 0.45459187248014943,
+      "learning_rate": 7.925390714331185e-06,
+      "loss": 0.5211,
+      "step": 2281
+    },
+    {
+      "epoch": 1.6576396613170499,
+      "grad_norm": 0.4232053463313131,
+      "learning_rate": 7.923454393790053e-06,
+      "loss": 0.5393,
+      "step": 2282
+    },
+    {
+      "epoch": 1.658366059065217,
+      "grad_norm": 0.44803108722347296,
+      "learning_rate": 7.921517406834675e-06,
+      "loss": 0.528,
+      "step": 2283
+    },
+    {
+      "epoch": 1.6590924568133838,
+      "grad_norm": 0.514049972418491,
+      "learning_rate": 7.919579753906595e-06,
+      "loss": 0.5276,
+      "step": 2284
+    },
+    {
+      "epoch": 1.6598188545615509,
+      "grad_norm": 0.36688233278611193,
+      "learning_rate": 7.917641435447508e-06,
+      "loss": 0.5106,
+      "step": 2285
+    },
+    {
+      "epoch": 1.660545252309718,
+      "grad_norm": 0.37359403943964403,
+      "learning_rate": 7.915702451899264e-06,
+      "loss": 0.5164,
+      "step": 2286
+    },
+    {
+      "epoch": 1.6612716500578848,
+      "grad_norm": 0.41369210429929865,
+      "learning_rate": 7.913762803703861e-06,
+      "loss": 0.521,
+      "step": 2287
+    },
+    {
+      "epoch": 1.6619980478060516,
+      "grad_norm": 0.4009306112880066,
+      "learning_rate": 7.911822491303453e-06,
+      "loss": 0.5239,
+      "step": 2288
+    },
+    {
+      "epoch": 1.6627244455542187,
+      "grad_norm": 0.7311945919702445,
+      "learning_rate": 7.90988151514034e-06,
+      "loss": 0.5308,
+      "step": 2289
+    },
+    {
+      "epoch": 1.6634508433023858,
+      "grad_norm": 0.43626534803107847,
+      "learning_rate": 7.907939875656978e-06,
+      "loss": 0.5288,
+      "step": 2290
+    },
+    {
+      "epoch": 1.6641772410505529,
+      "grad_norm": 0.35224352677829485,
+      "learning_rate": 7.90599757329597e-06,
+      "loss": 0.5175,
+      "step": 2291
+    },
+    {
+      "epoch": 1.6649036387987197,
+      "grad_norm": 0.5331153980860087,
+      "learning_rate": 7.904054608500075e-06,
+      "loss": 0.5116,
+      "step": 2292
+    },
+    {
+      "epoch": 1.6656300365468866,
+      "grad_norm": 0.6710273266272868,
+      "learning_rate": 7.902110981712198e-06,
+      "loss": 0.5107,
+      "step": 2293
+    },
+    {
+      "epoch": 1.6663564342950536,
+      "grad_norm": 0.3526683617151688,
+      "learning_rate": 7.9001666933754e-06,
+      "loss": 0.5233,
+      "step": 2294
+    },
+    {
+      "epoch": 1.6670828320432207,
+      "grad_norm": 0.36532635665255275,
+      "learning_rate": 7.898221743932887e-06,
+      "loss": 0.5169,
+      "step": 2295
+    },
+    {
+      "epoch": 1.6678092297913878,
+      "grad_norm": 0.3585418169457838,
+      "learning_rate": 7.896276133828023e-06,
+      "loss": 0.5082,
+      "step": 2296
+    },
+    {
+      "epoch": 1.6685356275395546,
+      "grad_norm": 0.3544722750485833,
+      "learning_rate": 7.894329863504314e-06,
+      "loss": 0.5072,
+      "step": 2297
+    },
+    {
+      "epoch": 1.6692620252877215,
+      "grad_norm": 0.3546616658809727,
+      "learning_rate": 7.89238293340542e-06,
+      "loss": 0.5185,
+      "step": 2298
+    },
+    {
+      "epoch": 1.6699884230358886,
+      "grad_norm": 0.3625346473206743,
+      "learning_rate": 7.890435343975158e-06,
+      "loss": 0.5204,
+      "step": 2299
+    },
+    {
+      "epoch": 1.6707148207840556,
+      "grad_norm": 1.0160843919774494,
+      "learning_rate": 7.888487095657484e-06,
+      "loss": 0.5225,
+      "step": 2300
+    },
+    {
+      "epoch": 1.6714412185322225,
+      "grad_norm": 0.4678669118863533,
+      "learning_rate": 7.886538188896511e-06,
+      "loss": 0.5134,
+      "step": 2301
+    },
+    {
+      "epoch": 1.6721676162803896,
+      "grad_norm": 0.4058252989421966,
+      "learning_rate": 7.884588624136505e-06,
+      "loss": 0.537,
+      "step": 2302
+    },
+    {
+      "epoch": 1.6728940140285564,
+      "grad_norm": 0.5993490154767858,
+      "learning_rate": 7.882638401821873e-06,
+      "loss": 0.5204,
+      "step": 2303
+    },
+    {
+      "epoch": 1.6736204117767235,
+      "grad_norm": 0.4303193883714196,
+      "learning_rate": 7.880687522397177e-06,
+      "loss": 0.5125,
+      "step": 2304
+    },
+    {
+      "epoch": 1.6743468095248906,
+      "grad_norm": 0.35404110515752385,
+      "learning_rate": 7.878735986307133e-06,
+      "loss": 0.5295,
+      "step": 2305
+    },
+    {
+      "epoch": 1.6750732072730574,
+      "grad_norm": 0.39844908796784556,
+      "learning_rate": 7.876783793996597e-06,
+      "loss": 0.5211,
+      "step": 2306
+    },
+    {
+      "epoch": 1.6757996050212243,
+      "grad_norm": 0.5182957330420442,
+      "learning_rate": 7.874830945910581e-06,
+      "loss": 0.5266,
+      "step": 2307
+    },
+    {
+      "epoch": 1.6765260027693913,
+      "grad_norm": 0.5028901218850625,
+      "learning_rate": 7.872877442494248e-06,
+      "loss": 0.5172,
+      "step": 2308
+    },
+    {
+      "epoch": 1.6772524005175584,
+      "grad_norm": 0.3663308757833073,
+      "learning_rate": 7.870923284192904e-06,
+      "loss": 0.525,
+      "step": 2309
+    },
+    {
+      "epoch": 1.6779787982657255,
+      "grad_norm": 0.3306688367658942,
+      "learning_rate": 7.868968471452012e-06,
+      "loss": 0.5097,
+      "step": 2310
+    },
+    {
+      "epoch": 1.6787051960138923,
+      "grad_norm": 0.38291734663388316,
+      "learning_rate": 7.867013004717177e-06,
+      "loss": 0.5036,
+      "step": 2311
+    },
+    {
+      "epoch": 1.6794315937620592,
+      "grad_norm": 0.37384892481056603,
+      "learning_rate": 7.865056884434156e-06,
+      "loss": 0.5048,
+      "step": 2312
+    },
+    {
+      "epoch": 1.6801579915102263,
+      "grad_norm": 0.40607943931890755,
+      "learning_rate": 7.863100111048858e-06,
+      "loss": 0.541,
+      "step": 2313
+    },
+    {
+      "epoch": 1.6808843892583933,
+      "grad_norm": 0.3674587638957362,
+      "learning_rate": 7.861142685007335e-06,
+      "loss": 0.5415,
+      "step": 2314
+    },
+    {
+      "epoch": 1.6816107870065604,
+      "grad_norm": 0.3971339138125036,
+      "learning_rate": 7.859184606755793e-06,
+      "loss": 0.5155,
+      "step": 2315
+    },
+    {
+      "epoch": 1.6823371847547273,
+      "grad_norm": 0.4463837403698997,
+      "learning_rate": 7.857225876740585e-06,
+      "loss": 0.5256,
+      "step": 2316
+    },
+    {
+      "epoch": 1.6830635825028941,
+      "grad_norm": 0.39645405562986596,
+      "learning_rate": 7.855266495408208e-06,
+      "loss": 0.5257,
+      "step": 2317
+    },
+    {
+      "epoch": 1.6837899802510612,
+      "grad_norm": 0.3647194703278608,
+      "learning_rate": 7.853306463205317e-06,
+      "loss": 0.5159,
+      "step": 2318
+    },
+    {
+      "epoch": 1.6845163779992283,
+      "grad_norm": 0.3589570080440384,
+      "learning_rate": 7.851345780578705e-06,
+      "loss": 0.5033,
+      "step": 2319
+    },
+    {
+      "epoch": 1.6852427757473953,
+      "grad_norm": 0.38129920749398744,
+      "learning_rate": 7.849384447975322e-06,
+      "loss": 0.5071,
+      "step": 2320
+    },
+    {
+      "epoch": 1.6859691734955622,
+      "grad_norm": 0.42195602445792807,
+      "learning_rate": 7.84742246584226e-06,
+      "loss": 0.518,
+      "step": 2321
+    },
+    {
+      "epoch": 1.686695571243729,
+      "grad_norm": 0.3687914244924957,
+      "learning_rate": 7.845459834626763e-06,
+      "loss": 0.5051,
+      "step": 2322
+    },
+    {
+      "epoch": 1.6874219689918961,
+      "grad_norm": 0.3685612246975215,
+      "learning_rate": 7.843496554776222e-06,
+      "loss": 0.5182,
+      "step": 2323
+    },
+    {
+      "epoch": 1.6881483667400632,
+      "grad_norm": 0.3708898771698278,
+      "learning_rate": 7.841532626738174e-06,
+      "loss": 0.5192,
+      "step": 2324
+    },
+    {
+      "epoch": 1.68887476448823,
+      "grad_norm": 0.3794685872043508,
+      "learning_rate": 7.839568050960302e-06,
+      "loss": 0.5147,
+      "step": 2325
+    },
+    {
+      "epoch": 1.6896011622363971,
+      "grad_norm": 0.4042012250094266,
+      "learning_rate": 7.837602827890445e-06,
+      "loss": 0.5179,
+      "step": 2326
+    },
+    {
+      "epoch": 1.690327559984564,
+      "grad_norm": 0.46794334682587546,
+      "learning_rate": 7.83563695797658e-06,
+      "loss": 0.5236,
+      "step": 2327
+    },
+    {
+      "epoch": 1.691053957732731,
+      "grad_norm": 0.401175275572764,
+      "learning_rate": 7.83367044166684e-06,
+      "loss": 0.5271,
+      "step": 2328
+    },
+    {
+      "epoch": 1.6917803554808981,
+      "grad_norm": 0.34682830425628947,
+      "learning_rate": 7.831703279409498e-06,
+      "loss": 0.5151,
+      "step": 2329
+    },
+    {
+      "epoch": 1.692506753229065,
+      "grad_norm": 0.43242771033611604,
+      "learning_rate": 7.829735471652978e-06,
+      "loss": 0.5253,
+      "step": 2330
+    },
+    {
+      "epoch": 1.6932331509772318,
+      "grad_norm": 0.3938986346700491,
+      "learning_rate": 7.827767018845847e-06,
+      "loss": 0.5106,
+      "step": 2331
+    },
+    {
+      "epoch": 1.693959548725399,
+      "grad_norm": 0.37655455018986816,
+      "learning_rate": 7.825797921436829e-06,
+      "loss": 0.5207,
+      "step": 2332
+    },
+    {
+      "epoch": 1.694685946473566,
+      "grad_norm": 0.42286674943570335,
+      "learning_rate": 7.823828179874784e-06,
+      "loss": 0.5227,
+      "step": 2333
+    },
+    {
+      "epoch": 1.695412344221733,
+      "grad_norm": 2.153158814997957,
+      "learning_rate": 7.821857794608724e-06,
+      "loss": 0.5097,
+      "step": 2334
+    },
+    {
+      "epoch": 1.6961387419698999,
+      "grad_norm": 0.3888254323572561,
+      "learning_rate": 7.819886766087808e-06,
+      "loss": 0.5073,
+      "step": 2335
+    },
+    {
+      "epoch": 1.6968651397180667,
+      "grad_norm": 0.4613541494505823,
+      "learning_rate": 7.81791509476134e-06,
+      "loss": 0.5133,
+      "step": 2336
+    },
+    {
+      "epoch": 1.6975915374662338,
+      "grad_norm": 0.39905679540451405,
+      "learning_rate": 7.815942781078772e-06,
+      "loss": 0.4997,
+      "step": 2337
+    },
+    {
+      "epoch": 1.6983179352144009,
+      "grad_norm": 0.42740584656532915,
+      "learning_rate": 7.813969825489698e-06,
+      "loss": 0.5125,
+      "step": 2338
+    },
+    {
+      "epoch": 1.699044332962568,
+      "grad_norm": 0.39208560533952436,
+      "learning_rate": 7.811996228443866e-06,
+      "loss": 0.5203,
+      "step": 2339
+    },
+    {
+      "epoch": 1.6997707307107348,
+      "grad_norm": 0.4894039824183243,
+      "learning_rate": 7.810021990391163e-06,
+      "loss": 0.5213,
+      "step": 2340
+    },
+    {
+      "epoch": 1.7004971284589017,
+      "grad_norm": 0.5137340338976691,
+      "learning_rate": 7.808047111781628e-06,
+      "loss": 0.5201,
+      "step": 2341
+    },
+    {
+      "epoch": 1.7012235262070687,
+      "grad_norm": 0.5554929278674202,
+      "learning_rate": 7.806071593065441e-06,
+      "loss": 0.5193,
+      "step": 2342
+    },
+    {
+      "epoch": 1.7019499239552358,
+      "grad_norm": 0.6216775677657115,
+      "learning_rate": 7.804095434692933e-06,
+      "loss": 0.5321,
+      "step": 2343
+    },
+    {
+      "epoch": 1.7026763217034029,
+      "grad_norm": 0.5283921114968795,
+      "learning_rate": 7.802118637114575e-06,
+      "loss": 0.5177,
+      "step": 2344
+    },
+    {
+      "epoch": 1.7034027194515697,
+      "grad_norm": 0.5158106759029534,
+      "learning_rate": 7.800141200780987e-06,
+      "loss": 0.5214,
+      "step": 2345
+    },
+    {
+      "epoch": 1.7041291171997366,
+      "grad_norm": 0.3910353597249199,
+      "learning_rate": 7.798163126142935e-06,
+      "loss": 0.5214,
+      "step": 2346
+    },
+    {
+      "epoch": 1.7048555149479037,
+      "grad_norm": 0.357667881957513,
+      "learning_rate": 7.796184413651331e-06,
+      "loss": 0.5209,
+      "step": 2347
+    },
+    {
+      "epoch": 1.7055819126960707,
+      "grad_norm": 0.3663707168426187,
+      "learning_rate": 7.79420506375723e-06,
+      "loss": 0.5095,
+      "step": 2348
+    },
+    {
+      "epoch": 1.7063083104442376,
+      "grad_norm": 0.34462530618496784,
+      "learning_rate": 7.792225076911833e-06,
+      "loss": 0.5144,
+      "step": 2349
+    },
+    {
+      "epoch": 1.7070347081924044,
+      "grad_norm": 0.43033742655038754,
+      "learning_rate": 7.790244453566486e-06,
+      "loss": 0.5156,
+      "step": 2350
+    },
+    {
+      "epoch": 1.7077611059405715,
+      "grad_norm": 0.39115934719776607,
+      "learning_rate": 7.788263194172684e-06,
+      "loss": 0.5296,
+      "step": 2351
+    },
+    {
+      "epoch": 1.7084875036887386,
+      "grad_norm": 0.3751449069052403,
+      "learning_rate": 7.78628129918206e-06,
+      "loss": 0.4988,
+      "step": 2352
+    },
+    {
+      "epoch": 1.7092139014369057,
+      "grad_norm": 2.2865318787607047,
+      "learning_rate": 7.784298769046399e-06,
+      "loss": 0.5238,
+      "step": 2353
+    },
+    {
+      "epoch": 1.7099402991850725,
+      "grad_norm": 0.3308255892497287,
+      "learning_rate": 7.782315604217623e-06,
+      "loss": 0.5084,
+      "step": 2354
+    },
+    {
+      "epoch": 1.7106666969332394,
+      "grad_norm": 0.42055271857264337,
+      "learning_rate": 7.780331805147809e-06,
+      "loss": 0.5109,
+      "step": 2355
+    },
+    {
+      "epoch": 1.7113930946814064,
+      "grad_norm": 0.3484879952270989,
+      "learning_rate": 7.778347372289168e-06,
+      "loss": 0.5209,
+      "step": 2356
+    },
+    {
+      "epoch": 1.7121194924295735,
+      "grad_norm": 0.36894370842792445,
+      "learning_rate": 7.776362306094063e-06,
+      "loss": 0.5071,
+      "step": 2357
+    },
+    {
+      "epoch": 1.7128458901777406,
+      "grad_norm": 0.4033215008387943,
+      "learning_rate": 7.774376607014995e-06,
+      "loss": 0.5123,
+      "step": 2358
+    },
+    {
+      "epoch": 1.7135722879259074,
+      "grad_norm": 0.34254343652123953,
+      "learning_rate": 7.772390275504618e-06,
+      "loss": 0.5221,
+      "step": 2359
+    },
+    {
+      "epoch": 1.7142986856740743,
+      "grad_norm": 0.6806500999901334,
+      "learning_rate": 7.77040331201572e-06,
+      "loss": 0.528,
+      "step": 2360
+    },
+    {
+      "epoch": 1.7150250834222414,
+      "grad_norm": 0.3957327094391318,
+      "learning_rate": 7.768415717001241e-06,
+      "loss": 0.5061,
+      "step": 2361
+    },
+    {
+      "epoch": 1.7157514811704084,
+      "grad_norm": 0.8887630735489043,
+      "learning_rate": 7.766427490914261e-06,
+      "loss": 0.5009,
+      "step": 2362
+    },
+    {
+      "epoch": 1.7164778789185755,
+      "grad_norm": 0.41414804407176514,
+      "learning_rate": 7.764438634208007e-06,
+      "loss": 0.5091,
+      "step": 2363
+    },
+    {
+      "epoch": 1.7172042766667424,
+      "grad_norm": 0.3997412655268966,
+      "learning_rate": 7.762449147335843e-06,
+      "loss": 0.5262,
+      "step": 2364
+    },
+    {
+      "epoch": 1.7179306744149092,
+      "grad_norm": 0.41272868045457767,
+      "learning_rate": 7.760459030751285e-06,
+      "loss": 0.5122,
+      "step": 2365
+    },
+    {
+      "epoch": 1.7186570721630763,
+      "grad_norm": 0.35094999229763957,
+      "learning_rate": 7.758468284907988e-06,
+      "loss": 0.5122,
+      "step": 2366
+    },
+    {
+      "epoch": 1.7193834699112434,
+      "grad_norm": 0.45851342079117346,
+      "learning_rate": 7.756476910259749e-06,
+      "loss": 0.5184,
+      "step": 2367
+    },
+    {
+      "epoch": 1.7201098676594102,
+      "grad_norm": 0.4626631942589889,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.523,
+      "step": 2368
+    },
+    {
+      "epoch": 1.7208362654075773,
+      "grad_norm": 0.3868549743980831,
+      "learning_rate": 7.752492276364368e-06,
+      "loss": 0.5176,
+      "step": 2369
+    },
+    {
+      "epoch": 1.7215626631557441,
+      "grad_norm": 0.3810781296920432,
+      "learning_rate": 7.750499018025537e-06,
+      "loss": 0.5064,
+      "step": 2370
+    },
+    {
+      "epoch": 1.7222890609039112,
+      "grad_norm": 0.5430177338034557,
+      "learning_rate": 7.748505132698396e-06,
+      "loss": 0.5297,
+      "step": 2371
+    },
+    {
+      "epoch": 1.7230154586520783,
+      "grad_norm": 0.3297386109193866,
+      "learning_rate": 7.74651062083746e-06,
+      "loss": 0.5241,
+      "step": 2372
+    },
+    {
+      "epoch": 1.7237418564002451,
+      "grad_norm": 0.3606346057085869,
+      "learning_rate": 7.744515482897386e-06,
+      "loss": 0.539,
+      "step": 2373
+    },
+    {
+      "epoch": 1.724468254148412,
+      "grad_norm": 0.34521028847501406,
+      "learning_rate": 7.742519719332972e-06,
+      "loss": 0.5191,
+      "step": 2374
+    },
+    {
+      "epoch": 1.725194651896579,
+      "grad_norm": 0.3755384059531963,
+      "learning_rate": 7.740523330599166e-06,
+      "loss": 0.5433,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7259210496447461,
+      "grad_norm": 0.4504796185974387,
+      "learning_rate": 7.73852631715105e-06,
+      "loss": 0.506,
+      "step": 2376
+    },
+    {
+      "epoch": 1.7266474473929132,
+      "grad_norm": 0.5107682186808763,
+      "learning_rate": 7.736528679443853e-06,
+      "loss": 0.5193,
+      "step": 2377
+    },
+    {
+      "epoch": 1.72737384514108,
+      "grad_norm": 0.4141904772610753,
+      "learning_rate": 7.734530417932947e-06,
+      "loss": 0.5169,
+      "step": 2378
+    },
+    {
+      "epoch": 1.728100242889247,
+      "grad_norm": 0.36227736569282754,
+      "learning_rate": 7.73253153307384e-06,
+      "loss": 0.5149,
+      "step": 2379
+    },
+    {
+      "epoch": 1.728826640637414,
+      "grad_norm": 0.4409459557738853,
+      "learning_rate": 7.73053202532219e-06,
+      "loss": 0.5239,
+      "step": 2380
+    },
+    {
+      "epoch": 1.729553038385581,
+      "grad_norm": 0.33578102943722477,
+      "learning_rate": 7.728531895133792e-06,
+      "loss": 0.5109,
+      "step": 2381
+    },
+    {
+      "epoch": 1.7302794361337481,
+      "grad_norm": 0.4250354978387207,
+      "learning_rate": 7.726531142964587e-06,
+      "loss": 0.5155,
+      "step": 2382
+    },
+    {
+      "epoch": 1.731005833881915,
+      "grad_norm": 0.4302601157882407,
+      "learning_rate": 7.724529769270652e-06,
+      "loss": 0.5141,
+      "step": 2383
+    },
+    {
+      "epoch": 1.7317322316300818,
+      "grad_norm": 0.43407033758643465,
+      "learning_rate": 7.722527774508212e-06,
+      "loss": 0.5511,
+      "step": 2384
+    },
+    {
+      "epoch": 1.732458629378249,
+      "grad_norm": 0.40151669451134714,
+      "learning_rate": 7.720525159133627e-06,
+      "loss": 0.5219,
+      "step": 2385
+    },
+    {
+      "epoch": 1.733185027126416,
+      "grad_norm": 0.35884136542676387,
+      "learning_rate": 7.718521923603404e-06,
+      "loss": 0.5093,
+      "step": 2386
+    },
+    {
+      "epoch": 1.733911424874583,
+      "grad_norm": 0.3816066124029688,
+      "learning_rate": 7.71651806837419e-06,
+      "loss": 0.5049,
+      "step": 2387
+    },
+    {
+      "epoch": 1.73463782262275,
+      "grad_norm": 0.3676995514155556,
+      "learning_rate": 7.714513593902771e-06,
+      "loss": 0.521,
+      "step": 2388
+    },
+    {
+      "epoch": 1.7353642203709168,
+      "grad_norm": 0.4667926009034362,
+      "learning_rate": 7.712508500646077e-06,
+      "loss": 0.5317,
+      "step": 2389
+    },
+    {
+      "epoch": 1.7360906181190838,
+      "grad_norm": 0.4129366004482653,
+      "learning_rate": 7.710502789061178e-06,
+      "loss": 0.5099,
+      "step": 2390
+    },
+    {
+      "epoch": 1.736817015867251,
+      "grad_norm": 0.3592680433090348,
+      "learning_rate": 7.708496459605283e-06,
+      "loss": 0.5212,
+      "step": 2391
+    },
+    {
+      "epoch": 1.7375434136154178,
+      "grad_norm": 0.34614013237021196,
+      "learning_rate": 7.706489512735745e-06,
+      "loss": 0.5303,
+      "step": 2392
+    },
+    {
+      "epoch": 1.7382698113635848,
+      "grad_norm": 0.3555021535589424,
+      "learning_rate": 7.704481948910057e-06,
+      "loss": 0.5277,
+      "step": 2393
+    },
+    {
+      "epoch": 1.7389962091117517,
+      "grad_norm": 0.37639501186215735,
+      "learning_rate": 7.702473768585847e-06,
+      "loss": 0.525,
+      "step": 2394
+    },
+    {
+      "epoch": 1.7397226068599188,
+      "grad_norm": 0.37361879877081183,
+      "learning_rate": 7.700464972220895e-06,
+      "loss": 0.5289,
+      "step": 2395
+    },
+    {
+      "epoch": 1.7404490046080858,
+      "grad_norm": 0.39992477600936793,
+      "learning_rate": 7.698455560273112e-06,
+      "loss": 0.5312,
+      "step": 2396
+    },
+    {
+      "epoch": 1.7411754023562527,
+      "grad_norm": 0.37406681284872073,
+      "learning_rate": 7.696445533200553e-06,
+      "loss": 0.5307,
+      "step": 2397
+    },
+    {
+      "epoch": 1.7419018001044195,
+      "grad_norm": 0.42683497458023184,
+      "learning_rate": 7.694434891461413e-06,
+      "loss": 0.5001,
+      "step": 2398
+    },
+    {
+      "epoch": 1.7426281978525866,
+      "grad_norm": 0.41158580911618886,
+      "learning_rate": 7.692423635514025e-06,
+      "loss": 0.5332,
+      "step": 2399
+    },
+    {
+      "epoch": 1.7433545956007537,
+      "grad_norm": 0.36702630887820314,
+      "learning_rate": 7.690411765816864e-06,
+      "loss": 0.5278,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7440809933489207,
+      "grad_norm": 0.37401723327421216,
+      "learning_rate": 7.688399282828544e-06,
+      "loss": 0.5189,
+      "step": 2401
+    },
+    {
+      "epoch": 1.7448073910970876,
+      "grad_norm": 0.4612737879621514,
+      "learning_rate": 7.686386187007822e-06,
+      "loss": 0.5219,
+      "step": 2402
+    },
+    {
+      "epoch": 1.7455337888452545,
+      "grad_norm": 0.36477017291251335,
+      "learning_rate": 7.68437247881359e-06,
+      "loss": 0.5103,
+      "step": 2403
+    },
+    {
+      "epoch": 1.7462601865934215,
+      "grad_norm": 0.3617003411712773,
+      "learning_rate": 7.682358158704882e-06,
+      "loss": 0.5025,
+      "step": 2404
+    },
+    {
+      "epoch": 1.7469865843415886,
+      "grad_norm": 0.37483282727802664,
+      "learning_rate": 7.68034322714087e-06,
+      "loss": 0.5273,
+      "step": 2405
+    },
+    {
+      "epoch": 1.7477129820897557,
+      "grad_norm": 0.3495224935981692,
+      "learning_rate": 7.678327684580867e-06,
+      "loss": 0.5176,
+      "step": 2406
+    },
+    {
+      "epoch": 1.7484393798379225,
+      "grad_norm": 0.3385413622434483,
+      "learning_rate": 7.676311531484324e-06,
+      "loss": 0.5274,
+      "step": 2407
+    },
+    {
+      "epoch": 1.7491657775860894,
+      "grad_norm": 0.3414632763254147,
+      "learning_rate": 7.674294768310836e-06,
+      "loss": 0.5324,
+      "step": 2408
+    },
+    {
+      "epoch": 1.7498921753342565,
+      "grad_norm": 0.3305838987379325,
+      "learning_rate": 7.672277395520127e-06,
+      "loss": 0.5109,
+      "step": 2409
+    },
+    {
+      "epoch": 1.7506185730824235,
+      "grad_norm": 0.38555130400427856,
+      "learning_rate": 7.670259413572073e-06,
+      "loss": 0.5024,
+      "step": 2410
+    },
+    {
+      "epoch": 1.7513449708305904,
+      "grad_norm": 0.3199334271176942,
+      "learning_rate": 7.668240822926674e-06,
+      "loss": 0.508,
+      "step": 2411
+    },
+    {
+      "epoch": 1.7520713685787574,
+      "grad_norm": 0.3747470270663056,
+      "learning_rate": 7.66622162404408e-06,
+      "loss": 0.5014,
+      "step": 2412
+    },
+    {
+      "epoch": 1.7527977663269243,
+      "grad_norm": 0.3682403245152217,
+      "learning_rate": 7.66420181738458e-06,
+      "loss": 0.506,
+      "step": 2413
+    },
+    {
+      "epoch": 1.7535241640750914,
+      "grad_norm": 0.38757289195608097,
+      "learning_rate": 7.662181403408593e-06,
+      "loss": 0.522,
+      "step": 2414
+    },
+    {
+      "epoch": 1.7542505618232584,
+      "grad_norm": 0.3894174612444933,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.5237,
+      "step": 2415
+    },
+    {
+      "epoch": 1.7549769595714253,
+      "grad_norm": 0.3553310739437953,
+      "learning_rate": 7.658138755349552e-06,
+      "loss": 0.5176,
+      "step": 2416
+    },
+    {
+      "epoch": 1.7557033573195922,
+      "grad_norm": 0.412183295644057,
+      "learning_rate": 7.656116522188034e-06,
+      "loss": 0.5145,
+      "step": 2417
+    },
+    {
+      "epoch": 1.7564297550677592,
+      "grad_norm": 0.4269731415373766,
+      "learning_rate": 7.654093683553111e-06,
+      "loss": 0.505,
+      "step": 2418
+    },
+    {
+      "epoch": 1.7571561528159263,
+      "grad_norm": 0.40445962828489695,
+      "learning_rate": 7.652070239905896e-06,
+      "loss": 0.5385,
+      "step": 2419
+    },
+    {
+      "epoch": 1.7578825505640934,
+      "grad_norm": 0.35438541605346585,
+      "learning_rate": 7.650046191707641e-06,
+      "loss": 0.5234,
+      "step": 2420
+    },
+    {
+      "epoch": 1.7586089483122602,
+      "grad_norm": 0.36150585874514246,
+      "learning_rate": 7.648021539419737e-06,
+      "loss": 0.5071,
+      "step": 2421
+    },
+    {
+      "epoch": 1.759335346060427,
+      "grad_norm": 0.43918572767685066,
+      "learning_rate": 7.645996283503713e-06,
+      "loss": 0.509,
+      "step": 2422
+    },
+    {
+      "epoch": 1.7600617438085941,
+      "grad_norm": 0.41897424823585466,
+      "learning_rate": 7.643970424421235e-06,
+      "loss": 0.5289,
+      "step": 2423
+    },
+    {
+      "epoch": 1.7607881415567612,
+      "grad_norm": 0.3721778531489851,
+      "learning_rate": 7.641943962634106e-06,
+      "loss": 0.5076,
+      "step": 2424
+    },
+    {
+      "epoch": 1.7615145393049283,
+      "grad_norm": 0.4098431311637956,
+      "learning_rate": 7.639916898604267e-06,
+      "loss": 0.5307,
+      "step": 2425
+    },
+    {
+      "epoch": 1.7622409370530951,
+      "grad_norm": 0.3773059345853294,
+      "learning_rate": 7.637889232793795e-06,
+      "loss": 0.5181,
+      "step": 2426
+    },
+    {
+      "epoch": 1.762967334801262,
+      "grad_norm": 0.35280772424239176,
+      "learning_rate": 7.63586096566491e-06,
+      "loss": 0.5145,
+      "step": 2427
+    },
+    {
+      "epoch": 1.763693732549429,
+      "grad_norm": 0.3423930857113253,
+      "learning_rate": 7.633832097679959e-06,
+      "loss": 0.53,
+      "step": 2428
+    },
+    {
+      "epoch": 1.7644201302975961,
+      "grad_norm": 0.41118629604911827,
+      "learning_rate": 7.631802629301433e-06,
+      "loss": 0.5146,
+      "step": 2429
+    },
+    {
+      "epoch": 1.7651465280457632,
+      "grad_norm": 0.3759577573061447,
+      "learning_rate": 7.62977256099196e-06,
+      "loss": 0.5162,
+      "step": 2430
+    },
+    {
+      "epoch": 1.76587292579393,
+      "grad_norm": 0.33315814587839315,
+      "learning_rate": 7.6277418932143e-06,
+      "loss": 0.5309,
+      "step": 2431
+    },
+    {
+      "epoch": 1.766599323542097,
+      "grad_norm": 0.37737468694841797,
+      "learning_rate": 7.625710626431354e-06,
+      "loss": 0.5203,
+      "step": 2432
+    },
+    {
+      "epoch": 1.767325721290264,
+      "grad_norm": 0.38578068172952584,
+      "learning_rate": 7.623678761106159e-06,
+      "loss": 0.5209,
+      "step": 2433
+    },
+    {
+      "epoch": 1.768052119038431,
+      "grad_norm": 0.3471234798763853,
+      "learning_rate": 7.621646297701886e-06,
+      "loss": 0.5175,
+      "step": 2434
+    },
+    {
+      "epoch": 1.768778516786598,
+      "grad_norm": 0.33769756380371013,
+      "learning_rate": 7.619613236681845e-06,
+      "loss": 0.5155,
+      "step": 2435
+    },
+    {
+      "epoch": 1.769504914534765,
+      "grad_norm": 0.3484262277106607,
+      "learning_rate": 7.617579578509481e-06,
+      "loss": 0.5151,
+      "step": 2436
+    },
+    {
+      "epoch": 1.7702313122829318,
+      "grad_norm": 0.3472024381402114,
+      "learning_rate": 7.6155453236483745e-06,
+      "loss": 0.5287,
+      "step": 2437
+    },
+    {
+      "epoch": 1.770957710031099,
+      "grad_norm": 0.4865050975301496,
+      "learning_rate": 7.613510472562245e-06,
+      "loss": 0.5106,
+      "step": 2438
+    },
+    {
+      "epoch": 1.771684107779266,
+      "grad_norm": 0.43499596061429086,
+      "learning_rate": 7.611475025714945e-06,
+      "loss": 0.5126,
+      "step": 2439
+    },
+    {
+      "epoch": 1.7724105055274328,
+      "grad_norm": 0.35984637513168016,
+      "learning_rate": 7.609438983570461e-06,
+      "loss": 0.5126,
+      "step": 2440
+    },
+    {
+      "epoch": 1.7731369032755997,
+      "grad_norm": 0.3777402269226685,
+      "learning_rate": 7.6074023465929206e-06,
+      "loss": 0.5063,
+      "step": 2441
+    },
+    {
+      "epoch": 1.7738633010237668,
+      "grad_norm": 0.3368264229441657,
+      "learning_rate": 7.605365115246581e-06,
+      "loss": 0.5253,
+      "step": 2442
+    },
+    {
+      "epoch": 1.7745896987719338,
+      "grad_norm": 0.36427617602043305,
+      "learning_rate": 7.603327289995843e-06,
+      "loss": 0.5067,
+      "step": 2443
+    },
+    {
+      "epoch": 1.775316096520101,
+      "grad_norm": 0.36829048675097775,
+      "learning_rate": 7.601288871305235e-06,
+      "loss": 0.4993,
+      "step": 2444
+    },
+    {
+      "epoch": 1.7760424942682678,
+      "grad_norm": 0.3274746744967745,
+      "learning_rate": 7.5992498596394215e-06,
+      "loss": 0.5023,
+      "step": 2445
+    },
+    {
+      "epoch": 1.7767688920164346,
+      "grad_norm": 0.3690904489859233,
+      "learning_rate": 7.597210255463206e-06,
+      "loss": 0.5089,
+      "step": 2446
+    },
+    {
+      "epoch": 1.7774952897646017,
+      "grad_norm": 0.39823221371440226,
+      "learning_rate": 7.595170059241527e-06,
+      "loss": 0.5216,
+      "step": 2447
+    },
+    {
+      "epoch": 1.7782216875127688,
+      "grad_norm": 0.4064508140087145,
+      "learning_rate": 7.593129271439454e-06,
+      "loss": 0.5179,
+      "step": 2448
+    },
+    {
+      "epoch": 1.7789480852609358,
+      "grad_norm": 0.3365068405492714,
+      "learning_rate": 7.591087892522193e-06,
+      "loss": 0.5069,
+      "step": 2449
+    },
+    {
+      "epoch": 1.7796744830091027,
+      "grad_norm": 0.368825203113945,
+      "learning_rate": 7.589045922955085e-06,
+      "loss": 0.509,
+      "step": 2450
+    },
+    {
+      "epoch": 1.7804008807572695,
+      "grad_norm": 0.3702408504105955,
+      "learning_rate": 7.587003363203609e-06,
+      "loss": 0.5239,
+      "step": 2451
+    },
+    {
+      "epoch": 1.7811272785054366,
+      "grad_norm": 0.54516242684235,
+      "learning_rate": 7.5849602137333745e-06,
+      "loss": 0.5017,
+      "step": 2452
+    },
+    {
+      "epoch": 1.7818536762536037,
+      "grad_norm": 0.3789259608047946,
+      "learning_rate": 7.582916475010125e-06,
+      "loss": 0.5291,
+      "step": 2453
+    },
+    {
+      "epoch": 1.7825800740017705,
+      "grad_norm": 0.3929344951040863,
+      "learning_rate": 7.580872147499738e-06,
+      "loss": 0.5099,
+      "step": 2454
+    },
+    {
+      "epoch": 1.7833064717499376,
+      "grad_norm": 0.4521458967644181,
+      "learning_rate": 7.578827231668231e-06,
+      "loss": 0.5137,
+      "step": 2455
+    },
+    {
+      "epoch": 1.7840328694981045,
+      "grad_norm": 0.3396557213768358,
+      "learning_rate": 7.5767817279817505e-06,
+      "loss": 0.5165,
+      "step": 2456
+    },
+    {
+      "epoch": 1.7847592672462715,
+      "grad_norm": 0.3591207918432544,
+      "learning_rate": 7.574735636906576e-06,
+      "loss": 0.5031,
+      "step": 2457
+    },
+    {
+      "epoch": 1.7854856649944386,
+      "grad_norm": 0.3280120304606946,
+      "learning_rate": 7.572688958909124e-06,
+      "loss": 0.5184,
+      "step": 2458
+    },
+    {
+      "epoch": 1.7862120627426055,
+      "grad_norm": 0.3910837966006328,
+      "learning_rate": 7.570641694455945e-06,
+      "loss": 0.5042,
+      "step": 2459
+    },
+    {
+      "epoch": 1.7869384604907723,
+      "grad_norm": 0.4156980995760029,
+      "learning_rate": 7.5685938440137185e-06,
+      "loss": 0.5105,
+      "step": 2460
+    },
+    {
+      "epoch": 1.7876648582389394,
+      "grad_norm": 0.37191043977151966,
+      "learning_rate": 7.566545408049264e-06,
+      "loss": 0.5194,
+      "step": 2461
+    },
+    {
+      "epoch": 1.7883912559871065,
+      "grad_norm": 0.41929543282454695,
+      "learning_rate": 7.564496387029532e-06,
+      "loss": 0.4994,
+      "step": 2462
+    },
+    {
+      "epoch": 1.7891176537352735,
+      "grad_norm": 0.5535216928962473,
+      "learning_rate": 7.562446781421604e-06,
+      "loss": 0.5113,
+      "step": 2463
+    },
+    {
+      "epoch": 1.7898440514834404,
+      "grad_norm": 0.4440265170284457,
+      "learning_rate": 7.560396591692696e-06,
+      "loss": 0.5071,
+      "step": 2464
+    },
+    {
+      "epoch": 1.7905704492316072,
+      "grad_norm": 0.31691196067500516,
+      "learning_rate": 7.558345818310159e-06,
+      "loss": 0.5133,
+      "step": 2465
+    },
+    {
+      "epoch": 1.7912968469797743,
+      "grad_norm": 0.3976398324340962,
+      "learning_rate": 7.556294461741476e-06,
+      "loss": 0.5091,
+      "step": 2466
+    },
+    {
+      "epoch": 1.7920232447279414,
+      "grad_norm": 0.47629872897047004,
+      "learning_rate": 7.554242522454262e-06,
+      "loss": 0.5077,
+      "step": 2467
+    },
+    {
+      "epoch": 1.7927496424761085,
+      "grad_norm": 0.3915447918725823,
+      "learning_rate": 7.552190000916267e-06,
+      "loss": 0.5142,
+      "step": 2468
+    },
+    {
+      "epoch": 1.7934760402242753,
+      "grad_norm": 0.34586851694938503,
+      "learning_rate": 7.55013689759537e-06,
+      "loss": 0.5197,
+      "step": 2469
+    },
+    {
+      "epoch": 1.7942024379724422,
+      "grad_norm": 0.39615903534776675,
+      "learning_rate": 7.548083212959588e-06,
+      "loss": 0.5102,
+      "step": 2470
+    },
+    {
+      "epoch": 1.7949288357206092,
+      "grad_norm": 0.3824072375603738,
+      "learning_rate": 7.5460289474770645e-06,
+      "loss": 0.5051,
+      "step": 2471
+    },
+    {
+      "epoch": 1.7956552334687763,
+      "grad_norm": 0.3984984887068014,
+      "learning_rate": 7.5439741016160785e-06,
+      "loss": 0.5264,
+      "step": 2472
+    },
+    {
+      "epoch": 1.7963816312169434,
+      "grad_norm": 0.3854979324219856,
+      "learning_rate": 7.541918675845044e-06,
+      "loss": 0.5112,
+      "step": 2473
+    },
+    {
+      "epoch": 1.7971080289651102,
+      "grad_norm": 0.4373313917377871,
+      "learning_rate": 7.539862670632501e-06,
+      "loss": 0.502,
+      "step": 2474
+    },
+    {
+      "epoch": 1.797834426713277,
+      "grad_norm": 0.392190615648265,
+      "learning_rate": 7.537806086447124e-06,
+      "loss": 0.5299,
+      "step": 2475
+    },
+    {
+      "epoch": 1.7985608244614442,
+      "grad_norm": 0.33446120395684537,
+      "learning_rate": 7.5357489237577246e-06,
+      "loss": 0.4955,
+      "step": 2476
+    },
+    {
+      "epoch": 1.7992872222096112,
+      "grad_norm": 0.3924100537430557,
+      "learning_rate": 7.53369118303324e-06,
+      "loss": 0.5157,
+      "step": 2477
+    },
+    {
+      "epoch": 1.800013619957778,
+      "grad_norm": 0.3947888140852665,
+      "learning_rate": 7.531632864742742e-06,
+      "loss": 0.5108,
+      "step": 2478
+    },
+    {
+      "epoch": 1.8007400177059452,
+      "grad_norm": 0.41413216803259373,
+      "learning_rate": 7.52957396935543e-06,
+      "loss": 0.5258,
+      "step": 2479
+    },
+    {
+      "epoch": 1.801466415454112,
+      "grad_norm": 0.3430398095552677,
+      "learning_rate": 7.527514497340642e-06,
+      "loss": 0.5216,
+      "step": 2480
+    },
+    {
+      "epoch": 1.802192813202279,
+      "grad_norm": 0.3803404160910686,
+      "learning_rate": 7.525454449167844e-06,
+      "loss": 0.5112,
+      "step": 2481
+    },
+    {
+      "epoch": 1.8029192109504462,
+      "grad_norm": 0.362708206016911,
+      "learning_rate": 7.52339382530663e-06,
+      "loss": 0.5031,
+      "step": 2482
+    },
+    {
+      "epoch": 1.803645608698613,
+      "grad_norm": 0.3874709070824654,
+      "learning_rate": 7.52133262622673e-06,
+      "loss": 0.5206,
+      "step": 2483
+    },
+    {
+      "epoch": 1.8043720064467799,
+      "grad_norm": 0.35370761771305603,
+      "learning_rate": 7.519270852398002e-06,
+      "loss": 0.5128,
+      "step": 2484
+    },
+    {
+      "epoch": 1.805098404194947,
+      "grad_norm": 0.3378750964273353,
+      "learning_rate": 7.517208504290438e-06,
+      "loss": 0.5139,
+      "step": 2485
+    },
+    {
+      "epoch": 1.805824801943114,
+      "grad_norm": 0.47739310413310937,
+      "learning_rate": 7.5151455823741605e-06,
+      "loss": 0.5325,
+      "step": 2486
+    },
+    {
+      "epoch": 1.806551199691281,
+      "grad_norm": 1.0220584924467446,
+      "learning_rate": 7.513082087119419e-06,
+      "loss": 0.5134,
+      "step": 2487
+    },
+    {
+      "epoch": 1.807277597439448,
+      "grad_norm": 0.34733949051067253,
+      "learning_rate": 7.511018018996597e-06,
+      "loss": 0.5258,
+      "step": 2488
+    },
+    {
+      "epoch": 1.8080039951876148,
+      "grad_norm": 0.43738246407528614,
+      "learning_rate": 7.508953378476207e-06,
+      "loss": 0.5223,
+      "step": 2489
+    },
+    {
+      "epoch": 1.8087303929357819,
+      "grad_norm": 0.3884686553974104,
+      "learning_rate": 7.506888166028893e-06,
+      "loss": 0.524,
+      "step": 2490
+    },
+    {
+      "epoch": 1.809456790683949,
+      "grad_norm": 0.3664437825331931,
+      "learning_rate": 7.504822382125432e-06,
+      "loss": 0.5057,
+      "step": 2491
+    },
+    {
+      "epoch": 1.810183188432116,
+      "grad_norm": 0.4718259906344069,
+      "learning_rate": 7.502756027236725e-06,
+      "loss": 0.5269,
+      "step": 2492
+    },
+    {
+      "epoch": 1.8109095861802829,
+      "grad_norm": 0.3578874000507274,
+      "learning_rate": 7.500689101833809e-06,
+      "loss": 0.4972,
+      "step": 2493
+    },
+    {
+      "epoch": 1.8116359839284497,
+      "grad_norm": 0.4239146637573705,
+      "learning_rate": 7.498621606387848e-06,
+      "loss": 0.5037,
+      "step": 2494
+    },
+    {
+      "epoch": 1.8123623816766168,
+      "grad_norm": 0.4527212658614285,
+      "learning_rate": 7.496553541370136e-06,
+      "loss": 0.5138,
+      "step": 2495
+    },
+    {
+      "epoch": 1.8130887794247839,
+      "grad_norm": 0.38027801188865673,
+      "learning_rate": 7.4944849072520994e-06,
+      "loss": 0.5002,
+      "step": 2496
+    },
+    {
+      "epoch": 1.8138151771729507,
+      "grad_norm": 0.4033291849795312,
+      "learning_rate": 7.492415704505291e-06,
+      "loss": 0.5226,
+      "step": 2497
+    },
+    {
+      "epoch": 1.8145415749211178,
+      "grad_norm": 0.3807979142898502,
+      "learning_rate": 7.490345933601395e-06,
+      "loss": 0.5204,
+      "step": 2498
+    },
+    {
+      "epoch": 1.8152679726692846,
+      "grad_norm": 0.36024441295502735,
+      "learning_rate": 7.488275595012222e-06,
+      "loss": 0.5123,
+      "step": 2499
+    },
+    {
+      "epoch": 1.8159943704174517,
+      "grad_norm": 0.34085085330306947,
+      "learning_rate": 7.486204689209719e-06,
+      "loss": 0.5172,
+      "step": 2500
+    },
+    {
+      "epoch": 1.8167207681656188,
+      "grad_norm": 0.384830670861468,
+      "learning_rate": 7.484133216665956e-06,
+      "loss": 0.4996,
+      "step": 2501
+    },
+    {
+      "epoch": 1.8174471659137856,
+      "grad_norm": 0.32600758570920274,
+      "learning_rate": 7.482061177853135e-06,
+      "loss": 0.5017,
+      "step": 2502
+    },
+    {
+      "epoch": 1.8181735636619525,
+      "grad_norm": 0.451413015847686,
+      "learning_rate": 7.479988573243586e-06,
+      "loss": 0.4986,
+      "step": 2503
+    },
+    {
+      "epoch": 1.8188999614101196,
+      "grad_norm": 0.36176694758470074,
+      "learning_rate": 7.477915403309768e-06,
+      "loss": 0.5233,
+      "step": 2504
+    },
+    {
+      "epoch": 1.8196263591582866,
+      "grad_norm": 0.4201765718524903,
+      "learning_rate": 7.475841668524268e-06,
+      "loss": 0.5158,
+      "step": 2505
+    },
+    {
+      "epoch": 1.8203527569064537,
+      "grad_norm": 0.349747935999797,
+      "learning_rate": 7.473767369359805e-06,
+      "loss": 0.5084,
+      "step": 2506
+    },
+    {
+      "epoch": 1.8210791546546206,
+      "grad_norm": 0.37715608223832664,
+      "learning_rate": 7.471692506289225e-06,
+      "loss": 0.5133,
+      "step": 2507
+    },
+    {
+      "epoch": 1.8218055524027874,
+      "grad_norm": 0.3781868222530327,
+      "learning_rate": 7.4696170797855005e-06,
+      "loss": 0.5018,
+      "step": 2508
+    },
+    {
+      "epoch": 1.8225319501509545,
+      "grad_norm": 0.3251394185869421,
+      "learning_rate": 7.467541090321735e-06,
+      "loss": 0.5001,
+      "step": 2509
+    },
+    {
+      "epoch": 1.8232583478991216,
+      "grad_norm": 0.3696746423083357,
+      "learning_rate": 7.465464538371159e-06,
+      "loss": 0.5067,
+      "step": 2510
+    },
+    {
+      "epoch": 1.8239847456472886,
+      "grad_norm": 0.359042070785887,
+      "learning_rate": 7.463387424407131e-06,
+      "loss": 0.5146,
+      "step": 2511
+    },
+    {
+      "epoch": 1.8247111433954555,
+      "grad_norm": 0.467326002019918,
+      "learning_rate": 7.461309748903138e-06,
+      "loss": 0.5228,
+      "step": 2512
+    },
+    {
+      "epoch": 1.8254375411436223,
+      "grad_norm": 0.43379456229117774,
+      "learning_rate": 7.459231512332799e-06,
+      "loss": 0.5165,
+      "step": 2513
+    },
+    {
+      "epoch": 1.8261639388917894,
+      "grad_norm": 0.39953052494671776,
+      "learning_rate": 7.4571527151698505e-06,
+      "loss": 0.5054,
+      "step": 2514
+    },
+    {
+      "epoch": 1.8268903366399565,
+      "grad_norm": 0.4293974894737732,
+      "learning_rate": 7.45507335788817e-06,
+      "loss": 0.5028,
+      "step": 2515
+    },
+    {
+      "epoch": 1.8276167343881236,
+      "grad_norm": 0.3691821182232711,
+      "learning_rate": 7.452993440961751e-06,
+      "loss": 0.5014,
+      "step": 2516
+    },
+    {
+      "epoch": 1.8283431321362904,
+      "grad_norm": 0.3315988208836827,
+      "learning_rate": 7.450912964864724e-06,
+      "loss": 0.5079,
+      "step": 2517
+    },
+    {
+      "epoch": 1.8290695298844573,
+      "grad_norm": 0.38090799522614277,
+      "learning_rate": 7.44883193007134e-06,
+      "loss": 0.5174,
+      "step": 2518
+    },
+    {
+      "epoch": 1.8297959276326243,
+      "grad_norm": 0.37861845120198107,
+      "learning_rate": 7.4467503370559806e-06,
+      "loss": 0.5203,
+      "step": 2519
+    },
+    {
+      "epoch": 1.8305223253807914,
+      "grad_norm": 0.3700730085426923,
+      "learning_rate": 7.444668186293153e-06,
+      "loss": 0.5108,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8312487231289583,
+      "grad_norm": 0.541136540489763,
+      "learning_rate": 7.4425854782574935e-06,
+      "loss": 0.5103,
+      "step": 2521
+    },
+    {
+      "epoch": 1.8319751208771253,
+      "grad_norm": 0.3980367064113061,
+      "learning_rate": 7.440502213423766e-06,
+      "loss": 0.5072,
+      "step": 2522
+    },
+    {
+      "epoch": 1.8327015186252922,
+      "grad_norm": 0.3341368904221697,
+      "learning_rate": 7.43841839226686e-06,
+      "loss": 0.5087,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8334279163734593,
+      "grad_norm": 0.6917196441784205,
+      "learning_rate": 7.436334015261787e-06,
+      "loss": 0.522,
+      "step": 2524
+    },
+    {
+      "epoch": 1.8341543141216263,
+      "grad_norm": 0.3987386461839978,
+      "learning_rate": 7.434249082883694e-06,
+      "loss": 0.5026,
+      "step": 2525
+    },
+    {
+      "epoch": 1.8348807118697932,
+      "grad_norm": 0.39046462611388316,
+      "learning_rate": 7.432163595607851e-06,
+      "loss": 0.5129,
+      "step": 2526
+    },
+    {
+      "epoch": 1.83560710961796,
+      "grad_norm": 0.3474126740060321,
+      "learning_rate": 7.430077553909651e-06,
+      "loss": 0.5111,
+      "step": 2527
+    },
+    {
+      "epoch": 1.836333507366127,
+      "grad_norm": 0.4391027195312627,
+      "learning_rate": 7.427990958264619e-06,
+      "loss": 0.5074,
+      "step": 2528
+    },
+    {
+      "epoch": 1.8370599051142942,
+      "grad_norm": 0.3674104817530332,
+      "learning_rate": 7.425903809148401e-06,
+      "loss": 0.5142,
+      "step": 2529
+    },
+    {
+      "epoch": 1.8377863028624613,
+      "grad_norm": 0.4269078954287343,
+      "learning_rate": 7.423816107036774e-06,
+      "loss": 0.5234,
+      "step": 2530
+    },
+    {
+      "epoch": 1.838512700610628,
+      "grad_norm": 0.3256909453235737,
+      "learning_rate": 7.421727852405639e-06,
+      "loss": 0.5067,
+      "step": 2531
+    },
+    {
+      "epoch": 1.839239098358795,
+      "grad_norm": 0.35553526649015293,
+      "learning_rate": 7.419639045731022e-06,
+      "loss": 0.5097,
+      "step": 2532
+    },
+    {
+      "epoch": 1.839965496106962,
+      "grad_norm": 0.370387646589311,
+      "learning_rate": 7.417549687489074e-06,
+      "loss": 0.5112,
+      "step": 2533
+    },
+    {
+      "epoch": 1.840691893855129,
+      "grad_norm": 0.4062361187580708,
+      "learning_rate": 7.415459778156075e-06,
+      "loss": 0.5266,
+      "step": 2534
+    },
+    {
+      "epoch": 1.8414182916032962,
+      "grad_norm": 0.38909458471907526,
+      "learning_rate": 7.413369318208431e-06,
+      "loss": 0.52,
+      "step": 2535
+    },
+    {
+      "epoch": 1.842144689351463,
+      "grad_norm": 0.3450664136417991,
+      "learning_rate": 7.411278308122669e-06,
+      "loss": 0.5115,
+      "step": 2536
+    },
+    {
+      "epoch": 1.8428710870996299,
+      "grad_norm": 0.3481338815009868,
+      "learning_rate": 7.409186748375443e-06,
+      "loss": 0.4961,
+      "step": 2537
+    },
+    {
+      "epoch": 1.843597484847797,
+      "grad_norm": 0.3794933531601418,
+      "learning_rate": 7.4070946394435364e-06,
+      "loss": 0.5213,
+      "step": 2538
+    },
+    {
+      "epoch": 1.844323882595964,
+      "grad_norm": 0.34680653454325094,
+      "learning_rate": 7.405001981803851e-06,
+      "loss": 0.4881,
+      "step": 2539
+    },
+    {
+      "epoch": 1.845050280344131,
+      "grad_norm": 0.38012119673509104,
+      "learning_rate": 7.402908775933419e-06,
+      "loss": 0.5223,
+      "step": 2540
+    },
+    {
+      "epoch": 1.845776678092298,
+      "grad_norm": 0.391126869682353,
+      "learning_rate": 7.4008150223093974e-06,
+      "loss": 0.5082,
+      "step": 2541
+    },
+    {
+      "epoch": 1.8465030758404648,
+      "grad_norm": 0.3950671051221416,
+      "learning_rate": 7.398720721409065e-06,
+      "loss": 0.5038,
+      "step": 2542
+    },
+    {
+      "epoch": 1.8472294735886319,
+      "grad_norm": 0.7518357674558912,
+      "learning_rate": 7.396625873709826e-06,
+      "loss": 0.524,
+      "step": 2543
+    },
+    {
+      "epoch": 1.847955871336799,
+      "grad_norm": 0.440563224216234,
+      "learning_rate": 7.394530479689211e-06,
+      "loss": 0.5075,
+      "step": 2544
+    },
+    {
+      "epoch": 1.8486822690849658,
+      "grad_norm": 0.37825420804775123,
+      "learning_rate": 7.392434539824874e-06,
+      "loss": 0.5067,
+      "step": 2545
+    },
+    {
+      "epoch": 1.8494086668331327,
+      "grad_norm": 0.4329732625380953,
+      "learning_rate": 7.390338054594595e-06,
+      "loss": 0.5144,
+      "step": 2546
+    },
+    {
+      "epoch": 1.8501350645812997,
+      "grad_norm": 0.37240168722065525,
+      "learning_rate": 7.388241024476276e-06,
+      "loss": 0.5066,
+      "step": 2547
+    },
+    {
+      "epoch": 1.8508614623294668,
+      "grad_norm": 0.39353747779853976,
+      "learning_rate": 7.386143449947945e-06,
+      "loss": 0.4925,
+      "step": 2548
+    },
+    {
+      "epoch": 1.8515878600776339,
+      "grad_norm": 0.3359778131925848,
+      "learning_rate": 7.3840453314877505e-06,
+      "loss": 0.505,
+      "step": 2549
+    },
+    {
+      "epoch": 1.8523142578258007,
+      "grad_norm": 0.4354222874826661,
+      "learning_rate": 7.381946669573971e-06,
+      "loss": 0.5012,
+      "step": 2550
+    },
+    {
+      "epoch": 1.8530406555739676,
+      "grad_norm": 0.36701161656457704,
+      "learning_rate": 7.3798474646850034e-06,
+      "loss": 0.5013,
+      "step": 2551
+    },
+    {
+      "epoch": 1.8537670533221347,
+      "grad_norm": 0.33308540835740597,
+      "learning_rate": 7.377747717299373e-06,
+      "loss": 0.5023,
+      "step": 2552
+    },
+    {
+      "epoch": 1.8544934510703017,
+      "grad_norm": 0.3908552653407151,
+      "learning_rate": 7.375647427895725e-06,
+      "loss": 0.5154,
+      "step": 2553
+    },
+    {
+      "epoch": 1.8552198488184688,
+      "grad_norm": 0.3887867886469977,
+      "learning_rate": 7.373546596952829e-06,
+      "loss": 0.5071,
+      "step": 2554
+    },
+    {
+      "epoch": 1.8559462465666356,
+      "grad_norm": 0.4647679307155514,
+      "learning_rate": 7.371445224949581e-06,
+      "loss": 0.5118,
+      "step": 2555
+    },
+    {
+      "epoch": 1.8566726443148025,
+      "grad_norm": 0.44954855385912307,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.5154,
+      "step": 2556
+    },
+    {
+      "epoch": 1.8573990420629696,
+      "grad_norm": 0.38040083522847523,
+      "learning_rate": 7.367240859678212e-06,
+      "loss": 0.5199,
+      "step": 2557
+    },
+    {
+      "epoch": 1.8581254398111366,
+      "grad_norm": 0.4046416379097866,
+      "learning_rate": 7.365137867368497e-06,
+      "loss": 0.5123,
+      "step": 2558
+    },
+    {
+      "epoch": 1.8588518375593037,
+      "grad_norm": 0.3872560720187017,
+      "learning_rate": 7.363034335915234e-06,
+      "loss": 0.5152,
+      "step": 2559
+    },
+    {
+      "epoch": 1.8595782353074706,
+      "grad_norm": 0.36183881336332757,
+      "learning_rate": 7.360930265797934e-06,
+      "loss": 0.505,
+      "step": 2560
+    },
+    {
+      "epoch": 1.8603046330556374,
+      "grad_norm": 0.4649911594672085,
+      "learning_rate": 7.358825657496228e-06,
+      "loss": 0.5054,
+      "step": 2561
+    },
+    {
+      "epoch": 1.8610310308038045,
+      "grad_norm": 0.3458025268789646,
+      "learning_rate": 7.356720511489873e-06,
+      "loss": 0.4952,
+      "step": 2562
+    },
+    {
+      "epoch": 1.8617574285519716,
+      "grad_norm": 0.45459208605712137,
+      "learning_rate": 7.354614828258741e-06,
+      "loss": 0.529,
+      "step": 2563
+    },
+    {
+      "epoch": 1.8624838263001384,
+      "grad_norm": 0.33773416586056954,
+      "learning_rate": 7.352508608282838e-06,
+      "loss": 0.5128,
+      "step": 2564
+    },
+    {
+      "epoch": 1.8632102240483055,
+      "grad_norm": 0.3808978498416382,
+      "learning_rate": 7.3504018520422825e-06,
+      "loss": 0.4986,
+      "step": 2565
+    },
+    {
+      "epoch": 1.8639366217964723,
+      "grad_norm": 0.37701174485289957,
+      "learning_rate": 7.3482945600173195e-06,
+      "loss": 0.512,
+      "step": 2566
+    },
+    {
+      "epoch": 1.8646630195446394,
+      "grad_norm": 0.38208771506041556,
+      "learning_rate": 7.346186732688314e-06,
+      "loss": 0.5125,
+      "step": 2567
+    },
+    {
+      "epoch": 1.8653894172928065,
+      "grad_norm": 0.3494274116064339,
+      "learning_rate": 7.344078370535757e-06,
+      "loss": 0.517,
+      "step": 2568
+    },
+    {
+      "epoch": 1.8661158150409733,
+      "grad_norm": 0.3780497325193841,
+      "learning_rate": 7.341969474040257e-06,
+      "loss": 0.4967,
+      "step": 2569
+    },
+    {
+      "epoch": 1.8668422127891402,
+      "grad_norm": 0.4486675122297359,
+      "learning_rate": 7.339860043682548e-06,
+      "loss": 0.5147,
+      "step": 2570
+    },
+    {
+      "epoch": 1.8675686105373073,
+      "grad_norm": 0.4138925018957802,
+      "learning_rate": 7.337750079943483e-06,
+      "loss": 0.4959,
+      "step": 2571
+    },
+    {
+      "epoch": 1.8682950082854743,
+      "grad_norm": 0.44735718137758135,
+      "learning_rate": 7.335639583304037e-06,
+      "loss": 0.5104,
+      "step": 2572
+    },
+    {
+      "epoch": 1.8690214060336414,
+      "grad_norm": 0.3372381434600118,
+      "learning_rate": 7.333528554245307e-06,
+      "loss": 0.5245,
+      "step": 2573
+    },
+    {
+      "epoch": 1.8697478037818083,
+      "grad_norm": 0.4449369599629364,
+      "learning_rate": 7.331416993248511e-06,
+      "loss": 0.5205,
+      "step": 2574
+    },
+    {
+      "epoch": 1.8704742015299751,
+      "grad_norm": 0.36217733835115423,
+      "learning_rate": 7.329304900794991e-06,
+      "loss": 0.499,
+      "step": 2575
+    },
+    {
+      "epoch": 1.8712005992781422,
+      "grad_norm": 0.3573225113196394,
+      "learning_rate": 7.327192277366206e-06,
+      "loss": 0.5075,
+      "step": 2576
+    },
+    {
+      "epoch": 1.8719269970263093,
+      "grad_norm": 0.3530999968916237,
+      "learning_rate": 7.325079123443737e-06,
+      "loss": 0.5111,
+      "step": 2577
+    },
+    {
+      "epoch": 1.8726533947744763,
+      "grad_norm": 0.381411425903793,
+      "learning_rate": 7.322965439509287e-06,
+      "loss": 0.5145,
+      "step": 2578
+    },
+    {
+      "epoch": 1.8733797925226432,
+      "grad_norm": 0.43926138695135614,
+      "learning_rate": 7.3208512260446805e-06,
+      "loss": 0.5094,
+      "step": 2579
+    },
+    {
+      "epoch": 1.87410619027081,
+      "grad_norm": 0.48230232367488945,
+      "learning_rate": 7.318736483531861e-06,
+      "loss": 0.4854,
+      "step": 2580
+    },
+    {
+      "epoch": 1.8748325880189771,
+      "grad_norm": 0.41277915585386155,
+      "learning_rate": 7.316621212452895e-06,
+      "loss": 0.4943,
+      "step": 2581
+    },
+    {
+      "epoch": 1.8755589857671442,
+      "grad_norm": 0.3723827383864031,
+      "learning_rate": 7.314505413289964e-06,
+      "loss": 0.5323,
+      "step": 2582
+    },
+    {
+      "epoch": 1.8762853835153113,
+      "grad_norm": 0.3684546635367408,
+      "learning_rate": 7.3123890865253765e-06,
+      "loss": 0.5198,
+      "step": 2583
+    },
+    {
+      "epoch": 1.8770117812634781,
+      "grad_norm": 0.4422550701185108,
+      "learning_rate": 7.310272232641559e-06,
+      "loss": 0.4991,
+      "step": 2584
+    },
+    {
+      "epoch": 1.877738179011645,
+      "grad_norm": 0.35634373073083137,
+      "learning_rate": 7.308154852121057e-06,
+      "loss": 0.5054,
+      "step": 2585
+    },
+    {
+      "epoch": 1.878464576759812,
+      "grad_norm": 0.3807604321924993,
+      "learning_rate": 7.306036945446535e-06,
+      "loss": 0.5041,
+      "step": 2586
+    },
+    {
+      "epoch": 1.8791909745079791,
+      "grad_norm": 0.43867415636051965,
+      "learning_rate": 7.30391851310078e-06,
+      "loss": 0.4928,
+      "step": 2587
+    },
+    {
+      "epoch": 1.879917372256146,
+      "grad_norm": 0.563585109354283,
+      "learning_rate": 7.301799555566701e-06,
+      "loss": 0.5268,
+      "step": 2588
+    },
+    {
+      "epoch": 1.8806437700043128,
+      "grad_norm": 0.45682805335053467,
+      "learning_rate": 7.2996800733273196e-06,
+      "loss": 0.5193,
+      "step": 2589
+    },
+    {
+      "epoch": 1.88137016775248,
+      "grad_norm": 0.4653425530708123,
+      "learning_rate": 7.297560066865782e-06,
+      "loss": 0.5199,
+      "step": 2590
+    },
+    {
+      "epoch": 1.882096565500647,
+      "grad_norm": 0.46759940311509823,
+      "learning_rate": 7.2954395366653545e-06,
+      "loss": 0.4955,
+      "step": 2591
+    },
+    {
+      "epoch": 1.882822963248814,
+      "grad_norm": 0.4788522864372003,
+      "learning_rate": 7.293318483209419e-06,
+      "loss": 0.52,
+      "step": 2592
+    },
+    {
+      "epoch": 1.883549360996981,
+      "grad_norm": 0.3746450129791861,
+      "learning_rate": 7.29119690698148e-06,
+      "loss": 0.5057,
+      "step": 2593
+    },
+    {
+      "epoch": 1.8842757587451477,
+      "grad_norm": 0.4550524021827555,
+      "learning_rate": 7.289074808465162e-06,
+      "loss": 0.507,
+      "step": 2594
+    },
+    {
+      "epoch": 1.8850021564933148,
+      "grad_norm": 0.36358004633730967,
+      "learning_rate": 7.286952188144204e-06,
+      "loss": 0.509,
+      "step": 2595
+    },
+    {
+      "epoch": 1.885728554241482,
+      "grad_norm": 0.3871569601109488,
+      "learning_rate": 7.284829046502467e-06,
+      "loss": 0.5017,
+      "step": 2596
+    },
+    {
+      "epoch": 1.886454951989649,
+      "grad_norm": 0.3555485582102289,
+      "learning_rate": 7.282705384023933e-06,
+      "loss": 0.5198,
+      "step": 2597
+    },
+    {
+      "epoch": 1.8871813497378158,
+      "grad_norm": 0.3741738055108657,
+      "learning_rate": 7.280581201192696e-06,
+      "loss": 0.501,
+      "step": 2598
+    },
+    {
+      "epoch": 1.8879077474859827,
+      "grad_norm": 0.3219865204438388,
+      "learning_rate": 7.278456498492975e-06,
+      "loss": 0.5235,
+      "step": 2599
+    },
+    {
+      "epoch": 1.8886341452341497,
+      "grad_norm": 0.40093648121874215,
+      "learning_rate": 7.2763312764091055e-06,
+      "loss": 0.4906,
+      "step": 2600
+    },
+    {
+      "epoch": 1.8893605429823168,
+      "grad_norm": 0.3942335836495865,
+      "learning_rate": 7.274205535425543e-06,
+      "loss": 0.5226,
+      "step": 2601
+    },
+    {
+      "epoch": 1.8900869407304839,
+      "grad_norm": 0.3743931565246306,
+      "learning_rate": 7.272079276026855e-06,
+      "loss": 0.5086,
+      "step": 2602
+    },
+    {
+      "epoch": 1.8908133384786507,
+      "grad_norm": 0.3921609659047992,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.4963,
+      "step": 2603
+    },
+    {
+      "epoch": 1.8915397362268176,
+      "grad_norm": 0.4079084279428097,
+      "learning_rate": 7.26782520392299e-06,
+      "loss": 0.4983,
+      "step": 2604
+    },
+    {
+      "epoch": 1.8922661339749847,
+      "grad_norm": 0.4573292211751414,
+      "learning_rate": 7.265697392187547e-06,
+      "loss": 0.5138,
+      "step": 2605
+    },
+    {
+      "epoch": 1.8929925317231517,
+      "grad_norm": 0.39025123088916785,
+      "learning_rate": 7.263569063976449e-06,
+      "loss": 0.5126,
+      "step": 2606
+    },
+    {
+      "epoch": 1.8937189294713186,
+      "grad_norm": 0.527907757792281,
+      "learning_rate": 7.261440219774858e-06,
+      "loss": 0.5029,
+      "step": 2607
+    },
+    {
+      "epoch": 1.8944453272194857,
+      "grad_norm": 0.3581702843772708,
+      "learning_rate": 7.2593108600680524e-06,
+      "loss": 0.5176,
+      "step": 2608
+    },
+    {
+      "epoch": 1.8951717249676525,
+      "grad_norm": 0.39048666471502463,
+      "learning_rate": 7.257180985341432e-06,
+      "loss": 0.5152,
+      "step": 2609
+    },
+    {
+      "epoch": 1.8958981227158196,
+      "grad_norm": 0.42110401723381574,
+      "learning_rate": 7.25505059608051e-06,
+      "loss": 0.5083,
+      "step": 2610
+    },
+    {
+      "epoch": 1.8966245204639867,
+      "grad_norm": 0.4571340129019425,
+      "learning_rate": 7.252919692770916e-06,
+      "loss": 0.5125,
+      "step": 2611
+    },
+    {
+      "epoch": 1.8973509182121535,
+      "grad_norm": 0.36420915076613447,
+      "learning_rate": 7.250788275898401e-06,
+      "loss": 0.5118,
+      "step": 2612
+    },
+    {
+      "epoch": 1.8980773159603204,
+      "grad_norm": 0.37701735727524127,
+      "learning_rate": 7.248656345948828e-06,
+      "loss": 0.5109,
+      "step": 2613
+    },
+    {
+      "epoch": 1.8988037137084874,
+      "grad_norm": 0.4371017222186262,
+      "learning_rate": 7.2465239034081835e-06,
+      "loss": 0.4916,
+      "step": 2614
+    },
+    {
+      "epoch": 1.8995301114566545,
+      "grad_norm": 0.3654714865890906,
+      "learning_rate": 7.244390948762566e-06,
+      "loss": 0.5172,
+      "step": 2615
+    },
+    {
+      "epoch": 1.9002565092048216,
+      "grad_norm": 0.47754877856771066,
+      "learning_rate": 7.242257482498191e-06,
+      "loss": 0.5287,
+      "step": 2616
+    },
+    {
+      "epoch": 1.9009829069529884,
+      "grad_norm": 0.36891648606279626,
+      "learning_rate": 7.2401235051013885e-06,
+      "loss": 0.5142,
+      "step": 2617
+    },
+    {
+      "epoch": 1.9017093047011553,
+      "grad_norm": 0.4223943980005817,
+      "learning_rate": 7.237989017058614e-06,
+      "loss": 0.4889,
+      "step": 2618
+    },
+    {
+      "epoch": 1.9024357024493224,
+      "grad_norm": 0.35807149966277446,
+      "learning_rate": 7.235854018856429e-06,
+      "loss": 0.509,
+      "step": 2619
+    },
+    {
+      "epoch": 1.9031621001974894,
+      "grad_norm": 0.4103889588551492,
+      "learning_rate": 7.23371851098152e-06,
+      "loss": 0.5074,
+      "step": 2620
+    },
+    {
+      "epoch": 1.9038884979456565,
+      "grad_norm": 0.47569057216314453,
+      "learning_rate": 7.23158249392068e-06,
+      "loss": 0.5167,
+      "step": 2621
+    },
+    {
+      "epoch": 1.9046148956938234,
+      "grad_norm": 0.7088526151457004,
+      "learning_rate": 7.2294459681608275e-06,
+      "loss": 0.5204,
+      "step": 2622
+    },
+    {
+      "epoch": 1.9053412934419902,
+      "grad_norm": 0.4343984003997984,
+      "learning_rate": 7.22730893418899e-06,
+      "loss": 0.5124,
+      "step": 2623
+    },
+    {
+      "epoch": 1.9060676911901573,
+      "grad_norm": 0.3905184752532141,
+      "learning_rate": 7.225171392492316e-06,
+      "loss": 0.5247,
+      "step": 2624
+    },
+    {
+      "epoch": 1.9067940889383244,
+      "grad_norm": 0.35839885716842484,
+      "learning_rate": 7.223033343558068e-06,
+      "loss": 0.52,
+      "step": 2625
+    },
+    {
+      "epoch": 1.9075204866864914,
+      "grad_norm": 0.40567822205512094,
+      "learning_rate": 7.220894787873621e-06,
+      "loss": 0.5013,
+      "step": 2626
+    },
+    {
+      "epoch": 1.9082468844346583,
+      "grad_norm": 0.49459869136202017,
+      "learning_rate": 7.218755725926471e-06,
+      "loss": 0.5093,
+      "step": 2627
+    },
+    {
+      "epoch": 1.9089732821828251,
+      "grad_norm": 0.3794644738069637,
+      "learning_rate": 7.216616158204223e-06,
+      "loss": 0.4999,
+      "step": 2628
+    },
+    {
+      "epoch": 1.9096996799309922,
+      "grad_norm": 0.4855621032827998,
+      "learning_rate": 7.214476085194605e-06,
+      "loss": 0.5087,
+      "step": 2629
+    },
+    {
+      "epoch": 1.9104260776791593,
+      "grad_norm": 0.36297925771622874,
+      "learning_rate": 7.212335507385453e-06,
+      "loss": 0.5063,
+      "step": 2630
+    },
+    {
+      "epoch": 1.9111524754273261,
+      "grad_norm": 0.3944990999511598,
+      "learning_rate": 7.210194425264723e-06,
+      "loss": 0.5068,
+      "step": 2631
+    },
+    {
+      "epoch": 1.9118788731754932,
+      "grad_norm": 0.4122143059291724,
+      "learning_rate": 7.208052839320481e-06,
+      "loss": 0.5162,
+      "step": 2632
+    },
+    {
+      "epoch": 1.91260527092366,
+      "grad_norm": 0.41692514963900895,
+      "learning_rate": 7.205910750040914e-06,
+      "loss": 0.4828,
+      "step": 2633
+    },
+    {
+      "epoch": 1.9133316686718271,
+      "grad_norm": 0.3663441755784621,
+      "learning_rate": 7.203768157914321e-06,
+      "loss": 0.494,
+      "step": 2634
+    },
+    {
+      "epoch": 1.9140580664199942,
+      "grad_norm": 0.4070037156583176,
+      "learning_rate": 7.201625063429113e-06,
+      "loss": 0.5121,
+      "step": 2635
+    },
+    {
+      "epoch": 1.914784464168161,
+      "grad_norm": 0.4380005820938112,
+      "learning_rate": 7.199481467073819e-06,
+      "loss": 0.5157,
+      "step": 2636
+    },
+    {
+      "epoch": 1.915510861916328,
+      "grad_norm": 0.39134139767667614,
+      "learning_rate": 7.197337369337081e-06,
+      "loss": 0.5115,
+      "step": 2637
+    },
+    {
+      "epoch": 1.916237259664495,
+      "grad_norm": 0.3780955623983413,
+      "learning_rate": 7.1951927707076545e-06,
+      "loss": 0.5138,
+      "step": 2638
+    },
+    {
+      "epoch": 1.916963657412662,
+      "grad_norm": 0.37420155415019496,
+      "learning_rate": 7.193047671674411e-06,
+      "loss": 0.5109,
+      "step": 2639
+    },
+    {
+      "epoch": 1.9176900551608291,
+      "grad_norm": 0.3975633927338851,
+      "learning_rate": 7.190902072726336e-06,
+      "loss": 0.5175,
+      "step": 2640
+    },
+    {
+      "epoch": 1.918416452908996,
+      "grad_norm": 0.41788467337021007,
+      "learning_rate": 7.188755974352528e-06,
+      "loss": 0.5034,
+      "step": 2641
+    },
+    {
+      "epoch": 1.9191428506571628,
+      "grad_norm": 0.36881038549681205,
+      "learning_rate": 7.186609377042199e-06,
+      "loss": 0.4932,
+      "step": 2642
+    },
+    {
+      "epoch": 1.91986924840533,
+      "grad_norm": 0.4923105516525616,
+      "learning_rate": 7.1844622812846745e-06,
+      "loss": 0.5074,
+      "step": 2643
+    },
+    {
+      "epoch": 1.920595646153497,
+      "grad_norm": 0.4675468904655994,
+      "learning_rate": 7.182314687569395e-06,
+      "loss": 0.4971,
+      "step": 2644
+    },
+    {
+      "epoch": 1.921322043901664,
+      "grad_norm": 0.33391764389643824,
+      "learning_rate": 7.180166596385915e-06,
+      "loss": 0.5141,
+      "step": 2645
+    },
+    {
+      "epoch": 1.922048441649831,
+      "grad_norm": 0.37264094779898216,
+      "learning_rate": 7.1780180082238994e-06,
+      "loss": 0.5082,
+      "step": 2646
+    },
+    {
+      "epoch": 1.9227748393979978,
+      "grad_norm": 0.33950669946292117,
+      "learning_rate": 7.175868923573129e-06,
+      "loss": 0.519,
+      "step": 2647
+    },
+    {
+      "epoch": 1.9235012371461648,
+      "grad_norm": 0.3907280044051392,
+      "learning_rate": 7.1737193429234985e-06,
+      "loss": 0.5033,
+      "step": 2648
+    },
+    {
+      "epoch": 1.924227634894332,
+      "grad_norm": 0.3981209785053024,
+      "learning_rate": 7.171569266765012e-06,
+      "loss": 0.5113,
+      "step": 2649
+    },
+    {
+      "epoch": 1.9249540326424988,
+      "grad_norm": 0.3552940618648031,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.5262,
+      "step": 2650
+    },
+    {
+      "epoch": 1.9256804303906658,
+      "grad_norm": 0.45671174402732945,
+      "learning_rate": 7.167267629882066e-06,
+      "loss": 0.5087,
+      "step": 2651
+    },
+    {
+      "epoch": 1.9264068281388327,
+      "grad_norm": 0.36036782434775116,
+      "learning_rate": 7.165116070138183e-06,
+      "loss": 0.5093,
+      "step": 2652
+    },
+    {
+      "epoch": 1.9271332258869998,
+      "grad_norm": 0.36770341112039956,
+      "learning_rate": 7.162964016846597e-06,
+      "loss": 0.5015,
+      "step": 2653
+    },
+    {
+      "epoch": 1.9278596236351668,
+      "grad_norm": 0.4171887047487715,
+      "learning_rate": 7.160811470497881e-06,
+      "loss": 0.4875,
+      "step": 2654
+    },
+    {
+      "epoch": 1.9285860213833337,
+      "grad_norm": 0.7588013368151402,
+      "learning_rate": 7.158658431582718e-06,
+      "loss": 0.4812,
+      "step": 2655
+    },
+    {
+      "epoch": 1.9293124191315005,
+      "grad_norm": 0.4838546627793747,
+      "learning_rate": 7.156504900591899e-06,
+      "loss": 0.5061,
+      "step": 2656
+    },
+    {
+      "epoch": 1.9300388168796676,
+      "grad_norm": 0.3893073278129443,
+      "learning_rate": 7.154350878016332e-06,
+      "loss": 0.4997,
+      "step": 2657
+    },
+    {
+      "epoch": 1.9307652146278347,
+      "grad_norm": 0.35598578094946026,
+      "learning_rate": 7.152196364347037e-06,
+      "loss": 0.4818,
+      "step": 2658
+    },
+    {
+      "epoch": 1.9314916123760018,
+      "grad_norm": 0.7028633945623345,
+      "learning_rate": 7.1500413600751465e-06,
+      "loss": 0.5064,
+      "step": 2659
+    },
+    {
+      "epoch": 1.9322180101241686,
+      "grad_norm": 0.5021000366038473,
+      "learning_rate": 7.147885865691899e-06,
+      "loss": 0.4953,
+      "step": 2660
+    },
+    {
+      "epoch": 1.9329444078723355,
+      "grad_norm": 0.4521851195168597,
+      "learning_rate": 7.145729881688651e-06,
+      "loss": 0.508,
+      "step": 2661
+    },
+    {
+      "epoch": 1.9336708056205025,
+      "grad_norm": 0.32339465479764373,
+      "learning_rate": 7.143573408556867e-06,
+      "loss": 0.4954,
+      "step": 2662
+    },
+    {
+      "epoch": 1.9343972033686696,
+      "grad_norm": 0.401677364787166,
+      "learning_rate": 7.141416446788128e-06,
+      "loss": 0.4913,
+      "step": 2663
+    },
+    {
+      "epoch": 1.9351236011168367,
+      "grad_norm": 0.41714933742992666,
+      "learning_rate": 7.139258996874122e-06,
+      "loss": 0.5232,
+      "step": 2664
+    },
+    {
+      "epoch": 1.9358499988650035,
+      "grad_norm": 0.4588630688440973,
+      "learning_rate": 7.137101059306647e-06,
+      "loss": 0.5167,
+      "step": 2665
+    },
+    {
+      "epoch": 1.9365763966131704,
+      "grad_norm": 0.5128126005709326,
+      "learning_rate": 7.134942634577615e-06,
+      "loss": 0.5073,
+      "step": 2666
+    },
+    {
+      "epoch": 1.9373027943613375,
+      "grad_norm": 0.3220816956148639,
+      "learning_rate": 7.1327837231790484e-06,
+      "loss": 0.4918,
+      "step": 2667
+    },
+    {
+      "epoch": 1.9380291921095045,
+      "grad_norm": 0.3864453720966872,
+      "learning_rate": 7.1306243256030815e-06,
+      "loss": 0.5051,
+      "step": 2668
+    },
+    {
+      "epoch": 1.9387555898576716,
+      "grad_norm": 0.3417311969733972,
+      "learning_rate": 7.128464442341958e-06,
+      "loss": 0.4992,
+      "step": 2669
+    },
+    {
+      "epoch": 1.9394819876058385,
+      "grad_norm": 0.3666997641598726,
+      "learning_rate": 7.126304073888033e-06,
+      "loss": 0.4915,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9402083853540053,
+      "grad_norm": 0.32912302124120496,
+      "learning_rate": 7.124143220733772e-06,
+      "loss": 0.5132,
+      "step": 2671
+    },
+    {
+      "epoch": 1.9409347831021724,
+      "grad_norm": 0.5894784767394627,
+      "learning_rate": 7.121981883371748e-06,
+      "loss": 0.5106,
+      "step": 2672
+    },
+    {
+      "epoch": 1.9416611808503395,
+      "grad_norm": 0.36134194753601506,
+      "learning_rate": 7.1198200622946516e-06,
+      "loss": 0.5322,
+      "step": 2673
+    },
+    {
+      "epoch": 1.9423875785985063,
+      "grad_norm": 0.5233500856029822,
+      "learning_rate": 7.117657757995277e-06,
+      "loss": 0.5006,
+      "step": 2674
+    },
+    {
+      "epoch": 1.9431139763466734,
+      "grad_norm": 0.45441916828563017,
+      "learning_rate": 7.115494970966532e-06,
+      "loss": 0.5015,
+      "step": 2675
+    },
+    {
+      "epoch": 1.9438403740948402,
+      "grad_norm": 0.42922725575066273,
+      "learning_rate": 7.113331701701433e-06,
+      "loss": 0.5118,
+      "step": 2676
+    },
+    {
+      "epoch": 1.9445667718430073,
+      "grad_norm": 0.37288209324065047,
+      "learning_rate": 7.111167950693106e-06,
+      "loss": 0.5133,
+      "step": 2677
+    },
+    {
+      "epoch": 1.9452931695911744,
+      "grad_norm": 0.5428767194900757,
+      "learning_rate": 7.109003718434788e-06,
+      "loss": 0.5109,
+      "step": 2678
+    },
+    {
+      "epoch": 1.9460195673393412,
+      "grad_norm": 0.48476276527501894,
+      "learning_rate": 7.106839005419825e-06,
+      "loss": 0.5195,
+      "step": 2679
+    },
+    {
+      "epoch": 1.946745965087508,
+      "grad_norm": 0.42779549179479115,
+      "learning_rate": 7.104673812141676e-06,
+      "loss": 0.5278,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9474723628356752,
+      "grad_norm": 0.4172577512072656,
+      "learning_rate": 7.102508139093902e-06,
+      "loss": 0.5225,
+      "step": 2681
+    },
+    {
+      "epoch": 1.9481987605838422,
+      "grad_norm": 0.6445236653546356,
+      "learning_rate": 7.100341986770178e-06,
+      "loss": 0.4956,
+      "step": 2682
+    },
+    {
+      "epoch": 1.9489251583320093,
+      "grad_norm": 0.36704041066233434,
+      "learning_rate": 7.09817535566429e-06,
+      "loss": 0.5056,
+      "step": 2683
+    },
+    {
+      "epoch": 1.9496515560801762,
+      "grad_norm": 0.42608468891834045,
+      "learning_rate": 7.0960082462701315e-06,
+      "loss": 0.4976,
+      "step": 2684
+    },
+    {
+      "epoch": 1.950377953828343,
+      "grad_norm": 0.3725547949696026,
+      "learning_rate": 7.093840659081702e-06,
+      "loss": 0.5209,
+      "step": 2685
+    },
+    {
+      "epoch": 1.95110435157651,
+      "grad_norm": 0.347960151425129,
+      "learning_rate": 7.091672594593114e-06,
+      "loss": 0.4956,
+      "step": 2686
+    },
+    {
+      "epoch": 1.9518307493246772,
+      "grad_norm": 0.3919601507999021,
+      "learning_rate": 7.089504053298587e-06,
+      "loss": 0.499,
+      "step": 2687
+    },
+    {
+      "epoch": 1.9525571470728442,
+      "grad_norm": 0.43156269797326463,
+      "learning_rate": 7.0873350356924495e-06,
+      "loss": 0.5023,
+      "step": 2688
+    },
+    {
+      "epoch": 1.953283544821011,
+      "grad_norm": 0.34776606928147463,
+      "learning_rate": 7.085165542269139e-06,
+      "loss": 0.4955,
+      "step": 2689
+    },
+    {
+      "epoch": 1.954009942569178,
+      "grad_norm": 0.7017977923984215,
+      "learning_rate": 7.082995573523202e-06,
+      "loss": 0.4999,
+      "step": 2690
+    },
+    {
+      "epoch": 1.954736340317345,
+      "grad_norm": 0.4815505166236814,
+      "learning_rate": 7.080825129949289e-06,
+      "loss": 0.5146,
+      "step": 2691
+    },
+    {
+      "epoch": 1.955462738065512,
+      "grad_norm": 0.43130480748007916,
+      "learning_rate": 7.0786542120421645e-06,
+      "loss": 0.5215,
+      "step": 2692
+    },
+    {
+      "epoch": 1.956189135813679,
+      "grad_norm": 0.38558580874476744,
+      "learning_rate": 7.0764828202966986e-06,
+      "loss": 0.5011,
+      "step": 2693
+    },
+    {
+      "epoch": 1.956915533561846,
+      "grad_norm": 0.4036358786260398,
+      "learning_rate": 7.074310955207869e-06,
+      "loss": 0.5061,
+      "step": 2694
+    },
+    {
+      "epoch": 1.9576419313100129,
+      "grad_norm": 0.33067727483602,
+      "learning_rate": 7.072138617270765e-06,
+      "loss": 0.4895,
+      "step": 2695
+    },
+    {
+      "epoch": 1.95836832905818,
+      "grad_norm": 0.5605426386928977,
+      "learning_rate": 7.069965806980574e-06,
+      "loss": 0.5062,
+      "step": 2696
+    },
+    {
+      "epoch": 1.959094726806347,
+      "grad_norm": 0.35218801878956085,
+      "learning_rate": 7.067792524832604e-06,
+      "loss": 0.5043,
+      "step": 2697
+    },
+    {
+      "epoch": 1.9598211245545138,
+      "grad_norm": 0.3887738969745063,
+      "learning_rate": 7.0656187713222615e-06,
+      "loss": 0.5165,
+      "step": 2698
+    },
+    {
+      "epoch": 1.9605475223026807,
+      "grad_norm": 0.3868917511130432,
+      "learning_rate": 7.063444546945063e-06,
+      "loss": 0.5117,
+      "step": 2699
+    },
+    {
+      "epoch": 1.9612739200508478,
+      "grad_norm": 0.3402687842463978,
+      "learning_rate": 7.061269852196633e-06,
+      "loss": 0.4885,
+      "step": 2700
+    },
+    {
+      "epoch": 1.9620003177990148,
+      "grad_norm": 0.36804469030096626,
+      "learning_rate": 7.059094687572701e-06,
+      "loss": 0.509,
+      "step": 2701
+    },
+    {
+      "epoch": 1.962726715547182,
+      "grad_norm": 0.39457259575266557,
+      "learning_rate": 7.056919053569109e-06,
+      "loss": 0.4954,
+      "step": 2702
+    },
+    {
+      "epoch": 1.9634531132953488,
+      "grad_norm": 0.3204056385239754,
+      "learning_rate": 7.0547429506817986e-06,
+      "loss": 0.512,
+      "step": 2703
+    },
+    {
+      "epoch": 1.9641795110435156,
+      "grad_norm": 0.3629638300335219,
+      "learning_rate": 7.052566379406824e-06,
+      "loss": 0.4915,
+      "step": 2704
+    },
+    {
+      "epoch": 1.9649059087916827,
+      "grad_norm": 0.4589761089027009,
+      "learning_rate": 7.050389340240345e-06,
+      "loss": 0.4938,
+      "step": 2705
+    },
+    {
+      "epoch": 1.9656323065398498,
+      "grad_norm": 0.5304412242646968,
+      "learning_rate": 7.048211833678624e-06,
+      "loss": 0.5124,
+      "step": 2706
+    },
+    {
+      "epoch": 1.9663587042880168,
+      "grad_norm": 0.36321087701659993,
+      "learning_rate": 7.046033860218036e-06,
+      "loss": 0.4953,
+      "step": 2707
+    },
+    {
+      "epoch": 1.9670851020361837,
+      "grad_norm": 0.4130645277458678,
+      "learning_rate": 7.04385542035506e-06,
+      "loss": 0.4892,
+      "step": 2708
+    },
+    {
+      "epoch": 1.9678114997843505,
+      "grad_norm": 0.35499258033837994,
+      "learning_rate": 7.041676514586279e-06,
+      "loss": 0.4795,
+      "step": 2709
+    },
+    {
+      "epoch": 1.9685378975325176,
+      "grad_norm": 0.3760642593459764,
+      "learning_rate": 7.039497143408384e-06,
+      "loss": 0.5112,
+      "step": 2710
+    },
+    {
+      "epoch": 1.9692642952806847,
+      "grad_norm": 0.39847541288606975,
+      "learning_rate": 7.037317307318172e-06,
+      "loss": 0.5131,
+      "step": 2711
+    },
+    {
+      "epoch": 1.9699906930288518,
+      "grad_norm": 0.3325102212417593,
+      "learning_rate": 7.035137006812548e-06,
+      "loss": 0.5095,
+      "step": 2712
+    },
+    {
+      "epoch": 1.9707170907770186,
+      "grad_norm": 0.41951767852434335,
+      "learning_rate": 7.03295624238852e-06,
+      "loss": 0.5029,
+      "step": 2713
+    },
+    {
+      "epoch": 1.9714434885251855,
+      "grad_norm": 0.3882568800021323,
+      "learning_rate": 7.030775014543204e-06,
+      "loss": 0.5161,
+      "step": 2714
+    },
+    {
+      "epoch": 1.9721698862733525,
+      "grad_norm": 0.3246175477769834,
+      "learning_rate": 7.028593323773819e-06,
+      "loss": 0.5173,
+      "step": 2715
+    },
+    {
+      "epoch": 1.9728962840215196,
+      "grad_norm": 0.4320032858507335,
+      "learning_rate": 7.026411170577691e-06,
+      "loss": 0.4991,
+      "step": 2716
+    },
+    {
+      "epoch": 1.9736226817696865,
+      "grad_norm": 0.3582516058922341,
+      "learning_rate": 7.024228555452253e-06,
+      "loss": 0.4979,
+      "step": 2717
+    },
+    {
+      "epoch": 1.9743490795178535,
+      "grad_norm": 0.3865649105180352,
+      "learning_rate": 7.022045478895038e-06,
+      "loss": 0.5052,
+      "step": 2718
+    },
+    {
+      "epoch": 1.9750754772660204,
+      "grad_norm": 0.3865623049303795,
+      "learning_rate": 7.019861941403693e-06,
+      "loss": 0.5102,
+      "step": 2719
+    },
+    {
+      "epoch": 1.9758018750141875,
+      "grad_norm": 0.3744366218022015,
+      "learning_rate": 7.017677943475962e-06,
+      "loss": 0.5037,
+      "step": 2720
+    },
+    {
+      "epoch": 1.9765282727623545,
+      "grad_norm": 0.4049135491697213,
+      "learning_rate": 7.015493485609697e-06,
+      "loss": 0.5103,
+      "step": 2721
+    },
+    {
+      "epoch": 1.9772546705105214,
+      "grad_norm": 0.41669353889570854,
+      "learning_rate": 7.013308568302855e-06,
+      "loss": 0.496,
+      "step": 2722
+    },
+    {
+      "epoch": 1.9779810682586882,
+      "grad_norm": 1.55599304714908,
+      "learning_rate": 7.011123192053497e-06,
+      "loss": 0.4885,
+      "step": 2723
+    },
+    {
+      "epoch": 1.9787074660068553,
+      "grad_norm": 0.3766075391392106,
+      "learning_rate": 7.00893735735979e-06,
+      "loss": 0.5091,
+      "step": 2724
+    },
+    {
+      "epoch": 1.9794338637550224,
+      "grad_norm": 0.39432289032043133,
+      "learning_rate": 7.006751064720003e-06,
+      "loss": 0.5043,
+      "step": 2725
+    },
+    {
+      "epoch": 1.9801602615031895,
+      "grad_norm": 0.42252648320613373,
+      "learning_rate": 7.004564314632514e-06,
+      "loss": 0.4971,
+      "step": 2726
+    },
+    {
+      "epoch": 1.9808866592513563,
+      "grad_norm": 0.5220318083615987,
+      "learning_rate": 7.002377107595799e-06,
+      "loss": 0.5131,
+      "step": 2727
+    },
+    {
+      "epoch": 1.9816130569995232,
+      "grad_norm": 0.439124427096527,
+      "learning_rate": 7.000189444108443e-06,
+      "loss": 0.4987,
+      "step": 2728
+    },
+    {
+      "epoch": 1.9823394547476902,
+      "grad_norm": 0.6087596878574801,
+      "learning_rate": 6.998001324669135e-06,
+      "loss": 0.5007,
+      "step": 2729
+    },
+    {
+      "epoch": 1.9830658524958573,
+      "grad_norm": 0.4689995890970593,
+      "learning_rate": 6.995812749776663e-06,
+      "loss": 0.4928,
+      "step": 2730
+    },
+    {
+      "epoch": 1.9837922502440244,
+      "grad_norm": 0.4303674417050366,
+      "learning_rate": 6.993623719929924e-06,
+      "loss": 0.4972,
+      "step": 2731
+    },
+    {
+      "epoch": 1.9845186479921912,
+      "grad_norm": 1.7806687410524744,
+      "learning_rate": 6.991434235627918e-06,
+      "loss": 0.5007,
+      "step": 2732
+    },
+    {
+      "epoch": 1.985245045740358,
+      "grad_norm": 0.45617373447370785,
+      "learning_rate": 6.989244297369746e-06,
+      "loss": 0.5038,
+      "step": 2733
+    },
+    {
+      "epoch": 1.9859714434885252,
+      "grad_norm": 0.39328968595405445,
+      "learning_rate": 6.9870539056546145e-06,
+      "loss": 0.5095,
+      "step": 2734
+    },
+    {
+      "epoch": 1.9866978412366922,
+      "grad_norm": 0.43161678194415737,
+      "learning_rate": 6.984863060981835e-06,
+      "loss": 0.5066,
+      "step": 2735
+    },
+    {
+      "epoch": 1.987424238984859,
+      "grad_norm": 0.3628627971741911,
+      "learning_rate": 6.982671763850814e-06,
+      "loss": 0.4919,
+      "step": 2736
+    },
+    {
+      "epoch": 1.9881506367330262,
+      "grad_norm": 0.5086084610925337,
+      "learning_rate": 6.980480014761074e-06,
+      "loss": 0.51,
+      "step": 2737
+    },
+    {
+      "epoch": 1.988877034481193,
+      "grad_norm": 0.3982110835234876,
+      "learning_rate": 6.9782878142122315e-06,
+      "loss": 0.5163,
+      "step": 2738
+    },
+    {
+      "epoch": 1.98960343222936,
+      "grad_norm": 0.3966667256287206,
+      "learning_rate": 6.976095162704006e-06,
+      "loss": 0.4951,
+      "step": 2739
+    },
+    {
+      "epoch": 1.9903298299775272,
+      "grad_norm": 0.5304184543224457,
+      "learning_rate": 6.973902060736226e-06,
+      "loss": 0.5118,
+      "step": 2740
+    },
+    {
+      "epoch": 1.991056227725694,
+      "grad_norm": 0.3981730445998376,
+      "learning_rate": 6.971708508808815e-06,
+      "loss": 0.5081,
+      "step": 2741
+    },
+    {
+      "epoch": 1.9917826254738609,
+      "grad_norm": 0.5015706091927886,
+      "learning_rate": 6.969514507421805e-06,
+      "loss": 0.5093,
+      "step": 2742
+    },
+    {
+      "epoch": 1.992509023222028,
+      "grad_norm": 0.37103098758948716,
+      "learning_rate": 6.967320057075329e-06,
+      "loss": 0.4965,
+      "step": 2743
+    },
+    {
+      "epoch": 1.993235420970195,
+      "grad_norm": 0.5781372172121526,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.5011,
+      "step": 2744
+    },
+    {
+      "epoch": 1.993961818718362,
+      "grad_norm": 0.4157387952447532,
+      "learning_rate": 6.962929811505013e-06,
+      "loss": 0.5058,
+      "step": 2745
+    },
+    {
+      "epoch": 1.994688216466529,
+      "grad_norm": 0.4072625761207944,
+      "learning_rate": 6.9607340172819495e-06,
+      "loss": 0.5087,
+      "step": 2746
+    },
+    {
+      "epoch": 1.9954146142146958,
+      "grad_norm": 0.4575323537660217,
+      "learning_rate": 6.958537776100971e-06,
+      "loss": 0.4713,
+      "step": 2747
+    },
+    {
+      "epoch": 1.9961410119628629,
+      "grad_norm": 0.4236304013107574,
+      "learning_rate": 6.9563410884627195e-06,
+      "loss": 0.4951,
+      "step": 2748
+    },
+    {
+      "epoch": 1.99686740971103,
+      "grad_norm": 0.3395318180298695,
+      "learning_rate": 6.9541439548679394e-06,
+      "loss": 0.5169,
+      "step": 2749
+    },
+    {
+      "epoch": 1.997593807459197,
+      "grad_norm": 0.40012977930638305,
+      "learning_rate": 6.9519463758174745e-06,
+      "loss": 0.4898,
+      "step": 2750
+    },
+    {
+      "epoch": 1.9983202052073639,
+      "grad_norm": 0.6044468828057461,
+      "learning_rate": 6.949748351812277e-06,
+      "loss": 0.4968,
+      "step": 2751
+    },
+    {
+      "epoch": 1.9990466029555307,
+      "grad_norm": 0.3290317646035132,
+      "learning_rate": 6.947549883353393e-06,
+      "loss": 0.4896,
+      "step": 2752
+    },
+    {
+      "epoch": 1.9997730007036978,
+      "grad_norm": 0.4191772501374024,
+      "learning_rate": 6.9453509709419754e-06,
+      "loss": 0.5034,
+      "step": 2753
+    },
+    {
+      "epoch": 2.000499398451865,
+      "grad_norm": 0.4383592732805349,
+      "learning_rate": 6.943151615079273e-06,
+      "loss": 0.4997,
+      "step": 2754
+    },
+    {
+      "epoch": 2.001225796200032,
+      "grad_norm": 0.4310380742661681,
+      "learning_rate": 6.9409518162666416e-06,
+      "loss": 0.5013,
+      "step": 2755
+    },
+    {
+      "epoch": 2.0019521939481986,
+      "grad_norm": 0.6634042092614508,
+      "learning_rate": 6.938751575005531e-06,
+      "loss": 0.5085,
+      "step": 2756
+    },
+    {
+      "epoch": 2.0026785916963656,
+      "grad_norm": 0.43186399669748915,
+      "learning_rate": 6.9365508917975e-06,
+      "loss": 0.5018,
+      "step": 2757
+    },
+    {
+      "epoch": 2.0034049894445327,
+      "grad_norm": 0.34231458896554495,
+      "learning_rate": 6.934349767144203e-06,
+      "loss": 0.4969,
+      "step": 2758
+    },
+    {
+      "epoch": 2.0041313871927,
+      "grad_norm": 0.3852057248556303,
+      "learning_rate": 6.932148201547395e-06,
+      "loss": 0.5099,
+      "step": 2759
+    },
+    {
+      "epoch": 2.004857784940867,
+      "grad_norm": 0.5008104603895283,
+      "learning_rate": 6.929946195508933e-06,
+      "loss": 0.5165,
+      "step": 2760
+    },
+    {
+      "epoch": 2.0055841826890335,
+      "grad_norm": 0.3317056532033946,
+      "learning_rate": 6.9277437495307745e-06,
+      "loss": 0.5002,
+      "step": 2761
+    },
+    {
+      "epoch": 2.0063105804372006,
+      "grad_norm": 0.5412828602249329,
+      "learning_rate": 6.925540864114975e-06,
+      "loss": 0.4962,
+      "step": 2762
+    },
+    {
+      "epoch": 2.0070369781853676,
+      "grad_norm": 0.3430388566201603,
+      "learning_rate": 6.923337539763693e-06,
+      "loss": 0.4937,
+      "step": 2763
+    },
+    {
+      "epoch": 2.0077633759335347,
+      "grad_norm": 0.3647393285999755,
+      "learning_rate": 6.921133776979186e-06,
+      "loss": 0.4975,
+      "step": 2764
+    },
+    {
+      "epoch": 2.008489773681702,
+      "grad_norm": 0.40829618718724847,
+      "learning_rate": 6.918929576263811e-06,
+      "loss": 0.4984,
+      "step": 2765
+    },
+    {
+      "epoch": 2.0092161714298684,
+      "grad_norm": 0.40555403400281487,
+      "learning_rate": 6.916724938120026e-06,
+      "loss": 0.5066,
+      "step": 2766
+    },
+    {
+      "epoch": 2.0099425691780355,
+      "grad_norm": 0.34064395872559766,
+      "learning_rate": 6.914519863050388e-06,
+      "loss": 0.5029,
+      "step": 2767
+    },
+    {
+      "epoch": 2.0106689669262026,
+      "grad_norm": 0.38967717514091643,
+      "learning_rate": 6.912314351557552e-06,
+      "loss": 0.5132,
+      "step": 2768
+    },
+    {
+      "epoch": 2.0113953646743696,
+      "grad_norm": 0.3360598328932662,
+      "learning_rate": 6.910108404144276e-06,
+      "loss": 0.4864,
+      "step": 2769
+    },
+    {
+      "epoch": 2.0121217624225367,
+      "grad_norm": 0.36912864206432744,
+      "learning_rate": 6.907902021313414e-06,
+      "loss": 0.4959,
+      "step": 2770
+    },
+    {
+      "epoch": 2.0128481601707033,
+      "grad_norm": 0.35828330181110757,
+      "learning_rate": 6.905695203567919e-06,
+      "loss": 0.4931,
+      "step": 2771
+    },
+    {
+      "epoch": 2.0135745579188704,
+      "grad_norm": 0.3555368715816114,
+      "learning_rate": 6.90348795141085e-06,
+      "loss": 0.5029,
+      "step": 2772
+    },
+    {
+      "epoch": 2.0143009556670375,
+      "grad_norm": 0.33659315686379043,
+      "learning_rate": 6.901280265345355e-06,
+      "loss": 0.4989,
+      "step": 2773
+    },
+    {
+      "epoch": 2.0150273534152046,
+      "grad_norm": 0.37550584227442796,
+      "learning_rate": 6.8990721458746875e-06,
+      "loss": 0.5201,
+      "step": 2774
+    },
+    {
+      "epoch": 2.015753751163371,
+      "grad_norm": 0.3783837186018702,
+      "learning_rate": 6.896863593502198e-06,
+      "loss": 0.5071,
+      "step": 2775
+    },
+    {
+      "epoch": 2.0164801489115383,
+      "grad_norm": 0.3258935438932034,
+      "learning_rate": 6.894654608731335e-06,
+      "loss": 0.5182,
+      "step": 2776
+    },
+    {
+      "epoch": 2.0172065466597053,
+      "grad_norm": 0.39315885025393604,
+      "learning_rate": 6.8924451920656475e-06,
+      "loss": 0.4923,
+      "step": 2777
+    },
+    {
+      "epoch": 2.0179329444078724,
+      "grad_norm": 0.3632809798562676,
+      "learning_rate": 6.890235344008781e-06,
+      "loss": 0.4815,
+      "step": 2778
+    },
+    {
+      "epoch": 2.0186593421560395,
+      "grad_norm": 0.36527444592888336,
+      "learning_rate": 6.88802506506448e-06,
+      "loss": 0.4956,
+      "step": 2779
+    },
+    {
+      "epoch": 2.019385739904206,
+      "grad_norm": 0.3295945817112341,
+      "learning_rate": 6.8858143557365865e-06,
+      "loss": 0.5091,
+      "step": 2780
+    },
+    {
+      "epoch": 2.020112137652373,
+      "grad_norm": 0.38931789216868123,
+      "learning_rate": 6.883603216529043e-06,
+      "loss": 0.5059,
+      "step": 2781
+    },
+    {
+      "epoch": 2.0208385354005403,
+      "grad_norm": 0.3801269142307661,
+      "learning_rate": 6.881391647945887e-06,
+      "loss": 0.4952,
+      "step": 2782
+    },
+    {
+      "epoch": 2.0215649331487073,
+      "grad_norm": 0.3763092423880549,
+      "learning_rate": 6.879179650491257e-06,
+      "loss": 0.4924,
+      "step": 2783
+    },
+    {
+      "epoch": 2.0222913308968744,
+      "grad_norm": 0.36364396731621496,
+      "learning_rate": 6.8769672246693865e-06,
+      "loss": 0.487,
+      "step": 2784
+    },
+    {
+      "epoch": 2.023017728645041,
+      "grad_norm": 0.48214570723774847,
+      "learning_rate": 6.8747543709846064e-06,
+      "loss": 0.4799,
+      "step": 2785
+    },
+    {
+      "epoch": 2.023744126393208,
+      "grad_norm": 0.3373087430212078,
+      "learning_rate": 6.872541089941347e-06,
+      "loss": 0.5032,
+      "step": 2786
+    },
+    {
+      "epoch": 2.024470524141375,
+      "grad_norm": 0.4246028661887048,
+      "learning_rate": 6.870327382044138e-06,
+      "loss": 0.514,
+      "step": 2787
+    },
+    {
+      "epoch": 2.0251969218895423,
+      "grad_norm": 0.3637838866454337,
+      "learning_rate": 6.868113247797601e-06,
+      "loss": 0.4958,
+      "step": 2788
+    },
+    {
+      "epoch": 2.0259233196377093,
+      "grad_norm": 0.3624459628075546,
+      "learning_rate": 6.865898687706458e-06,
+      "loss": 0.5183,
+      "step": 2789
+    },
+    {
+      "epoch": 2.026649717385876,
+      "grad_norm": 0.3497999660081715,
+      "learning_rate": 6.8636837022755275e-06,
+      "loss": 0.4969,
+      "step": 2790
+    },
+    {
+      "epoch": 2.027376115134043,
+      "grad_norm": 0.30925439334485977,
+      "learning_rate": 6.8614682920097265e-06,
+      "loss": 0.4905,
+      "step": 2791
+    },
+    {
+      "epoch": 2.02810251288221,
+      "grad_norm": 0.4779644594533042,
+      "learning_rate": 6.859252457414067e-06,
+      "loss": 0.4983,
+      "step": 2792
+    },
+    {
+      "epoch": 2.028828910630377,
+      "grad_norm": 0.4156555631857716,
+      "learning_rate": 6.857036198993658e-06,
+      "loss": 0.5111,
+      "step": 2793
+    },
+    {
+      "epoch": 2.029555308378544,
+      "grad_norm": 0.4231134051622082,
+      "learning_rate": 6.8548195172537045e-06,
+      "loss": 0.5015,
+      "step": 2794
+    },
+    {
+      "epoch": 2.030281706126711,
+      "grad_norm": 0.35883752557089954,
+      "learning_rate": 6.8526024126995096e-06,
+      "loss": 0.4901,
+      "step": 2795
+    },
+    {
+      "epoch": 2.031008103874878,
+      "grad_norm": 0.5176921934711561,
+      "learning_rate": 6.850384885836472e-06,
+      "loss": 0.4762,
+      "step": 2796
+    },
+    {
+      "epoch": 2.031734501623045,
+      "grad_norm": 0.47898011289447745,
+      "learning_rate": 6.8481669371700865e-06,
+      "loss": 0.5091,
+      "step": 2797
+    },
+    {
+      "epoch": 2.032460899371212,
+      "grad_norm": 0.38298304746184586,
+      "learning_rate": 6.845948567205945e-06,
+      "loss": 0.4839,
+      "step": 2798
+    },
+    {
+      "epoch": 2.0331872971193787,
+      "grad_norm": 0.49166223619894195,
+      "learning_rate": 6.843729776449734e-06,
+      "loss": 0.4936,
+      "step": 2799
+    },
+    {
+      "epoch": 2.033913694867546,
+      "grad_norm": 0.36732722633298054,
+      "learning_rate": 6.841510565407235e-06,
+      "loss": 0.5045,
+      "step": 2800
+    },
+    {
+      "epoch": 2.034640092615713,
+      "grad_norm": 0.5769337316874685,
+      "learning_rate": 6.839290934584332e-06,
+      "loss": 0.5023,
+      "step": 2801
+    },
+    {
+      "epoch": 2.03536649036388,
+      "grad_norm": 0.34290008976429126,
+      "learning_rate": 6.837070884486994e-06,
+      "loss": 0.504,
+      "step": 2802
+    },
+    {
+      "epoch": 2.036092888112047,
+      "grad_norm": 0.36473122595802115,
+      "learning_rate": 6.8348504156212925e-06,
+      "loss": 0.5153,
+      "step": 2803
+    },
+    {
+      "epoch": 2.0368192858602137,
+      "grad_norm": 0.4304895612429296,
+      "learning_rate": 6.832629528493395e-06,
+      "loss": 0.4997,
+      "step": 2804
+    },
+    {
+      "epoch": 2.0375456836083807,
+      "grad_norm": 0.4387726012724692,
+      "learning_rate": 6.830408223609559e-06,
+      "loss": 0.5105,
+      "step": 2805
+    },
+    {
+      "epoch": 2.038272081356548,
+      "grad_norm": 0.9160945864401482,
+      "learning_rate": 6.828186501476145e-06,
+      "loss": 0.4922,
+      "step": 2806
+    },
+    {
+      "epoch": 2.038998479104715,
+      "grad_norm": 0.36725867499766507,
+      "learning_rate": 6.8259643625996016e-06,
+      "loss": 0.5068,
+      "step": 2807
+    },
+    {
+      "epoch": 2.039724876852882,
+      "grad_norm": 0.36862502375795786,
+      "learning_rate": 6.8237418074864766e-06,
+      "loss": 0.5122,
+      "step": 2808
+    },
+    {
+      "epoch": 2.0404512746010486,
+      "grad_norm": 0.33923787542671696,
+      "learning_rate": 6.8215188366434104e-06,
+      "loss": 0.5058,
+      "step": 2809
+    },
+    {
+      "epoch": 2.0411776723492157,
+      "grad_norm": 0.44498404380627354,
+      "learning_rate": 6.819295450577138e-06,
+      "loss": 0.4926,
+      "step": 2810
+    },
+    {
+      "epoch": 2.0419040700973827,
+      "grad_norm": 0.33216610091469256,
+      "learning_rate": 6.817071649794491e-06,
+      "loss": 0.4973,
+      "step": 2811
+    },
+    {
+      "epoch": 2.04263046784555,
+      "grad_norm": 0.3335571997556061,
+      "learning_rate": 6.8148474348023954e-06,
+      "loss": 0.4831,
+      "step": 2812
+    },
+    {
+      "epoch": 2.043356865593717,
+      "grad_norm": 0.360857049482772,
+      "learning_rate": 6.812622806107869e-06,
+      "loss": 0.5019,
+      "step": 2813
+    },
+    {
+      "epoch": 2.0440832633418835,
+      "grad_norm": 0.4074737961264872,
+      "learning_rate": 6.810397764218027e-06,
+      "loss": 0.5036,
+      "step": 2814
+    },
+    {
+      "epoch": 2.0448096610900506,
+      "grad_norm": 0.334689546487221,
+      "learning_rate": 6.808172309640078e-06,
+      "loss": 0.5003,
+      "step": 2815
+    },
+    {
+      "epoch": 2.0455360588382177,
+      "grad_norm": 0.37500374203543435,
+      "learning_rate": 6.805946442881322e-06,
+      "loss": 0.5213,
+      "step": 2816
+    },
+    {
+      "epoch": 2.0462624565863847,
+      "grad_norm": 0.4206563647325095,
+      "learning_rate": 6.803720164449155e-06,
+      "loss": 0.4977,
+      "step": 2817
+    },
+    {
+      "epoch": 2.0469888543345514,
+      "grad_norm": 0.370548084274352,
+      "learning_rate": 6.801493474851069e-06,
+      "loss": 0.5225,
+      "step": 2818
+    },
+    {
+      "epoch": 2.0477152520827184,
+      "grad_norm": 0.354149466875892,
+      "learning_rate": 6.799266374594646e-06,
+      "loss": 0.4976,
+      "step": 2819
+    },
+    {
+      "epoch": 2.0484416498308855,
+      "grad_norm": 0.3426096343449712,
+      "learning_rate": 6.797038864187564e-06,
+      "loss": 0.5028,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0491680475790526,
+      "grad_norm": 0.3460308554783054,
+      "learning_rate": 6.794810944137595e-06,
+      "loss": 0.5117,
+      "step": 2821
+    },
+    {
+      "epoch": 2.0498944453272196,
+      "grad_norm": 0.3678978614040748,
+      "learning_rate": 6.792582614952602e-06,
+      "loss": 0.4935,
+      "step": 2822
+    },
+    {
+      "epoch": 2.0506208430753863,
+      "grad_norm": 0.3933345821668368,
+      "learning_rate": 6.790353877140542e-06,
+      "loss": 0.5033,
+      "step": 2823
+    },
+    {
+      "epoch": 2.0513472408235534,
+      "grad_norm": 0.3994192333516289,
+      "learning_rate": 6.788124731209466e-06,
+      "loss": 0.4953,
+      "step": 2824
+    },
+    {
+      "epoch": 2.0520736385717204,
+      "grad_norm": 0.36607993612051454,
+      "learning_rate": 6.785895177667516e-06,
+      "loss": 0.4963,
+      "step": 2825
+    },
+    {
+      "epoch": 2.0528000363198875,
+      "grad_norm": 0.36687689120176453,
+      "learning_rate": 6.78366521702293e-06,
+      "loss": 0.5003,
+      "step": 2826
+    },
+    {
+      "epoch": 2.0535264340680546,
+      "grad_norm": 0.42363155301268685,
+      "learning_rate": 6.781434849784039e-06,
+      "loss": 0.51,
+      "step": 2827
+    },
+    {
+      "epoch": 2.054252831816221,
+      "grad_norm": 0.3378823422646072,
+      "learning_rate": 6.779204076459264e-06,
+      "loss": 0.4923,
+      "step": 2828
+    },
+    {
+      "epoch": 2.0549792295643883,
+      "grad_norm": 0.41958106038730947,
+      "learning_rate": 6.776972897557117e-06,
+      "loss": 0.4943,
+      "step": 2829
+    },
+    {
+      "epoch": 2.0557056273125554,
+      "grad_norm": 0.3637963294712784,
+      "learning_rate": 6.774741313586206e-06,
+      "loss": 0.5042,
+      "step": 2830
+    },
+    {
+      "epoch": 2.0564320250607224,
+      "grad_norm": 0.35444308760596194,
+      "learning_rate": 6.772509325055233e-06,
+      "loss": 0.497,
+      "step": 2831
+    },
+    {
+      "epoch": 2.0571584228088895,
+      "grad_norm": 0.30668522235604445,
+      "learning_rate": 6.770276932472987e-06,
+      "loss": 0.5104,
+      "step": 2832
+    },
+    {
+      "epoch": 2.057884820557056,
+      "grad_norm": 0.5739622435217333,
+      "learning_rate": 6.768044136348353e-06,
+      "loss": 0.4828,
+      "step": 2833
+    },
+    {
+      "epoch": 2.058611218305223,
+      "grad_norm": 0.4356973061131128,
+      "learning_rate": 6.765810937190307e-06,
+      "loss": 0.5088,
+      "step": 2834
+    },
+    {
+      "epoch": 2.0593376160533903,
+      "grad_norm": 0.39513686409727394,
+      "learning_rate": 6.763577335507913e-06,
+      "loss": 0.4884,
+      "step": 2835
+    },
+    {
+      "epoch": 2.0600640138015573,
+      "grad_norm": 0.4080599817351704,
+      "learning_rate": 6.761343331810334e-06,
+      "loss": 0.4983,
+      "step": 2836
+    },
+    {
+      "epoch": 2.060790411549724,
+      "grad_norm": 0.4641360221646442,
+      "learning_rate": 6.759108926606821e-06,
+      "loss": 0.5074,
+      "step": 2837
+    },
+    {
+      "epoch": 2.061516809297891,
+      "grad_norm": 0.48283517031623097,
+      "learning_rate": 6.7568741204067145e-06,
+      "loss": 0.4941,
+      "step": 2838
+    },
+    {
+      "epoch": 2.062243207046058,
+      "grad_norm": 0.37677270766893173,
+      "learning_rate": 6.754638913719449e-06,
+      "loss": 0.495,
+      "step": 2839
+    },
+    {
+      "epoch": 2.062969604794225,
+      "grad_norm": 0.3897940565546679,
+      "learning_rate": 6.752403307054549e-06,
+      "loss": 0.4832,
+      "step": 2840
+    },
+    {
+      "epoch": 2.0636960025423923,
+      "grad_norm": 0.3706329128849379,
+      "learning_rate": 6.750167300921635e-06,
+      "loss": 0.4951,
+      "step": 2841
+    },
+    {
+      "epoch": 2.064422400290559,
+      "grad_norm": 0.3677737911519454,
+      "learning_rate": 6.747930895830409e-06,
+      "loss": 0.4971,
+      "step": 2842
+    },
+    {
+      "epoch": 2.065148798038726,
+      "grad_norm": 0.4806697321398137,
+      "learning_rate": 6.745694092290671e-06,
+      "loss": 0.5163,
+      "step": 2843
+    },
+    {
+      "epoch": 2.065875195786893,
+      "grad_norm": 0.3258014555766835,
+      "learning_rate": 6.74345689081231e-06,
+      "loss": 0.4869,
+      "step": 2844
+    },
+    {
+      "epoch": 2.06660159353506,
+      "grad_norm": 0.35430162341459304,
+      "learning_rate": 6.741219291905308e-06,
+      "loss": 0.5137,
+      "step": 2845
+    },
+    {
+      "epoch": 2.067327991283227,
+      "grad_norm": 0.3735414728864755,
+      "learning_rate": 6.738981296079734e-06,
+      "loss": 0.5199,
+      "step": 2846
+    },
+    {
+      "epoch": 2.068054389031394,
+      "grad_norm": 0.4005930008235301,
+      "learning_rate": 6.7367429038457485e-06,
+      "loss": 0.5184,
+      "step": 2847
+    },
+    {
+      "epoch": 2.068780786779561,
+      "grad_norm": 0.46490959194042725,
+      "learning_rate": 6.7345041157136035e-06,
+      "loss": 0.5035,
+      "step": 2848
+    },
+    {
+      "epoch": 2.069507184527728,
+      "grad_norm": 0.336522204114012,
+      "learning_rate": 6.7322649321936395e-06,
+      "loss": 0.4975,
+      "step": 2849
+    },
+    {
+      "epoch": 2.070233582275895,
+      "grad_norm": 0.3625662873299568,
+      "learning_rate": 6.7300253537962905e-06,
+      "loss": 0.5191,
+      "step": 2850
+    },
+    {
+      "epoch": 2.070959980024062,
+      "grad_norm": 0.33860160399452294,
+      "learning_rate": 6.727785381032076e-06,
+      "loss": 0.5,
+      "step": 2851
+    },
+    {
+      "epoch": 2.0716863777722287,
+      "grad_norm": 0.363185881586708,
+      "learning_rate": 6.725545014411608e-06,
+      "loss": 0.4773,
+      "step": 2852
+    },
+    {
+      "epoch": 2.072412775520396,
+      "grad_norm": 0.32039147732115986,
+      "learning_rate": 6.723304254445589e-06,
+      "loss": 0.5072,
+      "step": 2853
+    },
+    {
+      "epoch": 2.073139173268563,
+      "grad_norm": 0.37782007575870213,
+      "learning_rate": 6.72106310164481e-06,
+      "loss": 0.4984,
+      "step": 2854
+    },
+    {
+      "epoch": 2.07386557101673,
+      "grad_norm": 0.33014550752157024,
+      "learning_rate": 6.718821556520151e-06,
+      "loss": 0.4903,
+      "step": 2855
+    },
+    {
+      "epoch": 2.074591968764897,
+      "grad_norm": 0.4828649833387322,
+      "learning_rate": 6.716579619582581e-06,
+      "loss": 0.502,
+      "step": 2856
+    },
+    {
+      "epoch": 2.0753183665130637,
+      "grad_norm": 0.40367185786911486,
+      "learning_rate": 6.71433729134316e-06,
+      "loss": 0.506,
+      "step": 2857
+    },
+    {
+      "epoch": 2.0760447642612307,
+      "grad_norm": 0.3460938036307214,
+      "learning_rate": 6.712094572313038e-06,
+      "loss": 0.496,
+      "step": 2858
+    },
+    {
+      "epoch": 2.076771162009398,
+      "grad_norm": 0.3581135934064617,
+      "learning_rate": 6.70985146300345e-06,
+      "loss": 0.4855,
+      "step": 2859
+    },
+    {
+      "epoch": 2.077497559757565,
+      "grad_norm": 0.3518723354378461,
+      "learning_rate": 6.707607963925725e-06,
+      "loss": 0.5071,
+      "step": 2860
+    },
+    {
+      "epoch": 2.0782239575057315,
+      "grad_norm": 0.41437491459633446,
+      "learning_rate": 6.7053640755912786e-06,
+      "loss": 0.4892,
+      "step": 2861
+    },
+    {
+      "epoch": 2.0789503552538986,
+      "grad_norm": 0.40016810758491533,
+      "learning_rate": 6.703119798511612e-06,
+      "loss": 0.5168,
+      "step": 2862
+    },
+    {
+      "epoch": 2.0796767530020657,
+      "grad_norm": 0.4107704547854394,
+      "learning_rate": 6.700875133198321e-06,
+      "loss": 0.5094,
+      "step": 2863
+    },
+    {
+      "epoch": 2.0804031507502327,
+      "grad_norm": 0.3885733785737631,
+      "learning_rate": 6.698630080163086e-06,
+      "loss": 0.5,
+      "step": 2864
+    },
+    {
+      "epoch": 2.0811295484984,
+      "grad_norm": 0.5226559711677936,
+      "learning_rate": 6.696384639917677e-06,
+      "loss": 0.5024,
+      "step": 2865
+    },
+    {
+      "epoch": 2.0818559462465664,
+      "grad_norm": 0.3797436835719642,
+      "learning_rate": 6.6941388129739504e-06,
+      "loss": 0.4978,
+      "step": 2866
+    },
+    {
+      "epoch": 2.0825823439947335,
+      "grad_norm": 0.3646627564573875,
+      "learning_rate": 6.691892599843856e-06,
+      "loss": 0.5022,
+      "step": 2867
+    },
+    {
+      "epoch": 2.0833087417429006,
+      "grad_norm": 0.36932403717003065,
+      "learning_rate": 6.689646001039422e-06,
+      "loss": 0.4874,
+      "step": 2868
+    },
+    {
+      "epoch": 2.0840351394910677,
+      "grad_norm": 0.3689439163666255,
+      "learning_rate": 6.687399017072775e-06,
+      "loss": 0.4938,
+      "step": 2869
+    },
+    {
+      "epoch": 2.0847615372392347,
+      "grad_norm": 0.3554966085123132,
+      "learning_rate": 6.685151648456124e-06,
+      "loss": 0.4857,
+      "step": 2870
+    },
+    {
+      "epoch": 2.0854879349874014,
+      "grad_norm": 0.6683235515857348,
+      "learning_rate": 6.682903895701767e-06,
+      "loss": 0.4834,
+      "step": 2871
+    },
+    {
+      "epoch": 2.0862143327355684,
+      "grad_norm": 0.37256500006151105,
+      "learning_rate": 6.680655759322086e-06,
+      "loss": 0.5091,
+      "step": 2872
+    },
+    {
+      "epoch": 2.0869407304837355,
+      "grad_norm": 0.3597428273584394,
+      "learning_rate": 6.678407239829558e-06,
+      "loss": 0.5078,
+      "step": 2873
+    },
+    {
+      "epoch": 2.0876671282319026,
+      "grad_norm": 0.4260727973510339,
+      "learning_rate": 6.676158337736738e-06,
+      "loss": 0.5046,
+      "step": 2874
+    },
+    {
+      "epoch": 2.0883935259800697,
+      "grad_norm": 0.41474895800784195,
+      "learning_rate": 6.673909053556278e-06,
+      "loss": 0.489,
+      "step": 2875
+    },
+    {
+      "epoch": 2.0891199237282363,
+      "grad_norm": 0.40947917474288614,
+      "learning_rate": 6.67165938780091e-06,
+      "loss": 0.5022,
+      "step": 2876
+    },
+    {
+      "epoch": 2.0898463214764034,
+      "grad_norm": 0.39584157189048297,
+      "learning_rate": 6.669409340983455e-06,
+      "loss": 0.4945,
+      "step": 2877
+    },
+    {
+      "epoch": 2.0905727192245704,
+      "grad_norm": 0.3825280975330202,
+      "learning_rate": 6.66715891361682e-06,
+      "loss": 0.492,
+      "step": 2878
+    },
+    {
+      "epoch": 2.0912991169727375,
+      "grad_norm": 0.37703028844024256,
+      "learning_rate": 6.664908106214001e-06,
+      "loss": 0.4941,
+      "step": 2879
+    },
+    {
+      "epoch": 2.0920255147209046,
+      "grad_norm": 0.364638637636816,
+      "learning_rate": 6.66265691928808e-06,
+      "loss": 0.5158,
+      "step": 2880
+    },
+    {
+      "epoch": 2.092751912469071,
+      "grad_norm": 0.33601793792720597,
+      "learning_rate": 6.660405353352226e-06,
+      "loss": 0.5131,
+      "step": 2881
+    },
+    {
+      "epoch": 2.0934783102172383,
+      "grad_norm": 0.592614366219023,
+      "learning_rate": 6.658153408919689e-06,
+      "loss": 0.4827,
+      "step": 2882
+    },
+    {
+      "epoch": 2.0942047079654054,
+      "grad_norm": 0.3619686249243115,
+      "learning_rate": 6.6559010865038135e-06,
+      "loss": 0.4959,
+      "step": 2883
+    },
+    {
+      "epoch": 2.0949311057135724,
+      "grad_norm": 0.35243439705723756,
+      "learning_rate": 6.653648386618025e-06,
+      "loss": 0.4988,
+      "step": 2884
+    },
+    {
+      "epoch": 2.095657503461739,
+      "grad_norm": 0.3651217663787908,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.496,
+      "step": 2885
+    },
+    {
+      "epoch": 2.096383901209906,
+      "grad_norm": 0.3842440109628493,
+      "learning_rate": 6.649141856490846e-06,
+      "loss": 0.4979,
+      "step": 2886
+    },
+    {
+      "epoch": 2.097110298958073,
+      "grad_norm": 0.4581904771466546,
+      "learning_rate": 6.646888027276739e-06,
+      "loss": 0.5091,
+      "step": 2887
+    },
+    {
+      "epoch": 2.0978366967062403,
+      "grad_norm": 0.5193582269979015,
+      "learning_rate": 6.644633822647285e-06,
+      "loss": 0.5116,
+      "step": 2888
+    },
+    {
+      "epoch": 2.0985630944544074,
+      "grad_norm": 0.3359619709738642,
+      "learning_rate": 6.6423792431163395e-06,
+      "loss": 0.4923,
+      "step": 2889
+    },
+    {
+      "epoch": 2.099289492202574,
+      "grad_norm": 0.3597705367645605,
+      "learning_rate": 6.640124289197845e-06,
+      "loss": 0.4991,
+      "step": 2890
+    },
+    {
+      "epoch": 2.100015889950741,
+      "grad_norm": 0.407415043370364,
+      "learning_rate": 6.637868961405829e-06,
+      "loss": 0.487,
+      "step": 2891
+    },
+    {
+      "epoch": 2.100742287698908,
+      "grad_norm": 0.37796274922777096,
+      "learning_rate": 6.635613260254401e-06,
+      "loss": 0.4914,
+      "step": 2892
+    },
+    {
+      "epoch": 2.101468685447075,
+      "grad_norm": 0.31201147100352206,
+      "learning_rate": 6.633357186257759e-06,
+      "loss": 0.5067,
+      "step": 2893
+    },
+    {
+      "epoch": 2.1021950831952423,
+      "grad_norm": 0.3265575415617553,
+      "learning_rate": 6.6311007399301855e-06,
+      "loss": 0.4925,
+      "step": 2894
+    },
+    {
+      "epoch": 2.102921480943409,
+      "grad_norm": 0.38724063508599177,
+      "learning_rate": 6.628843921786045e-06,
+      "loss": 0.4985,
+      "step": 2895
+    },
+    {
+      "epoch": 2.103647878691576,
+      "grad_norm": 0.6051025075106355,
+      "learning_rate": 6.626586732339794e-06,
+      "loss": 0.5002,
+      "step": 2896
+    },
+    {
+      "epoch": 2.104374276439743,
+      "grad_norm": 0.38902900482916525,
+      "learning_rate": 6.624329172105964e-06,
+      "loss": 0.5102,
+      "step": 2897
+    },
+    {
+      "epoch": 2.10510067418791,
+      "grad_norm": 0.3551678176261896,
+      "learning_rate": 6.622071241599175e-06,
+      "loss": 0.5054,
+      "step": 2898
+    },
+    {
+      "epoch": 2.105827071936077,
+      "grad_norm": 0.3393281263547164,
+      "learning_rate": 6.619812941334136e-06,
+      "loss": 0.5078,
+      "step": 2899
+    },
+    {
+      "epoch": 2.106553469684244,
+      "grad_norm": 1.5832168634036479,
+      "learning_rate": 6.617554271825636e-06,
+      "loss": 0.4975,
+      "step": 2900
+    },
+    {
+      "epoch": 2.107279867432411,
+      "grad_norm": 0.3944324081013014,
+      "learning_rate": 6.615295233588546e-06,
+      "loss": 0.5181,
+      "step": 2901
+    },
+    {
+      "epoch": 2.108006265180578,
+      "grad_norm": 0.3668693933031106,
+      "learning_rate": 6.613035827137827e-06,
+      "loss": 0.5025,
+      "step": 2902
+    },
+    {
+      "epoch": 2.108732662928745,
+      "grad_norm": 0.35678377216491264,
+      "learning_rate": 6.610776052988519e-06,
+      "loss": 0.4916,
+      "step": 2903
+    },
+    {
+      "epoch": 2.1094590606769117,
+      "grad_norm": 0.3724762793133281,
+      "learning_rate": 6.608515911655744e-06,
+      "loss": 0.4914,
+      "step": 2904
+    },
+    {
+      "epoch": 2.1101854584250788,
+      "grad_norm": 0.29830214021032636,
+      "learning_rate": 6.606255403654717e-06,
+      "loss": 0.4919,
+      "step": 2905
+    },
+    {
+      "epoch": 2.110911856173246,
+      "grad_norm": 0.4018527850111275,
+      "learning_rate": 6.603994529500728e-06,
+      "loss": 0.5077,
+      "step": 2906
+    },
+    {
+      "epoch": 2.111638253921413,
+      "grad_norm": 0.4001560282776422,
+      "learning_rate": 6.601733289709154e-06,
+      "loss": 0.4812,
+      "step": 2907
+    },
+    {
+      "epoch": 2.11236465166958,
+      "grad_norm": 0.4123126231877299,
+      "learning_rate": 6.599471684795452e-06,
+      "loss": 0.4916,
+      "step": 2908
+    },
+    {
+      "epoch": 2.1130910494177466,
+      "grad_norm": 0.31888312898973853,
+      "learning_rate": 6.597209715275168e-06,
+      "loss": 0.4819,
+      "step": 2909
+    },
+    {
+      "epoch": 2.1138174471659137,
+      "grad_norm": 0.37376059043712934,
+      "learning_rate": 6.594947381663926e-06,
+      "loss": 0.4869,
+      "step": 2910
+    },
+    {
+      "epoch": 2.1145438449140808,
+      "grad_norm": 0.3313260091877316,
+      "learning_rate": 6.592684684477435e-06,
+      "loss": 0.5016,
+      "step": 2911
+    },
+    {
+      "epoch": 2.115270242662248,
+      "grad_norm": 0.3587303872117973,
+      "learning_rate": 6.590421624231487e-06,
+      "loss": 0.4815,
+      "step": 2912
+    },
+    {
+      "epoch": 2.115996640410415,
+      "grad_norm": 0.37481813006623715,
+      "learning_rate": 6.588158201441956e-06,
+      "loss": 0.4888,
+      "step": 2913
+    },
+    {
+      "epoch": 2.1167230381585815,
+      "grad_norm": 0.32510590971086595,
+      "learning_rate": 6.5858944166247994e-06,
+      "loss": 0.5049,
+      "step": 2914
+    },
+    {
+      "epoch": 2.1174494359067486,
+      "grad_norm": 0.4162742377307629,
+      "learning_rate": 6.583630270296057e-06,
+      "loss": 0.4937,
+      "step": 2915
+    },
+    {
+      "epoch": 2.1181758336549157,
+      "grad_norm": 0.4144903712969387,
+      "learning_rate": 6.581365762971853e-06,
+      "loss": 0.5029,
+      "step": 2916
+    },
+    {
+      "epoch": 2.1189022314030828,
+      "grad_norm": 0.3926017570855505,
+      "learning_rate": 6.579100895168389e-06,
+      "loss": 0.5063,
+      "step": 2917
+    },
+    {
+      "epoch": 2.11962862915125,
+      "grad_norm": 0.37089262361926517,
+      "learning_rate": 6.576835667401953e-06,
+      "loss": 0.494,
+      "step": 2918
+    },
+    {
+      "epoch": 2.1203550268994165,
+      "grad_norm": 0.414632796429618,
+      "learning_rate": 6.574570080188911e-06,
+      "loss": 0.4996,
+      "step": 2919
+    },
+    {
+      "epoch": 2.1210814246475835,
+      "grad_norm": 0.3823046012390911,
+      "learning_rate": 6.5723041340457175e-06,
+      "loss": 0.506,
+      "step": 2920
+    },
+    {
+      "epoch": 2.1218078223957506,
+      "grad_norm": 0.34052878156073463,
+      "learning_rate": 6.570037829488902e-06,
+      "loss": 0.4983,
+      "step": 2921
+    },
+    {
+      "epoch": 2.1225342201439177,
+      "grad_norm": 0.3739161848070694,
+      "learning_rate": 6.56777116703508e-06,
+      "loss": 0.4977,
+      "step": 2922
+    },
+    {
+      "epoch": 2.1232606178920843,
+      "grad_norm": 0.3433004278765078,
+      "learning_rate": 6.565504147200945e-06,
+      "loss": 0.5015,
+      "step": 2923
+    },
+    {
+      "epoch": 2.1239870156402514,
+      "grad_norm": 0.3546076096060906,
+      "learning_rate": 6.563236770503276e-06,
+      "loss": 0.4928,
+      "step": 2924
+    },
+    {
+      "epoch": 2.1247134133884185,
+      "grad_norm": 0.3522187399759698,
+      "learning_rate": 6.560969037458933e-06,
+      "loss": 0.5002,
+      "step": 2925
+    },
+    {
+      "epoch": 2.1254398111365855,
+      "grad_norm": 0.43583768519910443,
+      "learning_rate": 6.558700948584852e-06,
+      "loss": 0.4951,
+      "step": 2926
+    },
+    {
+      "epoch": 2.1261662088847526,
+      "grad_norm": 0.39854393558885376,
+      "learning_rate": 6.556432504398056e-06,
+      "loss": 0.5109,
+      "step": 2927
+    },
+    {
+      "epoch": 2.1268926066329192,
+      "grad_norm": 0.4879786970248183,
+      "learning_rate": 6.554163705415646e-06,
+      "loss": 0.4779,
+      "step": 2928
+    },
+    {
+      "epoch": 2.1276190043810863,
+      "grad_norm": 0.3796776631929523,
+      "learning_rate": 6.551894552154806e-06,
+      "loss": 0.5165,
+      "step": 2929
+    },
+    {
+      "epoch": 2.1283454021292534,
+      "grad_norm": 0.34666624815506336,
+      "learning_rate": 6.5496250451327996e-06,
+      "loss": 0.4924,
+      "step": 2930
+    },
+    {
+      "epoch": 2.1290717998774205,
+      "grad_norm": 0.41599288142856355,
+      "learning_rate": 6.547355184866968e-06,
+      "loss": 0.5063,
+      "step": 2931
+    },
+    {
+      "epoch": 2.1297981976255875,
+      "grad_norm": 0.3738382621240727,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.4959,
+      "step": 2932
+    },
+    {
+      "epoch": 2.130524595373754,
+      "grad_norm": 0.5175550892040155,
+      "learning_rate": 6.542814406673613e-06,
+      "loss": 0.5029,
+      "step": 2933
+    },
+    {
+      "epoch": 2.1312509931219212,
+      "grad_norm": 0.4203991259923369,
+      "learning_rate": 6.540543489781183e-06,
+      "loss": 0.5143,
+      "step": 2934
+    },
+    {
+      "epoch": 2.1319773908700883,
+      "grad_norm": 0.4217207006128095,
+      "learning_rate": 6.538272221715107e-06,
+      "loss": 0.5133,
+      "step": 2935
+    },
+    {
+      "epoch": 2.1327037886182554,
+      "grad_norm": 0.42445433984426745,
+      "learning_rate": 6.536000602993134e-06,
+      "loss": 0.5033,
+      "step": 2936
+    },
+    {
+      "epoch": 2.1334301863664225,
+      "grad_norm": 0.3376368181992335,
+      "learning_rate": 6.533728634133089e-06,
+      "loss": 0.5009,
+      "step": 2937
+    },
+    {
+      "epoch": 2.134156584114589,
+      "grad_norm": 0.5977879995318446,
+      "learning_rate": 6.531456315652878e-06,
+      "loss": 0.5002,
+      "step": 2938
+    },
+    {
+      "epoch": 2.134882981862756,
+      "grad_norm": 0.4754900561503473,
+      "learning_rate": 6.529183648070484e-06,
+      "loss": 0.5076,
+      "step": 2939
+    },
+    {
+      "epoch": 2.1356093796109232,
+      "grad_norm": 0.3782456325532489,
+      "learning_rate": 6.526910631903973e-06,
+      "loss": 0.5007,
+      "step": 2940
+    },
+    {
+      "epoch": 2.1363357773590903,
+      "grad_norm": 0.37297735671201765,
+      "learning_rate": 6.5246372676714895e-06,
+      "loss": 0.5082,
+      "step": 2941
+    },
+    {
+      "epoch": 2.137062175107257,
+      "grad_norm": 0.3580637523533214,
+      "learning_rate": 6.522363555891255e-06,
+      "loss": 0.5074,
+      "step": 2942
+    },
+    {
+      "epoch": 2.137788572855424,
+      "grad_norm": 0.46572760943088304,
+      "learning_rate": 6.520089497081574e-06,
+      "loss": 0.4881,
+      "step": 2943
+    },
+    {
+      "epoch": 2.138514970603591,
+      "grad_norm": 0.4387344136375126,
+      "learning_rate": 6.5178150917608265e-06,
+      "loss": 0.4914,
+      "step": 2944
+    },
+    {
+      "epoch": 2.139241368351758,
+      "grad_norm": 0.3956585802771223,
+      "learning_rate": 6.515540340447475e-06,
+      "loss": 0.5024,
+      "step": 2945
+    },
+    {
+      "epoch": 2.1399677660999252,
+      "grad_norm": 0.3692454181128716,
+      "learning_rate": 6.513265243660057e-06,
+      "loss": 0.4931,
+      "step": 2946
+    },
+    {
+      "epoch": 2.1406941638480923,
+      "grad_norm": 0.3357775537505742,
+      "learning_rate": 6.5109898019171924e-06,
+      "loss": 0.5035,
+      "step": 2947
+    },
+    {
+      "epoch": 2.141420561596259,
+      "grad_norm": 0.3150057320366553,
+      "learning_rate": 6.508714015737577e-06,
+      "loss": 0.5124,
+      "step": 2948
+    },
+    {
+      "epoch": 2.142146959344426,
+      "grad_norm": 0.3200772493002218,
+      "learning_rate": 6.506437885639988e-06,
+      "loss": 0.5045,
+      "step": 2949
+    },
+    {
+      "epoch": 2.142873357092593,
+      "grad_norm": 0.5384146504997512,
+      "learning_rate": 6.504161412143277e-06,
+      "loss": 0.5008,
+      "step": 2950
+    },
+    {
+      "epoch": 2.14359975484076,
+      "grad_norm": 0.34588162160118074,
+      "learning_rate": 6.5018845957663764e-06,
+      "loss": 0.5021,
+      "step": 2951
+    },
+    {
+      "epoch": 2.144326152588927,
+      "grad_norm": 0.38778620028106675,
+      "learning_rate": 6.499607437028298e-06,
+      "loss": 0.4979,
+      "step": 2952
+    },
+    {
+      "epoch": 2.145052550337094,
+      "grad_norm": 0.7015904343487487,
+      "learning_rate": 6.497329936448129e-06,
+      "loss": 0.4771,
+      "step": 2953
+    },
+    {
+      "epoch": 2.145778948085261,
+      "grad_norm": 0.32149055233561774,
+      "learning_rate": 6.495052094545036e-06,
+      "loss": 0.4944,
+      "step": 2954
+    },
+    {
+      "epoch": 2.146505345833428,
+      "grad_norm": 0.45082281024800525,
+      "learning_rate": 6.492773911838263e-06,
+      "loss": 0.4984,
+      "step": 2955
+    },
+    {
+      "epoch": 2.147231743581595,
+      "grad_norm": 0.4953862513462056,
+      "learning_rate": 6.490495388847131e-06,
+      "loss": 0.4928,
+      "step": 2956
+    },
+    {
+      "epoch": 2.1479581413297617,
+      "grad_norm": 0.3827423100083045,
+      "learning_rate": 6.488216526091042e-06,
+      "loss": 0.4853,
+      "step": 2957
+    },
+    {
+      "epoch": 2.148684539077929,
+      "grad_norm": 0.3300300050257459,
+      "learning_rate": 6.485937324089468e-06,
+      "loss": 0.4915,
+      "step": 2958
+    },
+    {
+      "epoch": 2.149410936826096,
+      "grad_norm": 0.3433551945936361,
+      "learning_rate": 6.483657783361968e-06,
+      "loss": 0.4973,
+      "step": 2959
+    },
+    {
+      "epoch": 2.150137334574263,
+      "grad_norm": 0.3816368521232514,
+      "learning_rate": 6.481377904428171e-06,
+      "loss": 0.4866,
+      "step": 2960
+    },
+    {
+      "epoch": 2.15086373232243,
+      "grad_norm": 0.4339166761827746,
+      "learning_rate": 6.479097687807785e-06,
+      "loss": 0.5095,
+      "step": 2961
+    },
+    {
+      "epoch": 2.1515901300705966,
+      "grad_norm": 0.34956120380265815,
+      "learning_rate": 6.476817134020596e-06,
+      "loss": 0.4938,
+      "step": 2962
+    },
+    {
+      "epoch": 2.1523165278187637,
+      "grad_norm": 0.4460892120279158,
+      "learning_rate": 6.474536243586466e-06,
+      "loss": 0.505,
+      "step": 2963
+    },
+    {
+      "epoch": 2.1530429255669308,
+      "grad_norm": 0.3513400937290252,
+      "learning_rate": 6.472255017025334e-06,
+      "loss": 0.4939,
+      "step": 2964
+    },
+    {
+      "epoch": 2.153769323315098,
+      "grad_norm": 0.36376467899952253,
+      "learning_rate": 6.469973454857217e-06,
+      "loss": 0.4992,
+      "step": 2965
+    },
+    {
+      "epoch": 2.154495721063265,
+      "grad_norm": 0.36673117211425776,
+      "learning_rate": 6.4676915576022045e-06,
+      "loss": 0.5045,
+      "step": 2966
+    },
+    {
+      "epoch": 2.1552221188114316,
+      "grad_norm": 0.3762124077379351,
+      "learning_rate": 6.465409325780467e-06,
+      "loss": 0.5015,
+      "step": 2967
+    },
+    {
+      "epoch": 2.1559485165595986,
+      "grad_norm": 0.35115455230934106,
+      "learning_rate": 6.463126759912248e-06,
+      "loss": 0.5098,
+      "step": 2968
+    },
+    {
+      "epoch": 2.1566749143077657,
+      "grad_norm": 0.35298907038203053,
+      "learning_rate": 6.4608438605178695e-06,
+      "loss": 0.485,
+      "step": 2969
+    },
+    {
+      "epoch": 2.1574013120559328,
+      "grad_norm": 0.3472801797067342,
+      "learning_rate": 6.458560628117727e-06,
+      "loss": 0.5066,
+      "step": 2970
+    },
+    {
+      "epoch": 2.1581277098040994,
+      "grad_norm": 0.36895758673806917,
+      "learning_rate": 6.456277063232296e-06,
+      "loss": 0.5113,
+      "step": 2971
+    },
+    {
+      "epoch": 2.1588541075522665,
+      "grad_norm": 0.6875475378019651,
+      "learning_rate": 6.453993166382122e-06,
+      "loss": 0.5131,
+      "step": 2972
+    },
+    {
+      "epoch": 2.1595805053004336,
+      "grad_norm": 0.3214418264033814,
+      "learning_rate": 6.451708938087831e-06,
+      "loss": 0.4929,
+      "step": 2973
+    },
+    {
+      "epoch": 2.1603069030486006,
+      "grad_norm": 0.34283067423716496,
+      "learning_rate": 6.449424378870123e-06,
+      "loss": 0.4839,
+      "step": 2974
+    },
+    {
+      "epoch": 2.1610333007967677,
+      "grad_norm": 0.30824117611470303,
+      "learning_rate": 6.4471394892497714e-06,
+      "loss": 0.4937,
+      "step": 2975
+    },
+    {
+      "epoch": 2.1617596985449343,
+      "grad_norm": 0.36744675557677187,
+      "learning_rate": 6.44485426974763e-06,
+      "loss": 0.484,
+      "step": 2976
+    },
+    {
+      "epoch": 2.1624860962931014,
+      "grad_norm": 0.36198374511775755,
+      "learning_rate": 6.442568720884621e-06,
+      "loss": 0.51,
+      "step": 2977
+    },
+    {
+      "epoch": 2.1632124940412685,
+      "grad_norm": 0.3395204712387428,
+      "learning_rate": 6.440282843181748e-06,
+      "loss": 0.5048,
+      "step": 2978
+    },
+    {
+      "epoch": 2.1639388917894355,
+      "grad_norm": 0.3586355283964876,
+      "learning_rate": 6.437996637160086e-06,
+      "loss": 0.4987,
+      "step": 2979
+    },
+    {
+      "epoch": 2.1646652895376026,
+      "grad_norm": 0.6935755788835902,
+      "learning_rate": 6.435710103340787e-06,
+      "loss": 0.5091,
+      "step": 2980
+    },
+    {
+      "epoch": 2.1653916872857693,
+      "grad_norm": 0.4025325375633823,
+      "learning_rate": 6.433423242245074e-06,
+      "loss": 0.4987,
+      "step": 2981
+    },
+    {
+      "epoch": 2.1661180850339363,
+      "grad_norm": 0.3298638497588873,
+      "learning_rate": 6.431136054394247e-06,
+      "loss": 0.5024,
+      "step": 2982
+    },
+    {
+      "epoch": 2.1668444827821034,
+      "grad_norm": 0.4156589971885642,
+      "learning_rate": 6.4288485403096825e-06,
+      "loss": 0.507,
+      "step": 2983
+    },
+    {
+      "epoch": 2.1675708805302705,
+      "grad_norm": 0.30768317468521594,
+      "learning_rate": 6.426560700512828e-06,
+      "loss": 0.4919,
+      "step": 2984
+    },
+    {
+      "epoch": 2.1682972782784375,
+      "grad_norm": 0.3163321608467066,
+      "learning_rate": 6.4242725355252075e-06,
+      "loss": 0.5029,
+      "step": 2985
+    },
+    {
+      "epoch": 2.169023676026604,
+      "grad_norm": 0.3998333975752197,
+      "learning_rate": 6.421984045868418e-06,
+      "loss": 0.5134,
+      "step": 2986
+    },
+    {
+      "epoch": 2.1697500737747712,
+      "grad_norm": 0.3610186546760319,
+      "learning_rate": 6.419695232064131e-06,
+      "loss": 0.4949,
+      "step": 2987
+    },
+    {
+      "epoch": 2.1704764715229383,
+      "grad_norm": 0.33557243978359264,
+      "learning_rate": 6.41740609463409e-06,
+      "loss": 0.4966,
+      "step": 2988
+    },
+    {
+      "epoch": 2.1712028692711054,
+      "grad_norm": 0.30979741916167625,
+      "learning_rate": 6.415116634100116e-06,
+      "loss": 0.4861,
+      "step": 2989
+    },
+    {
+      "epoch": 2.171929267019272,
+      "grad_norm": 0.36967468828561945,
+      "learning_rate": 6.412826850984099e-06,
+      "loss": 0.5139,
+      "step": 2990
+    },
+    {
+      "epoch": 2.172655664767439,
+      "grad_norm": 0.510837030995537,
+      "learning_rate": 6.410536745808009e-06,
+      "loss": 0.5146,
+      "step": 2991
+    },
+    {
+      "epoch": 2.173382062515606,
+      "grad_norm": 0.3706820249330839,
+      "learning_rate": 6.408246319093882e-06,
+      "loss": 0.5086,
+      "step": 2992
+    },
+    {
+      "epoch": 2.1741084602637732,
+      "grad_norm": 0.38337344616742824,
+      "learning_rate": 6.405955571363832e-06,
+      "loss": 0.5003,
+      "step": 2993
+    },
+    {
+      "epoch": 2.1748348580119403,
+      "grad_norm": 0.33710792016006064,
+      "learning_rate": 6.403664503140046e-06,
+      "loss": 0.4857,
+      "step": 2994
+    },
+    {
+      "epoch": 2.175561255760107,
+      "grad_norm": 0.3325991143206928,
+      "learning_rate": 6.401373114944781e-06,
+      "loss": 0.504,
+      "step": 2995
+    },
+    {
+      "epoch": 2.176287653508274,
+      "grad_norm": 0.3747987822017251,
+      "learning_rate": 6.399081407300372e-06,
+      "loss": 0.4865,
+      "step": 2996
+    },
+    {
+      "epoch": 2.177014051256441,
+      "grad_norm": 0.8042188756656197,
+      "learning_rate": 6.396789380729218e-06,
+      "loss": 0.5041,
+      "step": 2997
+    },
+    {
+      "epoch": 2.177740449004608,
+      "grad_norm": 0.32635741396114964,
+      "learning_rate": 6.394497035753804e-06,
+      "loss": 0.4877,
+      "step": 2998
+    },
+    {
+      "epoch": 2.1784668467527752,
+      "grad_norm": 0.5737248436672333,
+      "learning_rate": 6.392204372896676e-06,
+      "loss": 0.4965,
+      "step": 2999
+    },
+    {
+      "epoch": 2.179193244500942,
+      "grad_norm": 0.3958470679136034,
+      "learning_rate": 6.3899113926804565e-06,
+      "loss": 0.4748,
+      "step": 3000
+    },
+    {
+      "epoch": 2.179919642249109,
+      "grad_norm": 0.4448283621367889,
+      "learning_rate": 6.387618095627841e-06,
+      "loss": 0.4839,
+      "step": 3001
+    },
+    {
+      "epoch": 2.180646039997276,
+      "grad_norm": 0.5670955549609107,
+      "learning_rate": 6.385324482261597e-06,
+      "loss": 0.4843,
+      "step": 3002
+    },
+    {
+      "epoch": 2.181372437745443,
+      "grad_norm": 0.3817161520952691,
+      "learning_rate": 6.383030553104562e-06,
+      "loss": 0.4961,
+      "step": 3003
+    },
+    {
+      "epoch": 2.18209883549361,
+      "grad_norm": 0.3568277844701838,
+      "learning_rate": 6.380736308679649e-06,
+      "loss": 0.4869,
+      "step": 3004
+    },
+    {
+      "epoch": 2.182825233241777,
+      "grad_norm": 0.35591197709871963,
+      "learning_rate": 6.378441749509841e-06,
+      "loss": 0.4872,
+      "step": 3005
+    },
+    {
+      "epoch": 2.183551630989944,
+      "grad_norm": 0.45168722406676537,
+      "learning_rate": 6.376146876118193e-06,
+      "loss": 0.4842,
+      "step": 3006
+    },
+    {
+      "epoch": 2.184278028738111,
+      "grad_norm": 0.36618818991265534,
+      "learning_rate": 6.37385168902783e-06,
+      "loss": 0.5104,
+      "step": 3007
+    },
+    {
+      "epoch": 2.185004426486278,
+      "grad_norm": 0.4164922266504299,
+      "learning_rate": 6.37155618876195e-06,
+      "loss": 0.4961,
+      "step": 3008
+    },
+    {
+      "epoch": 2.1857308242344446,
+      "grad_norm": 0.33660435418460627,
+      "learning_rate": 6.369260375843825e-06,
+      "loss": 0.494,
+      "step": 3009
+    },
+    {
+      "epoch": 2.1864572219826117,
+      "grad_norm": 0.3661615173054258,
+      "learning_rate": 6.366964250796794e-06,
+      "loss": 0.5024,
+      "step": 3010
+    },
+    {
+      "epoch": 2.187183619730779,
+      "grad_norm": 0.3621135555183418,
+      "learning_rate": 6.364667814144267e-06,
+      "loss": 0.494,
+      "step": 3011
+    },
+    {
+      "epoch": 2.187910017478946,
+      "grad_norm": 0.39242459700593657,
+      "learning_rate": 6.362371066409727e-06,
+      "loss": 0.4949,
+      "step": 3012
+    },
+    {
+      "epoch": 2.188636415227113,
+      "grad_norm": 0.35575069031596895,
+      "learning_rate": 6.360074008116732e-06,
+      "loss": 0.5088,
+      "step": 3013
+    },
+    {
+      "epoch": 2.1893628129752796,
+      "grad_norm": 0.36556209399790124,
+      "learning_rate": 6.357776639788904e-06,
+      "loss": 0.5023,
+      "step": 3014
+    },
+    {
+      "epoch": 2.1900892107234466,
+      "grad_norm": 0.34006098095149784,
+      "learning_rate": 6.355478961949935e-06,
+      "loss": 0.5036,
+      "step": 3015
+    },
+    {
+      "epoch": 2.1908156084716137,
+      "grad_norm": 0.3649755041154573,
+      "learning_rate": 6.353180975123595e-06,
+      "loss": 0.5236,
+      "step": 3016
+    },
+    {
+      "epoch": 2.191542006219781,
+      "grad_norm": 0.5093498310975254,
+      "learning_rate": 6.3508826798337184e-06,
+      "loss": 0.4942,
+      "step": 3017
+    },
+    {
+      "epoch": 2.192268403967948,
+      "grad_norm": 0.3997941447173825,
+      "learning_rate": 6.348584076604213e-06,
+      "loss": 0.4917,
+      "step": 3018
+    },
+    {
+      "epoch": 2.1929948017161145,
+      "grad_norm": 0.3734740659212164,
+      "learning_rate": 6.346285165959053e-06,
+      "loss": 0.4966,
+      "step": 3019
+    },
+    {
+      "epoch": 2.1937211994642816,
+      "grad_norm": 0.5631358961409713,
+      "learning_rate": 6.3439859484222874e-06,
+      "loss": 0.4836,
+      "step": 3020
+    },
+    {
+      "epoch": 2.1944475972124486,
+      "grad_norm": 0.3562949020025143,
+      "learning_rate": 6.341686424518032e-06,
+      "loss": 0.4878,
+      "step": 3021
+    },
+    {
+      "epoch": 2.1951739949606157,
+      "grad_norm": 0.44992840420045366,
+      "learning_rate": 6.339386594770472e-06,
+      "loss": 0.5014,
+      "step": 3022
+    },
+    {
+      "epoch": 2.195900392708783,
+      "grad_norm": 0.34442915629672294,
+      "learning_rate": 6.337086459703867e-06,
+      "loss": 0.4986,
+      "step": 3023
+    },
+    {
+      "epoch": 2.1966267904569494,
+      "grad_norm": 0.39024905830114803,
+      "learning_rate": 6.33478601984254e-06,
+      "loss": 0.5024,
+      "step": 3024
+    },
+    {
+      "epoch": 2.1973531882051165,
+      "grad_norm": 0.3998500751389736,
+      "learning_rate": 6.332485275710889e-06,
+      "loss": 0.4949,
+      "step": 3025
+    },
+    {
+      "epoch": 2.1980795859532836,
+      "grad_norm": 0.3429294829411977,
+      "learning_rate": 6.330184227833376e-06,
+      "loss": 0.4807,
+      "step": 3026
+    },
+    {
+      "epoch": 2.1988059837014506,
+      "grad_norm": 0.690099298899355,
+      "learning_rate": 6.327882876734537e-06,
+      "loss": 0.4909,
+      "step": 3027
+    },
+    {
+      "epoch": 2.1995323814496177,
+      "grad_norm": 0.35195712782936517,
+      "learning_rate": 6.325581222938972e-06,
+      "loss": 0.48,
+      "step": 3028
+    },
+    {
+      "epoch": 2.2002587791977843,
+      "grad_norm": 0.328965359456782,
+      "learning_rate": 6.323279266971357e-06,
+      "loss": 0.4956,
+      "step": 3029
+    },
+    {
+      "epoch": 2.2009851769459514,
+      "grad_norm": 0.4625848240891544,
+      "learning_rate": 6.3209770093564315e-06,
+      "loss": 0.5042,
+      "step": 3030
+    },
+    {
+      "epoch": 2.2017115746941185,
+      "grad_norm": 0.4053912419568821,
+      "learning_rate": 6.318674450619002e-06,
+      "loss": 0.494,
+      "step": 3031
+    },
+    {
+      "epoch": 2.2024379724422856,
+      "grad_norm": 0.3512279925789144,
+      "learning_rate": 6.316371591283953e-06,
+      "loss": 0.4903,
+      "step": 3032
+    },
+    {
+      "epoch": 2.2031643701904526,
+      "grad_norm": 0.3636383991578603,
+      "learning_rate": 6.3140684318762265e-06,
+      "loss": 0.4762,
+      "step": 3033
+    },
+    {
+      "epoch": 2.2038907679386193,
+      "grad_norm": 0.3852602096351379,
+      "learning_rate": 6.31176497292084e-06,
+      "loss": 0.4977,
+      "step": 3034
+    },
+    {
+      "epoch": 2.2046171656867863,
+      "grad_norm": 0.3947804255450504,
+      "learning_rate": 6.309461214942877e-06,
+      "loss": 0.4976,
+      "step": 3035
+    },
+    {
+      "epoch": 2.2053435634349534,
+      "grad_norm": 0.3790065839004977,
+      "learning_rate": 6.307157158467487e-06,
+      "loss": 0.4863,
+      "step": 3036
+    },
+    {
+      "epoch": 2.2060699611831205,
+      "grad_norm": 0.41911269447716715,
+      "learning_rate": 6.304852804019892e-06,
+      "loss": 0.4962,
+      "step": 3037
+    },
+    {
+      "epoch": 2.206796358931287,
+      "grad_norm": 0.35734766993699474,
+      "learning_rate": 6.30254815212538e-06,
+      "loss": 0.5068,
+      "step": 3038
+    },
+    {
+      "epoch": 2.207522756679454,
+      "grad_norm": 0.6636967409043618,
+      "learning_rate": 6.3002432033093055e-06,
+      "loss": 0.4963,
+      "step": 3039
+    },
+    {
+      "epoch": 2.2082491544276213,
+      "grad_norm": 0.32487300685748505,
+      "learning_rate": 6.297937958097094e-06,
+      "loss": 0.4934,
+      "step": 3040
+    },
+    {
+      "epoch": 2.2089755521757883,
+      "grad_norm": 0.31860100445543343,
+      "learning_rate": 6.29563241701423e-06,
+      "loss": 0.5128,
+      "step": 3041
+    },
+    {
+      "epoch": 2.2097019499239554,
+      "grad_norm": 0.644224408832893,
+      "learning_rate": 6.293326580586278e-06,
+      "loss": 0.495,
+      "step": 3042
+    },
+    {
+      "epoch": 2.210428347672122,
+      "grad_norm": 0.3863384310540497,
+      "learning_rate": 6.29102044933886e-06,
+      "loss": 0.5091,
+      "step": 3043
+    },
+    {
+      "epoch": 2.211154745420289,
+      "grad_norm": 0.3246376834758291,
+      "learning_rate": 6.2887140237976714e-06,
+      "loss": 0.5011,
+      "step": 3044
+    },
+    {
+      "epoch": 2.211881143168456,
+      "grad_norm": 0.34220816049628433,
+      "learning_rate": 6.286407304488471e-06,
+      "loss": 0.4955,
+      "step": 3045
+    },
+    {
+      "epoch": 2.2126075409166233,
+      "grad_norm": 0.32757096583048706,
+      "learning_rate": 6.284100291937083e-06,
+      "loss": 0.5024,
+      "step": 3046
+    },
+    {
+      "epoch": 2.2133339386647903,
+      "grad_norm": 0.37712871027281575,
+      "learning_rate": 6.281792986669406e-06,
+      "loss": 0.4955,
+      "step": 3047
+    },
+    {
+      "epoch": 2.214060336412957,
+      "grad_norm": 0.4039253546929864,
+      "learning_rate": 6.279485389211396e-06,
+      "loss": 0.4844,
+      "step": 3048
+    },
+    {
+      "epoch": 2.214786734161124,
+      "grad_norm": 0.3442024462559989,
+      "learning_rate": 6.277177500089082e-06,
+      "loss": 0.5045,
+      "step": 3049
+    },
+    {
+      "epoch": 2.215513131909291,
+      "grad_norm": 0.36987896779171475,
+      "learning_rate": 6.274869319828558e-06,
+      "loss": 0.4955,
+      "step": 3050
+    },
+    {
+      "epoch": 2.216239529657458,
+      "grad_norm": 0.33664971222500834,
+      "learning_rate": 6.272560848955982e-06,
+      "loss": 0.4976,
+      "step": 3051
+    },
+    {
+      "epoch": 2.2169659274056253,
+      "grad_norm": 0.38221122818636305,
+      "learning_rate": 6.270252087997581e-06,
+      "loss": 0.4949,
+      "step": 3052
+    },
+    {
+      "epoch": 2.217692325153792,
+      "grad_norm": 0.3254009518810701,
+      "learning_rate": 6.267943037479647e-06,
+      "loss": 0.504,
+      "step": 3053
+    },
+    {
+      "epoch": 2.218418722901959,
+      "grad_norm": 0.3533296250544195,
+      "learning_rate": 6.265633697928539e-06,
+      "loss": 0.495,
+      "step": 3054
+    },
+    {
+      "epoch": 2.219145120650126,
+      "grad_norm": 0.4377959364901721,
+      "learning_rate": 6.26332406987068e-06,
+      "loss": 0.4848,
+      "step": 3055
+    },
+    {
+      "epoch": 2.219871518398293,
+      "grad_norm": 0.44974371152679193,
+      "learning_rate": 6.261014153832559e-06,
+      "loss": 0.5034,
+      "step": 3056
+    },
+    {
+      "epoch": 2.2205979161464597,
+      "grad_norm": 0.4524681508883102,
+      "learning_rate": 6.258703950340734e-06,
+      "loss": 0.5038,
+      "step": 3057
+    },
+    {
+      "epoch": 2.221324313894627,
+      "grad_norm": 0.45769261482601487,
+      "learning_rate": 6.256393459921824e-06,
+      "loss": 0.5083,
+      "step": 3058
+    },
+    {
+      "epoch": 2.222050711642794,
+      "grad_norm": 0.37051014434450674,
+      "learning_rate": 6.254082683102517e-06,
+      "loss": 0.4945,
+      "step": 3059
+    },
+    {
+      "epoch": 2.222777109390961,
+      "grad_norm": 0.3394404946299637,
+      "learning_rate": 6.251771620409563e-06,
+      "loss": 0.4961,
+      "step": 3060
+    },
+    {
+      "epoch": 2.223503507139128,
+      "grad_norm": 0.33392609888135194,
+      "learning_rate": 6.24946027236978e-06,
+      "loss": 0.4922,
+      "step": 3061
+    },
+    {
+      "epoch": 2.2242299048872947,
+      "grad_norm": 0.39803836210224014,
+      "learning_rate": 6.24714863951005e-06,
+      "loss": 0.5061,
+      "step": 3062
+    },
+    {
+      "epoch": 2.2249563026354617,
+      "grad_norm": 0.5053484367967763,
+      "learning_rate": 6.244836722357319e-06,
+      "loss": 0.4951,
+      "step": 3063
+    },
+    {
+      "epoch": 2.225682700383629,
+      "grad_norm": 0.3402645327719718,
+      "learning_rate": 6.2425245214386e-06,
+      "loss": 0.488,
+      "step": 3064
+    },
+    {
+      "epoch": 2.226409098131796,
+      "grad_norm": 0.363552088042168,
+      "learning_rate": 6.240212037280967e-06,
+      "loss": 0.487,
+      "step": 3065
+    },
+    {
+      "epoch": 2.227135495879963,
+      "grad_norm": 0.5809668434482768,
+      "learning_rate": 6.237899270411564e-06,
+      "loss": 0.4819,
+      "step": 3066
+    },
+    {
+      "epoch": 2.2278618936281296,
+      "grad_norm": 0.36027206956518293,
+      "learning_rate": 6.2355862213575935e-06,
+      "loss": 0.5039,
+      "step": 3067
+    },
+    {
+      "epoch": 2.2285882913762967,
+      "grad_norm": 0.3955705798934775,
+      "learning_rate": 6.233272890646327e-06,
+      "loss": 0.4914,
+      "step": 3068
+    },
+    {
+      "epoch": 2.2293146891244637,
+      "grad_norm": 0.36772804787972874,
+      "learning_rate": 6.230959278805098e-06,
+      "loss": 0.494,
+      "step": 3069
+    },
+    {
+      "epoch": 2.230041086872631,
+      "grad_norm": 0.3811689830122236,
+      "learning_rate": 6.228645386361304e-06,
+      "loss": 0.5019,
+      "step": 3070
+    },
+    {
+      "epoch": 2.230767484620798,
+      "grad_norm": 0.3383061085045838,
+      "learning_rate": 6.226331213842406e-06,
+      "loss": 0.4885,
+      "step": 3071
+    },
+    {
+      "epoch": 2.2314938823689645,
+      "grad_norm": 0.36887931278841335,
+      "learning_rate": 6.224016761775933e-06,
+      "loss": 0.493,
+      "step": 3072
+    },
+    {
+      "epoch": 2.2322202801171316,
+      "grad_norm": 0.47523629811023393,
+      "learning_rate": 6.2217020306894705e-06,
+      "loss": 0.4951,
+      "step": 3073
+    },
+    {
+      "epoch": 2.2329466778652987,
+      "grad_norm": 0.3919649192352115,
+      "learning_rate": 6.2193870211106745e-06,
+      "loss": 0.4818,
+      "step": 3074
+    },
+    {
+      "epoch": 2.2336730756134657,
+      "grad_norm": 0.33973722288683916,
+      "learning_rate": 6.217071733567261e-06,
+      "loss": 0.4911,
+      "step": 3075
+    },
+    {
+      "epoch": 2.2343994733616324,
+      "grad_norm": 0.39854275169930475,
+      "learning_rate": 6.214756168587009e-06,
+      "loss": 0.5012,
+      "step": 3076
+    },
+    {
+      "epoch": 2.2351258711097994,
+      "grad_norm": 0.44043656008659554,
+      "learning_rate": 6.212440326697762e-06,
+      "loss": 0.5017,
+      "step": 3077
+    },
+    {
+      "epoch": 2.2358522688579665,
+      "grad_norm": 0.3722796430104987,
+      "learning_rate": 6.210124208427427e-06,
+      "loss": 0.4996,
+      "step": 3078
+    },
+    {
+      "epoch": 2.2365786666061336,
+      "grad_norm": 0.3412876211388963,
+      "learning_rate": 6.207807814303973e-06,
+      "loss": 0.5017,
+      "step": 3079
+    },
+    {
+      "epoch": 2.2373050643543007,
+      "grad_norm": 0.32142078513383754,
+      "learning_rate": 6.205491144855432e-06,
+      "loss": 0.4852,
+      "step": 3080
+    },
+    {
+      "epoch": 2.2380314621024673,
+      "grad_norm": 0.4026214624360047,
+      "learning_rate": 6.203174200609899e-06,
+      "loss": 0.4937,
+      "step": 3081
+    },
+    {
+      "epoch": 2.2387578598506344,
+      "grad_norm": 0.35889235187183816,
+      "learning_rate": 6.200856982095532e-06,
+      "loss": 0.495,
+      "step": 3082
+    },
+    {
+      "epoch": 2.2394842575988014,
+      "grad_norm": 0.3788649730714019,
+      "learning_rate": 6.19853948984055e-06,
+      "loss": 0.4982,
+      "step": 3083
+    },
+    {
+      "epoch": 2.2402106553469685,
+      "grad_norm": 0.36923423941185074,
+      "learning_rate": 6.196221724373237e-06,
+      "loss": 0.4852,
+      "step": 3084
+    },
+    {
+      "epoch": 2.2409370530951356,
+      "grad_norm": 0.3499125392928531,
+      "learning_rate": 6.193903686221937e-06,
+      "loss": 0.4993,
+      "step": 3085
+    },
+    {
+      "epoch": 2.241663450843302,
+      "grad_norm": 0.41064362066578464,
+      "learning_rate": 6.191585375915056e-06,
+      "loss": 0.4944,
+      "step": 3086
+    },
+    {
+      "epoch": 2.2423898485914693,
+      "grad_norm": 0.38299816267520065,
+      "learning_rate": 6.189266793981064e-06,
+      "loss": 0.4726,
+      "step": 3087
+    },
+    {
+      "epoch": 2.2431162463396364,
+      "grad_norm": 0.3238281351522023,
+      "learning_rate": 6.186947940948494e-06,
+      "loss": 0.4907,
+      "step": 3088
+    },
+    {
+      "epoch": 2.2438426440878034,
+      "grad_norm": 0.38073181834522724,
+      "learning_rate": 6.184628817345936e-06,
+      "loss": 0.5025,
+      "step": 3089
+    },
+    {
+      "epoch": 2.2445690418359705,
+      "grad_norm": 0.31508364422589685,
+      "learning_rate": 6.182309423702044e-06,
+      "loss": 0.4906,
+      "step": 3090
+    },
+    {
+      "epoch": 2.245295439584137,
+      "grad_norm": 0.47437382669171824,
+      "learning_rate": 6.179989760545535e-06,
+      "loss": 0.4859,
+      "step": 3091
+    },
+    {
+      "epoch": 2.246021837332304,
+      "grad_norm": 0.3875849029709213,
+      "learning_rate": 6.177669828405186e-06,
+      "loss": 0.5086,
+      "step": 3092
+    },
+    {
+      "epoch": 2.2467482350804713,
+      "grad_norm": 0.3891730372212856,
+      "learning_rate": 6.175349627809839e-06,
+      "loss": 0.5044,
+      "step": 3093
+    },
+    {
+      "epoch": 2.2474746328286384,
+      "grad_norm": 0.43368525656394613,
+      "learning_rate": 6.173029159288388e-06,
+      "loss": 0.5012,
+      "step": 3094
+    },
+    {
+      "epoch": 2.248201030576805,
+      "grad_norm": 0.6038511756961967,
+      "learning_rate": 6.1707084233697974e-06,
+      "loss": 0.4894,
+      "step": 3095
+    },
+    {
+      "epoch": 2.248927428324972,
+      "grad_norm": 0.3066072645157146,
+      "learning_rate": 6.1683874205830905e-06,
+      "loss": 0.498,
+      "step": 3096
+    },
+    {
+      "epoch": 2.249653826073139,
+      "grad_norm": 0.3188905488105647,
+      "learning_rate": 6.166066151457347e-06,
+      "loss": 0.4794,
+      "step": 3097
+    },
+    {
+      "epoch": 2.250380223821306,
+      "grad_norm": 0.43514160901236826,
+      "learning_rate": 6.163744616521712e-06,
+      "loss": 0.4904,
+      "step": 3098
+    },
+    {
+      "epoch": 2.2511066215694733,
+      "grad_norm": 0.36486654258921625,
+      "learning_rate": 6.161422816305389e-06,
+      "loss": 0.4897,
+      "step": 3099
+    },
+    {
+      "epoch": 2.2518330193176403,
+      "grad_norm": 0.3480152596351159,
+      "learning_rate": 6.1591007513376425e-06,
+      "loss": 0.4867,
+      "step": 3100
+    },
+    {
+      "epoch": 2.252559417065807,
+      "grad_norm": 0.6364902568308674,
+      "learning_rate": 6.156778422147797e-06,
+      "loss": 0.5002,
+      "step": 3101
+    },
+    {
+      "epoch": 2.253285814813974,
+      "grad_norm": 0.3973215322038487,
+      "learning_rate": 6.154455829265239e-06,
+      "loss": 0.4754,
+      "step": 3102
+    },
+    {
+      "epoch": 2.254012212562141,
+      "grad_norm": 0.4008644063048734,
+      "learning_rate": 6.152132973219414e-06,
+      "loss": 0.5,
+      "step": 3103
+    },
+    {
+      "epoch": 2.254738610310308,
+      "grad_norm": 0.770372781416335,
+      "learning_rate": 6.1498098545398255e-06,
+      "loss": 0.4892,
+      "step": 3104
+    },
+    {
+      "epoch": 2.255465008058475,
+      "grad_norm": 0.3680821695745185,
+      "learning_rate": 6.147486473756039e-06,
+      "loss": 0.4938,
+      "step": 3105
+    },
+    {
+      "epoch": 2.256191405806642,
+      "grad_norm": 0.4193679580025291,
+      "learning_rate": 6.14516283139768e-06,
+      "loss": 0.4937,
+      "step": 3106
+    },
+    {
+      "epoch": 2.256917803554809,
+      "grad_norm": 0.35600056467080976,
+      "learning_rate": 6.142838927994432e-06,
+      "loss": 0.491,
+      "step": 3107
+    },
+    {
+      "epoch": 2.257644201302976,
+      "grad_norm": 0.3740725509927286,
+      "learning_rate": 6.1405147640760395e-06,
+      "loss": 0.4931,
+      "step": 3108
+    },
+    {
+      "epoch": 2.258370599051143,
+      "grad_norm": 0.34855260413863015,
+      "learning_rate": 6.138190340172308e-06,
+      "loss": 0.5016,
+      "step": 3109
+    },
+    {
+      "epoch": 2.2590969967993098,
+      "grad_norm": 0.5207683008045049,
+      "learning_rate": 6.135865656813095e-06,
+      "loss": 0.5182,
+      "step": 3110
+    },
+    {
+      "epoch": 2.259823394547477,
+      "grad_norm": 0.40710281188709585,
+      "learning_rate": 6.1335407145283265e-06,
+      "loss": 0.4872,
+      "step": 3111
+    },
+    {
+      "epoch": 2.260549792295644,
+      "grad_norm": 0.38524703767888147,
+      "learning_rate": 6.131215513847982e-06,
+      "loss": 0.4888,
+      "step": 3112
+    },
+    {
+      "epoch": 2.261276190043811,
+      "grad_norm": 0.3611442300937129,
+      "learning_rate": 6.128890055302103e-06,
+      "loss": 0.4971,
+      "step": 3113
+    },
+    {
+      "epoch": 2.2620025877919776,
+      "grad_norm": 0.3726264631189811,
+      "learning_rate": 6.126564339420784e-06,
+      "loss": 0.4982,
+      "step": 3114
+    },
+    {
+      "epoch": 2.2627289855401447,
+      "grad_norm": 0.35003659748054033,
+      "learning_rate": 6.124238366734185e-06,
+      "loss": 0.4857,
+      "step": 3115
+    },
+    {
+      "epoch": 2.2634553832883118,
+      "grad_norm": 0.44795422664526696,
+      "learning_rate": 6.121912137772521e-06,
+      "loss": 0.4933,
+      "step": 3116
+    },
+    {
+      "epoch": 2.264181781036479,
+      "grad_norm": 0.38836552057562496,
+      "learning_rate": 6.119585653066067e-06,
+      "loss": 0.4868,
+      "step": 3117
+    },
+    {
+      "epoch": 2.264908178784646,
+      "grad_norm": 0.3874958318464041,
+      "learning_rate": 6.117258913145153e-06,
+      "loss": 0.4896,
+      "step": 3118
+    },
+    {
+      "epoch": 2.265634576532813,
+      "grad_norm": 0.4184151509447441,
+      "learning_rate": 6.114931918540172e-06,
+      "loss": 0.4927,
+      "step": 3119
+    },
+    {
+      "epoch": 2.2663609742809796,
+      "grad_norm": 0.5489700776940928,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.4811,
+      "step": 3120
+    },
+    {
+      "epoch": 2.2670873720291467,
+      "grad_norm": 0.37173040857538353,
+      "learning_rate": 6.110277167399861e-06,
+      "loss": 0.4978,
+      "step": 3121
+    },
+    {
+      "epoch": 2.2678137697773137,
+      "grad_norm": 0.3215371108487471,
+      "learning_rate": 6.107949411925599e-06,
+      "loss": 0.4808,
+      "step": 3122
+    },
+    {
+      "epoch": 2.268540167525481,
+      "grad_norm": 0.4337388157888445,
+      "learning_rate": 6.105621403889411e-06,
+      "loss": 0.501,
+      "step": 3123
+    },
+    {
+      "epoch": 2.2692665652736475,
+      "grad_norm": 0.33974619025502883,
+      "learning_rate": 6.103293143821978e-06,
+      "loss": 0.4955,
+      "step": 3124
+    },
+    {
+      "epoch": 2.2699929630218145,
+      "grad_norm": 0.4190678472874704,
+      "learning_rate": 6.100964632254033e-06,
+      "loss": 0.489,
+      "step": 3125
+    },
+    {
+      "epoch": 2.2707193607699816,
+      "grad_norm": 0.4983709975295048,
+      "learning_rate": 6.098635869716375e-06,
+      "loss": 0.4962,
+      "step": 3126
+    },
+    {
+      "epoch": 2.2714457585181487,
+      "grad_norm": 0.4360393265527959,
+      "learning_rate": 6.0963068567398535e-06,
+      "loss": 0.49,
+      "step": 3127
+    },
+    {
+      "epoch": 2.2721721562663157,
+      "grad_norm": 0.347103768192224,
+      "learning_rate": 6.093977593855376e-06,
+      "loss": 0.4925,
+      "step": 3128
+    },
+    {
+      "epoch": 2.2728985540144824,
+      "grad_norm": 0.4549595372358685,
+      "learning_rate": 6.0916480815939095e-06,
+      "loss": 0.4792,
+      "step": 3129
+    },
+    {
+      "epoch": 2.2736249517626494,
+      "grad_norm": 0.42519053661063105,
+      "learning_rate": 6.089318320486477e-06,
+      "loss": 0.5004,
+      "step": 3130
+    },
+    {
+      "epoch": 2.2743513495108165,
+      "grad_norm": 0.34401001903742934,
+      "learning_rate": 6.086988311064157e-06,
+      "loss": 0.4918,
+      "step": 3131
+    },
+    {
+      "epoch": 2.2750777472589836,
+      "grad_norm": 0.3675422068036801,
+      "learning_rate": 6.084658053858086e-06,
+      "loss": 0.4946,
+      "step": 3132
+    },
+    {
+      "epoch": 2.2758041450071507,
+      "grad_norm": 0.6373015019480114,
+      "learning_rate": 6.082327549399456e-06,
+      "loss": 0.4968,
+      "step": 3133
+    },
+    {
+      "epoch": 2.2765305427553173,
+      "grad_norm": 0.9060836261092253,
+      "learning_rate": 6.079996798219516e-06,
+      "loss": 0.5049,
+      "step": 3134
+    },
+    {
+      "epoch": 2.2772569405034844,
+      "grad_norm": 0.4086724137891712,
+      "learning_rate": 6.077665800849568e-06,
+      "loss": 0.4936,
+      "step": 3135
+    },
+    {
+      "epoch": 2.2779833382516514,
+      "grad_norm": 0.42824112282521215,
+      "learning_rate": 6.075334557820977e-06,
+      "loss": 0.5069,
+      "step": 3136
+    },
+    {
+      "epoch": 2.2787097359998185,
+      "grad_norm": 0.4370864620437589,
+      "learning_rate": 6.073003069665161e-06,
+      "loss": 0.4845,
+      "step": 3137
+    },
+    {
+      "epoch": 2.2794361337479856,
+      "grad_norm": 0.34589256703550214,
+      "learning_rate": 6.070671336913588e-06,
+      "loss": 0.4836,
+      "step": 3138
+    },
+    {
+      "epoch": 2.2801625314961522,
+      "grad_norm": 0.44081977847163717,
+      "learning_rate": 6.068339360097791e-06,
+      "loss": 0.499,
+      "step": 3139
+    },
+    {
+      "epoch": 2.2808889292443193,
+      "grad_norm": 0.3693827842084408,
+      "learning_rate": 6.066007139749351e-06,
+      "loss": 0.4819,
+      "step": 3140
+    },
+    {
+      "epoch": 2.2816153269924864,
+      "grad_norm": 0.4114801998919762,
+      "learning_rate": 6.063674676399911e-06,
+      "loss": 0.4971,
+      "step": 3141
+    },
+    {
+      "epoch": 2.2823417247406534,
+      "grad_norm": 0.5174468598265398,
+      "learning_rate": 6.061341970581165e-06,
+      "loss": 0.5008,
+      "step": 3142
+    },
+    {
+      "epoch": 2.28306812248882,
+      "grad_norm": 0.7341098293975072,
+      "learning_rate": 6.0590090228248625e-06,
+      "loss": 0.4929,
+      "step": 3143
+    },
+    {
+      "epoch": 2.283794520236987,
+      "grad_norm": 0.3505216954061479,
+      "learning_rate": 6.056675833662811e-06,
+      "loss": 0.5036,
+      "step": 3144
+    },
+    {
+      "epoch": 2.284520917985154,
+      "grad_norm": 0.3540500347732656,
+      "learning_rate": 6.0543424036268675e-06,
+      "loss": 0.4915,
+      "step": 3145
+    },
+    {
+      "epoch": 2.2852473157333213,
+      "grad_norm": 0.44413442338960246,
+      "learning_rate": 6.052008733248954e-06,
+      "loss": 0.5071,
+      "step": 3146
+    },
+    {
+      "epoch": 2.2859737134814884,
+      "grad_norm": 0.3900031444781187,
+      "learning_rate": 6.049674823061036e-06,
+      "loss": 0.5175,
+      "step": 3147
+    },
+    {
+      "epoch": 2.286700111229655,
+      "grad_norm": 0.32819673521829124,
+      "learning_rate": 6.047340673595139e-06,
+      "loss": 0.4848,
+      "step": 3148
+    },
+    {
+      "epoch": 2.287426508977822,
+      "grad_norm": 0.37858795931699807,
+      "learning_rate": 6.045006285383342e-06,
+      "loss": 0.4929,
+      "step": 3149
+    },
+    {
+      "epoch": 2.288152906725989,
+      "grad_norm": 0.3666616444696444,
+      "learning_rate": 6.042671658957783e-06,
+      "loss": 0.5038,
+      "step": 3150
+    },
+    {
+      "epoch": 2.288879304474156,
+      "grad_norm": 0.3633303205358319,
+      "learning_rate": 6.040336794850646e-06,
+      "loss": 0.4951,
+      "step": 3151
+    },
+    {
+      "epoch": 2.2896057022223233,
+      "grad_norm": 1.1115070192327328,
+      "learning_rate": 6.038001693594176e-06,
+      "loss": 0.4958,
+      "step": 3152
+    },
+    {
+      "epoch": 2.29033209997049,
+      "grad_norm": 0.326420594941503,
+      "learning_rate": 6.03566635572067e-06,
+      "loss": 0.4869,
+      "step": 3153
+    },
+    {
+      "epoch": 2.291058497718657,
+      "grad_norm": 0.3963449158303792,
+      "learning_rate": 6.033330781762476e-06,
+      "loss": 0.4788,
+      "step": 3154
+    },
+    {
+      "epoch": 2.291784895466824,
+      "grad_norm": 0.5105806805590508,
+      "learning_rate": 6.0309949722519986e-06,
+      "loss": 0.5032,
+      "step": 3155
+    },
+    {
+      "epoch": 2.292511293214991,
+      "grad_norm": 0.4155693537256211,
+      "learning_rate": 6.028658927721698e-06,
+      "loss": 0.4902,
+      "step": 3156
+    },
+    {
+      "epoch": 2.293237690963158,
+      "grad_norm": 0.4138282695922685,
+      "learning_rate": 6.0263226487040845e-06,
+      "loss": 0.4809,
+      "step": 3157
+    },
+    {
+      "epoch": 2.293964088711325,
+      "grad_norm": 0.37462980188091954,
+      "learning_rate": 6.023986135731724e-06,
+      "loss": 0.4942,
+      "step": 3158
+    },
+    {
+      "epoch": 2.294690486459492,
+      "grad_norm": 0.4289202950128778,
+      "learning_rate": 6.021649389337234e-06,
+      "loss": 0.498,
+      "step": 3159
+    },
+    {
+      "epoch": 2.295416884207659,
+      "grad_norm": 0.5635091191246396,
+      "learning_rate": 6.019312410053286e-06,
+      "loss": 0.489,
+      "step": 3160
+    },
+    {
+      "epoch": 2.296143281955826,
+      "grad_norm": 0.4607311853871575,
+      "learning_rate": 6.016975198412606e-06,
+      "loss": 0.4946,
+      "step": 3161
+    },
+    {
+      "epoch": 2.2968696797039927,
+      "grad_norm": 0.6935083323300786,
+      "learning_rate": 6.014637754947969e-06,
+      "loss": 0.4811,
+      "step": 3162
+    },
+    {
+      "epoch": 2.2975960774521598,
+      "grad_norm": 0.41357546263443523,
+      "learning_rate": 6.012300080192208e-06,
+      "loss": 0.5075,
+      "step": 3163
+    },
+    {
+      "epoch": 2.298322475200327,
+      "grad_norm": 0.3542243619660625,
+      "learning_rate": 6.0099621746782035e-06,
+      "loss": 0.4971,
+      "step": 3164
+    },
+    {
+      "epoch": 2.299048872948494,
+      "grad_norm": 0.44134295561503295,
+      "learning_rate": 6.007624038938895e-06,
+      "loss": 0.4888,
+      "step": 3165
+    },
+    {
+      "epoch": 2.299775270696661,
+      "grad_norm": 0.39417830167581025,
+      "learning_rate": 6.0052856735072685e-06,
+      "loss": 0.4984,
+      "step": 3166
+    },
+    {
+      "epoch": 2.300501668444828,
+      "grad_norm": 0.33506843695028593,
+      "learning_rate": 6.002947078916365e-06,
+      "loss": 0.4846,
+      "step": 3167
+    },
+    {
+      "epoch": 2.3012280661929947,
+      "grad_norm": 0.394726807159148,
+      "learning_rate": 6.000608255699277e-06,
+      "loss": 0.5015,
+      "step": 3168
+    },
+    {
+      "epoch": 2.3019544639411618,
+      "grad_norm": 0.35592888824186264,
+      "learning_rate": 5.99826920438915e-06,
+      "loss": 0.4928,
+      "step": 3169
+    },
+    {
+      "epoch": 2.302680861689329,
+      "grad_norm": 0.3741228112311038,
+      "learning_rate": 5.995929925519181e-06,
+      "loss": 0.4748,
+      "step": 3170
+    },
+    {
+      "epoch": 2.303407259437496,
+      "grad_norm": 0.4118521684904315,
+      "learning_rate": 5.993590419622619e-06,
+      "loss": 0.4916,
+      "step": 3171
+    },
+    {
+      "epoch": 2.3041336571856625,
+      "grad_norm": 0.4161129936032844,
+      "learning_rate": 5.991250687232764e-06,
+      "loss": 0.4841,
+      "step": 3172
+    },
+    {
+      "epoch": 2.3048600549338296,
+      "grad_norm": 0.42952160504666864,
+      "learning_rate": 5.98891072888297e-06,
+      "loss": 0.4884,
+      "step": 3173
+    },
+    {
+      "epoch": 2.3055864526819967,
+      "grad_norm": 0.39432337272336604,
+      "learning_rate": 5.986570545106638e-06,
+      "loss": 0.4956,
+      "step": 3174
+    },
+    {
+      "epoch": 2.3063128504301638,
+      "grad_norm": 0.3996450987725706,
+      "learning_rate": 5.984230136437226e-06,
+      "loss": 0.4968,
+      "step": 3175
+    },
+    {
+      "epoch": 2.307039248178331,
+      "grad_norm": 0.3391473825126834,
+      "learning_rate": 5.981889503408238e-06,
+      "loss": 0.4883,
+      "step": 3176
+    },
+    {
+      "epoch": 2.3077656459264975,
+      "grad_norm": 0.3267696126670061,
+      "learning_rate": 5.9795486465532325e-06,
+      "loss": 0.4978,
+      "step": 3177
+    },
+    {
+      "epoch": 2.3084920436746645,
+      "grad_norm": 0.3922857284244964,
+      "learning_rate": 5.977207566405818e-06,
+      "loss": 0.4833,
+      "step": 3178
+    },
+    {
+      "epoch": 2.3092184414228316,
+      "grad_norm": 0.42615109149541347,
+      "learning_rate": 5.974866263499655e-06,
+      "loss": 0.4861,
+      "step": 3179
+    },
+    {
+      "epoch": 2.3099448391709987,
+      "grad_norm": 0.39246348259809566,
+      "learning_rate": 5.972524738368452e-06,
+      "loss": 0.499,
+      "step": 3180
+    },
+    {
+      "epoch": 2.3106712369191653,
+      "grad_norm": 0.3811353423363143,
+      "learning_rate": 5.970182991545972e-06,
+      "loss": 0.4878,
+      "step": 3181
+    },
+    {
+      "epoch": 2.3113976346673324,
+      "grad_norm": 0.3538069167085555,
+      "learning_rate": 5.967841023566025e-06,
+      "loss": 0.4999,
+      "step": 3182
+    },
+    {
+      "epoch": 2.3121240324154995,
+      "grad_norm": 0.451875857319256,
+      "learning_rate": 5.965498834962472e-06,
+      "loss": 0.4811,
+      "step": 3183
+    },
+    {
+      "epoch": 2.3128504301636665,
+      "grad_norm": 0.5972179413766493,
+      "learning_rate": 5.963156426269228e-06,
+      "loss": 0.4903,
+      "step": 3184
+    },
+    {
+      "epoch": 2.3135768279118336,
+      "grad_norm": 0.41997194324350673,
+      "learning_rate": 5.960813798020252e-06,
+      "loss": 0.4753,
+      "step": 3185
+    },
+    {
+      "epoch": 2.3143032256600007,
+      "grad_norm": 0.6097372572334416,
+      "learning_rate": 5.9584709507495594e-06,
+      "loss": 0.5037,
+      "step": 3186
+    },
+    {
+      "epoch": 2.3150296234081673,
+      "grad_norm": 0.5162731967709743,
+      "learning_rate": 5.956127884991211e-06,
+      "loss": 0.4946,
+      "step": 3187
+    },
+    {
+      "epoch": 2.3157560211563344,
+      "grad_norm": 0.41135839595878126,
+      "learning_rate": 5.9537846012793184e-06,
+      "loss": 0.497,
+      "step": 3188
+    },
+    {
+      "epoch": 2.3164824189045015,
+      "grad_norm": 0.5225429394800519,
+      "learning_rate": 5.9514411001480435e-06,
+      "loss": 0.5023,
+      "step": 3189
+    },
+    {
+      "epoch": 2.3172088166526685,
+      "grad_norm": 0.36058624696826763,
+      "learning_rate": 5.949097382131599e-06,
+      "loss": 0.4968,
+      "step": 3190
+    },
+    {
+      "epoch": 2.317935214400835,
+      "grad_norm": 0.4896109590425641,
+      "learning_rate": 5.946753447764245e-06,
+      "loss": 0.4831,
+      "step": 3191
+    },
+    {
+      "epoch": 2.3186616121490022,
+      "grad_norm": 0.37739851530243934,
+      "learning_rate": 5.9444092975802925e-06,
+      "loss": 0.5023,
+      "step": 3192
+    },
+    {
+      "epoch": 2.3193880098971693,
+      "grad_norm": 0.3958899188666041,
+      "learning_rate": 5.942064932114099e-06,
+      "loss": 0.49,
+      "step": 3193
+    },
+    {
+      "epoch": 2.3201144076453364,
+      "grad_norm": 0.5308468797720521,
+      "learning_rate": 5.939720351900072e-06,
+      "loss": 0.497,
+      "step": 3194
+    },
+    {
+      "epoch": 2.3208408053935035,
+      "grad_norm": 0.3649823573866849,
+      "learning_rate": 5.9373755574726735e-06,
+      "loss": 0.4998,
+      "step": 3195
+    },
+    {
+      "epoch": 2.32156720314167,
+      "grad_norm": 0.3904577934507057,
+      "learning_rate": 5.935030549366405e-06,
+      "loss": 0.485,
+      "step": 3196
+    },
+    {
+      "epoch": 2.322293600889837,
+      "grad_norm": 0.34420467341162336,
+      "learning_rate": 5.932685328115823e-06,
+      "loss": 0.4921,
+      "step": 3197
+    },
+    {
+      "epoch": 2.3230199986380042,
+      "grad_norm": 0.3332275063766719,
+      "learning_rate": 5.930339894255533e-06,
+      "loss": 0.495,
+      "step": 3198
+    },
+    {
+      "epoch": 2.3237463963861713,
+      "grad_norm": 0.38028556409357467,
+      "learning_rate": 5.927994248320183e-06,
+      "loss": 0.4818,
+      "step": 3199
+    },
+    {
+      "epoch": 2.324472794134338,
+      "grad_norm": 0.42536308376316817,
+      "learning_rate": 5.925648390844476e-06,
+      "loss": 0.4848,
+      "step": 3200
+    },
+    {
+      "epoch": 2.1526435235773724,
+      "grad_norm": 0.36400339398627185,
+      "learning_rate": 6.446353231581457e-06,
+      "loss": 0.4911,
+      "step": 3201
+    },
+    {
+      "epoch": 2.1533159620072286,
+      "grad_norm": 0.37954296733354875,
+      "learning_rate": 6.4442456764159814e-06,
+      "loss": 0.4976,
+      "step": 3202
+    },
+    {
+      "epoch": 2.153988400437085,
+      "grad_norm": 0.45941645823821337,
+      "learning_rate": 6.44213784125384e-06,
+      "loss": 0.4858,
+      "step": 3203
+    },
+    {
+      "epoch": 2.154660838866941,
+      "grad_norm": 0.40495850868238187,
+      "learning_rate": 6.4400297265036795e-06,
+      "loss": 0.4881,
+      "step": 3204
+    },
+    {
+      "epoch": 2.1553332772967977,
+      "grad_norm": 0.3544217596686645,
+      "learning_rate": 6.437921332574203e-06,
+      "loss": 0.4921,
+      "step": 3205
+    },
+    {
+      "epoch": 2.156005715726654,
+      "grad_norm": 0.35855590079318694,
+      "learning_rate": 6.435812659874163e-06,
+      "loss": 0.4878,
+      "step": 3206
+    },
+    {
+      "epoch": 2.15667815415651,
+      "grad_norm": 0.4257310659468136,
+      "learning_rate": 6.4337037088123734e-06,
+      "loss": 0.4784,
+      "step": 3207
+    },
+    {
+      "epoch": 2.1573505925863663,
+      "grad_norm": 0.359797194091946,
+      "learning_rate": 6.431594479797693e-06,
+      "loss": 0.477,
+      "step": 3208
+    },
+    {
+      "epoch": 2.1580230310162225,
+      "grad_norm": 0.5311983101781117,
+      "learning_rate": 6.429484973239041e-06,
+      "loss": 0.4895,
+      "step": 3209
+    },
+    {
+      "epoch": 2.1586954694460787,
+      "grad_norm": 0.3738818622798392,
+      "learning_rate": 6.4273751895453894e-06,
+      "loss": 0.4743,
+      "step": 3210
+    },
+    {
+      "epoch": 2.159367907875935,
+      "grad_norm": 0.3760099340466295,
+      "learning_rate": 6.425265129125762e-06,
+      "loss": 0.4886,
+      "step": 3211
+    },
+    {
+      "epoch": 2.1600403463057916,
+      "grad_norm": 0.39293130543136123,
+      "learning_rate": 6.423154792389237e-06,
+      "loss": 0.4735,
+      "step": 3212
+    },
+    {
+      "epoch": 2.1607127847356478,
+      "grad_norm": 0.3819149221012835,
+      "learning_rate": 6.421044179744946e-06,
+      "loss": 0.4881,
+      "step": 3213
+    },
+    {
+      "epoch": 2.161385223165504,
+      "grad_norm": 0.3526992849837529,
+      "learning_rate": 6.418933291602079e-06,
+      "loss": 0.4905,
+      "step": 3214
+    },
+    {
+      "epoch": 2.16205766159536,
+      "grad_norm": 0.39242237563550325,
+      "learning_rate": 6.4168221283698696e-06,
+      "loss": 0.478,
+      "step": 3215
+    },
+    {
+      "epoch": 2.1627301000252164,
+      "grad_norm": 0.5052242268359664,
+      "learning_rate": 6.414710690457613e-06,
+      "loss": 0.4813,
+      "step": 3216
+    },
+    {
+      "epoch": 2.1634025384550726,
+      "grad_norm": 0.3974416216109638,
+      "learning_rate": 6.4125989782746535e-06,
+      "loss": 0.4613,
+      "step": 3217
+    },
+    {
+      "epoch": 2.164074976884929,
+      "grad_norm": 0.3953775362044758,
+      "learning_rate": 6.4104869922303905e-06,
+      "loss": 0.474,
+      "step": 3218
+    },
+    {
+      "epoch": 2.1647474153147854,
+      "grad_norm": 0.9970467066795999,
+      "learning_rate": 6.408374732734275e-06,
+      "loss": 0.4636,
+      "step": 3219
+    },
+    {
+      "epoch": 2.1654198537446416,
+      "grad_norm": 0.3449898830543444,
+      "learning_rate": 6.406262200195812e-06,
+      "loss": 0.4636,
+      "step": 3220
+    },
+    {
+      "epoch": 2.166092292174498,
+      "grad_norm": 0.3320632740312974,
+      "learning_rate": 6.4041493950245605e-06,
+      "loss": 0.4874,
+      "step": 3221
+    },
+    {
+      "epoch": 2.166764730604354,
+      "grad_norm": 0.3599970105192158,
+      "learning_rate": 6.40203631763013e-06,
+      "loss": 0.452,
+      "step": 3222
+    },
+    {
+      "epoch": 2.1674371690342102,
+      "grad_norm": 0.49644697700894996,
+      "learning_rate": 6.399922968422186e-06,
+      "loss": 0.4668,
+      "step": 3223
+    },
+    {
+      "epoch": 2.1681096074640664,
+      "grad_norm": 0.42659876148794784,
+      "learning_rate": 6.397809347810441e-06,
+      "loss": 0.4592,
+      "step": 3224
+    },
+    {
+      "epoch": 2.1687820458939226,
+      "grad_norm": 0.39713710109686995,
+      "learning_rate": 6.395695456204667e-06,
+      "loss": 0.4839,
+      "step": 3225
+    },
+    {
+      "epoch": 2.1694544843237793,
+      "grad_norm": 0.44749879329318243,
+      "learning_rate": 6.3935812940146835e-06,
+      "loss": 0.4735,
+      "step": 3226
+    },
+    {
+      "epoch": 2.1701269227536355,
+      "grad_norm": 0.4321525163181309,
+      "learning_rate": 6.391466861650365e-06,
+      "loss": 0.4775,
+      "step": 3227
+    },
+    {
+      "epoch": 2.1707993611834917,
+      "grad_norm": 0.3855482525351591,
+      "learning_rate": 6.389352159521636e-06,
+      "loss": 0.461,
+      "step": 3228
+    },
+    {
+      "epoch": 2.171471799613348,
+      "grad_norm": 0.44598796268875185,
+      "learning_rate": 6.3872371880384776e-06,
+      "loss": 0.4644,
+      "step": 3229
+    },
+    {
+      "epoch": 2.172144238043204,
+      "grad_norm": 0.3776752834861266,
+      "learning_rate": 6.3851219476109184e-06,
+      "loss": 0.487,
+      "step": 3230
+    },
+    {
+      "epoch": 2.1728166764730603,
+      "grad_norm": 0.3441671467053756,
+      "learning_rate": 6.383006438649042e-06,
+      "loss": 0.4682,
+      "step": 3231
+    },
+    {
+      "epoch": 2.1734891149029165,
+      "grad_norm": 0.38288668022048156,
+      "learning_rate": 6.380890661562984e-06,
+      "loss": 0.4876,
+      "step": 3232
+    },
+    {
+      "epoch": 2.174161553332773,
+      "grad_norm": 0.4303974490863329,
+      "learning_rate": 6.3787746167629286e-06,
+      "loss": 0.4628,
+      "step": 3233
+    },
+    {
+      "epoch": 2.1748339917626294,
+      "grad_norm": 0.44345318856219873,
+      "learning_rate": 6.376658304659118e-06,
+      "loss": 0.4638,
+      "step": 3234
+    },
+    {
+      "epoch": 2.1755064301924856,
+      "grad_norm": 0.4281777031529893,
+      "learning_rate": 6.37454172566184e-06,
+      "loss": 0.4844,
+      "step": 3235
+    },
+    {
+      "epoch": 2.1761788686223418,
+      "grad_norm": 0.3941670465285334,
+      "learning_rate": 6.372424880181437e-06,
+      "loss": 0.4864,
+      "step": 3236
+    },
+    {
+      "epoch": 2.176851307052198,
+      "grad_norm": 0.4568583459931907,
+      "learning_rate": 6.3703077686283045e-06,
+      "loss": 0.4735,
+      "step": 3237
+    },
+    {
+      "epoch": 2.177523745482054,
+      "grad_norm": 0.5070676742381048,
+      "learning_rate": 6.368190391412887e-06,
+      "loss": 0.4655,
+      "step": 3238
+    },
+    {
+      "epoch": 2.1781961839119104,
+      "grad_norm": 0.34743038619990396,
+      "learning_rate": 6.3660727489456815e-06,
+      "loss": 0.4652,
+      "step": 3239
+    },
+    {
+      "epoch": 2.178868622341767,
+      "grad_norm": 0.39638695582927597,
+      "learning_rate": 6.3639548416372385e-06,
+      "loss": 0.4702,
+      "step": 3240
+    },
+    {
+      "epoch": 2.1795410607716232,
+      "grad_norm": 0.40020178726324046,
+      "learning_rate": 6.3618366698981545e-06,
+      "loss": 0.4537,
+      "step": 3241
+    },
+    {
+      "epoch": 2.1802134992014794,
+      "grad_norm": 0.37629287034991654,
+      "learning_rate": 6.359718234139082e-06,
+      "loss": 0.4581,
+      "step": 3242
+    },
+    {
+      "epoch": 2.1808859376313356,
+      "grad_norm": 0.4216261561384884,
+      "learning_rate": 6.357599534770724e-06,
+      "loss": 0.4898,
+      "step": 3243
+    },
+    {
+      "epoch": 2.181558376061192,
+      "grad_norm": 0.33402757128286764,
+      "learning_rate": 6.355480572203834e-06,
+      "loss": 0.4574,
+      "step": 3244
+    },
+    {
+      "epoch": 2.182230814491048,
+      "grad_norm": 0.365971405847702,
+      "learning_rate": 6.353361346849214e-06,
+      "loss": 0.4571,
+      "step": 3245
+    },
+    {
+      "epoch": 2.1829032529209043,
+      "grad_norm": 0.3427010018820965,
+      "learning_rate": 6.35124185911772e-06,
+      "loss": 0.48,
+      "step": 3246
+    },
+    {
+      "epoch": 2.183575691350761,
+      "grad_norm": 0.36662548856695765,
+      "learning_rate": 6.349122109420262e-06,
+      "loss": 0.4668,
+      "step": 3247
+    },
+    {
+      "epoch": 2.184248129780617,
+      "grad_norm": 0.3478504185173162,
+      "learning_rate": 6.3470020981677915e-06,
+      "loss": 0.4723,
+      "step": 3248
+    },
+    {
+      "epoch": 2.1849205682104733,
+      "grad_norm": 0.31574374838479446,
+      "learning_rate": 6.3448818257713205e-06,
+      "loss": 0.4635,
+      "step": 3249
+    },
+    {
+      "epoch": 2.1855930066403295,
+      "grad_norm": 0.43950264307355086,
+      "learning_rate": 6.3427612926419056e-06,
+      "loss": 0.4791,
+      "step": 3250
+    },
+    {
+      "epoch": 2.1862654450701857,
+      "grad_norm": 0.3558974965207327,
+      "learning_rate": 6.340640499190656e-06,
+      "loss": 0.4614,
+      "step": 3251
+    },
+    {
+      "epoch": 2.186937883500042,
+      "grad_norm": 0.6481493882892034,
+      "learning_rate": 6.338519445828728e-06,
+      "loss": 0.4551,
+      "step": 3252
+    },
+    {
+      "epoch": 2.187610321929898,
+      "grad_norm": 0.39982021103181276,
+      "learning_rate": 6.336398132967337e-06,
+      "loss": 0.4529,
+      "step": 3253
+    },
+    {
+      "epoch": 2.1882827603597548,
+      "grad_norm": 0.3361374722833279,
+      "learning_rate": 6.334276561017738e-06,
+      "loss": 0.475,
+      "step": 3254
+    },
+    {
+      "epoch": 2.188955198789611,
+      "grad_norm": 0.3661649872244156,
+      "learning_rate": 6.3321547303912445e-06,
+      "loss": 0.4701,
+      "step": 3255
+    },
+    {
+      "epoch": 2.189627637219467,
+      "grad_norm": 0.38467033502442166,
+      "learning_rate": 6.330032641499216e-06,
+      "loss": 0.4654,
+      "step": 3256
+    },
+    {
+      "epoch": 2.1903000756493234,
+      "grad_norm": 0.33990424949533915,
+      "learning_rate": 6.3279102947530626e-06,
+      "loss": 0.4636,
+      "step": 3257
+    },
+    {
+      "epoch": 2.1909725140791796,
+      "grad_norm": 0.5088859601290591,
+      "learning_rate": 6.325787690564246e-06,
+      "loss": 0.4741,
+      "step": 3258
+    },
+    {
+      "epoch": 2.191644952509036,
+      "grad_norm": 0.4521884277447263,
+      "learning_rate": 6.3236648293442735e-06,
+      "loss": 0.4617,
+      "step": 3259
+    },
+    {
+      "epoch": 2.192317390938892,
+      "grad_norm": 0.3272684919643106,
+      "learning_rate": 6.321541711504708e-06,
+      "loss": 0.4675,
+      "step": 3260
+    },
+    {
+      "epoch": 2.1929898293687486,
+      "grad_norm": 0.3178252049326452,
+      "learning_rate": 6.319418337457158e-06,
+      "loss": 0.4601,
+      "step": 3261
+    },
+    {
+      "epoch": 2.193662267798605,
+      "grad_norm": 0.343212213649888,
+      "learning_rate": 6.317294707613287e-06,
+      "loss": 0.4823,
+      "step": 3262
+    },
+    {
+      "epoch": 2.194334706228461,
+      "grad_norm": 0.36111254446942886,
+      "learning_rate": 6.315170822384797e-06,
+      "loss": 0.4666,
+      "step": 3263
+    },
+    {
+      "epoch": 2.1950071446583173,
+      "grad_norm": 0.3247937151282206,
+      "learning_rate": 6.313046682183452e-06,
+      "loss": 0.4479,
+      "step": 3264
+    },
+    {
+      "epoch": 2.1956795830881735,
+      "grad_norm": 0.32402685278862575,
+      "learning_rate": 6.310922287421062e-06,
+      "loss": 0.4686,
+      "step": 3265
+    },
+    {
+      "epoch": 2.1963520215180297,
+      "grad_norm": 0.3085536264866017,
+      "learning_rate": 6.308797638509478e-06,
+      "loss": 0.4792,
+      "step": 3266
+    },
+    {
+      "epoch": 2.197024459947886,
+      "grad_norm": 0.3181533538129706,
+      "learning_rate": 6.3066727358606126e-06,
+      "loss": 0.4721,
+      "step": 3267
+    },
+    {
+      "epoch": 2.1976968983777425,
+      "grad_norm": 0.3312265223041896,
+      "learning_rate": 6.304547579886419e-06,
+      "loss": 0.4547,
+      "step": 3268
+    },
+    {
+      "epoch": 2.1983693368075987,
+      "grad_norm": 0.328345337965804,
+      "learning_rate": 6.302422170998903e-06,
+      "loss": 0.4789,
+      "step": 3269
+    },
+    {
+      "epoch": 2.199041775237455,
+      "grad_norm": 0.39821230084937137,
+      "learning_rate": 6.300296509610117e-06,
+      "loss": 0.4763,
+      "step": 3270
+    },
+    {
+      "epoch": 2.199714213667311,
+      "grad_norm": 0.45659507362333,
+      "learning_rate": 6.298170596132168e-06,
+      "loss": 0.4832,
+      "step": 3271
+    },
+    {
+      "epoch": 2.2003866520971673,
+      "grad_norm": 0.3582687048929397,
+      "learning_rate": 6.296044430977205e-06,
+      "loss": 0.4933,
+      "step": 3272
+    },
+    {
+      "epoch": 2.2010590905270235,
+      "grad_norm": 0.3347341994137685,
+      "learning_rate": 6.293918014557429e-06,
+      "loss": 0.4704,
+      "step": 3273
+    },
+    {
+      "epoch": 2.2017315289568797,
+      "grad_norm": 0.3215277495026712,
+      "learning_rate": 6.2917913472850915e-06,
+      "loss": 0.4629,
+      "step": 3274
+    },
+    {
+      "epoch": 2.2024039673867364,
+      "grad_norm": 0.3515898713741847,
+      "learning_rate": 6.289664429572486e-06,
+      "loss": 0.4675,
+      "step": 3275
+    },
+    {
+      "epoch": 2.2030764058165926,
+      "grad_norm": 0.40158213129557646,
+      "learning_rate": 6.287537261831965e-06,
+      "loss": 0.4712,
+      "step": 3276
+    },
+    {
+      "epoch": 2.203748844246449,
+      "grad_norm": 0.36658677429253866,
+      "learning_rate": 6.285409844475918e-06,
+      "loss": 0.4657,
+      "step": 3277
+    },
+    {
+      "epoch": 2.204421282676305,
+      "grad_norm": 0.40253283718196536,
+      "learning_rate": 6.283282177916794e-06,
+      "loss": 0.4712,
+      "step": 3278
+    },
+    {
+      "epoch": 2.205093721106161,
+      "grad_norm": 0.35863478731973064,
+      "learning_rate": 6.28115426256708e-06,
+      "loss": 0.4657,
+      "step": 3279
+    },
+    {
+      "epoch": 2.2057661595360174,
+      "grad_norm": 0.3305351152359045,
+      "learning_rate": 6.279026098839321e-06,
+      "loss": 0.4557,
+      "step": 3280
+    },
+    {
+      "epoch": 2.2064385979658736,
+      "grad_norm": 0.3318879526526031,
+      "learning_rate": 6.2768976871460985e-06,
+      "loss": 0.4601,
+      "step": 3281
+    },
+    {
+      "epoch": 2.20711103639573,
+      "grad_norm": 0.35760047494368197,
+      "learning_rate": 6.274769027900056e-06,
+      "loss": 0.478,
+      "step": 3282
+    },
+    {
+      "epoch": 2.2077834748255865,
+      "grad_norm": 0.34678732685220764,
+      "learning_rate": 6.272640121513872e-06,
+      "loss": 0.471,
+      "step": 3283
+    },
+    {
+      "epoch": 2.2084559132554427,
+      "grad_norm": 0.32565899960537,
+      "learning_rate": 6.270510968400283e-06,
+      "loss": 0.4814,
+      "step": 3284
+    },
+    {
+      "epoch": 2.209128351685299,
+      "grad_norm": 0.3208575439821756,
+      "learning_rate": 6.268381568972065e-06,
+      "loss": 0.4651,
+      "step": 3285
+    },
+    {
+      "epoch": 2.209800790115155,
+      "grad_norm": 0.4131626036588382,
+      "learning_rate": 6.266251923642049e-06,
+      "loss": 0.4898,
+      "step": 3286
+    },
+    {
+      "epoch": 2.2104732285450113,
+      "grad_norm": 0.3806906259045866,
+      "learning_rate": 6.264122032823111e-06,
+      "loss": 0.4727,
+      "step": 3287
+    },
+    {
+      "epoch": 2.2111456669748675,
+      "grad_norm": 0.7017980432113152,
+      "learning_rate": 6.26199189692817e-06,
+      "loss": 0.4762,
+      "step": 3288
+    },
+    {
+      "epoch": 2.2118181054047237,
+      "grad_norm": 0.3308519344551586,
+      "learning_rate": 6.259861516370201e-06,
+      "loss": 0.4555,
+      "step": 3289
+    },
+    {
+      "epoch": 2.2124905438345803,
+      "grad_norm": 0.3946936336332011,
+      "learning_rate": 6.2577308915622196e-06,
+      "loss": 0.4517,
+      "step": 3290
+    },
+    {
+      "epoch": 2.2131629822644365,
+      "grad_norm": 0.41976183541831025,
+      "learning_rate": 6.255600022917292e-06,
+      "loss": 0.4712,
+      "step": 3291
+    },
+    {
+      "epoch": 2.2138354206942927,
+      "grad_norm": 0.3441105184459733,
+      "learning_rate": 6.253468910848529e-06,
+      "loss": 0.4409,
+      "step": 3292
+    },
+    {
+      "epoch": 2.214507859124149,
+      "grad_norm": 0.440199416455677,
+      "learning_rate": 6.251337555769093e-06,
+      "loss": 0.4681,
+      "step": 3293
+    },
+    {
+      "epoch": 2.215180297554005,
+      "grad_norm": 0.45599533834389516,
+      "learning_rate": 6.2492059580921886e-06,
+      "loss": 0.4509,
+      "step": 3294
+    },
+    {
+      "epoch": 2.2158527359838613,
+      "grad_norm": 0.6261088005455595,
+      "learning_rate": 6.2470741182310725e-06,
+      "loss": 0.4504,
+      "step": 3295
+    },
+    {
+      "epoch": 2.2165251744137175,
+      "grad_norm": 0.32376267307506235,
+      "learning_rate": 6.244942036599042e-06,
+      "loss": 0.4565,
+      "step": 3296
+    },
+    {
+      "epoch": 2.217197612843574,
+      "grad_norm": 0.4449321605914357,
+      "learning_rate": 6.2428097136094476e-06,
+      "loss": 0.4837,
+      "step": 3297
+    },
+    {
+      "epoch": 2.2178700512734304,
+      "grad_norm": 0.4372822395750341,
+      "learning_rate": 6.240677149675684e-06,
+      "loss": 0.4713,
+      "step": 3298
+    },
+    {
+      "epoch": 2.2185424897032866,
+      "grad_norm": 0.7497825332115169,
+      "learning_rate": 6.2385443452111915e-06,
+      "loss": 0.4561,
+      "step": 3299
+    },
+    {
+      "epoch": 2.219214928133143,
+      "grad_norm": 0.3561248007825388,
+      "learning_rate": 6.236411300629458e-06,
+      "loss": 0.4714,
+      "step": 3300
+    },
+    {
+      "epoch": 2.219887366562999,
+      "grad_norm": 0.355500848926921,
+      "learning_rate": 6.234278016344018e-06,
+      "loss": 0.4625,
+      "step": 3301
+    },
+    {
+      "epoch": 2.220559804992855,
+      "grad_norm": 0.30501535700416,
+      "learning_rate": 6.232144492768451e-06,
+      "loss": 0.4719,
+      "step": 3302
+    },
+    {
+      "epoch": 2.2212322434227114,
+      "grad_norm": 0.42608821946659853,
+      "learning_rate": 6.230010730316388e-06,
+      "loss": 0.48,
+      "step": 3303
+    },
+    {
+      "epoch": 2.221904681852568,
+      "grad_norm": 0.45846152556150416,
+      "learning_rate": 6.227876729401501e-06,
+      "loss": 0.4938,
+      "step": 3304
+    },
+    {
+      "epoch": 2.2225771202824243,
+      "grad_norm": 0.4631201581405841,
+      "learning_rate": 6.225742490437507e-06,
+      "loss": 0.4493,
+      "step": 3305
+    },
+    {
+      "epoch": 2.2232495587122805,
+      "grad_norm": 0.3319109235846109,
+      "learning_rate": 6.223608013838177e-06,
+      "loss": 0.4509,
+      "step": 3306
+    },
+    {
+      "epoch": 2.2239219971421367,
+      "grad_norm": 0.4319123676573331,
+      "learning_rate": 6.221473300017319e-06,
+      "loss": 0.4613,
+      "step": 3307
+    },
+    {
+      "epoch": 2.224594435571993,
+      "grad_norm": 0.5502446668762675,
+      "learning_rate": 6.219338349388792e-06,
+      "loss": 0.4665,
+      "step": 3308
+    },
+    {
+      "epoch": 2.225266874001849,
+      "grad_norm": 0.3383954086793452,
+      "learning_rate": 6.217203162366502e-06,
+      "loss": 0.4773,
+      "step": 3309
+    },
+    {
+      "epoch": 2.2259393124317053,
+      "grad_norm": 0.3253169022175458,
+      "learning_rate": 6.215067739364397e-06,
+      "loss": 0.4457,
+      "step": 3310
+    },
+    {
+      "epoch": 2.226611750861562,
+      "grad_norm": 0.38283019608837743,
+      "learning_rate": 6.212932080796473e-06,
+      "loss": 0.4683,
+      "step": 3311
+    },
+    {
+      "epoch": 2.227284189291418,
+      "grad_norm": 0.3627285206977748,
+      "learning_rate": 6.2107961870767706e-06,
+      "loss": 0.4643,
+      "step": 3312
+    },
+    {
+      "epoch": 2.2279566277212743,
+      "grad_norm": 0.37400335315235345,
+      "learning_rate": 6.20866005861938e-06,
+      "loss": 0.4774,
+      "step": 3313
+    },
+    {
+      "epoch": 2.2286290661511305,
+      "grad_norm": 0.4257019090989514,
+      "learning_rate": 6.206523695838428e-06,
+      "loss": 0.4554,
+      "step": 3314
+    },
+    {
+      "epoch": 2.2293015045809867,
+      "grad_norm": 0.4793611486308495,
+      "learning_rate": 6.204387099148097e-06,
+      "loss": 0.4524,
+      "step": 3315
+    },
+    {
+      "epoch": 2.229973943010843,
+      "grad_norm": 0.38225534048231125,
+      "learning_rate": 6.2022502689626075e-06,
+      "loss": 0.4689,
+      "step": 3316
+    },
+    {
+      "epoch": 2.230646381440699,
+      "grad_norm": 0.32799250015071885,
+      "learning_rate": 6.200113205696228e-06,
+      "loss": 0.4682,
+      "step": 3317
+    },
+    {
+      "epoch": 2.231318819870556,
+      "grad_norm": 0.47480492674234276,
+      "learning_rate": 6.197975909763273e-06,
+      "loss": 0.4702,
+      "step": 3318
+    },
+    {
+      "epoch": 2.231991258300412,
+      "grad_norm": 0.3494408626924014,
+      "learning_rate": 6.195838381578101e-06,
+      "loss": 0.4562,
+      "step": 3319
+    },
+    {
+      "epoch": 2.232663696730268,
+      "grad_norm": 0.37782094497467594,
+      "learning_rate": 6.1937006215551176e-06,
+      "loss": 0.4707,
+      "step": 3320
+    },
+    {
+      "epoch": 2.2333361351601244,
+      "grad_norm": 0.37275458351555435,
+      "learning_rate": 6.191562630108767e-06,
+      "loss": 0.4572,
+      "step": 3321
+    },
+    {
+      "epoch": 2.2340085735899806,
+      "grad_norm": 0.3533869889329662,
+      "learning_rate": 6.189424407653548e-06,
+      "loss": 0.4544,
+      "step": 3322
+    },
+    {
+      "epoch": 2.234681012019837,
+      "grad_norm": 0.30999442867104227,
+      "learning_rate": 6.187285954603994e-06,
+      "loss": 0.472,
+      "step": 3323
+    },
+    {
+      "epoch": 2.235353450449693,
+      "grad_norm": 0.33558462774882347,
+      "learning_rate": 6.185147271374692e-06,
+      "loss": 0.4644,
+      "step": 3324
+    },
+    {
+      "epoch": 2.2360258888795497,
+      "grad_norm": 0.328979183614441,
+      "learning_rate": 6.183008358380266e-06,
+      "loss": 0.4646,
+      "step": 3325
+    },
+    {
+      "epoch": 2.236698327309406,
+      "grad_norm": 0.33931223367496843,
+      "learning_rate": 6.1808692160353904e-06,
+      "loss": 0.4729,
+      "step": 3326
+    },
+    {
+      "epoch": 2.237370765739262,
+      "grad_norm": 0.36880270279008426,
+      "learning_rate": 6.178729844754782e-06,
+      "loss": 0.4598,
+      "step": 3327
+    },
+    {
+      "epoch": 2.2380432041691183,
+      "grad_norm": 0.3645994812221974,
+      "learning_rate": 6.176590244953201e-06,
+      "loss": 0.4583,
+      "step": 3328
+    },
+    {
+      "epoch": 2.2387156425989745,
+      "grad_norm": 0.4262042189460906,
+      "learning_rate": 6.174450417045453e-06,
+      "loss": 0.4535,
+      "step": 3329
+    },
+    {
+      "epoch": 2.2393880810288307,
+      "grad_norm": 0.3778285311948193,
+      "learning_rate": 6.172310361446387e-06,
+      "loss": 0.451,
+      "step": 3330
+    },
+    {
+      "epoch": 2.240060519458687,
+      "grad_norm": 0.31217943304610896,
+      "learning_rate": 6.170170078570898e-06,
+      "loss": 0.4764,
+      "step": 3331
+    },
+    {
+      "epoch": 2.240732957888543,
+      "grad_norm": 0.43079020556899783,
+      "learning_rate": 6.168029568833923e-06,
+      "loss": 0.4607,
+      "step": 3332
+    },
+    {
+      "epoch": 2.2414053963183997,
+      "grad_norm": 0.3395066927764587,
+      "learning_rate": 6.165888832650444e-06,
+      "loss": 0.47,
+      "step": 3333
+    },
+    {
+      "epoch": 2.242077834748256,
+      "grad_norm": 0.36339912533861557,
+      "learning_rate": 6.163747870435486e-06,
+      "loss": 0.4749,
+      "step": 3334
+    },
+    {
+      "epoch": 2.242750273178112,
+      "grad_norm": 0.32776696036895525,
+      "learning_rate": 6.16160668260412e-06,
+      "loss": 0.4597,
+      "step": 3335
+    },
+    {
+      "epoch": 2.2434227116079684,
+      "grad_norm": 0.34177905036225076,
+      "learning_rate": 6.159465269571456e-06,
+      "loss": 0.4552,
+      "step": 3336
+    },
+    {
+      "epoch": 2.2440951500378246,
+      "grad_norm": 0.3385665299076455,
+      "learning_rate": 6.157323631752655e-06,
+      "loss": 0.4684,
+      "step": 3337
+    },
+    {
+      "epoch": 2.2447675884676808,
+      "grad_norm": 0.3009145657215783,
+      "learning_rate": 6.155181769562915e-06,
+      "loss": 0.4478,
+      "step": 3338
+    },
+    {
+      "epoch": 2.245440026897537,
+      "grad_norm": 0.3583058788615868,
+      "learning_rate": 6.15303968341748e-06,
+      "loss": 0.4798,
+      "step": 3339
+    },
+    {
+      "epoch": 2.2461124653273936,
+      "grad_norm": 0.38765323534294566,
+      "learning_rate": 6.150897373731637e-06,
+      "loss": 0.4591,
+      "step": 3340
+    },
+    {
+      "epoch": 2.24678490375725,
+      "grad_norm": 0.32351350562173625,
+      "learning_rate": 6.148754840920718e-06,
+      "loss": 0.4665,
+      "step": 3341
+    },
+    {
+      "epoch": 2.247457342187106,
+      "grad_norm": 0.3610813084026684,
+      "learning_rate": 6.146612085400096e-06,
+      "loss": 0.4749,
+      "step": 3342
+    },
+    {
+      "epoch": 2.2481297806169622,
+      "grad_norm": 0.346156630363326,
+      "learning_rate": 6.144469107585189e-06,
+      "loss": 0.4721,
+      "step": 3343
+    },
+    {
+      "epoch": 2.2488022190468184,
+      "grad_norm": 0.5827995420298924,
+      "learning_rate": 6.142325907891458e-06,
+      "loss": 0.4634,
+      "step": 3344
+    },
+    {
+      "epoch": 2.2494746574766746,
+      "grad_norm": 0.3453503009767493,
+      "learning_rate": 6.140182486734403e-06,
+      "loss": 0.4577,
+      "step": 3345
+    },
+    {
+      "epoch": 2.250147095906531,
+      "grad_norm": 0.39578599271575676,
+      "learning_rate": 6.138038844529574e-06,
+      "loss": 0.4555,
+      "step": 3346
+    },
+    {
+      "epoch": 2.2508195343363875,
+      "grad_norm": 0.3610888497577412,
+      "learning_rate": 6.1358949816925565e-06,
+      "loss": 0.4697,
+      "step": 3347
+    },
+    {
+      "epoch": 2.2514919727662437,
+      "grad_norm": 0.3602710216475961,
+      "learning_rate": 6.133750898638985e-06,
+      "loss": 0.4692,
+      "step": 3348
+    },
+    {
+      "epoch": 2.2521644111961,
+      "grad_norm": 0.36150798187522293,
+      "learning_rate": 6.131606595784531e-06,
+      "loss": 0.4515,
+      "step": 3349
+    },
+    {
+      "epoch": 2.252836849625956,
+      "grad_norm": 0.32433360019611446,
+      "learning_rate": 6.129462073544916e-06,
+      "loss": 0.4559,
+      "step": 3350
+    },
+    {
+      "epoch": 2.2535092880558123,
+      "grad_norm": 0.33643243164387887,
+      "learning_rate": 6.1273173323358955e-06,
+      "loss": 0.4524,
+      "step": 3351
+    },
+    {
+      "epoch": 2.2541817264856685,
+      "grad_norm": 0.3717075416643779,
+      "learning_rate": 6.125172372573275e-06,
+      "loss": 0.4674,
+      "step": 3352
+    },
+    {
+      "epoch": 2.2548541649155247,
+      "grad_norm": 0.4768402626852065,
+      "learning_rate": 6.123027194672897e-06,
+      "loss": 0.4533,
+      "step": 3353
+    },
+    {
+      "epoch": 2.2555266033453814,
+      "grad_norm": 0.4102465952984789,
+      "learning_rate": 6.120881799050648e-06,
+      "loss": 0.448,
+      "step": 3354
+    },
+    {
+      "epoch": 2.2561990417752376,
+      "grad_norm": 0.34922910900027687,
+      "learning_rate": 6.118736186122459e-06,
+      "loss": 0.4596,
+      "step": 3355
+    },
+    {
+      "epoch": 2.2568714802050938,
+      "grad_norm": 0.3605208529073994,
+      "learning_rate": 6.116590356304299e-06,
+      "loss": 0.4621,
+      "step": 3356
+    },
+    {
+      "epoch": 2.25754391863495,
+      "grad_norm": 0.4704991433860005,
+      "learning_rate": 6.114444310012181e-06,
+      "loss": 0.4639,
+      "step": 3357
+    },
+    {
+      "epoch": 2.258216357064806,
+      "grad_norm": 0.3261559815874674,
+      "learning_rate": 6.112298047662162e-06,
+      "loss": 0.4575,
+      "step": 3358
+    },
+    {
+      "epoch": 2.2588887954946624,
+      "grad_norm": 0.41003180273468104,
+      "learning_rate": 6.110151569670338e-06,
+      "loss": 0.4613,
+      "step": 3359
+    },
+    {
+      "epoch": 2.2595612339245186,
+      "grad_norm": 0.32332479510461204,
+      "learning_rate": 6.108004876452847e-06,
+      "loss": 0.4455,
+      "step": 3360
+    },
+    {
+      "epoch": 2.2602336723543752,
+      "grad_norm": 0.3221377433290264,
+      "learning_rate": 6.1058579684258715e-06,
+      "loss": 0.458,
+      "step": 3361
+    },
+    {
+      "epoch": 2.2609061107842314,
+      "grad_norm": 0.31199433964100104,
+      "learning_rate": 6.103710846005631e-06,
+      "loss": 0.4612,
+      "step": 3362
+    },
+    {
+      "epoch": 2.2615785492140876,
+      "grad_norm": 0.36232928587137264,
+      "learning_rate": 6.101563509608391e-06,
+      "loss": 0.4578,
+      "step": 3363
+    },
+    {
+      "epoch": 2.262250987643944,
+      "grad_norm": 0.4503992437491697,
+      "learning_rate": 6.099415959650459e-06,
+      "loss": 0.4629,
+      "step": 3364
+    },
+    {
+      "epoch": 2.2629234260738,
+      "grad_norm": 0.3428325411929094,
+      "learning_rate": 6.097268196548176e-06,
+      "loss": 0.461,
+      "step": 3365
+    },
+    {
+      "epoch": 2.2635958645036562,
+      "grad_norm": 0.3724685334484763,
+      "learning_rate": 6.095120220717934e-06,
+      "loss": 0.4531,
+      "step": 3366
+    },
+    {
+      "epoch": 2.2642683029335124,
+      "grad_norm": 0.3771343863955035,
+      "learning_rate": 6.092972032576161e-06,
+      "loss": 0.4652,
+      "step": 3367
+    },
+    {
+      "epoch": 2.264940741363369,
+      "grad_norm": 0.49327174181998573,
+      "learning_rate": 6.09082363253933e-06,
+      "loss": 0.4527,
+      "step": 3368
+    },
+    {
+      "epoch": 2.2656131797932253,
+      "grad_norm": 0.3807007587697114,
+      "learning_rate": 6.088675021023948e-06,
+      "loss": 0.4629,
+      "step": 3369
+    },
+    {
+      "epoch": 2.2662856182230815,
+      "grad_norm": 0.3282162633067282,
+      "learning_rate": 6.08652619844657e-06,
+      "loss": 0.4729,
+      "step": 3370
+    },
+    {
+      "epoch": 2.2669580566529377,
+      "grad_norm": 0.3615931019391284,
+      "learning_rate": 6.0843771652237884e-06,
+      "loss": 0.478,
+      "step": 3371
+    },
+    {
+      "epoch": 2.267630495082794,
+      "grad_norm": 0.7113957644212862,
+      "learning_rate": 6.082227921772239e-06,
+      "loss": 0.4586,
+      "step": 3372
+    },
+    {
+      "epoch": 2.26830293351265,
+      "grad_norm": 0.3374916536736294,
+      "learning_rate": 6.080078468508595e-06,
+      "loss": 0.4684,
+      "step": 3373
+    },
+    {
+      "epoch": 2.2689753719425063,
+      "grad_norm": 0.33119172696213184,
+      "learning_rate": 6.077928805849571e-06,
+      "loss": 0.4681,
+      "step": 3374
+    },
+    {
+      "epoch": 2.269647810372363,
+      "grad_norm": 0.3473668156048032,
+      "learning_rate": 6.075778934211926e-06,
+      "loss": 0.4488,
+      "step": 3375
+    },
+    {
+      "epoch": 2.270320248802219,
+      "grad_norm": 0.37623319413269796,
+      "learning_rate": 6.073628854012457e-06,
+      "loss": 0.4698,
+      "step": 3376
+    },
+    {
+      "epoch": 2.2709926872320754,
+      "grad_norm": 0.38767667636713227,
+      "learning_rate": 6.0714785656679984e-06,
+      "loss": 0.456,
+      "step": 3377
+    },
+    {
+      "epoch": 2.2716651256619316,
+      "grad_norm": 0.329376270538414,
+      "learning_rate": 6.069328069595431e-06,
+      "loss": 0.4644,
+      "step": 3378
+    },
+    {
+      "epoch": 2.2723375640917878,
+      "grad_norm": 0.3312790537268965,
+      "learning_rate": 6.06717736621167e-06,
+      "loss": 0.4416,
+      "step": 3379
+    },
+    {
+      "epoch": 2.273010002521644,
+      "grad_norm": 0.3267748945523019,
+      "learning_rate": 6.065026455933673e-06,
+      "loss": 0.4578,
+      "step": 3380
+    },
+    {
+      "epoch": 2.2736824409515,
+      "grad_norm": 0.4400018699156518,
+      "learning_rate": 6.06287533917844e-06,
+      "loss": 0.4839,
+      "step": 3381
+    },
+    {
+      "epoch": 2.274354879381357,
+      "grad_norm": 0.33881760895521307,
+      "learning_rate": 6.060724016363006e-06,
+      "loss": 0.4499,
+      "step": 3382
+    },
+    {
+      "epoch": 2.275027317811213,
+      "grad_norm": 0.3591199272872496,
+      "learning_rate": 6.058572487904454e-06,
+      "loss": 0.4621,
+      "step": 3383
+    },
+    {
+      "epoch": 2.2756997562410692,
+      "grad_norm": 0.43687174063020634,
+      "learning_rate": 6.056420754219898e-06,
+      "loss": 0.4558,
+      "step": 3384
+    },
+    {
+      "epoch": 2.2763721946709254,
+      "grad_norm": 0.6033053826657983,
+      "learning_rate": 6.054268815726498e-06,
+      "loss": 0.4604,
+      "step": 3385
+    },
+    {
+      "epoch": 2.2770446331007816,
+      "grad_norm": 0.3598084090904599,
+      "learning_rate": 6.052116672841451e-06,
+      "loss": 0.4625,
+      "step": 3386
+    },
+    {
+      "epoch": 2.277717071530638,
+      "grad_norm": 0.4066941227589285,
+      "learning_rate": 6.049964325981994e-06,
+      "loss": 0.4718,
+      "step": 3387
+    },
+    {
+      "epoch": 2.278389509960494,
+      "grad_norm": 0.37321132978250743,
+      "learning_rate": 6.047811775565403e-06,
+      "loss": 0.4598,
+      "step": 3388
+    },
+    {
+      "epoch": 2.2790619483903507,
+      "grad_norm": 0.36319732850319875,
+      "learning_rate": 6.045659022008993e-06,
+      "loss": 0.4669,
+      "step": 3389
+    },
+    {
+      "epoch": 2.279734386820207,
+      "grad_norm": 0.42909122433100383,
+      "learning_rate": 6.043506065730121e-06,
+      "loss": 0.4616,
+      "step": 3390
+    },
+    {
+      "epoch": 2.280406825250063,
+      "grad_norm": 0.3331074209108524,
+      "learning_rate": 6.041352907146182e-06,
+      "loss": 0.4412,
+      "step": 3391
+    },
+    {
+      "epoch": 2.2810792636799193,
+      "grad_norm": 0.3376033994060931,
+      "learning_rate": 6.0391995466746105e-06,
+      "loss": 0.4671,
+      "step": 3392
+    },
+    {
+      "epoch": 2.2817517021097755,
+      "grad_norm": 0.3632299799785127,
+      "learning_rate": 6.037045984732877e-06,
+      "loss": 0.4602,
+      "step": 3393
+    },
+    {
+      "epoch": 2.2824241405396317,
+      "grad_norm": 0.3367098095856448,
+      "learning_rate": 6.034892221738498e-06,
+      "loss": 0.4681,
+      "step": 3394
+    },
+    {
+      "epoch": 2.283096578969488,
+      "grad_norm": 0.3349169481214063,
+      "learning_rate": 6.032738258109019e-06,
+      "loss": 0.4693,
+      "step": 3395
+    },
+    {
+      "epoch": 2.2837690173993446,
+      "grad_norm": 0.43376510365880344,
+      "learning_rate": 6.030584094262034e-06,
+      "loss": 0.4651,
+      "step": 3396
+    },
+    {
+      "epoch": 2.2844414558292008,
+      "grad_norm": 0.3622779425031837,
+      "learning_rate": 6.028429730615172e-06,
+      "loss": 0.4505,
+      "step": 3397
+    },
+    {
+      "epoch": 2.285113894259057,
+      "grad_norm": 0.8038609794552343,
+      "learning_rate": 6.026275167586099e-06,
+      "loss": 0.455,
+      "step": 3398
+    },
+    {
+      "epoch": 2.285786332688913,
+      "grad_norm": 0.31065295877150795,
+      "learning_rate": 6.024120405592521e-06,
+      "loss": 0.4607,
+      "step": 3399
+    },
+    {
+      "epoch": 2.2864587711187694,
+      "grad_norm": 0.3118730137489182,
+      "learning_rate": 6.021965445052182e-06,
+      "loss": 0.4613,
+      "step": 3400
+    },
+    {
+      "epoch": 2.2871312095486256,
+      "grad_norm": 0.31927555976031363,
+      "learning_rate": 6.019810286382871e-06,
+      "loss": 0.4642,
+      "step": 3401
+    },
+    {
+      "epoch": 2.287803647978482,
+      "grad_norm": 0.3290263416853223,
+      "learning_rate": 6.017654930002404e-06,
+      "loss": 0.4467,
+      "step": 3402
+    },
+    {
+      "epoch": 2.2884760864083384,
+      "grad_norm": 0.41316779038138807,
+      "learning_rate": 6.015499376328642e-06,
+      "loss": 0.467,
+      "step": 3403
+    },
+    {
+      "epoch": 2.2891485248381946,
+      "grad_norm": 0.31784747734763796,
+      "learning_rate": 6.0133436257794845e-06,
+      "loss": 0.4578,
+      "step": 3404
+    },
+    {
+      "epoch": 2.289820963268051,
+      "grad_norm": 0.3090216314364643,
+      "learning_rate": 6.011187678772868e-06,
+      "loss": 0.463,
+      "step": 3405
+    },
+    {
+      "epoch": 2.290493401697907,
+      "grad_norm": 0.8887177266878555,
+      "learning_rate": 6.009031535726766e-06,
+      "loss": 0.4587,
+      "step": 3406
+    },
+    {
+      "epoch": 2.2911658401277633,
+      "grad_norm": 0.32367775021751516,
+      "learning_rate": 6.006875197059191e-06,
+      "loss": 0.4501,
+      "step": 3407
+    },
+    {
+      "epoch": 2.2918382785576195,
+      "grad_norm": 0.40011093762620686,
+      "learning_rate": 6.004718663188196e-06,
+      "loss": 0.4481,
+      "step": 3408
+    },
+    {
+      "epoch": 2.2925107169874757,
+      "grad_norm": 0.33581182782298286,
+      "learning_rate": 6.0025619345318665e-06,
+      "loss": 0.4477,
+      "step": 3409
+    },
+    {
+      "epoch": 2.2931831554173323,
+      "grad_norm": 0.35930830236800326,
+      "learning_rate": 6.000405011508331e-06,
+      "loss": 0.4642,
+      "step": 3410
+    },
+    {
+      "epoch": 2.2938555938471885,
+      "grad_norm": 0.38802853702385304,
+      "learning_rate": 5.99824789453575e-06,
+      "loss": 0.4627,
+      "step": 3411
+    },
+    {
+      "epoch": 2.2945280322770447,
+      "grad_norm": 0.3984723628092719,
+      "learning_rate": 5.996090584032329e-06,
+      "loss": 0.4616,
+      "step": 3412
+    },
+    {
+      "epoch": 2.295200470706901,
+      "grad_norm": 0.3867437885588343,
+      "learning_rate": 5.993933080416303e-06,
+      "loss": 0.461,
+      "step": 3413
+    },
+    {
+      "epoch": 2.295872909136757,
+      "grad_norm": 0.39906088363451553,
+      "learning_rate": 5.9917753841059516e-06,
+      "loss": 0.4793,
+      "step": 3414
+    },
+    {
+      "epoch": 2.2965453475666133,
+      "grad_norm": 0.3425550182660963,
+      "learning_rate": 5.989617495519586e-06,
+      "loss": 0.457,
+      "step": 3415
+    },
+    {
+      "epoch": 2.2972177859964695,
+      "grad_norm": 0.33384643461177643,
+      "learning_rate": 5.987459415075559e-06,
+      "loss": 0.4755,
+      "step": 3416
+    },
+    {
+      "epoch": 2.297890224426326,
+      "grad_norm": 0.38703558717300707,
+      "learning_rate": 5.985301143192258e-06,
+      "loss": 0.4556,
+      "step": 3417
+    },
+    {
+      "epoch": 2.2985626628561824,
+      "grad_norm": 0.4168459939748986,
+      "learning_rate": 5.983142680288109e-06,
+      "loss": 0.4587,
+      "step": 3418
+    },
+    {
+      "epoch": 2.2992351012860386,
+      "grad_norm": 0.3324296279399844,
+      "learning_rate": 5.980984026781572e-06,
+      "loss": 0.4543,
+      "step": 3419
+    },
+    {
+      "epoch": 2.299907539715895,
+      "grad_norm": 0.3950144902415733,
+      "learning_rate": 5.978825183091148e-06,
+      "loss": 0.4772,
+      "step": 3420
+    },
+    {
+      "epoch": 2.300579978145751,
+      "grad_norm": 0.4183703742331098,
+      "learning_rate": 5.976666149635375e-06,
+      "loss": 0.455,
+      "step": 3421
+    },
+    {
+      "epoch": 2.301252416575607,
+      "grad_norm": 0.3388616076348071,
+      "learning_rate": 5.974506926832822e-06,
+      "loss": 0.4559,
+      "step": 3422
+    },
+    {
+      "epoch": 2.3019248550054634,
+      "grad_norm": 0.343276092749269,
+      "learning_rate": 5.972347515102102e-06,
+      "loss": 0.4581,
+      "step": 3423
+    },
+    {
+      "epoch": 2.30259729343532,
+      "grad_norm": 0.3380498032264536,
+      "learning_rate": 5.97018791486186e-06,
+      "loss": 0.4748,
+      "step": 3424
+    },
+    {
+      "epoch": 2.3032697318651763,
+      "grad_norm": 0.35002944027653216,
+      "learning_rate": 5.96802812653078e-06,
+      "loss": 0.4524,
+      "step": 3425
+    },
+    {
+      "epoch": 2.3039421702950325,
+      "grad_norm": 0.34512713571252573,
+      "learning_rate": 5.9658681505275785e-06,
+      "loss": 0.457,
+      "step": 3426
+    },
+    {
+      "epoch": 2.3046146087248887,
+      "grad_norm": 0.380751421190636,
+      "learning_rate": 5.963707987271014e-06,
+      "loss": 0.4629,
+      "step": 3427
+    },
+    {
+      "epoch": 2.305287047154745,
+      "grad_norm": 0.44507606560359714,
+      "learning_rate": 5.961547637179875e-06,
+      "loss": 0.4526,
+      "step": 3428
+    },
+    {
+      "epoch": 2.305959485584601,
+      "grad_norm": 0.4277887184641815,
+      "learning_rate": 5.959387100672994e-06,
+      "loss": 0.4551,
+      "step": 3429
+    },
+    {
+      "epoch": 2.3066319240144573,
+      "grad_norm": 0.2969333291788095,
+      "learning_rate": 5.957226378169233e-06,
+      "loss": 0.4567,
+      "step": 3430
+    },
+    {
+      "epoch": 2.307304362444314,
+      "grad_norm": 0.3232796193905546,
+      "learning_rate": 5.955065470087491e-06,
+      "loss": 0.4579,
+      "step": 3431
+    },
+    {
+      "epoch": 2.30797680087417,
+      "grad_norm": 0.4424788971209814,
+      "learning_rate": 5.9529043768467085e-06,
+      "loss": 0.4559,
+      "step": 3432
+    },
+    {
+      "epoch": 2.3086492393040263,
+      "grad_norm": 0.4020983687858228,
+      "learning_rate": 5.950743098865854e-06,
+      "loss": 0.4594,
+      "step": 3433
+    },
+    {
+      "epoch": 2.3093216777338825,
+      "grad_norm": 0.33850722955430224,
+      "learning_rate": 5.948581636563939e-06,
+      "loss": 0.4516,
+      "step": 3434
+    },
+    {
+      "epoch": 2.3099941161637387,
+      "grad_norm": 0.34400717979285295,
+      "learning_rate": 5.946419990360004e-06,
+      "loss": 0.4634,
+      "step": 3435
+    },
+    {
+      "epoch": 2.310666554593595,
+      "grad_norm": 0.3447090933374568,
+      "learning_rate": 5.94425816067313e-06,
+      "loss": 0.4685,
+      "step": 3436
+    },
+    {
+      "epoch": 2.311338993023451,
+      "grad_norm": 0.3135329848238476,
+      "learning_rate": 5.9420961479224315e-06,
+      "loss": 0.4433,
+      "step": 3437
+    },
+    {
+      "epoch": 2.312011431453308,
+      "grad_norm": 0.38012311880815325,
+      "learning_rate": 5.9399339525270595e-06,
+      "loss": 0.4714,
+      "step": 3438
+    },
+    {
+      "epoch": 2.312683869883164,
+      "grad_norm": 0.3890841552249759,
+      "learning_rate": 5.9377715749062e-06,
+      "loss": 0.4611,
+      "step": 3439
+    },
+    {
+      "epoch": 2.31335630831302,
+      "grad_norm": 0.3188847168667241,
+      "learning_rate": 5.935609015479075e-06,
+      "loss": 0.4671,
+      "step": 3440
+    },
+    {
+      "epoch": 2.3140287467428764,
+      "grad_norm": 0.49850677405706556,
+      "learning_rate": 5.9334462746649375e-06,
+      "loss": 0.4444,
+      "step": 3441
+    },
+    {
+      "epoch": 2.3147011851727326,
+      "grad_norm": 0.35339646675873393,
+      "learning_rate": 5.9312833528830835e-06,
+      "loss": 0.4606,
+      "step": 3442
+    },
+    {
+      "epoch": 2.315373623602589,
+      "grad_norm": 0.3545069415592496,
+      "learning_rate": 5.92912025055284e-06,
+      "loss": 0.4423,
+      "step": 3443
+    },
+    {
+      "epoch": 2.316046062032445,
+      "grad_norm": 0.3661937143726214,
+      "learning_rate": 5.926956968093565e-06,
+      "loss": 0.46,
+      "step": 3444
+    },
+    {
+      "epoch": 2.3167185004623017,
+      "grad_norm": 0.3305650067264564,
+      "learning_rate": 5.924793505924658e-06,
+      "loss": 0.46,
+      "step": 3445
+    },
+    {
+      "epoch": 2.317390938892158,
+      "grad_norm": 0.36544706722770126,
+      "learning_rate": 5.9226298644655485e-06,
+      "loss": 0.4495,
+      "step": 3446
+    },
+    {
+      "epoch": 2.318063377322014,
+      "grad_norm": 0.3967132330665242,
+      "learning_rate": 5.920466044135704e-06,
+      "loss": 0.4619,
+      "step": 3447
+    },
+    {
+      "epoch": 2.3187358157518703,
+      "grad_norm": 0.47266167596193176,
+      "learning_rate": 5.918302045354626e-06,
+      "loss": 0.4483,
+      "step": 3448
+    },
+    {
+      "epoch": 2.3194082541817265,
+      "grad_norm": 0.3989441553994751,
+      "learning_rate": 5.91613786854185e-06,
+      "loss": 0.4614,
+      "step": 3449
+    },
+    {
+      "epoch": 2.3200806926115827,
+      "grad_norm": 0.519376934857127,
+      "learning_rate": 5.9139735141169455e-06,
+      "loss": 0.4338,
+      "step": 3450
+    },
+    {
+      "epoch": 2.320753131041439,
+      "grad_norm": 0.416453104062567,
+      "learning_rate": 5.911808982499519e-06,
+      "loss": 0.4563,
+      "step": 3451
+    },
+    {
+      "epoch": 2.3214255694712955,
+      "grad_norm": 0.36849376223410835,
+      "learning_rate": 5.909644274109204e-06,
+      "loss": 0.4497,
+      "step": 3452
+    },
+    {
+      "epoch": 2.3220980079011517,
+      "grad_norm": 0.33518302342042033,
+      "learning_rate": 5.907479389365678e-06,
+      "loss": 0.4594,
+      "step": 3453
+    },
+    {
+      "epoch": 2.322770446331008,
+      "grad_norm": 0.38748584452757723,
+      "learning_rate": 5.905314328688649e-06,
+      "loss": 0.4528,
+      "step": 3454
+    },
+    {
+      "epoch": 2.323442884760864,
+      "grad_norm": 0.44889342511131064,
+      "learning_rate": 5.903149092497856e-06,
+      "loss": 0.4646,
+      "step": 3455
+    },
+    {
+      "epoch": 2.3241153231907203,
+      "grad_norm": 0.39168359932896285,
+      "learning_rate": 5.900983681213075e-06,
+      "loss": 0.4451,
+      "step": 3456
+    },
+    {
+      "epoch": 2.3247877616205765,
+      "grad_norm": 0.33063191767706884,
+      "learning_rate": 5.898818095254116e-06,
+      "loss": 0.4552,
+      "step": 3457
+    },
+    {
+      "epoch": 2.3254602000504327,
+      "grad_norm": 0.4200611590444878,
+      "learning_rate": 5.896652335040825e-06,
+      "loss": 0.4658,
+      "step": 3458
+    },
+    {
+      "epoch": 2.3261326384802894,
+      "grad_norm": 0.3506150399617044,
+      "learning_rate": 5.894486400993072e-06,
+      "loss": 0.4623,
+      "step": 3459
+    },
+    {
+      "epoch": 2.3268050769101456,
+      "grad_norm": 0.4364839190318039,
+      "learning_rate": 5.8923202935307755e-06,
+      "loss": 0.4401,
+      "step": 3460
+    },
+    {
+      "epoch": 2.327477515340002,
+      "grad_norm": 0.3506604314068755,
+      "learning_rate": 5.890154013073875e-06,
+      "loss": 0.4564,
+      "step": 3461
+    },
+    {
+      "epoch": 2.328149953769858,
+      "grad_norm": 0.3195057224021962,
+      "learning_rate": 5.887987560042348e-06,
+      "loss": 0.4447,
+      "step": 3462
+    },
+    {
+      "epoch": 2.328822392199714,
+      "grad_norm": 0.4081013352389473,
+      "learning_rate": 5.88582093485621e-06,
+      "loss": 0.4575,
+      "step": 3463
+    },
+    {
+      "epoch": 2.3294948306295704,
+      "grad_norm": 0.3509868634510859,
+      "learning_rate": 5.883654137935501e-06,
+      "loss": 0.4437,
+      "step": 3464
+    },
+    {
+      "epoch": 2.3301672690594266,
+      "grad_norm": 0.3227618446614624,
+      "learning_rate": 5.881487169700305e-06,
+      "loss": 0.4718,
+      "step": 3465
+    },
+    {
+      "epoch": 2.330839707489283,
+      "grad_norm": 0.7520884280657425,
+      "learning_rate": 5.879320030570728e-06,
+      "loss": 0.4716,
+      "step": 3466
+    },
+    {
+      "epoch": 2.3315121459191395,
+      "grad_norm": 0.3747673865418648,
+      "learning_rate": 5.877152720966917e-06,
+      "loss": 0.4656,
+      "step": 3467
+    },
+    {
+      "epoch": 2.3321845843489957,
+      "grad_norm": 0.36206969058631283,
+      "learning_rate": 5.8749852413090465e-06,
+      "loss": 0.4562,
+      "step": 3468
+    },
+    {
+      "epoch": 2.332857022778852,
+      "grad_norm": 0.4123862983586637,
+      "learning_rate": 5.872817592017331e-06,
+      "loss": 0.4469,
+      "step": 3469
+    },
+    {
+      "epoch": 2.333529461208708,
+      "grad_norm": 0.33733669496642177,
+      "learning_rate": 5.870649773512011e-06,
+      "loss": 0.4565,
+      "step": 3470
+    },
+    {
+      "epoch": 2.3342018996385643,
+      "grad_norm": 0.42636616448320297,
+      "learning_rate": 5.868481786213364e-06,
+      "loss": 0.4468,
+      "step": 3471
+    },
+    {
+      "epoch": 2.3348743380684205,
+      "grad_norm": 0.384115666486269,
+      "learning_rate": 5.866313630541698e-06,
+      "loss": 0.4437,
+      "step": 3472
+    },
+    {
+      "epoch": 2.3355467764982767,
+      "grad_norm": 0.41716083426210254,
+      "learning_rate": 5.864145306917355e-06,
+      "loss": 0.4298,
+      "step": 3473
+    },
+    {
+      "epoch": 2.3362192149281333,
+      "grad_norm": 0.3513270477111295,
+      "learning_rate": 5.861976815760708e-06,
+      "loss": 0.4543,
+      "step": 3474
+    },
+    {
+      "epoch": 2.3368916533579895,
+      "grad_norm": 0.3658591809502371,
+      "learning_rate": 5.859808157492164e-06,
+      "loss": 0.4586,
+      "step": 3475
+    },
+    {
+      "epoch": 2.3375640917878457,
+      "grad_norm": 0.3776366683658066,
+      "learning_rate": 5.857639332532163e-06,
+      "loss": 0.4575,
+      "step": 3476
+    },
+    {
+      "epoch": 2.338236530217702,
+      "grad_norm": 0.36436135486161964,
+      "learning_rate": 5.855470341301176e-06,
+      "loss": 0.4386,
+      "step": 3477
+    },
+    {
+      "epoch": 2.338908968647558,
+      "grad_norm": 0.3240074843175133,
+      "learning_rate": 5.853301184219706e-06,
+      "loss": 0.4498,
+      "step": 3478
+    },
+    {
+      "epoch": 2.3395814070774144,
+      "grad_norm": 0.4090670479530033,
+      "learning_rate": 5.851131861708288e-06,
+      "loss": 0.473,
+      "step": 3479
+    },
+    {
+      "epoch": 2.3402538455072706,
+      "grad_norm": 0.388472246416491,
+      "learning_rate": 5.84896237418749e-06,
+      "loss": 0.4737,
+      "step": 3480
+    },
+    {
+      "epoch": 2.3409262839371268,
+      "grad_norm": 0.3360497237320795,
+      "learning_rate": 5.846792722077911e-06,
+      "loss": 0.4552,
+      "step": 3481
+    },
+    {
+      "epoch": 2.3415987223669834,
+      "grad_norm": 0.341364864082169,
+      "learning_rate": 5.8446229058001856e-06,
+      "loss": 0.4576,
+      "step": 3482
+    },
+    {
+      "epoch": 2.3422711607968396,
+      "grad_norm": 0.32697453039951657,
+      "learning_rate": 5.842452925774974e-06,
+      "loss": 0.4392,
+      "step": 3483
+    },
+    {
+      "epoch": 2.342943599226696,
+      "grad_norm": 0.3980001762154398,
+      "learning_rate": 5.840282782422972e-06,
+      "loss": 0.4551,
+      "step": 3484
+    },
+    {
+      "epoch": 2.343616037656552,
+      "grad_norm": 0.31000431629498004,
+      "learning_rate": 5.838112476164908e-06,
+      "loss": 0.4521,
+      "step": 3485
+    },
+    {
+      "epoch": 2.3442884760864082,
+      "grad_norm": 0.31545657709084923,
+      "learning_rate": 5.835942007421538e-06,
+      "loss": 0.4641,
+      "step": 3486
+    },
+    {
+      "epoch": 2.3449609145162644,
+      "grad_norm": 0.3679552244830917,
+      "learning_rate": 5.833771376613654e-06,
+      "loss": 0.4593,
+      "step": 3487
+    },
+    {
+      "epoch": 2.3456333529461206,
+      "grad_norm": 0.31157320097469177,
+      "learning_rate": 5.831600584162076e-06,
+      "loss": 0.4561,
+      "step": 3488
+    },
+    {
+      "epoch": 2.3463057913759773,
+      "grad_norm": 0.468879910189735,
+      "learning_rate": 5.829429630487659e-06,
+      "loss": 0.4376,
+      "step": 3489
+    },
+    {
+      "epoch": 2.3469782298058335,
+      "grad_norm": 0.3771545389043398,
+      "learning_rate": 5.827258516011284e-06,
+      "loss": 0.4609,
+      "step": 3490
+    },
+    {
+      "epoch": 2.3476506682356897,
+      "grad_norm": 0.39098431423034563,
+      "learning_rate": 5.825087241153867e-06,
+      "loss": 0.4572,
+      "step": 3491
+    },
+    {
+      "epoch": 2.348323106665546,
+      "grad_norm": 0.32933039095880423,
+      "learning_rate": 5.822915806336355e-06,
+      "loss": 0.4492,
+      "step": 3492
+    },
+    {
+      "epoch": 2.348995545095402,
+      "grad_norm": 0.33183804809411027,
+      "learning_rate": 5.820744211979725e-06,
+      "loss": 0.4514,
+      "step": 3493
+    },
+    {
+      "epoch": 2.3496679835252583,
+      "grad_norm": 0.33945276045807116,
+      "learning_rate": 5.818572458504983e-06,
+      "loss": 0.442,
+      "step": 3494
+    },
+    {
+      "epoch": 2.3503404219551145,
+      "grad_norm": 0.3534888950933072,
+      "learning_rate": 5.816400546333171e-06,
+      "loss": 0.4535,
+      "step": 3495
+    },
+    {
+      "epoch": 2.351012860384971,
+      "grad_norm": 0.3096377373774302,
+      "learning_rate": 5.814228475885357e-06,
+      "loss": 0.4546,
+      "step": 3496
+    },
+    {
+      "epoch": 2.3516852988148274,
+      "grad_norm": 0.32904977790765544,
+      "learning_rate": 5.812056247582643e-06,
+      "loss": 0.458,
+      "step": 3497
+    },
+    {
+      "epoch": 2.3523577372446836,
+      "grad_norm": 0.38029547051000695,
+      "learning_rate": 5.809883861846159e-06,
+      "loss": 0.4387,
+      "step": 3498
+    },
+    {
+      "epoch": 2.3530301756745398,
+      "grad_norm": 0.33342204975446,
+      "learning_rate": 5.8077113190970644e-06,
+      "loss": 0.4563,
+      "step": 3499
+    },
+    {
+      "epoch": 2.353702614104396,
+      "grad_norm": 0.34230708427497747,
+      "learning_rate": 5.8055386197565564e-06,
+      "loss": 0.4583,
+      "step": 3500
+    },
+    {
+      "epoch": 2.354375052534252,
+      "grad_norm": 0.2951783161108756,
+      "learning_rate": 5.803365764245852e-06,
+      "loss": 0.4583,
+      "step": 3501
+    },
+    {
+      "epoch": 2.3550474909641084,
+      "grad_norm": 0.3079953132771035,
+      "learning_rate": 5.801192752986208e-06,
+      "loss": 0.4503,
+      "step": 3502
+    },
+    {
+      "epoch": 2.355719929393965,
+      "grad_norm": 0.3570046765039688,
+      "learning_rate": 5.799019586398904e-06,
+      "loss": 0.4578,
+      "step": 3503
+    },
+    {
+      "epoch": 2.356392367823821,
+      "grad_norm": 0.3177125091323911,
+      "learning_rate": 5.796846264905255e-06,
+      "loss": 0.4648,
+      "step": 3504
+    },
+    {
+      "epoch": 2.3570648062536774,
+      "grad_norm": 0.3363125140497039,
+      "learning_rate": 5.794672788926602e-06,
+      "loss": 0.4623,
+      "step": 3505
+    },
+    {
+      "epoch": 2.3577372446835336,
+      "grad_norm": 0.3167853911375675,
+      "learning_rate": 5.792499158884322e-06,
+      "loss": 0.4667,
+      "step": 3506
+    },
+    {
+      "epoch": 2.35840968311339,
+      "grad_norm": 0.6921775759842348,
+      "learning_rate": 5.790325375199812e-06,
+      "loss": 0.4543,
+      "step": 3507
+    },
+    {
+      "epoch": 2.359082121543246,
+      "grad_norm": 0.30236339661586337,
+      "learning_rate": 5.788151438294509e-06,
+      "loss": 0.4433,
+      "step": 3508
+    },
+    {
+      "epoch": 2.3597545599731022,
+      "grad_norm": 0.40941230596207956,
+      "learning_rate": 5.785977348589875e-06,
+      "loss": 0.4474,
+      "step": 3509
+    },
+    {
+      "epoch": 2.360426998402959,
+      "grad_norm": 0.5196512787688972,
+      "learning_rate": 5.783803106507399e-06,
+      "loss": 0.4373,
+      "step": 3510
+    },
+    {
+      "epoch": 2.361099436832815,
+      "grad_norm": 0.2991559649141668,
+      "learning_rate": 5.781628712468605e-06,
+      "loss": 0.4538,
+      "step": 3511
+    },
+    {
+      "epoch": 2.3617718752626713,
+      "grad_norm": 0.3825417809422422,
+      "learning_rate": 5.779454166895042e-06,
+      "loss": 0.447,
+      "step": 3512
+    },
+    {
+      "epoch": 2.3624443136925275,
+      "grad_norm": 0.4769351696146306,
+      "learning_rate": 5.777279470208294e-06,
+      "loss": 0.4713,
+      "step": 3513
+    },
+    {
+      "epoch": 2.3631167521223837,
+      "grad_norm": 0.30023152393422464,
+      "learning_rate": 5.775104622829965e-06,
+      "loss": 0.446,
+      "step": 3514
+    },
+    {
+      "epoch": 2.36378919055224,
+      "grad_norm": 0.33153694221794716,
+      "learning_rate": 5.7729296251817e-06,
+      "loss": 0.4776,
+      "step": 3515
+    },
+    {
+      "epoch": 2.364461628982096,
+      "grad_norm": 0.4047692204296205,
+      "learning_rate": 5.770754477685162e-06,
+      "loss": 0.4573,
+      "step": 3516
+    },
+    {
+      "epoch": 2.3651340674119528,
+      "grad_norm": 0.40042975658998164,
+      "learning_rate": 5.7685791807620505e-06,
+      "loss": 0.4606,
+      "step": 3517
+    },
+    {
+      "epoch": 2.365806505841809,
+      "grad_norm": 0.327532117308796,
+      "learning_rate": 5.766403734834089e-06,
+      "loss": 0.445,
+      "step": 3518
+    },
+    {
+      "epoch": 2.366478944271665,
+      "grad_norm": 0.2704078193848059,
+      "learning_rate": 5.764228140323036e-06,
+      "loss": 0.4386,
+      "step": 3519
+    },
+    {
+      "epoch": 2.3671513827015214,
+      "grad_norm": 0.3404895554400373,
+      "learning_rate": 5.7620523976506715e-06,
+      "loss": 0.4511,
+      "step": 3520
+    },
+    {
+      "epoch": 2.3678238211313776,
+      "grad_norm": 0.4333037773933191,
+      "learning_rate": 5.759876507238811e-06,
+      "loss": 0.4511,
+      "step": 3521
+    },
+    {
+      "epoch": 2.3684962595612338,
+      "grad_norm": 1.309105942254137,
+      "learning_rate": 5.7577004695092945e-06,
+      "loss": 0.4603,
+      "step": 3522
+    },
+    {
+      "epoch": 2.36916869799109,
+      "grad_norm": 0.32328978955846294,
+      "learning_rate": 5.755524284883989e-06,
+      "loss": 0.4632,
+      "step": 3523
+    },
+    {
+      "epoch": 2.3698411364209466,
+      "grad_norm": 0.37849843215721984,
+      "learning_rate": 5.753347953784797e-06,
+      "loss": 0.4409,
+      "step": 3524
+    },
+    {
+      "epoch": 2.370513574850803,
+      "grad_norm": 0.3476793256077204,
+      "learning_rate": 5.751171476633641e-06,
+      "loss": 0.4486,
+      "step": 3525
+    },
+    {
+      "epoch": 2.371186013280659,
+      "grad_norm": 0.3750262990064418,
+      "learning_rate": 5.748994853852479e-06,
+      "loss": 0.4647,
+      "step": 3526
+    },
+    {
+      "epoch": 2.3718584517105152,
+      "grad_norm": 0.34635074214628625,
+      "learning_rate": 5.746818085863292e-06,
+      "loss": 0.4416,
+      "step": 3527
+    },
+    {
+      "epoch": 2.3725308901403714,
+      "grad_norm": 0.38434110558426926,
+      "learning_rate": 5.74464117308809e-06,
+      "loss": 0.4564,
+      "step": 3528
+    },
+    {
+      "epoch": 2.3732033285702276,
+      "grad_norm": 0.34561027779242004,
+      "learning_rate": 5.742464115948915e-06,
+      "loss": 0.4424,
+      "step": 3529
+    },
+    {
+      "epoch": 2.373875767000084,
+      "grad_norm": 0.33266273991326856,
+      "learning_rate": 5.740286914867832e-06,
+      "loss": 0.4387,
+      "step": 3530
+    },
+    {
+      "epoch": 2.3745482054299405,
+      "grad_norm": 0.42008457441870206,
+      "learning_rate": 5.738109570266939e-06,
+      "loss": 0.4511,
+      "step": 3531
+    },
+    {
+      "epoch": 2.3752206438597967,
+      "grad_norm": 0.3368066077329263,
+      "learning_rate": 5.735932082568356e-06,
+      "loss": 0.4416,
+      "step": 3532
+    },
+    {
+      "epoch": 2.375893082289653,
+      "grad_norm": 0.30953007129649845,
+      "learning_rate": 5.733754452194235e-06,
+      "loss": 0.4515,
+      "step": 3533
+    },
+    {
+      "epoch": 2.376565520719509,
+      "grad_norm": 0.3584541636729375,
+      "learning_rate": 5.731576679566754e-06,
+      "loss": 0.4529,
+      "step": 3534
+    },
+    {
+      "epoch": 2.3772379591493653,
+      "grad_norm": 0.3819730806896943,
+      "learning_rate": 5.729398765108118e-06,
+      "loss": 0.4545,
+      "step": 3535
+    },
+    {
+      "epoch": 2.3779103975792215,
+      "grad_norm": 0.3393313272383021,
+      "learning_rate": 5.727220709240563e-06,
+      "loss": 0.4607,
+      "step": 3536
+    },
+    {
+      "epoch": 2.3785828360090777,
+      "grad_norm": 0.3407277498003918,
+      "learning_rate": 5.725042512386347e-06,
+      "loss": 0.4556,
+      "step": 3537
+    },
+    {
+      "epoch": 2.3792552744389344,
+      "grad_norm": 0.44335690258193794,
+      "learning_rate": 5.72286417496776e-06,
+      "loss": 0.4492,
+      "step": 3538
+    },
+    {
+      "epoch": 2.3799277128687906,
+      "grad_norm": 0.360155250476559,
+      "learning_rate": 5.720685697407118e-06,
+      "loss": 0.4658,
+      "step": 3539
+    },
+    {
+      "epoch": 2.3806001512986468,
+      "grad_norm": 0.4081717614851412,
+      "learning_rate": 5.718507080126761e-06,
+      "loss": 0.453,
+      "step": 3540
+    },
+    {
+      "epoch": 2.381272589728503,
+      "grad_norm": 0.32668892308941555,
+      "learning_rate": 5.71632832354906e-06,
+      "loss": 0.4399,
+      "step": 3541
+    },
+    {
+      "epoch": 2.381945028158359,
+      "grad_norm": 0.4516247995817048,
+      "learning_rate": 5.714149428096413e-06,
+      "loss": 0.4543,
+      "step": 3542
+    },
+    {
+      "epoch": 2.3826174665882154,
+      "grad_norm": 0.35628091371878984,
+      "learning_rate": 5.711970394191241e-06,
+      "loss": 0.4426,
+      "step": 3543
+    },
+    {
+      "epoch": 2.3832899050180716,
+      "grad_norm": 0.3202215438454765,
+      "learning_rate": 5.709791222255996e-06,
+      "loss": 0.4556,
+      "step": 3544
+    },
+    {
+      "epoch": 2.3839623434479282,
+      "grad_norm": 0.3217462913974043,
+      "learning_rate": 5.7076119127131545e-06,
+      "loss": 0.4578,
+      "step": 3545
+    },
+    {
+      "epoch": 2.3846347818777844,
+      "grad_norm": 0.3800753121649306,
+      "learning_rate": 5.705432465985224e-06,
+      "loss": 0.4543,
+      "step": 3546
+    },
+    {
+      "epoch": 2.3853072203076406,
+      "grad_norm": 0.3745986832533898,
+      "learning_rate": 5.703252882494728e-06,
+      "loss": 0.4564,
+      "step": 3547
+    },
+    {
+      "epoch": 2.385979658737497,
+      "grad_norm": 0.39388160698750074,
+      "learning_rate": 5.70107316266423e-06,
+      "loss": 0.4553,
+      "step": 3548
+    },
+    {
+      "epoch": 2.386652097167353,
+      "grad_norm": 0.3693568996979446,
+      "learning_rate": 5.698893306916309e-06,
+      "loss": 0.4545,
+      "step": 3549
+    },
+    {
+      "epoch": 2.3873245355972093,
+      "grad_norm": 0.3358204292471544,
+      "learning_rate": 5.6967133156735785e-06,
+      "loss": 0.4654,
+      "step": 3550
+    },
+    {
+      "epoch": 2.3879969740270655,
+      "grad_norm": 0.3879923145296747,
+      "learning_rate": 5.6945331893586705e-06,
+      "loss": 0.4497,
+      "step": 3551
+    },
+    {
+      "epoch": 2.388669412456922,
+      "grad_norm": 0.4702166619666767,
+      "learning_rate": 5.69235292839425e-06,
+      "loss": 0.4523,
+      "step": 3552
+    },
+    {
+      "epoch": 2.3893418508867783,
+      "grad_norm": 0.35151582679612603,
+      "learning_rate": 5.690172533203005e-06,
+      "loss": 0.4548,
+      "step": 3553
+    },
+    {
+      "epoch": 2.3900142893166345,
+      "grad_norm": 0.33623006086909196,
+      "learning_rate": 5.68799200420765e-06,
+      "loss": 0.4395,
+      "step": 3554
+    },
+    {
+      "epoch": 2.3906867277464907,
+      "grad_norm": 0.3180533268004615,
+      "learning_rate": 5.685811341830924e-06,
+      "loss": 0.4521,
+      "step": 3555
+    },
+    {
+      "epoch": 2.391359166176347,
+      "grad_norm": 0.3120770132369895,
+      "learning_rate": 5.6836305464955945e-06,
+      "loss": 0.4563,
+      "step": 3556
+    },
+    {
+      "epoch": 2.392031604606203,
+      "grad_norm": 0.3313207875580303,
+      "learning_rate": 5.681449618624454e-06,
+      "loss": 0.4555,
+      "step": 3557
+    },
+    {
+      "epoch": 2.3927040430360593,
+      "grad_norm": 0.34906684447334524,
+      "learning_rate": 5.679268558640318e-06,
+      "loss": 0.4399,
+      "step": 3558
+    },
+    {
+      "epoch": 2.393376481465916,
+      "grad_norm": 0.3359379283214766,
+      "learning_rate": 5.677087366966031e-06,
+      "loss": 0.4499,
+      "step": 3559
+    },
+    {
+      "epoch": 2.394048919895772,
+      "grad_norm": 0.34212649051370675,
+      "learning_rate": 5.67490604402446e-06,
+      "loss": 0.4536,
+      "step": 3560
+    },
+    {
+      "epoch": 2.3947213583256284,
+      "grad_norm": 0.3968833539019725,
+      "learning_rate": 5.672724590238502e-06,
+      "loss": 0.4363,
+      "step": 3561
+    },
+    {
+      "epoch": 2.3953937967554846,
+      "grad_norm": 0.33982599660113577,
+      "learning_rate": 5.670543006031075e-06,
+      "loss": 0.4588,
+      "step": 3562
+    },
+    {
+      "epoch": 2.396066235185341,
+      "grad_norm": 0.3126079592635059,
+      "learning_rate": 5.668361291825124e-06,
+      "loss": 0.4459,
+      "step": 3563
+    },
+    {
+      "epoch": 2.396738673615197,
+      "grad_norm": 0.3645127404080134,
+      "learning_rate": 5.666179448043621e-06,
+      "loss": 0.4499,
+      "step": 3564
+    },
+    {
+      "epoch": 2.397411112045053,
+      "grad_norm": 0.3110482341885458,
+      "learning_rate": 5.663997475109558e-06,
+      "loss": 0.4678,
+      "step": 3565
+    },
+    {
+      "epoch": 2.39808355047491,
+      "grad_norm": 0.32199104525287026,
+      "learning_rate": 5.661815373445959e-06,
+      "loss": 0.4614,
+      "step": 3566
+    },
+    {
+      "epoch": 2.398755988904766,
+      "grad_norm": 0.8639742510000906,
+      "learning_rate": 5.659633143475864e-06,
+      "loss": 0.4617,
+      "step": 3567
+    },
+    {
+      "epoch": 2.3994284273346222,
+      "grad_norm": 0.40995524737792327,
+      "learning_rate": 5.6574507856223474e-06,
+      "loss": 0.4533,
+      "step": 3568
+    },
+    {
+      "epoch": 2.4001008657644785,
+      "grad_norm": 0.36762151659653114,
+      "learning_rate": 5.655268300308502e-06,
+      "loss": 0.4489,
+      "step": 3569
+    },
+    {
+      "epoch": 2.4007733041943347,
+      "grad_norm": 0.33318188755161604,
+      "learning_rate": 5.653085687957449e-06,
+      "loss": 0.4482,
+      "step": 3570
+    },
+    {
+      "epoch": 2.401445742624191,
+      "grad_norm": 0.28687255302183695,
+      "learning_rate": 5.650902948992332e-06,
+      "loss": 0.4448,
+      "step": 3571
+    },
+    {
+      "epoch": 2.402118181054047,
+      "grad_norm": 0.3976726902254766,
+      "learning_rate": 5.648720083836319e-06,
+      "loss": 0.4601,
+      "step": 3572
+    },
+    {
+      "epoch": 2.4027906194839037,
+      "grad_norm": 0.3193218278392546,
+      "learning_rate": 5.646537092912603e-06,
+      "loss": 0.4585,
+      "step": 3573
+    },
+    {
+      "epoch": 2.40346305791376,
+      "grad_norm": 0.36282104810944193,
+      "learning_rate": 5.644353976644402e-06,
+      "loss": 0.4362,
+      "step": 3574
+    },
+    {
+      "epoch": 2.404135496343616,
+      "grad_norm": 0.39340217861985266,
+      "learning_rate": 5.642170735454959e-06,
+      "loss": 0.4593,
+      "step": 3575
+    },
+    {
+      "epoch": 2.4048079347734723,
+      "grad_norm": 0.3427481067975898,
+      "learning_rate": 5.6399873697675385e-06,
+      "loss": 0.4485,
+      "step": 3576
+    },
+    {
+      "epoch": 2.4054803732033285,
+      "grad_norm": 0.405186505645815,
+      "learning_rate": 5.637803880005431e-06,
+      "loss": 0.4441,
+      "step": 3577
+    },
+    {
+      "epoch": 2.4061528116331847,
+      "grad_norm": 0.3167912422521965,
+      "learning_rate": 5.635620266591953e-06,
+      "loss": 0.4543,
+      "step": 3578
+    },
+    {
+      "epoch": 2.406825250063041,
+      "grad_norm": 0.35430272710261224,
+      "learning_rate": 5.63343652995044e-06,
+      "loss": 0.4806,
+      "step": 3579
+    },
+    {
+      "epoch": 2.4074976884928976,
+      "grad_norm": 0.39827430646088413,
+      "learning_rate": 5.631252670504255e-06,
+      "loss": 0.4627,
+      "step": 3580
+    },
+    {
+      "epoch": 2.408170126922754,
+      "grad_norm": 0.3536213396008051,
+      "learning_rate": 5.629068688676785e-06,
+      "loss": 0.4375,
+      "step": 3581
+    },
+    {
+      "epoch": 2.40884256535261,
+      "grad_norm": 0.5413875717151687,
+      "learning_rate": 5.626884584891438e-06,
+      "loss": 0.4451,
+      "step": 3582
+    },
+    {
+      "epoch": 2.409515003782466,
+      "grad_norm": 0.41928945425001374,
+      "learning_rate": 5.624700359571649e-06,
+      "loss": 0.4569,
+      "step": 3583
+    },
+    {
+      "epoch": 2.4101874422123224,
+      "grad_norm": 0.3354908193449054,
+      "learning_rate": 5.622516013140874e-06,
+      "loss": 0.4486,
+      "step": 3584
+    },
+    {
+      "epoch": 2.4108598806421786,
+      "grad_norm": 0.3648121009986923,
+      "learning_rate": 5.620331546022591e-06,
+      "loss": 0.4332,
+      "step": 3585
+    },
+    {
+      "epoch": 2.411532319072035,
+      "grad_norm": 0.35255253553402593,
+      "learning_rate": 5.618146958640309e-06,
+      "loss": 0.4381,
+      "step": 3586
+    },
+    {
+      "epoch": 2.4122047575018915,
+      "grad_norm": 0.333954692659307,
+      "learning_rate": 5.615962251417551e-06,
+      "loss": 0.4526,
+      "step": 3587
+    },
+    {
+      "epoch": 2.4128771959317477,
+      "grad_norm": 0.37607079229227774,
+      "learning_rate": 5.613777424777871e-06,
+      "loss": 0.4626,
+      "step": 3588
+    },
+    {
+      "epoch": 2.413549634361604,
+      "grad_norm": 0.38095934546167926,
+      "learning_rate": 5.6115924791448385e-06,
+      "loss": 0.4528,
+      "step": 3589
+    },
+    {
+      "epoch": 2.41422207279146,
+      "grad_norm": 0.3993076314402837,
+      "learning_rate": 5.609407414942053e-06,
+      "loss": 0.4643,
+      "step": 3590
+    },
+    {
+      "epoch": 2.4148945112213163,
+      "grad_norm": 0.39223329828299025,
+      "learning_rate": 5.607222232593131e-06,
+      "loss": 0.4641,
+      "step": 3591
+    },
+    {
+      "epoch": 2.4155669496511725,
+      "grad_norm": 0.3591949686390438,
+      "learning_rate": 5.605036932521718e-06,
+      "loss": 0.4494,
+      "step": 3592
+    },
+    {
+      "epoch": 2.4162393880810287,
+      "grad_norm": 0.3133711489094439,
+      "learning_rate": 5.602851515151477e-06,
+      "loss": 0.4429,
+      "step": 3593
+    },
+    {
+      "epoch": 2.4169118265108853,
+      "grad_norm": 0.33802182001683545,
+      "learning_rate": 5.600665980906099e-06,
+      "loss": 0.4481,
+      "step": 3594
+    },
+    {
+      "epoch": 2.4175842649407415,
+      "grad_norm": 0.35171603486105313,
+      "learning_rate": 5.59848033020929e-06,
+      "loss": 0.4791,
+      "step": 3595
+    },
+    {
+      "epoch": 2.4182567033705977,
+      "grad_norm": 0.3591522629231598,
+      "learning_rate": 5.596294563484787e-06,
+      "loss": 0.4724,
+      "step": 3596
+    },
+    {
+      "epoch": 2.418929141800454,
+      "grad_norm": 0.447799486858325,
+      "learning_rate": 5.594108681156347e-06,
+      "loss": 0.4495,
+      "step": 3597
+    },
+    {
+      "epoch": 2.41960158023031,
+      "grad_norm": 0.3290264637523835,
+      "learning_rate": 5.591922683647744e-06,
+      "loss": 0.4495,
+      "step": 3598
+    },
+    {
+      "epoch": 2.4202740186601663,
+      "grad_norm": 0.3379787091837133,
+      "learning_rate": 5.589736571382781e-06,
+      "loss": 0.4592,
+      "step": 3599
+    },
+    {
+      "epoch": 2.4209464570900225,
+      "grad_norm": 0.32521436538097526,
+      "learning_rate": 5.587550344785278e-06,
+      "loss": 0.4619,
+      "step": 3600
+    },
+    {
+      "epoch": 2.421618895519879,
+      "grad_norm": 0.3701978164103939,
+      "learning_rate": 5.585364004279083e-06,
+      "loss": 0.4436,
+      "step": 3601
+    },
+    {
+      "epoch": 2.4222913339497354,
+      "grad_norm": 0.3847282583347662,
+      "learning_rate": 5.583177550288063e-06,
+      "loss": 0.457,
+      "step": 3602
+    },
+    {
+      "epoch": 2.4229637723795916,
+      "grad_norm": 0.31382599299470876,
+      "learning_rate": 5.580990983236105e-06,
+      "loss": 0.4668,
+      "step": 3603
+    },
+    {
+      "epoch": 2.423636210809448,
+      "grad_norm": 0.3389055913276392,
+      "learning_rate": 5.5788043035471205e-06,
+      "loss": 0.4558,
+      "step": 3604
+    },
+    {
+      "epoch": 2.424308649239304,
+      "grad_norm": 1.0617375171131134,
+      "learning_rate": 5.576617511645044e-06,
+      "loss": 0.4664,
+      "step": 3605
+    },
+    {
+      "epoch": 2.42498108766916,
+      "grad_norm": 0.3532471892301109,
+      "learning_rate": 5.574430607953827e-06,
+      "loss": 0.4551,
+      "step": 3606
+    },
+    {
+      "epoch": 2.4256535260990164,
+      "grad_norm": 0.3346822353524649,
+      "learning_rate": 5.572243592897449e-06,
+      "loss": 0.4539,
+      "step": 3607
+    },
+    {
+      "epoch": 2.426325964528873,
+      "grad_norm": 0.36237163805522626,
+      "learning_rate": 5.570056466899904e-06,
+      "loss": 0.4612,
+      "step": 3608
+    },
+    {
+      "epoch": 2.4269984029587293,
+      "grad_norm": 0.3128470088619246,
+      "learning_rate": 5.567869230385214e-06,
+      "loss": 0.4469,
+      "step": 3609
+    },
+    {
+      "epoch": 2.4276708413885855,
+      "grad_norm": 0.34861094074937193,
+      "learning_rate": 5.565681883777418e-06,
+      "loss": 0.4551,
+      "step": 3610
+    },
+    {
+      "epoch": 2.4283432798184417,
+      "grad_norm": 0.4297843609897647,
+      "learning_rate": 5.56349442750058e-06,
+      "loss": 0.4449,
+      "step": 3611
+    },
+    {
+      "epoch": 2.429015718248298,
+      "grad_norm": 0.3477250631144471,
+      "learning_rate": 5.561306861978783e-06,
+      "loss": 0.4398,
+      "step": 3612
+    },
+    {
+      "epoch": 2.429688156678154,
+      "grad_norm": 0.3509821263654013,
+      "learning_rate": 5.559119187636129e-06,
+      "loss": 0.4441,
+      "step": 3613
+    },
+    {
+      "epoch": 2.4303605951080103,
+      "grad_norm": 0.35737277401120054,
+      "learning_rate": 5.556931404896747e-06,
+      "loss": 0.4494,
+      "step": 3614
+    },
+    {
+      "epoch": 2.431033033537867,
+      "grad_norm": 0.3625938554539421,
+      "learning_rate": 5.554743514184781e-06,
+      "loss": 0.4558,
+      "step": 3615
+    },
+    {
+      "epoch": 2.431705471967723,
+      "grad_norm": 0.36630770612755065,
+      "learning_rate": 5.552555515924398e-06,
+      "loss": 0.4443,
+      "step": 3616
+    },
+    {
+      "epoch": 2.4323779103975793,
+      "grad_norm": 0.33239617734472793,
+      "learning_rate": 5.550367410539788e-06,
+      "loss": 0.4478,
+      "step": 3617
+    },
+    {
+      "epoch": 2.4330503488274355,
+      "grad_norm": 0.7626069649045053,
+      "learning_rate": 5.5481791984551614e-06,
+      "loss": 0.4503,
+      "step": 3618
+    },
+    {
+      "epoch": 2.4337227872572917,
+      "grad_norm": 0.3158254591145617,
+      "learning_rate": 5.545990880094747e-06,
+      "loss": 0.4589,
+      "step": 3619
+    },
+    {
+      "epoch": 2.434395225687148,
+      "grad_norm": 0.3522408038054907,
+      "learning_rate": 5.543802455882793e-06,
+      "loss": 0.4539,
+      "step": 3620
+    },
+    {
+      "epoch": 2.435067664117004,
+      "grad_norm": 0.35852926157212767,
+      "learning_rate": 5.541613926243573e-06,
+      "loss": 0.4552,
+      "step": 3621
+    },
+    {
+      "epoch": 2.435740102546861,
+      "grad_norm": 0.3039993342004631,
+      "learning_rate": 5.539425291601378e-06,
+      "loss": 0.4565,
+      "step": 3622
+    },
+    {
+      "epoch": 2.436412540976717,
+      "grad_norm": 0.34123599831642026,
+      "learning_rate": 5.537236552380518e-06,
+      "loss": 0.4699,
+      "step": 3623
+    },
+    {
+      "epoch": 2.437084979406573,
+      "grad_norm": 0.34602669178981266,
+      "learning_rate": 5.535047709005327e-06,
+      "loss": 0.4612,
+      "step": 3624
+    },
+    {
+      "epoch": 2.4377574178364294,
+      "grad_norm": 0.38467285537375934,
+      "learning_rate": 5.532858761900156e-06,
+      "loss": 0.4437,
+      "step": 3625
+    },
+    {
+      "epoch": 2.4384298562662856,
+      "grad_norm": 0.3498196149534507,
+      "learning_rate": 5.5306697114893785e-06,
+      "loss": 0.4524,
+      "step": 3626
+    },
+    {
+      "epoch": 2.439102294696142,
+      "grad_norm": 0.3548262072706039,
+      "learning_rate": 5.5284805581973854e-06,
+      "loss": 0.4547,
+      "step": 3627
+    },
+    {
+      "epoch": 2.439774733125998,
+      "grad_norm": 0.3799419694111064,
+      "learning_rate": 5.526291302448589e-06,
+      "loss": 0.4424,
+      "step": 3628
+    },
+    {
+      "epoch": 2.4404471715558547,
+      "grad_norm": 0.9342712842749369,
+      "learning_rate": 5.5241019446674235e-06,
+      "loss": 0.4591,
+      "step": 3629
+    },
+    {
+      "epoch": 2.441119609985711,
+      "grad_norm": 0.3836099289086837,
+      "learning_rate": 5.52191248527834e-06,
+      "loss": 0.46,
+      "step": 3630
+    },
+    {
+      "epoch": 2.441792048415567,
+      "grad_norm": 0.3213034553405655,
+      "learning_rate": 5.519722924705808e-06,
+      "loss": 0.4524,
+      "step": 3631
+    },
+    {
+      "epoch": 2.4424644868454233,
+      "grad_norm": 0.3457962788868057,
+      "learning_rate": 5.517533263374322e-06,
+      "loss": 0.4429,
+      "step": 3632
+    },
+    {
+      "epoch": 2.4431369252752795,
+      "grad_norm": 0.30079475711553777,
+      "learning_rate": 5.51534350170839e-06,
+      "loss": 0.4608,
+      "step": 3633
+    },
+    {
+      "epoch": 2.4438093637051357,
+      "grad_norm": 0.33057774220755387,
+      "learning_rate": 5.513153640132543e-06,
+      "loss": 0.4751,
+      "step": 3634
+    },
+    {
+      "epoch": 2.444481802134992,
+      "grad_norm": 0.6793162709862243,
+      "learning_rate": 5.5109636790713315e-06,
+      "loss": 0.4369,
+      "step": 3635
+    },
+    {
+      "epoch": 2.445154240564848,
+      "grad_norm": 0.30836110440964787,
+      "learning_rate": 5.508773618949326e-06,
+      "loss": 0.4536,
+      "step": 3636
+    },
+    {
+      "epoch": 2.4458266789947047,
+      "grad_norm": 0.3242160993777931,
+      "learning_rate": 5.50658346019111e-06,
+      "loss": 0.4322,
+      "step": 3637
+    },
+    {
+      "epoch": 2.446499117424561,
+      "grad_norm": 0.4923792011120225,
+      "learning_rate": 5.504393203221294e-06,
+      "loss": 0.451,
+      "step": 3638
+    },
+    {
+      "epoch": 2.447171555854417,
+      "grad_norm": 0.3273613311767987,
+      "learning_rate": 5.502202848464504e-06,
+      "loss": 0.4545,
+      "step": 3639
+    },
+    {
+      "epoch": 2.4478439942842734,
+      "grad_norm": 0.37767840103116296,
+      "learning_rate": 5.500012396345385e-06,
+      "loss": 0.4412,
+      "step": 3640
+    },
+    {
+      "epoch": 2.4485164327141296,
+      "grad_norm": 0.39168523194271326,
+      "learning_rate": 5.497821847288599e-06,
+      "loss": 0.4675,
+      "step": 3641
+    },
+    {
+      "epoch": 2.4491888711439858,
+      "grad_norm": 0.38593499549881277,
+      "learning_rate": 5.4956312017188315e-06,
+      "loss": 0.4667,
+      "step": 3642
+    },
+    {
+      "epoch": 2.449861309573842,
+      "grad_norm": 0.3615940397816671,
+      "learning_rate": 5.493440460060785e-06,
+      "loss": 0.4512,
+      "step": 3643
+    },
+    {
+      "epoch": 2.450533748003698,
+      "grad_norm": 0.3281921881336059,
+      "learning_rate": 5.491249622739177e-06,
+      "loss": 0.4531,
+      "step": 3644
+    },
+    {
+      "epoch": 2.451206186433555,
+      "grad_norm": 0.34808784454053615,
+      "learning_rate": 5.489058690178748e-06,
+      "loss": 0.4414,
+      "step": 3645
+    },
+    {
+      "epoch": 2.451878624863411,
+      "grad_norm": 0.35169534590584306,
+      "learning_rate": 5.4868676628042524e-06,
+      "loss": 0.4396,
+      "step": 3646
+    },
+    {
+      "epoch": 2.452551063293267,
+      "grad_norm": 0.32127016672844627,
+      "learning_rate": 5.484676541040471e-06,
+      "loss": 0.4441,
+      "step": 3647
+    },
+    {
+      "epoch": 2.4532235017231234,
+      "grad_norm": 0.42630039919354706,
+      "learning_rate": 5.482485325312192e-06,
+      "loss": 0.453,
+      "step": 3648
+    },
+    {
+      "epoch": 2.4538959401529796,
+      "grad_norm": 0.3802912671793287,
+      "learning_rate": 5.480294016044232e-06,
+      "loss": 0.4543,
+      "step": 3649
+    },
+    {
+      "epoch": 2.454568378582836,
+      "grad_norm": 0.44748190503299723,
+      "learning_rate": 5.478102613661419e-06,
+      "loss": 0.4687,
+      "step": 3650
+    },
+    {
+      "epoch": 2.455240817012692,
+      "grad_norm": 0.3925107294918871,
+      "learning_rate": 5.475911118588603e-06,
+      "loss": 0.4542,
+      "step": 3651
+    },
+    {
+      "epoch": 2.4559132554425487,
+      "grad_norm": 0.30168919377227377,
+      "learning_rate": 5.473719531250648e-06,
+      "loss": 0.4424,
+      "step": 3652
+    },
+    {
+      "epoch": 2.456585693872405,
+      "grad_norm": 0.3977451541372308,
+      "learning_rate": 5.471527852072439e-06,
+      "loss": 0.4498,
+      "step": 3653
+    },
+    {
+      "epoch": 2.457258132302261,
+      "grad_norm": 0.3721355148940516,
+      "learning_rate": 5.46933608147888e-06,
+      "loss": 0.4476,
+      "step": 3654
+    },
+    {
+      "epoch": 2.4579305707321173,
+      "grad_norm": 0.35096306924828724,
+      "learning_rate": 5.467144219894888e-06,
+      "loss": 0.4413,
+      "step": 3655
+    },
+    {
+      "epoch": 2.4586030091619735,
+      "grad_norm": 0.3745171435882103,
+      "learning_rate": 5.4649522677454025e-06,
+      "loss": 0.4475,
+      "step": 3656
+    },
+    {
+      "epoch": 2.4592754475918297,
+      "grad_norm": 0.34740641576534054,
+      "learning_rate": 5.4627602254553756e-06,
+      "loss": 0.4531,
+      "step": 3657
+    },
+    {
+      "epoch": 2.459947886021686,
+      "grad_norm": 0.31338871132628493,
+      "learning_rate": 5.460568093449782e-06,
+      "loss": 0.454,
+      "step": 3658
+    },
+    {
+      "epoch": 2.4606203244515426,
+      "grad_norm": 0.3593929671118801,
+      "learning_rate": 5.458375872153611e-06,
+      "loss": 0.4474,
+      "step": 3659
+    },
+    {
+      "epoch": 2.4612927628813988,
+      "grad_norm": 0.33493349981955667,
+      "learning_rate": 5.45618356199187e-06,
+      "loss": 0.4511,
+      "step": 3660
+    },
+    {
+      "epoch": 2.461965201311255,
+      "grad_norm": 0.3787649160828917,
+      "learning_rate": 5.453991163389583e-06,
+      "loss": 0.4399,
+      "step": 3661
+    },
+    {
+      "epoch": 2.462637639741111,
+      "grad_norm": 0.32480355204924527,
+      "learning_rate": 5.451798676771793e-06,
+      "loss": 0.4438,
+      "step": 3662
+    },
+    {
+      "epoch": 2.4633100781709674,
+      "grad_norm": 0.31279420879956654,
+      "learning_rate": 5.449606102563554e-06,
+      "loss": 0.4452,
+      "step": 3663
+    },
+    {
+      "epoch": 2.4639825166008236,
+      "grad_norm": 0.28327007386162856,
+      "learning_rate": 5.447413441189945e-06,
+      "loss": 0.4472,
+      "step": 3664
+    },
+    {
+      "epoch": 2.4646549550306798,
+      "grad_norm": 0.3486201881078406,
+      "learning_rate": 5.445220693076058e-06,
+      "loss": 0.4563,
+      "step": 3665
+    },
+    {
+      "epoch": 2.4653273934605364,
+      "grad_norm": 0.36849582646330364,
+      "learning_rate": 5.443027858647002e-06,
+      "loss": 0.4522,
+      "step": 3666
+    },
+    {
+      "epoch": 2.4659998318903926,
+      "grad_norm": 0.3301145167435137,
+      "learning_rate": 5.440834938327905e-06,
+      "loss": 0.4584,
+      "step": 3667
+    },
+    {
+      "epoch": 2.466672270320249,
+      "grad_norm": 0.3003337613365606,
+      "learning_rate": 5.438641932543905e-06,
+      "loss": 0.4464,
+      "step": 3668
+    },
+    {
+      "epoch": 2.467344708750105,
+      "grad_norm": 0.3347995636249391,
+      "learning_rate": 5.436448841720166e-06,
+      "loss": 0.435,
+      "step": 3669
+    },
+    {
+      "epoch": 2.4680171471799612,
+      "grad_norm": 0.30100475327165044,
+      "learning_rate": 5.43425566628186e-06,
+      "loss": 0.4565,
+      "step": 3670
+    },
+    {
+      "epoch": 2.4686895856098174,
+      "grad_norm": 0.38738774674073373,
+      "learning_rate": 5.432062406654182e-06,
+      "loss": 0.4513,
+      "step": 3671
+    },
+    {
+      "epoch": 2.4693620240396736,
+      "grad_norm": 0.3242614621730323,
+      "learning_rate": 5.4298690632623375e-06,
+      "loss": 0.454,
+      "step": 3672
+    },
+    {
+      "epoch": 2.4700344624695303,
+      "grad_norm": 0.3905562361155058,
+      "learning_rate": 5.427675636531552e-06,
+      "loss": 0.444,
+      "step": 3673
+    },
+    {
+      "epoch": 2.4707069008993865,
+      "grad_norm": 0.34570382106032044,
+      "learning_rate": 5.425482126887067e-06,
+      "loss": 0.4396,
+      "step": 3674
+    },
+    {
+      "epoch": 2.4713793393292427,
+      "grad_norm": 0.3403550891713755,
+      "learning_rate": 5.423288534754141e-06,
+      "loss": 0.4522,
+      "step": 3675
+    },
+    {
+      "epoch": 2.472051777759099,
+      "grad_norm": 0.3318834277282211,
+      "learning_rate": 5.421094860558045e-06,
+      "loss": 0.4388,
+      "step": 3676
+    },
+    {
+      "epoch": 2.472724216188955,
+      "grad_norm": 0.29745058737843016,
+      "learning_rate": 5.4189011047240655e-06,
+      "loss": 0.4707,
+      "step": 3677
+    },
+    {
+      "epoch": 2.4733966546188113,
+      "grad_norm": 0.3149220065122759,
+      "learning_rate": 5.416707267677512e-06,
+      "loss": 0.4372,
+      "step": 3678
+    },
+    {
+      "epoch": 2.4740690930486675,
+      "grad_norm": 0.32808004579387384,
+      "learning_rate": 5.414513349843702e-06,
+      "loss": 0.4561,
+      "step": 3679
+    },
+    {
+      "epoch": 2.474741531478524,
+      "grad_norm": 0.38005276723555714,
+      "learning_rate": 5.412319351647971e-06,
+      "loss": 0.4572,
+      "step": 3680
+    },
+    {
+      "epoch": 2.4754139699083804,
+      "grad_norm": 0.32430615453417455,
+      "learning_rate": 5.410125273515672e-06,
+      "loss": 0.4377,
+      "step": 3681
+    },
+    {
+      "epoch": 2.4760864083382366,
+      "grad_norm": 0.29363910335500876,
+      "learning_rate": 5.407931115872169e-06,
+      "loss": 0.4345,
+      "step": 3682
+    },
+    {
+      "epoch": 2.4767588467680928,
+      "grad_norm": 0.390361054176332,
+      "learning_rate": 5.405736879142849e-06,
+      "loss": 0.45,
+      "step": 3683
+    },
+    {
+      "epoch": 2.477431285197949,
+      "grad_norm": 0.36684773614787974,
+      "learning_rate": 5.403542563753108e-06,
+      "loss": 0.4363,
+      "step": 3684
+    },
+    {
+      "epoch": 2.478103723627805,
+      "grad_norm": 0.3920056483082055,
+      "learning_rate": 5.401348170128359e-06,
+      "loss": 0.4523,
+      "step": 3685
+    },
+    {
+      "epoch": 2.4787761620576614,
+      "grad_norm": 1.4669592281148975,
+      "learning_rate": 5.399153698694029e-06,
+      "loss": 0.4544,
+      "step": 3686
+    },
+    {
+      "epoch": 2.479448600487518,
+      "grad_norm": 0.33742803366775703,
+      "learning_rate": 5.396959149875564e-06,
+      "loss": 0.4525,
+      "step": 3687
+    },
+    {
+      "epoch": 2.4801210389173742,
+      "grad_norm": 0.43895102427182286,
+      "learning_rate": 5.3947645240984185e-06,
+      "loss": 0.4398,
+      "step": 3688
+    },
+    {
+      "epoch": 2.4807934773472304,
+      "grad_norm": 0.3213947832025373,
+      "learning_rate": 5.392569821788069e-06,
+      "loss": 0.4261,
+      "step": 3689
+    },
+    {
+      "epoch": 2.4814659157770866,
+      "grad_norm": 0.35836800990492956,
+      "learning_rate": 5.390375043370002e-06,
+      "loss": 0.4289,
+      "step": 3690
+    },
+    {
+      "epoch": 2.482138354206943,
+      "grad_norm": 0.41082932360377455,
+      "learning_rate": 5.388180189269722e-06,
+      "loss": 0.466,
+      "step": 3691
+    },
+    {
+      "epoch": 2.482810792636799,
+      "grad_norm": 0.3616481356253951,
+      "learning_rate": 5.385985259912745e-06,
+      "loss": 0.4434,
+      "step": 3692
+    },
+    {
+      "epoch": 2.4834832310666553,
+      "grad_norm": 0.3233270418518311,
+      "learning_rate": 5.383790255724604e-06,
+      "loss": 0.4624,
+      "step": 3693
+    },
+    {
+      "epoch": 2.484155669496512,
+      "grad_norm": 0.3773964032169205,
+      "learning_rate": 5.381595177130844e-06,
+      "loss": 0.4633,
+      "step": 3694
+    },
+    {
+      "epoch": 2.484828107926368,
+      "grad_norm": 0.40606903638666175,
+      "learning_rate": 5.379400024557029e-06,
+      "loss": 0.4504,
+      "step": 3695
+    },
+    {
+      "epoch": 2.4855005463562243,
+      "grad_norm": 0.35668009511381027,
+      "learning_rate": 5.37720479842873e-06,
+      "loss": 0.4459,
+      "step": 3696
+    },
+    {
+      "epoch": 2.4861729847860805,
+      "grad_norm": 0.34781885195436796,
+      "learning_rate": 5.37500949917154e-06,
+      "loss": 0.4368,
+      "step": 3697
+    },
+    {
+      "epoch": 2.4868454232159367,
+      "grad_norm": 0.3977243510705404,
+      "learning_rate": 5.3728141272110625e-06,
+      "loss": 0.4397,
+      "step": 3698
+    },
+    {
+      "epoch": 2.487517861645793,
+      "grad_norm": 0.38920376735507745,
+      "learning_rate": 5.370618682972914e-06,
+      "loss": 0.4699,
+      "step": 3699
+    },
+    {
+      "epoch": 2.488190300075649,
+      "grad_norm": 0.535365852410878,
+      "learning_rate": 5.368423166882729e-06,
+      "loss": 0.4579,
+      "step": 3700
+    },
+    {
+      "epoch": 2.4888627385055058,
+      "grad_norm": 0.3562149949820014,
+      "learning_rate": 5.366227579366151e-06,
+      "loss": 0.4407,
+      "step": 3701
+    },
+    {
+      "epoch": 2.489535176935362,
+      "grad_norm": 0.3354089482329607,
+      "learning_rate": 5.364031920848841e-06,
+      "loss": 0.4611,
+      "step": 3702
+    },
+    {
+      "epoch": 2.490207615365218,
+      "grad_norm": 0.4049203957942453,
+      "learning_rate": 5.361836191756471e-06,
+      "loss": 0.4558,
+      "step": 3703
+    },
+    {
+      "epoch": 2.4908800537950744,
+      "grad_norm": 0.3010025011844406,
+      "learning_rate": 5.35964039251473e-06,
+      "loss": 0.4443,
+      "step": 3704
+    },
+    {
+      "epoch": 2.4915524922249306,
+      "grad_norm": 0.3885268762782305,
+      "learning_rate": 5.3574445235493165e-06,
+      "loss": 0.4522,
+      "step": 3705
+    },
+    {
+      "epoch": 2.492224930654787,
+      "grad_norm": 0.9568793935382124,
+      "learning_rate": 5.355248585285946e-06,
+      "loss": 0.4513,
+      "step": 3706
+    },
+    {
+      "epoch": 2.492897369084643,
+      "grad_norm": 0.4198708586420078,
+      "learning_rate": 5.353052578150346e-06,
+      "loss": 0.454,
+      "step": 3707
+    },
+    {
+      "epoch": 2.4935698075144996,
+      "grad_norm": 0.3116920558206328,
+      "learning_rate": 5.350856502568258e-06,
+      "loss": 0.4531,
+      "step": 3708
+    },
+    {
+      "epoch": 2.494242245944356,
+      "grad_norm": 0.3530569433463639,
+      "learning_rate": 5.348660358965438e-06,
+      "loss": 0.4373,
+      "step": 3709
+    },
+    {
+      "epoch": 2.494914684374212,
+      "grad_norm": 0.6219596914149117,
+      "learning_rate": 5.346464147767649e-06,
+      "loss": 0.47,
+      "step": 3710
+    },
+    {
+      "epoch": 2.4955871228040682,
+      "grad_norm": 0.7200298627812961,
+      "learning_rate": 5.344267869400676e-06,
+      "loss": 0.4562,
+      "step": 3711
+    },
+    {
+      "epoch": 2.4962595612339245,
+      "grad_norm": 0.3831329382165244,
+      "learning_rate": 5.342071524290311e-06,
+      "loss": 0.4503,
+      "step": 3712
+    },
+    {
+      "epoch": 2.4969319996637807,
+      "grad_norm": 0.3730820369626398,
+      "learning_rate": 5.339875112862361e-06,
+      "loss": 0.4544,
+      "step": 3713
+    },
+    {
+      "epoch": 2.497604438093637,
+      "grad_norm": 0.9308006610674574,
+      "learning_rate": 5.337678635542641e-06,
+      "loss": 0.4455,
+      "step": 3714
+    },
+    {
+      "epoch": 2.4982768765234935,
+      "grad_norm": 0.3900152218356268,
+      "learning_rate": 5.335482092756991e-06,
+      "loss": 0.4519,
+      "step": 3715
+    },
+    {
+      "epoch": 2.4989493149533497,
+      "grad_norm": 0.34510470883456346,
+      "learning_rate": 5.33328548493125e-06,
+      "loss": 0.4614,
+      "step": 3716
+    },
+    {
+      "epoch": 2.499621753383206,
+      "grad_norm": 0.39301377708235846,
+      "learning_rate": 5.331088812491278e-06,
+      "loss": 0.4376,
+      "step": 3717
+    },
+    {
+      "epoch": 2.500294191813062,
+      "grad_norm": 0.4132391443171161,
+      "learning_rate": 5.3288920758629445e-06,
+      "loss": 0.4611,
+      "step": 3718
+    },
+    {
+      "epoch": 2.5009666302429183,
+      "grad_norm": 0.30379365709876277,
+      "learning_rate": 5.326695275472132e-06,
+      "loss": 0.462,
+      "step": 3719
+    },
+    {
+      "epoch": 2.5016390686727745,
+      "grad_norm": 0.3063310686320014,
+      "learning_rate": 5.324498411744737e-06,
+      "loss": 0.4438,
+      "step": 3720
+    },
+    {
+      "epoch": 2.5023115071026307,
+      "grad_norm": 0.3687412953586924,
+      "learning_rate": 5.322301485106663e-06,
+      "loss": 0.4485,
+      "step": 3721
+    },
+    {
+      "epoch": 2.5029839455324874,
+      "grad_norm": 0.30038806642290977,
+      "learning_rate": 5.320104495983831e-06,
+      "loss": 0.448,
+      "step": 3722
+    },
+    {
+      "epoch": 2.5036563839623436,
+      "grad_norm": 0.4042132931618547,
+      "learning_rate": 5.317907444802174e-06,
+      "loss": 0.4486,
+      "step": 3723
+    },
+    {
+      "epoch": 2.5043288223922,
+      "grad_norm": 0.38294548405814394,
+      "learning_rate": 5.315710331987634e-06,
+      "loss": 0.459,
+      "step": 3724
+    },
+    {
+      "epoch": 2.505001260822056,
+      "grad_norm": 0.2917400221339506,
+      "learning_rate": 5.313513157966166e-06,
+      "loss": 0.4398,
+      "step": 3725
+    },
+    {
+      "epoch": 2.505673699251912,
+      "grad_norm": 0.30874946783443197,
+      "learning_rate": 5.311315923163739e-06,
+      "loss": 0.4611,
+      "step": 3726
+    },
+    {
+      "epoch": 2.5063461376817684,
+      "grad_norm": 0.32034127394020057,
+      "learning_rate": 5.30911862800633e-06,
+      "loss": 0.4564,
+      "step": 3727
+    },
+    {
+      "epoch": 2.5070185761116246,
+      "grad_norm": 0.393814502245678,
+      "learning_rate": 5.306921272919931e-06,
+      "loss": 0.4548,
+      "step": 3728
+    },
+    {
+      "epoch": 2.5076910145414812,
+      "grad_norm": 0.3589397375038359,
+      "learning_rate": 5.304723858330544e-06,
+      "loss": 0.4525,
+      "step": 3729
+    },
+    {
+      "epoch": 2.5083634529713374,
+      "grad_norm": 0.3886193975048878,
+      "learning_rate": 5.302526384664182e-06,
+      "loss": 0.4326,
+      "step": 3730
+    },
+    {
+      "epoch": 2.5090358914011937,
+      "grad_norm": 0.3619242724868831,
+      "learning_rate": 5.300328852346872e-06,
+      "loss": 0.4603,
+      "step": 3731
+    },
+    {
+      "epoch": 2.50970832983105,
+      "grad_norm": 0.3808986117863621,
+      "learning_rate": 5.29813126180465e-06,
+      "loss": 0.4587,
+      "step": 3732
+    },
+    {
+      "epoch": 2.510380768260906,
+      "grad_norm": 0.3818801429897,
+      "learning_rate": 5.295933613463565e-06,
+      "loss": 0.4528,
+      "step": 3733
+    },
+    {
+      "epoch": 2.5110532066907623,
+      "grad_norm": 0.3185246625786046,
+      "learning_rate": 5.2937359077496756e-06,
+      "loss": 0.4388,
+      "step": 3734
+    },
+    {
+      "epoch": 2.5117256451206185,
+      "grad_norm": 0.33417761556505876,
+      "learning_rate": 5.291538145089051e-06,
+      "loss": 0.4422,
+      "step": 3735
+    },
+    {
+      "epoch": 2.512398083550475,
+      "grad_norm": 0.3543133452008666,
+      "learning_rate": 5.289340325907774e-06,
+      "loss": 0.4375,
+      "step": 3736
+    },
+    {
+      "epoch": 2.5130705219803313,
+      "grad_norm": 0.45034041931345337,
+      "learning_rate": 5.287142450631937e-06,
+      "loss": 0.4511,
+      "step": 3737
+    },
+    {
+      "epoch": 2.5137429604101875,
+      "grad_norm": 0.3341749624172436,
+      "learning_rate": 5.284944519687641e-06,
+      "loss": 0.4419,
+      "step": 3738
+    },
+    {
+      "epoch": 2.5144153988400437,
+      "grad_norm": 0.31416689317385305,
+      "learning_rate": 5.2827465335010034e-06,
+      "loss": 0.4485,
+      "step": 3739
+    },
+    {
+      "epoch": 2.5150878372699,
+      "grad_norm": 0.32352161394960716,
+      "learning_rate": 5.280548492498145e-06,
+      "loss": 0.4637,
+      "step": 3740
+    },
+    {
+      "epoch": 2.515760275699756,
+      "grad_norm": 0.3596611050259769,
+      "learning_rate": 5.278350397105205e-06,
+      "loss": 0.4492,
+      "step": 3741
+    },
+    {
+      "epoch": 2.5164327141296123,
+      "grad_norm": 0.38533801199013906,
+      "learning_rate": 5.276152247748329e-06,
+      "loss": 0.4591,
+      "step": 3742
+    },
+    {
+      "epoch": 2.517105152559469,
+      "grad_norm": 0.39893128818506984,
+      "learning_rate": 5.27395404485367e-06,
+      "loss": 0.4515,
+      "step": 3743
+    },
+    {
+      "epoch": 2.517777590989325,
+      "grad_norm": 0.3601291730744921,
+      "learning_rate": 5.271755788847399e-06,
+      "loss": 0.4331,
+      "step": 3744
+    },
+    {
+      "epoch": 2.5184500294191814,
+      "grad_norm": 0.35341859376319246,
+      "learning_rate": 5.269557480155689e-06,
+      "loss": 0.4467,
+      "step": 3745
+    },
+    {
+      "epoch": 2.5191224678490376,
+      "grad_norm": 0.34460484487452686,
+      "learning_rate": 5.267359119204729e-06,
+      "loss": 0.4619,
+      "step": 3746
+    },
+    {
+      "epoch": 2.519794906278894,
+      "grad_norm": 0.3163956703666074,
+      "learning_rate": 5.2651607064207174e-06,
+      "loss": 0.4456,
+      "step": 3747
+    },
+    {
+      "epoch": 2.52046734470875,
+      "grad_norm": 0.38107463823791704,
+      "learning_rate": 5.262962242229861e-06,
+      "loss": 0.4604,
+      "step": 3748
+    },
+    {
+      "epoch": 2.521139783138606,
+      "grad_norm": 0.354401592955316,
+      "learning_rate": 5.260763727058377e-06,
+      "loss": 0.4598,
+      "step": 3749
+    },
+    {
+      "epoch": 2.521812221568463,
+      "grad_norm": 0.30038110187928374,
+      "learning_rate": 5.258565161332493e-06,
+      "loss": 0.4312,
+      "step": 3750
+    },
+    {
+      "epoch": 2.522484659998319,
+      "grad_norm": 0.35161551300999627,
+      "learning_rate": 5.256366545478444e-06,
+      "loss": 0.4369,
+      "step": 3751
+    },
+    {
+      "epoch": 2.5231570984281753,
+      "grad_norm": 0.4154141265683533,
+      "learning_rate": 5.2541678799224795e-06,
+      "loss": 0.4545,
+      "step": 3752
+    },
+    {
+      "epoch": 2.5238295368580315,
+      "grad_norm": 0.2843520409523974,
+      "learning_rate": 5.251969165090857e-06,
+      "loss": 0.4477,
+      "step": 3753
+    },
+    {
+      "epoch": 2.5245019752878877,
+      "grad_norm": 0.3356459667390433,
+      "learning_rate": 5.2497704014098375e-06,
+      "loss": 0.4437,
+      "step": 3754
+    },
+    {
+      "epoch": 2.525174413717744,
+      "grad_norm": 0.40533334967123485,
+      "learning_rate": 5.2475715893057e-06,
+      "loss": 0.4461,
+      "step": 3755
+    },
+    {
+      "epoch": 2.5258468521476,
+      "grad_norm": 0.351820305830887,
+      "learning_rate": 5.24537272920473e-06,
+      "loss": 0.4497,
+      "step": 3756
+    },
+    {
+      "epoch": 2.5265192905774567,
+      "grad_norm": 0.36700532615869463,
+      "learning_rate": 5.243173821533221e-06,
+      "loss": 0.4603,
+      "step": 3757
+    },
+    {
+      "epoch": 2.527191729007313,
+      "grad_norm": 0.36421125558057277,
+      "learning_rate": 5.2409748667174746e-06,
+      "loss": 0.4508,
+      "step": 3758
+    },
+    {
+      "epoch": 2.527864167437169,
+      "grad_norm": 0.3593637659685638,
+      "learning_rate": 5.238775865183805e-06,
+      "loss": 0.4319,
+      "step": 3759
+    },
+    {
+      "epoch": 2.5285366058670253,
+      "grad_norm": 0.451127830436243,
+      "learning_rate": 5.236576817358533e-06,
+      "loss": 0.4504,
+      "step": 3760
+    },
+    {
+      "epoch": 2.5292090442968815,
+      "grad_norm": 0.3612158552582543,
+      "learning_rate": 5.2343777236679905e-06,
+      "loss": 0.4432,
+      "step": 3761
+    },
+    {
+      "epoch": 2.5298814827267377,
+      "grad_norm": 0.35963170197322686,
+      "learning_rate": 5.232178584538515e-06,
+      "loss": 0.4586,
+      "step": 3762
+    },
+    {
+      "epoch": 2.530553921156594,
+      "grad_norm": 0.3513240604985171,
+      "learning_rate": 5.229979400396455e-06,
+      "loss": 0.4506,
+      "step": 3763
+    },
+    {
+      "epoch": 2.5312263595864506,
+      "grad_norm": 0.40705859055602794,
+      "learning_rate": 5.227780171668169e-06,
+      "loss": 0.4413,
+      "step": 3764
+    },
+    {
+      "epoch": 2.531898798016307,
+      "grad_norm": 0.38056720078199147,
+      "learning_rate": 5.225580898780022e-06,
+      "loss": 0.449,
+      "step": 3765
+    },
+    {
+      "epoch": 2.532571236446163,
+      "grad_norm": 0.3344126266364116,
+      "learning_rate": 5.223381582158388e-06,
+      "loss": 0.4499,
+      "step": 3766
+    },
+    {
+      "epoch": 2.533243674876019,
+      "grad_norm": 0.3450151129401091,
+      "learning_rate": 5.2211822222296495e-06,
+      "loss": 0.4584,
+      "step": 3767
+    },
+    {
+      "epoch": 2.5339161133058754,
+      "grad_norm": 0.3125769248548215,
+      "learning_rate": 5.218982819420199e-06,
+      "loss": 0.4434,
+      "step": 3768
+    },
+    {
+      "epoch": 2.5345885517357316,
+      "grad_norm": 0.3459655360483346,
+      "learning_rate": 5.216783374156432e-06,
+      "loss": 0.4543,
+      "step": 3769
+    },
+    {
+      "epoch": 2.535260990165588,
+      "grad_norm": 0.33743900715413994,
+      "learning_rate": 5.21458388686476e-06,
+      "loss": 0.4444,
+      "step": 3770
+    },
+    {
+      "epoch": 2.5359334285954445,
+      "grad_norm": 0.34409415630074597,
+      "learning_rate": 5.212384357971599e-06,
+      "loss": 0.4341,
+      "step": 3771
+    },
+    {
+      "epoch": 2.5366058670253,
+      "grad_norm": 0.3347872267650656,
+      "learning_rate": 5.21018478790337e-06,
+      "loss": 0.4525,
+      "step": 3772
+    },
+    {
+      "epoch": 2.537278305455157,
+      "grad_norm": 0.32985605985546435,
+      "learning_rate": 5.207985177086506e-06,
+      "loss": 0.4302,
+      "step": 3773
+    },
+    {
+      "epoch": 2.537950743885013,
+      "grad_norm": 0.3319653933091309,
+      "learning_rate": 5.205785525947445e-06,
+      "loss": 0.4317,
+      "step": 3774
+    },
+    {
+      "epoch": 2.5386231823148693,
+      "grad_norm": 0.3318287561536541,
+      "learning_rate": 5.20358583491264e-06,
+      "loss": 0.4572,
+      "step": 3775
+    },
+    {
+      "epoch": 2.5392956207447255,
+      "grad_norm": 0.2725473101671124,
+      "learning_rate": 5.20138610440854e-06,
+      "loss": 0.4362,
+      "step": 3776
+    },
+    {
+      "epoch": 2.5399680591745817,
+      "grad_norm": 0.6860141812033361,
+      "learning_rate": 5.199186334861612e-06,
+      "loss": 0.4323,
+      "step": 3777
+    },
+    {
+      "epoch": 2.5406404976044383,
+      "grad_norm": 0.3415301890154741,
+      "learning_rate": 5.1969865266983245e-06,
+      "loss": 0.4566,
+      "step": 3778
+    },
+    {
+      "epoch": 2.541312936034294,
+      "grad_norm": 0.3686557719090204,
+      "learning_rate": 5.1947866803451555e-06,
+      "loss": 0.4478,
+      "step": 3779
+    },
+    {
+      "epoch": 2.5419853744641507,
+      "grad_norm": 0.35319179579920057,
+      "learning_rate": 5.192586796228589e-06,
+      "loss": 0.4488,
+      "step": 3780
+    },
+    {
+      "epoch": 2.542657812894007,
+      "grad_norm": 0.3668355608939105,
+      "learning_rate": 5.190386874775123e-06,
+      "loss": 0.4484,
+      "step": 3781
+    },
+    {
+      "epoch": 2.543330251323863,
+      "grad_norm": 0.34965355223710987,
+      "learning_rate": 5.1881869164112516e-06,
+      "loss": 0.4337,
+      "step": 3782
+    },
+    {
+      "epoch": 2.5440026897537193,
+      "grad_norm": 0.35308599479483654,
+      "learning_rate": 5.185986921563485e-06,
+      "loss": 0.4373,
+      "step": 3783
+    },
+    {
+      "epoch": 2.5446751281835756,
+      "grad_norm": 0.324205088399401,
+      "learning_rate": 5.183786890658337e-06,
+      "loss": 0.4378,
+      "step": 3784
+    },
+    {
+      "epoch": 2.545347566613432,
+      "grad_norm": 0.380163684655463,
+      "learning_rate": 5.181586824122327e-06,
+      "loss": 0.4706,
+      "step": 3785
+    },
+    {
+      "epoch": 2.546020005043288,
+      "grad_norm": 0.33074437109441635,
+      "learning_rate": 5.1793867223819846e-06,
+      "loss": 0.4704,
+      "step": 3786
+    },
+    {
+      "epoch": 2.5466924434731446,
+      "grad_norm": 0.3292609650828323,
+      "learning_rate": 5.177186585863845e-06,
+      "loss": 0.45,
+      "step": 3787
+    },
+    {
+      "epoch": 2.547364881903001,
+      "grad_norm": 0.4139902078750704,
+      "learning_rate": 5.1749864149944485e-06,
+      "loss": 0.4398,
+      "step": 3788
+    },
+    {
+      "epoch": 2.548037320332857,
+      "grad_norm": 0.5993488931079183,
+      "learning_rate": 5.172786210200343e-06,
+      "loss": 0.4486,
+      "step": 3789
+    },
+    {
+      "epoch": 2.548709758762713,
+      "grad_norm": 0.42042569553752296,
+      "learning_rate": 5.170585971908087e-06,
+      "loss": 0.4456,
+      "step": 3790
+    },
+    {
+      "epoch": 2.5493821971925694,
+      "grad_norm": 0.3823238256484114,
+      "learning_rate": 5.168385700544239e-06,
+      "loss": 0.4457,
+      "step": 3791
+    },
+    {
+      "epoch": 2.550054635622426,
+      "grad_norm": 0.37123219611674746,
+      "learning_rate": 5.166185396535366e-06,
+      "loss": 0.4599,
+      "step": 3792
+    },
+    {
+      "epoch": 2.550727074052282,
+      "grad_norm": 0.36751036038349477,
+      "learning_rate": 5.1639850603080435e-06,
+      "loss": 0.4412,
+      "step": 3793
+    },
+    {
+      "epoch": 2.5513995124821385,
+      "grad_norm": 0.28656948143169914,
+      "learning_rate": 5.16178469228885e-06,
+      "loss": 0.4557,
+      "step": 3794
+    },
+    {
+      "epoch": 2.5520719509119947,
+      "grad_norm": 0.32446866213313197,
+      "learning_rate": 5.159584292904375e-06,
+      "loss": 0.4583,
+      "step": 3795
+    },
+    {
+      "epoch": 2.552744389341851,
+      "grad_norm": 0.5964162794538592,
+      "learning_rate": 5.157383862581209e-06,
+      "loss": 0.4458,
+      "step": 3796
+    },
+    {
+      "epoch": 2.553416827771707,
+      "grad_norm": 0.3297847044616623,
+      "learning_rate": 5.155183401745953e-06,
+      "loss": 0.4544,
+      "step": 3797
+    },
+    {
+      "epoch": 2.5540892662015633,
+      "grad_norm": 0.3327996968439367,
+      "learning_rate": 5.152982910825207e-06,
+      "loss": 0.437,
+      "step": 3798
+    },
+    {
+      "epoch": 2.55476170463142,
+      "grad_norm": 0.3178463020120286,
+      "learning_rate": 5.150782390245586e-06,
+      "loss": 0.4384,
+      "step": 3799
+    },
+    {
+      "epoch": 2.5554341430612757,
+      "grad_norm": 0.41892183789374043,
+      "learning_rate": 5.148581840433703e-06,
+      "loss": 0.447,
+      "step": 3800
+    },
+    {
+      "epoch": 2.5561065814911323,
+      "grad_norm": 0.39518782274915,
+      "learning_rate": 5.1463812618161815e-06,
+      "loss": 0.4528,
+      "step": 3801
+    },
+    {
+      "epoch": 2.5567790199209885,
+      "grad_norm": 0.32549332411081666,
+      "learning_rate": 5.144180654819647e-06,
+      "loss": 0.4519,
+      "step": 3802
+    },
+    {
+      "epoch": 2.5574514583508448,
+      "grad_norm": 0.38308949144270205,
+      "learning_rate": 5.1419800198707335e-06,
+      "loss": 0.4513,
+      "step": 3803
+    },
+    {
+      "epoch": 2.558123896780701,
+      "grad_norm": 0.3117899154840055,
+      "learning_rate": 5.13977935739608e-06,
+      "loss": 0.4426,
+      "step": 3804
+    },
+    {
+      "epoch": 2.558796335210557,
+      "grad_norm": 0.38078454486838426,
+      "learning_rate": 5.13757866782233e-06,
+      "loss": 0.4416,
+      "step": 3805
+    },
+    {
+      "epoch": 2.559468773640414,
+      "grad_norm": 0.3892915045158023,
+      "learning_rate": 5.135377951576129e-06,
+      "loss": 0.4484,
+      "step": 3806
+    },
+    {
+      "epoch": 2.5601412120702696,
+      "grad_norm": 0.3922351660486267,
+      "learning_rate": 5.133177209084135e-06,
+      "loss": 0.4734,
+      "step": 3807
+    },
+    {
+      "epoch": 2.560813650500126,
+      "grad_norm": 0.3145125925952972,
+      "learning_rate": 5.130976440773007e-06,
+      "loss": 0.4418,
+      "step": 3808
+    },
+    {
+      "epoch": 2.5614860889299824,
+      "grad_norm": 0.33503272386293115,
+      "learning_rate": 5.128775647069406e-06,
+      "loss": 0.4403,
+      "step": 3809
+    },
+    {
+      "epoch": 2.5621585273598386,
+      "grad_norm": 0.3206081131155908,
+      "learning_rate": 5.126574828400005e-06,
+      "loss": 0.4657,
+      "step": 3810
+    },
+    {
+      "epoch": 2.562830965789695,
+      "grad_norm": 0.3677827674996036,
+      "learning_rate": 5.124373985191473e-06,
+      "loss": 0.4607,
+      "step": 3811
+    },
+    {
+      "epoch": 2.563503404219551,
+      "grad_norm": 0.3313057068776185,
+      "learning_rate": 5.122173117870493e-06,
+      "loss": 0.439,
+      "step": 3812
+    },
+    {
+      "epoch": 2.5641758426494077,
+      "grad_norm": 0.33844242490608817,
+      "learning_rate": 5.119972226863746e-06,
+      "loss": 0.4511,
+      "step": 3813
+    },
+    {
+      "epoch": 2.5648482810792634,
+      "grad_norm": 0.46303435953046407,
+      "learning_rate": 5.117771312597921e-06,
+      "loss": 0.4489,
+      "step": 3814
+    },
+    {
+      "epoch": 2.56552071950912,
+      "grad_norm": 0.3947544136560684,
+      "learning_rate": 5.11557037549971e-06,
+      "loss": 0.4497,
+      "step": 3815
+    },
+    {
+      "epoch": 2.5661931579389763,
+      "grad_norm": 0.32707548217635574,
+      "learning_rate": 5.113369415995811e-06,
+      "loss": 0.4528,
+      "step": 3816
+    },
+    {
+      "epoch": 2.5668655963688325,
+      "grad_norm": 0.3493213925953552,
+      "learning_rate": 5.1111684345129216e-06,
+      "loss": 0.4442,
+      "step": 3817
+    },
+    {
+      "epoch": 2.5675380347986887,
+      "grad_norm": 0.32379197184450675,
+      "learning_rate": 5.10896743147775e-06,
+      "loss": 0.4392,
+      "step": 3818
+    },
+    {
+      "epoch": 2.568210473228545,
+      "grad_norm": 0.3734254339295873,
+      "learning_rate": 5.106766407317005e-06,
+      "loss": 0.4531,
+      "step": 3819
+    },
+    {
+      "epoch": 2.5688829116584015,
+      "grad_norm": 0.40420746016090003,
+      "learning_rate": 5.104565362457402e-06,
+      "loss": 0.4493,
+      "step": 3820
+    },
+    {
+      "epoch": 2.5695553500882573,
+      "grad_norm": 0.35353410656708223,
+      "learning_rate": 5.102364297325658e-06,
+      "loss": 0.4496,
+      "step": 3821
+    },
+    {
+      "epoch": 2.570227788518114,
+      "grad_norm": 0.3325958370960057,
+      "learning_rate": 5.100163212348492e-06,
+      "loss": 0.4439,
+      "step": 3822
+    },
+    {
+      "epoch": 2.57090022694797,
+      "grad_norm": 0.3045565716634522,
+      "learning_rate": 5.097962107952634e-06,
+      "loss": 0.4547,
+      "step": 3823
+    },
+    {
+      "epoch": 2.5715726653778264,
+      "grad_norm": 0.32827460477458903,
+      "learning_rate": 5.0957609845648095e-06,
+      "loss": 0.4599,
+      "step": 3824
+    },
+    {
+      "epoch": 2.5722451038076826,
+      "grad_norm": 0.3896500200822667,
+      "learning_rate": 5.093559842611753e-06,
+      "loss": 0.4547,
+      "step": 3825
+    },
+    {
+      "epoch": 2.5729175422375388,
+      "grad_norm": 0.3558175143222838,
+      "learning_rate": 5.091358682520201e-06,
+      "loss": 0.4546,
+      "step": 3826
+    },
+    {
+      "epoch": 2.5735899806673954,
+      "grad_norm": 0.49023083575365556,
+      "learning_rate": 5.089157504716892e-06,
+      "loss": 0.4364,
+      "step": 3827
+    },
+    {
+      "epoch": 2.574262419097251,
+      "grad_norm": 0.46450442763910227,
+      "learning_rate": 5.086956309628571e-06,
+      "loss": 0.4422,
+      "step": 3828
+    },
+    {
+      "epoch": 2.574934857527108,
+      "grad_norm": 0.2988745132917568,
+      "learning_rate": 5.084755097681986e-06,
+      "loss": 0.4425,
+      "step": 3829
+    },
+    {
+      "epoch": 2.575607295956964,
+      "grad_norm": 0.4269834171435638,
+      "learning_rate": 5.0825538693038855e-06,
+      "loss": 0.451,
+      "step": 3830
+    },
+    {
+      "epoch": 2.5762797343868202,
+      "grad_norm": 0.3509948140622454,
+      "learning_rate": 5.080352624921024e-06,
+      "loss": 0.4457,
+      "step": 3831
+    },
+    {
+      "epoch": 2.5769521728166764,
+      "grad_norm": 0.3429616383589686,
+      "learning_rate": 5.078151364960155e-06,
+      "loss": 0.4581,
+      "step": 3832
+    },
+    {
+      "epoch": 2.5776246112465326,
+      "grad_norm": 0.3287023293291114,
+      "learning_rate": 5.075950089848042e-06,
+      "loss": 0.4388,
+      "step": 3833
+    },
+    {
+      "epoch": 2.578297049676389,
+      "grad_norm": 0.39127126728235045,
+      "learning_rate": 5.0737488000114445e-06,
+      "loss": 0.4593,
+      "step": 3834
+    },
+    {
+      "epoch": 2.578969488106245,
+      "grad_norm": 0.3211215536959106,
+      "learning_rate": 5.0715474958771285e-06,
+      "loss": 0.4402,
+      "step": 3835
+    },
+    {
+      "epoch": 2.5796419265361017,
+      "grad_norm": 0.4274164835536321,
+      "learning_rate": 5.069346177871861e-06,
+      "loss": 0.432,
+      "step": 3836
+    },
+    {
+      "epoch": 2.580314364965958,
+      "grad_norm": 0.314409178497314,
+      "learning_rate": 5.067144846422414e-06,
+      "loss": 0.4412,
+      "step": 3837
+    },
+    {
+      "epoch": 2.580986803395814,
+      "grad_norm": 0.30381499647440013,
+      "learning_rate": 5.064943501955561e-06,
+      "loss": 0.4449,
+      "step": 3838
+    },
+    {
+      "epoch": 2.5816592418256703,
+      "grad_norm": 0.3075225485816905,
+      "learning_rate": 5.062742144898077e-06,
+      "loss": 0.4333,
+      "step": 3839
+    },
+    {
+      "epoch": 2.5823316802555265,
+      "grad_norm": 0.31533603082366796,
+      "learning_rate": 5.060540775676741e-06,
+      "loss": 0.4426,
+      "step": 3840
+    },
+    {
+      "epoch": 2.5830041186853827,
+      "grad_norm": 0.40525331303601203,
+      "learning_rate": 5.058339394718334e-06,
+      "loss": 0.4429,
+      "step": 3841
+    },
+    {
+      "epoch": 2.583676557115239,
+      "grad_norm": 0.3258861688476787,
+      "learning_rate": 5.056138002449637e-06,
+      "loss": 0.4559,
+      "step": 3842
+    },
+    {
+      "epoch": 2.5843489955450956,
+      "grad_norm": 0.4413250887046823,
+      "learning_rate": 5.053936599297434e-06,
+      "loss": 0.4413,
+      "step": 3843
+    },
+    {
+      "epoch": 2.5850214339749518,
+      "grad_norm": 0.34107229576023107,
+      "learning_rate": 5.051735185688517e-06,
+      "loss": 0.4546,
+      "step": 3844
+    },
+    {
+      "epoch": 2.585693872404808,
+      "grad_norm": 0.35051767639704207,
+      "learning_rate": 5.049533762049672e-06,
+      "loss": 0.429,
+      "step": 3845
+    },
+    {
+      "epoch": 2.586366310834664,
+      "grad_norm": 0.31965601826054163,
+      "learning_rate": 5.04733232880769e-06,
+      "loss": 0.4484,
+      "step": 3846
+    },
+    {
+      "epoch": 2.5870387492645204,
+      "grad_norm": 0.31771321814884435,
+      "learning_rate": 5.045130886389366e-06,
+      "loss": 0.4491,
+      "step": 3847
+    },
+    {
+      "epoch": 2.5877111876943766,
+      "grad_norm": 0.4950577405834247,
+      "learning_rate": 5.04292943522149e-06,
+      "loss": 0.4632,
+      "step": 3848
+    },
+    {
+      "epoch": 2.588383626124233,
+      "grad_norm": 0.3789943252896637,
+      "learning_rate": 5.040727975730866e-06,
+      "loss": 0.4484,
+      "step": 3849
+    },
+    {
+      "epoch": 2.5890560645540894,
+      "grad_norm": 0.3349219315480771,
+      "learning_rate": 5.038526508344286e-06,
+      "loss": 0.4377,
+      "step": 3850
+    },
+    {
+      "epoch": 2.5897285029839456,
+      "grad_norm": 0.29139705497556456,
+      "learning_rate": 5.036325033488552e-06,
+      "loss": 0.4329,
+      "step": 3851
+    },
+    {
+      "epoch": 2.590400941413802,
+      "grad_norm": 0.31799540729427056,
+      "learning_rate": 5.034123551590464e-06,
+      "loss": 0.4499,
+      "step": 3852
+    },
+    {
+      "epoch": 2.591073379843658,
+      "grad_norm": 0.46667337205344145,
+      "learning_rate": 5.031922063076825e-06,
+      "loss": 0.4435,
+      "step": 3853
+    },
+    {
+      "epoch": 2.5917458182735142,
+      "grad_norm": 0.3173954396303586,
+      "learning_rate": 5.029720568374441e-06,
+      "loss": 0.4434,
+      "step": 3854
+    },
+    {
+      "epoch": 2.5924182567033704,
+      "grad_norm": 0.4038778883302914,
+      "learning_rate": 5.027519067910113e-06,
+      "loss": 0.4562,
+      "step": 3855
+    },
+    {
+      "epoch": 2.5930906951332267,
+      "grad_norm": 0.3334517592595017,
+      "learning_rate": 5.0253175621106496e-06,
+      "loss": 0.4641,
+      "step": 3856
+    },
+    {
+      "epoch": 2.5937631335630833,
+      "grad_norm": 0.34674178762730656,
+      "learning_rate": 5.023116051402857e-06,
+      "loss": 0.4506,
+      "step": 3857
+    },
+    {
+      "epoch": 2.5944355719929395,
+      "grad_norm": 0.33680663953221074,
+      "learning_rate": 5.020914536213544e-06,
+      "loss": 0.4601,
+      "step": 3858
+    },
+    {
+      "epoch": 2.5951080104227957,
+      "grad_norm": 0.29991687118698995,
+      "learning_rate": 5.018713016969518e-06,
+      "loss": 0.4384,
+      "step": 3859
+    },
+    {
+      "epoch": 2.595780448852652,
+      "grad_norm": 0.6532795206230446,
+      "learning_rate": 5.016511494097589e-06,
+      "loss": 0.444,
+      "step": 3860
+    },
+    {
+      "epoch": 2.596452887282508,
+      "grad_norm": 0.40399908418384883,
+      "learning_rate": 5.014309968024569e-06,
+      "loss": 0.4498,
+      "step": 3861
+    },
+    {
+      "epoch": 2.5971253257123643,
+      "grad_norm": 0.4697959446174812,
+      "learning_rate": 5.0121084391772675e-06,
+      "loss": 0.4368,
+      "step": 3862
+    },
+    {
+      "epoch": 2.5977977641422205,
+      "grad_norm": 0.33135513650808096,
+      "learning_rate": 5.009906907982498e-06,
+      "loss": 0.4527,
+      "step": 3863
+    },
+    {
+      "epoch": 2.598470202572077,
+      "grad_norm": 0.4101478794356378,
+      "learning_rate": 5.007705374867069e-06,
+      "loss": 0.46,
+      "step": 3864
+    },
+    {
+      "epoch": 2.5991426410019334,
+      "grad_norm": 0.38900473343673664,
+      "learning_rate": 5.005503840257797e-06,
+      "loss": 0.4541,
+      "step": 3865
+    },
+    {
+      "epoch": 2.5998150794317896,
+      "grad_norm": 0.35719574910883917,
+      "learning_rate": 5.003302304581491e-06,
+      "loss": 0.4462,
+      "step": 3866
+    },
+    {
+      "epoch": 2.600487517861646,
+      "grad_norm": 0.3964272408590807,
+      "learning_rate": 5.001100768264967e-06,
+      "loss": 0.4498,
+      "step": 3867
+    },
+    {
+      "epoch": 2.601159956291502,
+      "grad_norm": 0.45572001774437354,
+      "learning_rate": 4.998899231735036e-06,
+      "loss": 0.4397,
+      "step": 3868
+    },
+    {
+      "epoch": 2.601832394721358,
+      "grad_norm": 0.3854538081021313,
+      "learning_rate": 4.996697695418509e-06,
+      "loss": 0.4431,
+      "step": 3869
+    },
+    {
+      "epoch": 2.6025048331512144,
+      "grad_norm": 0.46467833571113215,
+      "learning_rate": 4.9944961597422045e-06,
+      "loss": 0.4464,
+      "step": 3870
+    },
+    {
+      "epoch": 2.603177271581071,
+      "grad_norm": 0.49315106615561277,
+      "learning_rate": 4.992294625132931e-06,
+      "loss": 0.4454,
+      "step": 3871
+    },
+    {
+      "epoch": 2.6038497100109272,
+      "grad_norm": 0.41126821224026316,
+      "learning_rate": 4.990093092017505e-06,
+      "loss": 0.4275,
+      "step": 3872
+    },
+    {
+      "epoch": 2.6045221484407834,
+      "grad_norm": 0.33764530796775627,
+      "learning_rate": 4.987891560822734e-06,
+      "loss": 0.4448,
+      "step": 3873
+    },
+    {
+      "epoch": 2.6051945868706397,
+      "grad_norm": 0.3960123488009799,
+      "learning_rate": 4.9856900319754325e-06,
+      "loss": 0.4433,
+      "step": 3874
+    },
+    {
+      "epoch": 2.605867025300496,
+      "grad_norm": 0.40780345359310755,
+      "learning_rate": 4.983488505902412e-06,
+      "loss": 0.4566,
+      "step": 3875
+    },
+    {
+      "epoch": 2.606539463730352,
+      "grad_norm": 0.7557178612501396,
+      "learning_rate": 4.981286983030483e-06,
+      "loss": 0.4555,
+      "step": 3876
+    },
+    {
+      "epoch": 2.6072119021602083,
+      "grad_norm": 0.42171713653694304,
+      "learning_rate": 4.979085463786458e-06,
+      "loss": 0.4435,
+      "step": 3877
+    },
+    {
+      "epoch": 2.607884340590065,
+      "grad_norm": 0.3340723064181383,
+      "learning_rate": 4.976883948597144e-06,
+      "loss": 0.4303,
+      "step": 3878
+    },
+    {
+      "epoch": 2.608556779019921,
+      "grad_norm": 0.29293861353368855,
+      "learning_rate": 4.974682437889351e-06,
+      "loss": 0.4565,
+      "step": 3879
+    },
+    {
+      "epoch": 2.6092292174497773,
+      "grad_norm": 0.31253385788814575,
+      "learning_rate": 4.972480932089887e-06,
+      "loss": 0.4412,
+      "step": 3880
+    },
+    {
+      "epoch": 2.6099016558796335,
+      "grad_norm": 0.42975688666696954,
+      "learning_rate": 4.970279431625562e-06,
+      "loss": 0.4285,
+      "step": 3881
+    },
+    {
+      "epoch": 2.6105740943094897,
+      "grad_norm": 0.33551292043635644,
+      "learning_rate": 4.9680779369231765e-06,
+      "loss": 0.4223,
+      "step": 3882
+    },
+    {
+      "epoch": 2.611246532739346,
+      "grad_norm": 0.35565215137443845,
+      "learning_rate": 4.965876448409538e-06,
+      "loss": 0.4392,
+      "step": 3883
+    },
+    {
+      "epoch": 2.611918971169202,
+      "grad_norm": 0.3795433031972901,
+      "learning_rate": 4.96367496651145e-06,
+      "loss": 0.4491,
+      "step": 3884
+    },
+    {
+      "epoch": 2.612591409599059,
+      "grad_norm": 0.33398281760717086,
+      "learning_rate": 4.961473491655716e-06,
+      "loss": 0.4491,
+      "step": 3885
+    },
+    {
+      "epoch": 2.613263848028915,
+      "grad_norm": 0.34789460060253946,
+      "learning_rate": 4.9592720242691365e-06,
+      "loss": 0.4449,
+      "step": 3886
+    },
+    {
+      "epoch": 2.613936286458771,
+      "grad_norm": 0.7521759371904863,
+      "learning_rate": 4.95707056477851e-06,
+      "loss": 0.462,
+      "step": 3887
+    },
+    {
+      "epoch": 2.6146087248886274,
+      "grad_norm": 0.39615418310057654,
+      "learning_rate": 4.954869113610636e-06,
+      "loss": 0.4371,
+      "step": 3888
+    },
+    {
+      "epoch": 2.6152811633184836,
+      "grad_norm": 0.3072143702625055,
+      "learning_rate": 4.9526676711923105e-06,
+      "loss": 0.4308,
+      "step": 3889
+    },
+    {
+      "epoch": 2.61595360174834,
+      "grad_norm": 0.4958232349535628,
+      "learning_rate": 4.950466237950331e-06,
+      "loss": 0.4411,
+      "step": 3890
+    },
+    {
+      "epoch": 2.616626040178196,
+      "grad_norm": 0.3206297035074455,
+      "learning_rate": 4.948264814311485e-06,
+      "loss": 0.4503,
+      "step": 3891
+    },
+    {
+      "epoch": 2.6172984786080526,
+      "grad_norm": 0.34153060861871193,
+      "learning_rate": 4.9460634007025666e-06,
+      "loss": 0.4371,
+      "step": 3892
+    },
+    {
+      "epoch": 2.617970917037909,
+      "grad_norm": 0.34692001072221584,
+      "learning_rate": 4.943861997550364e-06,
+      "loss": 0.4316,
+      "step": 3893
+    },
+    {
+      "epoch": 2.618643355467765,
+      "grad_norm": 0.30897086469776225,
+      "learning_rate": 4.941660605281669e-06,
+      "loss": 0.44,
+      "step": 3894
+    },
+    {
+      "epoch": 2.6193157938976213,
+      "grad_norm": 0.2964363692863969,
+      "learning_rate": 4.93945922432326e-06,
+      "loss": 0.4264,
+      "step": 3895
+    },
+    {
+      "epoch": 2.6199882323274775,
+      "grad_norm": 0.3451149147036572,
+      "learning_rate": 4.937257855101924e-06,
+      "loss": 0.4494,
+      "step": 3896
+    },
+    {
+      "epoch": 2.6206606707573337,
+      "grad_norm": 0.3723792970190891,
+      "learning_rate": 4.9350564980444395e-06,
+      "loss": 0.4466,
+      "step": 3897
+    },
+    {
+      "epoch": 2.62133310918719,
+      "grad_norm": 0.49347025859965554,
+      "learning_rate": 4.932855153577586e-06,
+      "loss": 0.4522,
+      "step": 3898
+    },
+    {
+      "epoch": 2.6220055476170465,
+      "grad_norm": 0.3619643702258718,
+      "learning_rate": 4.9306538221281405e-06,
+      "loss": 0.435,
+      "step": 3899
+    },
+    {
+      "epoch": 2.6226779860469027,
+      "grad_norm": 0.6146626855230981,
+      "learning_rate": 4.928452504122873e-06,
+      "loss": 0.445,
+      "step": 3900
+    },
+    {
+      "epoch": 2.623350424476759,
+      "grad_norm": 0.38166015510603984,
+      "learning_rate": 4.926251199988557e-06,
+      "loss": 0.4572,
+      "step": 3901
+    },
+    {
+      "epoch": 2.624022862906615,
+      "grad_norm": 0.3070758076618173,
+      "learning_rate": 4.924049910151959e-06,
+      "loss": 0.4391,
+      "step": 3902
+    },
+    {
+      "epoch": 2.6246953013364713,
+      "grad_norm": 0.40800701682714846,
+      "learning_rate": 4.9218486350398465e-06,
+      "loss": 0.4461,
+      "step": 3903
+    },
+    {
+      "epoch": 2.6253677397663275,
+      "grad_norm": 0.2973151821861155,
+      "learning_rate": 4.919647375078978e-06,
+      "loss": 0.4479,
+      "step": 3904
+    },
+    {
+      "epoch": 2.6260401781961837,
+      "grad_norm": 0.38336864739259563,
+      "learning_rate": 4.917446130696115e-06,
+      "loss": 0.4475,
+      "step": 3905
+    },
+    {
+      "epoch": 2.6267126166260404,
+      "grad_norm": 0.3214570330346468,
+      "learning_rate": 4.915244902318015e-06,
+      "loss": 0.4498,
+      "step": 3906
+    },
+    {
+      "epoch": 2.6273850550558966,
+      "grad_norm": 0.6332904874395928,
+      "learning_rate": 4.913043690371428e-06,
+      "loss": 0.4369,
+      "step": 3907
+    },
+    {
+      "epoch": 2.628057493485753,
+      "grad_norm": 0.38335987959677537,
+      "learning_rate": 4.910842495283109e-06,
+      "loss": 0.4459,
+      "step": 3908
+    },
+    {
+      "epoch": 2.628729931915609,
+      "grad_norm": 0.6059650455962325,
+      "learning_rate": 4.908641317479801e-06,
+      "loss": 0.4444,
+      "step": 3909
+    },
+    {
+      "epoch": 2.629402370345465,
+      "grad_norm": 0.3583108368754653,
+      "learning_rate": 4.906440157388248e-06,
+      "loss": 0.4574,
+      "step": 3910
+    },
+    {
+      "epoch": 2.6300748087753214,
+      "grad_norm": 0.4759206529064839,
+      "learning_rate": 4.904239015435191e-06,
+      "loss": 0.4556,
+      "step": 3911
+    },
+    {
+      "epoch": 2.6307472472051776,
+      "grad_norm": 0.2995614698468876,
+      "learning_rate": 4.902037892047368e-06,
+      "loss": 0.4504,
+      "step": 3912
+    },
+    {
+      "epoch": 2.6314196856350343,
+      "grad_norm": 0.44317698613266354,
+      "learning_rate": 4.899836787651509e-06,
+      "loss": 0.4493,
+      "step": 3913
+    },
+    {
+      "epoch": 2.6320921240648905,
+      "grad_norm": 0.349065702885983,
+      "learning_rate": 4.897635702674344e-06,
+      "loss": 0.4488,
+      "step": 3914
+    },
+    {
+      "epoch": 2.6327645624947467,
+      "grad_norm": 0.8225983390548947,
+      "learning_rate": 4.895434637542598e-06,
+      "loss": 0.4612,
+      "step": 3915
+    },
+    {
+      "epoch": 2.633437000924603,
+      "grad_norm": 0.3849712777720729,
+      "learning_rate": 4.893233592682996e-06,
+      "loss": 0.4457,
+      "step": 3916
+    },
+    {
+      "epoch": 2.634109439354459,
+      "grad_norm": 0.37120607081678014,
+      "learning_rate": 4.8910325685222515e-06,
+      "loss": 0.4511,
+      "step": 3917
+    },
+    {
+      "epoch": 2.6347818777843153,
+      "grad_norm": 0.3827365226913108,
+      "learning_rate": 4.88883156548708e-06,
+      "loss": 0.4467,
+      "step": 3918
+    },
+    {
+      "epoch": 2.6354543162141715,
+      "grad_norm": 0.3254878222098606,
+      "learning_rate": 4.886630584004191e-06,
+      "loss": 0.4584,
+      "step": 3919
+    },
+    {
+      "epoch": 2.636126754644028,
+      "grad_norm": 0.771189071779267,
+      "learning_rate": 4.88442962450029e-06,
+      "loss": 0.44,
+      "step": 3920
+    },
+    {
+      "epoch": 2.6367991930738843,
+      "grad_norm": 0.3757486231448364,
+      "learning_rate": 4.882228687402081e-06,
+      "loss": 0.4406,
+      "step": 3921
+    },
+    {
+      "epoch": 2.6374716315037405,
+      "grad_norm": 0.3503419583682442,
+      "learning_rate": 4.880027773136255e-06,
+      "loss": 0.4483,
+      "step": 3922
+    },
+    {
+      "epoch": 2.6381440699335967,
+      "grad_norm": 0.3443790198819944,
+      "learning_rate": 4.877826882129509e-06,
+      "loss": 0.4395,
+      "step": 3923
+    },
+    {
+      "epoch": 2.638816508363453,
+      "grad_norm": 0.3113332727795207,
+      "learning_rate": 4.875626014808528e-06,
+      "loss": 0.456,
+      "step": 3924
+    },
+    {
+      "epoch": 2.639488946793309,
+      "grad_norm": 0.3685464703821919,
+      "learning_rate": 4.873425171599998e-06,
+      "loss": 0.4355,
+      "step": 3925
+    },
+    {
+      "epoch": 2.6401613852231653,
+      "grad_norm": 0.36662723975171313,
+      "learning_rate": 4.871224352930596e-06,
+      "loss": 0.4294,
+      "step": 3926
+    },
+    {
+      "epoch": 2.640833823653022,
+      "grad_norm": 0.3117729797433047,
+      "learning_rate": 4.8690235592269946e-06,
+      "loss": 0.4317,
+      "step": 3927
+    },
+    {
+      "epoch": 2.641506262082878,
+      "grad_norm": 0.35979033394621746,
+      "learning_rate": 4.866822790915865e-06,
+      "loss": 0.4415,
+      "step": 3928
+    },
+    {
+      "epoch": 2.6421787005127344,
+      "grad_norm": 0.37009695722503977,
+      "learning_rate": 4.864622048423871e-06,
+      "loss": 0.4302,
+      "step": 3929
+    },
+    {
+      "epoch": 2.6428511389425906,
+      "grad_norm": 0.3568595977172907,
+      "learning_rate": 4.862421332177674e-06,
+      "loss": 0.4406,
+      "step": 3930
+    },
+    {
+      "epoch": 2.643523577372447,
+      "grad_norm": 0.41973788949292573,
+      "learning_rate": 4.860220642603922e-06,
+      "loss": 0.4427,
+      "step": 3931
+    },
+    {
+      "epoch": 2.644196015802303,
+      "grad_norm": 0.36700531410496684,
+      "learning_rate": 4.858019980129267e-06,
+      "loss": 0.4534,
+      "step": 3932
+    },
+    {
+      "epoch": 2.644868454232159,
+      "grad_norm": 0.30269861413872146,
+      "learning_rate": 4.855819345180354e-06,
+      "loss": 0.4497,
+      "step": 3933
+    },
+    {
+      "epoch": 2.645540892662016,
+      "grad_norm": 0.6152798719083605,
+      "learning_rate": 4.853618738183821e-06,
+      "loss": 0.4531,
+      "step": 3934
+    },
+    {
+      "epoch": 2.6462133310918716,
+      "grad_norm": 0.30401248236622574,
+      "learning_rate": 4.8514181595662984e-06,
+      "loss": 0.4488,
+      "step": 3935
+    },
+    {
+      "epoch": 2.6468857695217283,
+      "grad_norm": 0.37429738252667466,
+      "learning_rate": 4.849217609754415e-06,
+      "loss": 0.4411,
+      "step": 3936
+    },
+    {
+      "epoch": 2.6475582079515845,
+      "grad_norm": 0.36301547372734627,
+      "learning_rate": 4.847017089174793e-06,
+      "loss": 0.4536,
+      "step": 3937
+    },
+    {
+      "epoch": 2.6482306463814407,
+      "grad_norm": 0.4981606026371317,
+      "learning_rate": 4.84481659825405e-06,
+      "loss": 0.4422,
+      "step": 3938
+    },
+    {
+      "epoch": 2.648903084811297,
+      "grad_norm": 0.34595349036346,
+      "learning_rate": 4.842616137418792e-06,
+      "loss": 0.4351,
+      "step": 3939
+    },
+    {
+      "epoch": 2.649575523241153,
+      "grad_norm": 0.34087659069215315,
+      "learning_rate": 4.840415707095626e-06,
+      "loss": 0.4605,
+      "step": 3940
+    },
+    {
+      "epoch": 2.6502479616710097,
+      "grad_norm": 0.4107782880869281,
+      "learning_rate": 4.8382153077111505e-06,
+      "loss": 0.4493,
+      "step": 3941
+    },
+    {
+      "epoch": 2.6509204001008655,
+      "grad_norm": 0.32892418309754395,
+      "learning_rate": 4.836014939691957e-06,
+      "loss": 0.4371,
+      "step": 3942
+    },
+    {
+      "epoch": 2.651592838530722,
+      "grad_norm": 0.35482669054847,
+      "learning_rate": 4.8338146034646365e-06,
+      "loss": 0.4512,
+      "step": 3943
+    },
+    {
+      "epoch": 2.6522652769605783,
+      "grad_norm": 0.4993082701462041,
+      "learning_rate": 4.831614299455763e-06,
+      "loss": 0.4514,
+      "step": 3944
+    },
+    {
+      "epoch": 2.6529377153904345,
+      "grad_norm": 0.4296586018836682,
+      "learning_rate": 4.829414028091914e-06,
+      "loss": 0.4436,
+      "step": 3945
+    },
+    {
+      "epoch": 2.6536101538202908,
+      "grad_norm": 0.3528586999832636,
+      "learning_rate": 4.827213789799656e-06,
+      "loss": 0.4454,
+      "step": 3946
+    },
+    {
+      "epoch": 2.654282592250147,
+      "grad_norm": 0.3524200784847369,
+      "learning_rate": 4.825013585005554e-06,
+      "loss": 0.4505,
+      "step": 3947
+    },
+    {
+      "epoch": 2.6549550306800036,
+      "grad_norm": 0.3637313457903069,
+      "learning_rate": 4.822813414136157e-06,
+      "loss": 0.4193,
+      "step": 3948
+    },
+    {
+      "epoch": 2.6556274691098594,
+      "grad_norm": 0.30803439934798615,
+      "learning_rate": 4.820613277618016e-06,
+      "loss": 0.4236,
+      "step": 3949
+    },
+    {
+      "epoch": 2.656299907539716,
+      "grad_norm": 0.32356597567643813,
+      "learning_rate": 4.818413175877674e-06,
+      "loss": 0.4343,
+      "step": 3950
+    },
+    {
+      "epoch": 2.656972345969572,
+      "grad_norm": 0.3117213382948113,
+      "learning_rate": 4.816213109341664e-06,
+      "loss": 0.4297,
+      "step": 3951
+    },
+    {
+      "epoch": 2.6576447843994284,
+      "grad_norm": 0.30292637566183234,
+      "learning_rate": 4.814013078436517e-06,
+      "loss": 0.4388,
+      "step": 3952
+    },
+    {
+      "epoch": 2.6583172228292846,
+      "grad_norm": 0.4441875471980197,
+      "learning_rate": 4.81181308358875e-06,
+      "loss": 0.4477,
+      "step": 3953
+    },
+    {
+      "epoch": 2.658989661259141,
+      "grad_norm": 0.35145162184961165,
+      "learning_rate": 4.8096131252248785e-06,
+      "loss": 0.4514,
+      "step": 3954
+    },
+    {
+      "epoch": 2.6596620996889975,
+      "grad_norm": 0.32937856083291966,
+      "learning_rate": 4.80741320377141e-06,
+      "loss": 0.4514,
+      "step": 3955
+    },
+    {
+      "epoch": 2.6603345381188532,
+      "grad_norm": 0.3354037013121662,
+      "learning_rate": 4.805213319654847e-06,
+      "loss": 0.4409,
+      "step": 3956
+    },
+    {
+      "epoch": 2.66100697654871,
+      "grad_norm": 0.354286594317942,
+      "learning_rate": 4.803013473301677e-06,
+      "loss": 0.449,
+      "step": 3957
+    },
+    {
+      "epoch": 2.661679414978566,
+      "grad_norm": 0.3923920764655487,
+      "learning_rate": 4.800813665138389e-06,
+      "loss": 0.4352,
+      "step": 3958
+    },
+    {
+      "epoch": 2.6623518534084223,
+      "grad_norm": 0.4069089239628588,
+      "learning_rate": 4.7986138955914604e-06,
+      "loss": 0.4469,
+      "step": 3959
+    },
+    {
+      "epoch": 2.6630242918382785,
+      "grad_norm": 0.40021340268552197,
+      "learning_rate": 4.796414165087363e-06,
+      "loss": 0.4566,
+      "step": 3960
+    },
+    {
+      "epoch": 2.6636967302681347,
+      "grad_norm": 0.4094469334606063,
+      "learning_rate": 4.794214474052555e-06,
+      "loss": 0.4343,
+      "step": 3961
+    },
+    {
+      "epoch": 2.6643691686979913,
+      "grad_norm": 0.3149962928170417,
+      "learning_rate": 4.792014822913496e-06,
+      "loss": 0.4337,
+      "step": 3962
+    },
+    {
+      "epoch": 2.665041607127847,
+      "grad_norm": 0.34533606893544566,
+      "learning_rate": 4.789815212096632e-06,
+      "loss": 0.4378,
+      "step": 3963
+    },
+    {
+      "epoch": 2.6657140455577037,
+      "grad_norm": 0.42166294954305816,
+      "learning_rate": 4.787615642028402e-06,
+      "loss": 0.441,
+      "step": 3964
+    },
+    {
+      "epoch": 2.66638648398756,
+      "grad_norm": 0.3720335576158297,
+      "learning_rate": 4.785416113135241e-06,
+      "loss": 0.454,
+      "step": 3965
+    },
+    {
+      "epoch": 2.667058922417416,
+      "grad_norm": 0.34103569866637795,
+      "learning_rate": 4.7832166258435684e-06,
+      "loss": 0.4489,
+      "step": 3966
+    },
+    {
+      "epoch": 2.6677313608472724,
+      "grad_norm": 0.3144751920124584,
+      "learning_rate": 4.781017180579803e-06,
+      "loss": 0.4367,
+      "step": 3967
+    },
+    {
+      "epoch": 2.6684037992771286,
+      "grad_norm": 0.2934768252750471,
+      "learning_rate": 4.7788177777703505e-06,
+      "loss": 0.451,
+      "step": 3968
+    },
+    {
+      "epoch": 2.669076237706985,
+      "grad_norm": 0.44173091051097085,
+      "learning_rate": 4.776618417841615e-06,
+      "loss": 0.4555,
+      "step": 3969
+    },
+    {
+      "epoch": 2.669748676136841,
+      "grad_norm": 0.3621724409828,
+      "learning_rate": 4.7744191012199805e-06,
+      "loss": 0.4462,
+      "step": 3970
+    },
+    {
+      "epoch": 2.6704211145666976,
+      "grad_norm": 0.4217030265201206,
+      "learning_rate": 4.772219828331833e-06,
+      "loss": 0.4401,
+      "step": 3971
+    },
+    {
+      "epoch": 2.671093552996554,
+      "grad_norm": 0.3258720085996176,
+      "learning_rate": 4.7700205996035465e-06,
+      "loss": 0.4438,
+      "step": 3972
+    },
+    {
+      "epoch": 2.67176599142641,
+      "grad_norm": 0.3286042825081745,
+      "learning_rate": 4.767821415461487e-06,
+      "loss": 0.4423,
+      "step": 3973
+    },
+    {
+      "epoch": 2.6724384298562662,
+      "grad_norm": 0.37240805371753094,
+      "learning_rate": 4.765622276332013e-06,
+      "loss": 0.4476,
+      "step": 3974
+    },
+    {
+      "epoch": 2.6731108682861224,
+      "grad_norm": 0.46705034090135494,
+      "learning_rate": 4.763423182641469e-06,
+      "loss": 0.4555,
+      "step": 3975
+    },
+    {
+      "epoch": 2.673783306715979,
+      "grad_norm": 0.31248097026521715,
+      "learning_rate": 4.761224134816197e-06,
+      "loss": 0.4265,
+      "step": 3976
+    },
+    {
+      "epoch": 2.674455745145835,
+      "grad_norm": 0.3701567562637434,
+      "learning_rate": 4.759025133282526e-06,
+      "loss": 0.4493,
+      "step": 3977
+    },
+    {
+      "epoch": 2.6751281835756915,
+      "grad_norm": 0.334949784468544,
+      "learning_rate": 4.756826178466783e-06,
+      "loss": 0.447,
+      "step": 3978
+    },
+    {
+      "epoch": 2.6758006220055477,
+      "grad_norm": 0.3687722696307002,
+      "learning_rate": 4.754627270795272e-06,
+      "loss": 0.4402,
+      "step": 3979
+    },
+    {
+      "epoch": 2.676473060435404,
+      "grad_norm": 0.33745162534262396,
+      "learning_rate": 4.752428410694301e-06,
+      "loss": 0.4341,
+      "step": 3980
+    },
+    {
+      "epoch": 2.67714549886526,
+      "grad_norm": 0.3030594134472027,
+      "learning_rate": 4.750229598590163e-06,
+      "loss": 0.4352,
+      "step": 3981
+    },
+    {
+      "epoch": 2.6778179372951163,
+      "grad_norm": 0.34207218560456,
+      "learning_rate": 4.748030834909147e-06,
+      "loss": 0.4322,
+      "step": 3982
+    },
+    {
+      "epoch": 2.678490375724973,
+      "grad_norm": 0.518303876053175,
+      "learning_rate": 4.745832120077521e-06,
+      "loss": 0.4557,
+      "step": 3983
+    },
+    {
+      "epoch": 2.6791628141548287,
+      "grad_norm": 0.3883315083460478,
+      "learning_rate": 4.7436334545215565e-06,
+      "loss": 0.4499,
+      "step": 3984
+    },
+    {
+      "epoch": 2.6798352525846854,
+      "grad_norm": 0.33195023039056065,
+      "learning_rate": 4.741434838667508e-06,
+      "loss": 0.4194,
+      "step": 3985
+    },
+    {
+      "epoch": 2.6805076910145416,
+      "grad_norm": 0.3431102125311453,
+      "learning_rate": 4.739236272941623e-06,
+      "loss": 0.4675,
+      "step": 3986
+    },
+    {
+      "epoch": 2.6811801294443978,
+      "grad_norm": 0.3330249236624501,
+      "learning_rate": 4.737037757770141e-06,
+      "loss": 0.4536,
+      "step": 3987
+    },
+    {
+      "epoch": 2.681852567874254,
+      "grad_norm": 0.7319725476689913,
+      "learning_rate": 4.734839293579284e-06,
+      "loss": 0.4496,
+      "step": 3988
+    },
+    {
+      "epoch": 2.68252500630411,
+      "grad_norm": 0.374642251156238,
+      "learning_rate": 4.732640880795272e-06,
+      "loss": 0.4486,
+      "step": 3989
+    },
+    {
+      "epoch": 2.683197444733967,
+      "grad_norm": 0.37675456514832767,
+      "learning_rate": 4.7304425198443125e-06,
+      "loss": 0.4308,
+      "step": 3990
+    },
+    {
+      "epoch": 2.6838698831638226,
+      "grad_norm": 0.3512707357224236,
+      "learning_rate": 4.728244211152604e-06,
+      "loss": 0.4269,
+      "step": 3991
+    },
+    {
+      "epoch": 2.6845423215936792,
+      "grad_norm": 0.4393990814568386,
+      "learning_rate": 4.726045955146331e-06,
+      "loss": 0.4512,
+      "step": 3992
+    },
+    {
+      "epoch": 2.6852147600235354,
+      "grad_norm": 0.35171319923804023,
+      "learning_rate": 4.723847752251673e-06,
+      "loss": 0.4468,
+      "step": 3993
+    },
+    {
+      "epoch": 2.6858871984533916,
+      "grad_norm": 0.4101286118141828,
+      "learning_rate": 4.721649602894795e-06,
+      "loss": 0.4436,
+      "step": 3994
+    },
+    {
+      "epoch": 2.686559636883248,
+      "grad_norm": 0.33289556695129713,
+      "learning_rate": 4.719451507501857e-06,
+      "loss": 0.4577,
+      "step": 3995
+    },
+    {
+      "epoch": 2.687232075313104,
+      "grad_norm": 0.3934530287799647,
+      "learning_rate": 4.717253466499e-06,
+      "loss": 0.4391,
+      "step": 3996
+    },
+    {
+      "epoch": 2.6879045137429602,
+      "grad_norm": 0.30849467174103273,
+      "learning_rate": 4.71505548031236e-06,
+      "loss": 0.4435,
+      "step": 3997
+    },
+    {
+      "epoch": 2.6885769521728164,
+      "grad_norm": 0.46370010026944075,
+      "learning_rate": 4.712857549368065e-06,
+      "loss": 0.4509,
+      "step": 3998
+    },
+    {
+      "epoch": 2.689249390602673,
+      "grad_norm": 0.3206942755628773,
+      "learning_rate": 4.710659674092226e-06,
+      "loss": 0.448,
+      "step": 3999
+    },
+    {
+      "epoch": 2.6899218290325293,
+      "grad_norm": 0.46281544863600615,
+      "learning_rate": 4.7084618549109515e-06,
+      "loss": 0.4525,
+      "step": 4000
+    },
+    {
+      "epoch": 2.6905942674623855,
+      "grad_norm": 0.34585129329342007,
+      "learning_rate": 4.706264092250327e-06,
+      "loss": 0.4379,
+      "step": 4001
+    },
+    {
+      "epoch": 2.6912667058922417,
+      "grad_norm": 0.3354546534002248,
+      "learning_rate": 4.704066386536436e-06,
+      "loss": 0.4447,
+      "step": 4002
+    },
+    {
+      "epoch": 2.691939144322098,
+      "grad_norm": 0.4748817548224659,
+      "learning_rate": 4.70186873819535e-06,
+      "loss": 0.4485,
+      "step": 4003
+    },
+    {
+      "epoch": 2.692611582751954,
+      "grad_norm": 0.33706271599700643,
+      "learning_rate": 4.699671147653129e-06,
+      "loss": 0.4599,
+      "step": 4004
+    },
+    {
+      "epoch": 2.6932840211818103,
+      "grad_norm": 0.41419105569679066,
+      "learning_rate": 4.697473615335819e-06,
+      "loss": 0.4354,
+      "step": 4005
+    },
+    {
+      "epoch": 2.693956459611667,
+      "grad_norm": 0.48293969803601555,
+      "learning_rate": 4.695276141669458e-06,
+      "loss": 0.443,
+      "step": 4006
+    },
+    {
+      "epoch": 2.694628898041523,
+      "grad_norm": 0.3852684945316683,
+      "learning_rate": 4.6930787270800705e-06,
+      "loss": 0.4492,
+      "step": 4007
+    },
+    {
+      "epoch": 2.6953013364713794,
+      "grad_norm": 0.35153149593918026,
+      "learning_rate": 4.690881371993671e-06,
+      "loss": 0.4336,
+      "step": 4008
+    },
+    {
+      "epoch": 2.6959737749012356,
+      "grad_norm": 0.31717053577082666,
+      "learning_rate": 4.688684076836264e-06,
+      "loss": 0.4458,
+      "step": 4009
+    },
+    {
+      "epoch": 2.696646213331092,
+      "grad_norm": 0.4152306709052566,
+      "learning_rate": 4.686486842033836e-06,
+      "loss": 0.4491,
+      "step": 4010
+    },
+    {
+      "epoch": 2.697318651760948,
+      "grad_norm": 0.3764797379680534,
+      "learning_rate": 4.684289668012367e-06,
+      "loss": 0.4335,
+      "step": 4011
+    },
+    {
+      "epoch": 2.697991090190804,
+      "grad_norm": 0.32045573355406193,
+      "learning_rate": 4.682092555197827e-06,
+      "loss": 0.4396,
+      "step": 4012
+    },
+    {
+      "epoch": 2.698663528620661,
+      "grad_norm": 0.36799773366081795,
+      "learning_rate": 4.67989550401617e-06,
+      "loss": 0.4408,
+      "step": 4013
+    },
+    {
+      "epoch": 2.699335967050517,
+      "grad_norm": 0.3558258065277898,
+      "learning_rate": 4.677698514893338e-06,
+      "loss": 0.4302,
+      "step": 4014
+    },
+    {
+      "epoch": 2.7000084054803732,
+      "grad_norm": 0.3554089068118138,
+      "learning_rate": 4.675501588255264e-06,
+      "loss": 0.4367,
+      "step": 4015
+    },
+    {
+      "epoch": 2.7006808439102294,
+      "grad_norm": 0.3475852676857237,
+      "learning_rate": 4.673304724527868e-06,
+      "loss": 0.4486,
+      "step": 4016
+    },
+    {
+      "epoch": 2.7013532823400856,
+      "grad_norm": 0.3515982513660669,
+      "learning_rate": 4.671107924137057e-06,
+      "loss": 0.4401,
+      "step": 4017
+    },
+    {
+      "epoch": 2.702025720769942,
+      "grad_norm": 0.39299387084846193,
+      "learning_rate": 4.668911187508723e-06,
+      "loss": 0.4507,
+      "step": 4018
+    },
+    {
+      "epoch": 2.702698159199798,
+      "grad_norm": 0.3191081249376199,
+      "learning_rate": 4.666714515068751e-06,
+      "loss": 0.4387,
+      "step": 4019
+    },
+    {
+      "epoch": 2.7033705976296547,
+      "grad_norm": 0.515033568771394,
+      "learning_rate": 4.664517907243011e-06,
+      "loss": 0.4438,
+      "step": 4020
+    },
+    {
+      "epoch": 2.704043036059511,
+      "grad_norm": 0.5144595273589103,
+      "learning_rate": 4.662321364457359e-06,
+      "loss": 0.4285,
+      "step": 4021
+    },
+    {
+      "epoch": 2.704715474489367,
+      "grad_norm": 0.3848515807272575,
+      "learning_rate": 4.660124887137643e-06,
+      "loss": 0.4335,
+      "step": 4022
+    },
+    {
+      "epoch": 2.7053879129192233,
+      "grad_norm": 0.3245188178348665,
+      "learning_rate": 4.657928475709691e-06,
+      "loss": 0.4494,
+      "step": 4023
+    },
+    {
+      "epoch": 2.7060603513490795,
+      "grad_norm": 0.32339811165861526,
+      "learning_rate": 4.6557321305993246e-06,
+      "loss": 0.4481,
+      "step": 4024
+    },
+    {
+      "epoch": 2.7067327897789357,
+      "grad_norm": 0.3965925056893796,
+      "learning_rate": 4.65353585223235e-06,
+      "loss": 0.4409,
+      "step": 4025
+    },
+    {
+      "epoch": 2.707405228208792,
+      "grad_norm": 0.35551982751089395,
+      "learning_rate": 4.651339641034565e-06,
+      "loss": 0.4524,
+      "step": 4026
+    },
+    {
+      "epoch": 2.7080776666386486,
+      "grad_norm": 0.48379414670857646,
+      "learning_rate": 4.649143497431743e-06,
+      "loss": 0.4272,
+      "step": 4027
+    },
+    {
+      "epoch": 2.7087501050685048,
+      "grad_norm": 0.41530286231090446,
+      "learning_rate": 4.646947421849655e-06,
+      "loss": 0.4315,
+      "step": 4028
+    },
+    {
+      "epoch": 2.709422543498361,
+      "grad_norm": 0.439133912286961,
+      "learning_rate": 4.644751414714056e-06,
+      "loss": 0.4314,
+      "step": 4029
+    },
+    {
+      "epoch": 2.710094981928217,
+      "grad_norm": 0.3371496209827778,
+      "learning_rate": 4.642555476450684e-06,
+      "loss": 0.4443,
+      "step": 4030
+    },
+    {
+      "epoch": 2.7107674203580734,
+      "grad_norm": 0.4041486073180736,
+      "learning_rate": 4.640359607485272e-06,
+      "loss": 0.4438,
+      "step": 4031
+    },
+    {
+      "epoch": 2.7114398587879296,
+      "grad_norm": 0.3215829040360271,
+      "learning_rate": 4.63816380824353e-06,
+      "loss": 0.4231,
+      "step": 4032
+    },
+    {
+      "epoch": 2.712112297217786,
+      "grad_norm": 0.3355336267006405,
+      "learning_rate": 4.63596807915116e-06,
+      "loss": 0.4408,
+      "step": 4033
+    },
+    {
+      "epoch": 2.7127847356476424,
+      "grad_norm": 0.374171806207746,
+      "learning_rate": 4.633772420633849e-06,
+      "loss": 0.4273,
+      "step": 4034
+    },
+    {
+      "epoch": 2.7134571740774986,
+      "grad_norm": 0.30441595052025694,
+      "learning_rate": 4.6315768331172725e-06,
+      "loss": 0.4489,
+      "step": 4035
+    },
+    {
+      "epoch": 2.714129612507355,
+      "grad_norm": 0.35620556437115636,
+      "learning_rate": 4.629381317027086e-06,
+      "loss": 0.4401,
+      "step": 4036
+    },
+    {
+      "epoch": 2.714802050937211,
+      "grad_norm": 0.35531137314619227,
+      "learning_rate": 4.627185872788938e-06,
+      "loss": 0.4385,
+      "step": 4037
+    },
+    {
+      "epoch": 2.7154744893670673,
+      "grad_norm": 0.4110343598892571,
+      "learning_rate": 4.6249905008284605e-06,
+      "loss": 0.4344,
+      "step": 4038
+    },
+    {
+      "epoch": 2.7161469277969235,
+      "grad_norm": 0.3394545822035457,
+      "learning_rate": 4.622795201571272e-06,
+      "loss": 0.4559,
+      "step": 4039
+    },
+    {
+      "epoch": 2.7168193662267797,
+      "grad_norm": 0.3676872378283041,
+      "learning_rate": 4.620599975442974e-06,
+      "loss": 0.4232,
+      "step": 4040
+    },
+    {
+      "epoch": 2.7174918046566363,
+      "grad_norm": 0.4307709856361871,
+      "learning_rate": 4.618404822869157e-06,
+      "loss": 0.4552,
+      "step": 4041
+    },
+    {
+      "epoch": 2.7181642430864925,
+      "grad_norm": 0.39071616092546935,
+      "learning_rate": 4.616209744275398e-06,
+      "loss": 0.4494,
+      "step": 4042
+    },
+    {
+      "epoch": 2.7188366815163487,
+      "grad_norm": 0.3350976454381335,
+      "learning_rate": 4.614014740087256e-06,
+      "loss": 0.4365,
+      "step": 4043
+    },
+    {
+      "epoch": 2.719509119946205,
+      "grad_norm": 0.35175623390334937,
+      "learning_rate": 4.611819810730281e-06,
+      "loss": 0.4249,
+      "step": 4044
+    },
+    {
+      "epoch": 2.720181558376061,
+      "grad_norm": 0.44179988158296624,
+      "learning_rate": 4.60962495663e-06,
+      "loss": 0.446,
+      "step": 4045
+    },
+    {
+      "epoch": 2.7208539968059173,
+      "grad_norm": 0.31338487083319794,
+      "learning_rate": 4.607430178211933e-06,
+      "loss": 0.4322,
+      "step": 4046
+    },
+    {
+      "epoch": 2.7215264352357735,
+      "grad_norm": 0.3153451800447677,
+      "learning_rate": 4.6052354759015815e-06,
+      "loss": 0.4327,
+      "step": 4047
+    },
+    {
+      "epoch": 2.72219887366563,
+      "grad_norm": 0.3394799354136693,
+      "learning_rate": 4.60304085012444e-06,
+      "loss": 0.4256,
+      "step": 4048
+    },
+    {
+      "epoch": 2.7228713120954864,
+      "grad_norm": 0.30607813581454435,
+      "learning_rate": 4.600846301305973e-06,
+      "loss": 0.4539,
+      "step": 4049
+    },
+    {
+      "epoch": 2.7235437505253426,
+      "grad_norm": 0.3282820415912827,
+      "learning_rate": 4.598651829871643e-06,
+      "loss": 0.4416,
+      "step": 4050
+    },
+    {
+      "epoch": 2.724216188955199,
+      "grad_norm": 0.324333649191008,
+      "learning_rate": 4.5964574362468925e-06,
+      "loss": 0.4425,
+      "step": 4051
+    },
+    {
+      "epoch": 2.724888627385055,
+      "grad_norm": 0.34630251064694667,
+      "learning_rate": 4.594263120857151e-06,
+      "loss": 0.451,
+      "step": 4052
+    },
+    {
+      "epoch": 2.725561065814911,
+      "grad_norm": 0.5041706542960473,
+      "learning_rate": 4.5920688841278315e-06,
+      "loss": 0.4337,
+      "step": 4053
+    },
+    {
+      "epoch": 2.7262335042447674,
+      "grad_norm": 0.35387917754802173,
+      "learning_rate": 4.58987472648433e-06,
+      "loss": 0.4352,
+      "step": 4054
+    },
+    {
+      "epoch": 2.726905942674624,
+      "grad_norm": 0.41760733333223016,
+      "learning_rate": 4.5876806483520305e-06,
+      "loss": 0.4278,
+      "step": 4055
+    },
+    {
+      "epoch": 2.7275783811044803,
+      "grad_norm": 0.3397217293758815,
+      "learning_rate": 4.585486650156299e-06,
+      "loss": 0.4257,
+      "step": 4056
+    },
+    {
+      "epoch": 2.7282508195343365,
+      "grad_norm": 0.36874491179924024,
+      "learning_rate": 4.58329273232249e-06,
+      "loss": 0.4438,
+      "step": 4057
+    },
+    {
+      "epoch": 2.7289232579641927,
+      "grad_norm": 0.3774561271535653,
+      "learning_rate": 4.581098895275935e-06,
+      "loss": 0.448,
+      "step": 4058
+    },
+    {
+      "epoch": 2.729595696394049,
+      "grad_norm": 0.33630915462567423,
+      "learning_rate": 4.578905139441957e-06,
+      "loss": 0.4407,
+      "step": 4059
+    },
+    {
+      "epoch": 2.730268134823905,
+      "grad_norm": 0.3126494069121454,
+      "learning_rate": 4.57671146524586e-06,
+      "loss": 0.4402,
+      "step": 4060
+    },
+    {
+      "epoch": 2.7309405732537613,
+      "grad_norm": 0.3582877240176243,
+      "learning_rate": 4.574517873112934e-06,
+      "loss": 0.445,
+      "step": 4061
+    },
+    {
+      "epoch": 2.731613011683618,
+      "grad_norm": 0.3211952898034876,
+      "learning_rate": 4.572324363468449e-06,
+      "loss": 0.4338,
+      "step": 4062
+    },
+    {
+      "epoch": 2.732285450113474,
+      "grad_norm": 0.3384746843259489,
+      "learning_rate": 4.570130936737664e-06,
+      "loss": 0.4574,
+      "step": 4063
+    },
+    {
+      "epoch": 2.7329578885433303,
+      "grad_norm": 0.38074821139775816,
+      "learning_rate": 4.56793759334582e-06,
+      "loss": 0.4278,
+      "step": 4064
+    },
+    {
+      "epoch": 2.7336303269731865,
+      "grad_norm": 0.3958921748871837,
+      "learning_rate": 4.565744333718141e-06,
+      "loss": 0.4445,
+      "step": 4065
+    },
+    {
+      "epoch": 2.7343027654030427,
+      "grad_norm": 0.3406667599130005,
+      "learning_rate": 4.563551158279837e-06,
+      "loss": 0.4668,
+      "step": 4066
+    },
+    {
+      "epoch": 2.734975203832899,
+      "grad_norm": 0.3326943427455027,
+      "learning_rate": 4.5613580674560964e-06,
+      "loss": 0.4531,
+      "step": 4067
+    },
+    {
+      "epoch": 2.735647642262755,
+      "grad_norm": 0.3079837541279532,
+      "learning_rate": 4.5591650616720975e-06,
+      "loss": 0.45,
+      "step": 4068
+    },
+    {
+      "epoch": 2.736320080692612,
+      "grad_norm": 0.37775753040507126,
+      "learning_rate": 4.556972141352999e-06,
+      "loss": 0.4235,
+      "step": 4069
+    },
+    {
+      "epoch": 2.736992519122468,
+      "grad_norm": 0.3191006036475625,
+      "learning_rate": 4.554779306923943e-06,
+      "loss": 0.4478,
+      "step": 4070
+    },
+    {
+      "epoch": 2.737664957552324,
+      "grad_norm": 0.3218784168108929,
+      "learning_rate": 4.5525865588100566e-06,
+      "loss": 0.4358,
+      "step": 4071
+    },
+    {
+      "epoch": 2.7383373959821804,
+      "grad_norm": 0.3934979110896771,
+      "learning_rate": 4.550393897436447e-06,
+      "loss": 0.4494,
+      "step": 4072
+    },
+    {
+      "epoch": 2.7390098344120366,
+      "grad_norm": 0.43781924786854065,
+      "learning_rate": 4.54820132322821e-06,
+      "loss": 0.4463,
+      "step": 4073
+    },
+    {
+      "epoch": 2.739682272841893,
+      "grad_norm": 0.3497296661145289,
+      "learning_rate": 4.546008836610417e-06,
+      "loss": 0.4373,
+      "step": 4074
+    },
+    {
+      "epoch": 2.740354711271749,
+      "grad_norm": 1.4186194019979426,
+      "learning_rate": 4.543816438008132e-06,
+      "loss": 0.449,
+      "step": 4075
+    },
+    {
+      "epoch": 2.7410271497016057,
+      "grad_norm": 0.39392153382582556,
+      "learning_rate": 4.54162412784639e-06,
+      "loss": 0.4452,
+      "step": 4076
+    },
+    {
+      "epoch": 2.741699588131462,
+      "grad_norm": 0.6244952823781512,
+      "learning_rate": 4.539431906550219e-06,
+      "loss": 0.451,
+      "step": 4077
+    },
+    {
+      "epoch": 2.742372026561318,
+      "grad_norm": 0.36142281857296416,
+      "learning_rate": 4.537239774544625e-06,
+      "loss": 0.4408,
+      "step": 4078
+    },
+    {
+      "epoch": 2.7430444649911743,
+      "grad_norm": 0.5173482600389665,
+      "learning_rate": 4.535047732254601e-06,
+      "loss": 0.4525,
+      "step": 4079
+    },
+    {
+      "epoch": 2.7437169034210305,
+      "grad_norm": 0.4456350759158602,
+      "learning_rate": 4.532855780105114e-06,
+      "loss": 0.4397,
+      "step": 4080
+    },
+    {
+      "epoch": 2.7443893418508867,
+      "grad_norm": 0.3776474889137343,
+      "learning_rate": 4.530663918521121e-06,
+      "loss": 0.4316,
+      "step": 4081
+    },
+    {
+      "epoch": 2.745061780280743,
+      "grad_norm": 0.4622116773312596,
+      "learning_rate": 4.528472147927561e-06,
+      "loss": 0.4318,
+      "step": 4082
+    },
+    {
+      "epoch": 2.7457342187105995,
+      "grad_norm": 0.42715846688985476,
+      "learning_rate": 4.526280468749355e-06,
+      "loss": 0.4494,
+      "step": 4083
+    },
+    {
+      "epoch": 2.7464066571404557,
+      "grad_norm": 0.33742122624911886,
+      "learning_rate": 4.5240888814114e-06,
+      "loss": 0.4285,
+      "step": 4084
+    },
+    {
+      "epoch": 2.747079095570312,
+      "grad_norm": 0.3831275524199066,
+      "learning_rate": 4.521897386338582e-06,
+      "loss": 0.4361,
+      "step": 4085
+    },
+    {
+      "epoch": 2.747751534000168,
+      "grad_norm": 0.34803028860558366,
+      "learning_rate": 4.5197059839557694e-06,
+      "loss": 0.4382,
+      "step": 4086
+    },
+    {
+      "epoch": 2.7484239724300243,
+      "grad_norm": 0.3358652039154297,
+      "learning_rate": 4.5175146746878086e-06,
+      "loss": 0.4402,
+      "step": 4087
+    },
+    {
+      "epoch": 2.7490964108598805,
+      "grad_norm": 0.3596305282095028,
+      "learning_rate": 4.515323458959532e-06,
+      "loss": 0.4499,
+      "step": 4088
+    },
+    {
+      "epoch": 2.7497688492897367,
+      "grad_norm": 0.3511275929318625,
+      "learning_rate": 4.513132337195748e-06,
+      "loss": 0.4516,
+      "step": 4089
+    },
+    {
+      "epoch": 2.7504412877195934,
+      "grad_norm": 0.31670305524414344,
+      "learning_rate": 4.510941309821254e-06,
+      "loss": 0.4397,
+      "step": 4090
+    },
+    {
+      "epoch": 2.7511137261494496,
+      "grad_norm": 0.32598957982308097,
+      "learning_rate": 4.508750377260824e-06,
+      "loss": 0.4392,
+      "step": 4091
+    },
+    {
+      "epoch": 2.751786164579306,
+      "grad_norm": 0.4135011005395795,
+      "learning_rate": 4.506559539939218e-06,
+      "loss": 0.4308,
+      "step": 4092
+    },
+    {
+      "epoch": 2.752458603009162,
+      "grad_norm": 0.35335225307008133,
+      "learning_rate": 4.50436879828117e-06,
+      "loss": 0.444,
+      "step": 4093
+    },
+    {
+      "epoch": 2.753131041439018,
+      "grad_norm": 0.339990579672587,
+      "learning_rate": 4.502178152711403e-06,
+      "loss": 0.4444,
+      "step": 4094
+    },
+    {
+      "epoch": 2.7538034798688744,
+      "grad_norm": 1.0034503263282213,
+      "learning_rate": 4.499987603654618e-06,
+      "loss": 0.4327,
+      "step": 4095
+    },
+    {
+      "epoch": 2.7544759182987306,
+      "grad_norm": 0.37428166492973364,
+      "learning_rate": 4.497797151535496e-06,
+      "loss": 0.4485,
+      "step": 4096
+    },
+    {
+      "epoch": 2.7551483567285873,
+      "grad_norm": 0.38282626112707663,
+      "learning_rate": 4.495606796778707e-06,
+      "loss": 0.4579,
+      "step": 4097
+    },
+    {
+      "epoch": 2.7558207951584435,
+      "grad_norm": 0.45749960797959477,
+      "learning_rate": 4.493416539808891e-06,
+      "loss": 0.4277,
+      "step": 4098
+    },
+    {
+      "epoch": 2.7564932335882997,
+      "grad_norm": 0.38810401286703083,
+      "learning_rate": 4.4912263810506765e-06,
+      "loss": 0.4263,
+      "step": 4099
+    },
+    {
+      "epoch": 2.757165672018156,
+      "grad_norm": 0.3944196941682623,
+      "learning_rate": 4.4890363209286685e-06,
+      "loss": 0.4492,
+      "step": 4100
+    },
+    {
+      "epoch": 2.757838110448012,
+      "grad_norm": 0.40649470910613944,
+      "learning_rate": 4.486846359867459e-06,
+      "loss": 0.4535,
+      "step": 4101
+    },
+    {
+      "epoch": 2.7585105488778683,
+      "grad_norm": 0.44185871010693245,
+      "learning_rate": 4.484656498291611e-06,
+      "loss": 0.4385,
+      "step": 4102
+    },
+    {
+      "epoch": 2.7591829873077245,
+      "grad_norm": 0.3637943966579024,
+      "learning_rate": 4.48246673662568e-06,
+      "loss": 0.4483,
+      "step": 4103
+    },
+    {
+      "epoch": 2.759855425737581,
+      "grad_norm": 0.37969279378764087,
+      "learning_rate": 4.480277075294192e-06,
+      "loss": 0.4399,
+      "step": 4104
+    },
+    {
+      "epoch": 2.760527864167437,
+      "grad_norm": 0.3138671963424064,
+      "learning_rate": 4.4780875147216625e-06,
+      "loss": 0.4369,
+      "step": 4105
+    },
+    {
+      "epoch": 2.7612003025972935,
+      "grad_norm": 0.3263125250207868,
+      "learning_rate": 4.475898055332578e-06,
+      "loss": 0.4395,
+      "step": 4106
+    },
+    {
+      "epoch": 2.7618727410271497,
+      "grad_norm": 0.5257770684119895,
+      "learning_rate": 4.473708697551411e-06,
+      "loss": 0.4347,
+      "step": 4107
+    },
+    {
+      "epoch": 2.762545179457006,
+      "grad_norm": 0.3248907365494214,
+      "learning_rate": 4.471519441802616e-06,
+      "loss": 0.4613,
+      "step": 4108
+    },
+    {
+      "epoch": 2.763217617886862,
+      "grad_norm": 0.32027788836591803,
+      "learning_rate": 4.469330288510622e-06,
+      "loss": 0.4398,
+      "step": 4109
+    },
+    {
+      "epoch": 2.7638900563167184,
+      "grad_norm": 0.3472818383721879,
+      "learning_rate": 4.4671412380998456e-06,
+      "loss": 0.45,
+      "step": 4110
+    },
+    {
+      "epoch": 2.764562494746575,
+      "grad_norm": 0.32326005111267875,
+      "learning_rate": 4.464952290994674e-06,
+      "loss": 0.4559,
+      "step": 4111
+    },
+    {
+      "epoch": 2.7652349331764308,
+      "grad_norm": 0.39181182928822017,
+      "learning_rate": 4.4627634476194826e-06,
+      "loss": 0.4388,
+      "step": 4112
+    },
+    {
+      "epoch": 2.7659073716062874,
+      "grad_norm": 0.31195199944135227,
+      "learning_rate": 4.4605747083986225e-06,
+      "loss": 0.4416,
+      "step": 4113
+    },
+    {
+      "epoch": 2.7665798100361436,
+      "grad_norm": 0.6106923654849684,
+      "learning_rate": 4.4583860737564285e-06,
+      "loss": 0.4526,
+      "step": 4114
+    },
+    {
+      "epoch": 2.767252248466,
+      "grad_norm": 0.3188243064392307,
+      "learning_rate": 4.456197544117208e-06,
+      "loss": 0.4384,
+      "step": 4115
+    },
+    {
+      "epoch": 2.767924686895856,
+      "grad_norm": 0.4046926030229404,
+      "learning_rate": 4.454009119905255e-06,
+      "loss": 0.4408,
+      "step": 4116
+    },
+    {
+      "epoch": 2.7685971253257122,
+      "grad_norm": 0.31559735785847665,
+      "learning_rate": 4.4518208015448385e-06,
+      "loss": 0.4272,
+      "step": 4117
+    },
+    {
+      "epoch": 2.769269563755569,
+      "grad_norm": 0.517474290715874,
+      "learning_rate": 4.449632589460212e-06,
+      "loss": 0.4266,
+      "step": 4118
+    },
+    {
+      "epoch": 2.7699420021854246,
+      "grad_norm": 0.5023251432358923,
+      "learning_rate": 4.447444484075603e-06,
+      "loss": 0.4456,
+      "step": 4119
+    },
+    {
+      "epoch": 2.7706144406152813,
+      "grad_norm": 0.32524193502553655,
+      "learning_rate": 4.445256485815222e-06,
+      "loss": 0.4409,
+      "step": 4120
+    },
+    {
+      "epoch": 2.7712868790451375,
+      "grad_norm": 0.35927764141343127,
+      "learning_rate": 4.4430685951032546e-06,
+      "loss": 0.4466,
+      "step": 4121
+    },
+    {
+      "epoch": 2.7719593174749937,
+      "grad_norm": 0.31890742397512795,
+      "learning_rate": 4.440880812363871e-06,
+      "loss": 0.4247,
+      "step": 4122
+    },
+    {
+      "epoch": 2.77263175590485,
+      "grad_norm": 0.3545582580996859,
+      "learning_rate": 4.43869313802122e-06,
+      "loss": 0.436,
+      "step": 4123
+    },
+    {
+      "epoch": 2.773304194334706,
+      "grad_norm": 0.4796722667815604,
+      "learning_rate": 4.436505572499422e-06,
+      "loss": 0.4489,
+      "step": 4124
+    },
+    {
+      "epoch": 2.7739766327645627,
+      "grad_norm": 0.30288716390684745,
+      "learning_rate": 4.434318116222583e-06,
+      "loss": 0.44,
+      "step": 4125
+    },
+    {
+      "epoch": 2.7746490711944185,
+      "grad_norm": 0.3279704264872385,
+      "learning_rate": 4.432130769614787e-06,
+      "loss": 0.4327,
+      "step": 4126
+    },
+    {
+      "epoch": 2.775321509624275,
+      "grad_norm": 0.45918742036048055,
+      "learning_rate": 4.429943533100098e-06,
+      "loss": 0.4224,
+      "step": 4127
+    },
+    {
+      "epoch": 2.7759939480541314,
+      "grad_norm": 0.3816536568297676,
+      "learning_rate": 4.427756407102554e-06,
+      "loss": 0.4453,
+      "step": 4128
+    },
+    {
+      "epoch": 2.7766663864839876,
+      "grad_norm": 0.3130350219354691,
+      "learning_rate": 4.425569392046174e-06,
+      "loss": 0.4357,
+      "step": 4129
+    },
+    {
+      "epoch": 2.7773388249138438,
+      "grad_norm": 0.31236615723388095,
+      "learning_rate": 4.423382488354957e-06,
+      "loss": 0.4327,
+      "step": 4130
+    },
+    {
+      "epoch": 2.7780112633437,
+      "grad_norm": 0.33840814755852966,
+      "learning_rate": 4.4211956964528795e-06,
+      "loss": 0.4528,
+      "step": 4131
+    },
+    {
+      "epoch": 2.7786837017735566,
+      "grad_norm": 0.3498544989130826,
+      "learning_rate": 4.4190090167638976e-06,
+      "loss": 0.436,
+      "step": 4132
+    },
+    {
+      "epoch": 2.7793561402034124,
+      "grad_norm": 0.359188442813231,
+      "learning_rate": 4.416822449711939e-06,
+      "loss": 0.4328,
+      "step": 4133
+    },
+    {
+      "epoch": 2.780028578633269,
+      "grad_norm": 0.4924888241936742,
+      "learning_rate": 4.414635995720918e-06,
+      "loss": 0.4611,
+      "step": 4134
+    },
+    {
+      "epoch": 2.7807010170631252,
+      "grad_norm": 0.3342595068619039,
+      "learning_rate": 4.4124496552147225e-06,
+      "loss": 0.4302,
+      "step": 4135
+    },
+    {
+      "epoch": 2.7813734554929814,
+      "grad_norm": 0.30432160255455765,
+      "learning_rate": 4.410263428617222e-06,
+      "loss": 0.4408,
+      "step": 4136
+    },
+    {
+      "epoch": 2.7820458939228376,
+      "grad_norm": 0.34090270895012315,
+      "learning_rate": 4.408077316352258e-06,
+      "loss": 0.4579,
+      "step": 4137
+    },
+    {
+      "epoch": 2.782718332352694,
+      "grad_norm": 0.342366852309439,
+      "learning_rate": 4.405891318843655e-06,
+      "loss": 0.4321,
+      "step": 4138
+    },
+    {
+      "epoch": 2.7833907707825505,
+      "grad_norm": 0.300326428255073,
+      "learning_rate": 4.403705436515212e-06,
+      "loss": 0.4386,
+      "step": 4139
+    },
+    {
+      "epoch": 2.7840632092124062,
+      "grad_norm": 0.4374188529908497,
+      "learning_rate": 4.401519669790709e-06,
+      "loss": 0.4617,
+      "step": 4140
+    },
+    {
+      "epoch": 2.784735647642263,
+      "grad_norm": 0.36903156045600943,
+      "learning_rate": 4.399334019093904e-06,
+      "loss": 0.4501,
+      "step": 4141
+    },
+    {
+      "epoch": 2.785408086072119,
+      "grad_norm": 0.3469467309885719,
+      "learning_rate": 4.3971484848485245e-06,
+      "loss": 0.4469,
+      "step": 4142
+    },
+    {
+      "epoch": 2.7860805245019753,
+      "grad_norm": 0.3529204570178575,
+      "learning_rate": 4.394963067478283e-06,
+      "loss": 0.4364,
+      "step": 4143
+    },
+    {
+      "epoch": 2.7867529629318315,
+      "grad_norm": 0.3156208007068114,
+      "learning_rate": 4.39277776740687e-06,
+      "loss": 0.4405,
+      "step": 4144
+    },
+    {
+      "epoch": 2.7874254013616877,
+      "grad_norm": 0.3589296480700081,
+      "learning_rate": 4.390592585057949e-06,
+      "loss": 0.444,
+      "step": 4145
+    },
+    {
+      "epoch": 2.7880978397915444,
+      "grad_norm": 0.4860533872791004,
+      "learning_rate": 4.388407520855162e-06,
+      "loss": 0.4198,
+      "step": 4146
+    },
+    {
+      "epoch": 2.7887702782214,
+      "grad_norm": 0.31359406004259816,
+      "learning_rate": 4.38622257522213e-06,
+      "loss": 0.444,
+      "step": 4147
+    },
+    {
+      "epoch": 2.7894427166512568,
+      "grad_norm": 0.3628458626927062,
+      "learning_rate": 4.384037748582448e-06,
+      "loss": 0.4365,
+      "step": 4148
+    },
+    {
+      "epoch": 2.790115155081113,
+      "grad_norm": 0.3971586747602741,
+      "learning_rate": 4.381853041359693e-06,
+      "loss": 0.4303,
+      "step": 4149
+    },
+    {
+      "epoch": 2.790787593510969,
+      "grad_norm": 0.3762934083513401,
+      "learning_rate": 4.37966845397741e-06,
+      "loss": 0.4448,
+      "step": 4150
+    },
+    {
+      "epoch": 2.7914600319408254,
+      "grad_norm": 0.3429231449205495,
+      "learning_rate": 4.377483986859128e-06,
+      "loss": 0.442,
+      "step": 4151
+    },
+    {
+      "epoch": 2.7921324703706816,
+      "grad_norm": 0.5394162778507079,
+      "learning_rate": 4.375299640428352e-06,
+      "loss": 0.4355,
+      "step": 4152
+    },
+    {
+      "epoch": 2.792804908800538,
+      "grad_norm": 0.32743745427999177,
+      "learning_rate": 4.373115415108563e-06,
+      "loss": 0.4447,
+      "step": 4153
+    },
+    {
+      "epoch": 2.793477347230394,
+      "grad_norm": 0.4000900373043025,
+      "learning_rate": 4.370931311323217e-06,
+      "loss": 0.4482,
+      "step": 4154
+    },
+    {
+      "epoch": 2.7941497856602506,
+      "grad_norm": 0.35385970730396854,
+      "learning_rate": 4.368747329495747e-06,
+      "loss": 0.4324,
+      "step": 4155
+    },
+    {
+      "epoch": 2.794822224090107,
+      "grad_norm": 0.3965033179606159,
+      "learning_rate": 4.366563470049561e-06,
+      "loss": 0.443,
+      "step": 4156
+    },
+    {
+      "epoch": 2.795494662519963,
+      "grad_norm": 0.34703061683707376,
+      "learning_rate": 4.364379733408048e-06,
+      "loss": 0.4407,
+      "step": 4157
+    },
+    {
+      "epoch": 2.7961671009498192,
+      "grad_norm": 0.31031118789611056,
+      "learning_rate": 4.36219611999457e-06,
+      "loss": 0.4496,
+      "step": 4158
+    },
+    {
+      "epoch": 2.7968395393796754,
+      "grad_norm": 0.356068190864218,
+      "learning_rate": 4.360012630232463e-06,
+      "loss": 0.4386,
+      "step": 4159
+    },
+    {
+      "epoch": 2.797511977809532,
+      "grad_norm": 0.3053172018438645,
+      "learning_rate": 4.357829264545042e-06,
+      "loss": 0.4418,
+      "step": 4160
+    },
+    {
+      "epoch": 2.798184416239388,
+      "grad_norm": 0.34260686402603957,
+      "learning_rate": 4.355646023355599e-06,
+      "loss": 0.4234,
+      "step": 4161
+    },
+    {
+      "epoch": 2.7988568546692445,
+      "grad_norm": 0.37957637036473135,
+      "learning_rate": 4.353462907087397e-06,
+      "loss": 0.4349,
+      "step": 4162
+    },
+    {
+      "epoch": 2.7995292930991007,
+      "grad_norm": 0.41495400538925514,
+      "learning_rate": 4.351279916163684e-06,
+      "loss": 0.4217,
+      "step": 4163
+    },
+    {
+      "epoch": 2.800201731528957,
+      "grad_norm": 0.39348273057249017,
+      "learning_rate": 4.34909705100767e-06,
+      "loss": 0.4421,
+      "step": 4164
+    },
+    {
+      "epoch": 2.800874169958813,
+      "grad_norm": 0.32721102101693095,
+      "learning_rate": 4.346914312042552e-06,
+      "loss": 0.4373,
+      "step": 4165
+    },
+    {
+      "epoch": 2.8015466083886693,
+      "grad_norm": 0.4929749225276872,
+      "learning_rate": 4.344731699691498e-06,
+      "loss": 0.4325,
+      "step": 4166
+    },
+    {
+      "epoch": 2.8022190468185255,
+      "grad_norm": 0.3944224671487195,
+      "learning_rate": 4.342549214377653e-06,
+      "loss": 0.4428,
+      "step": 4167
+    },
+    {
+      "epoch": 2.8028914852483817,
+      "grad_norm": 0.32805134167892863,
+      "learning_rate": 4.340366856524137e-06,
+      "loss": 0.4339,
+      "step": 4168
+    },
+    {
+      "epoch": 2.8035639236782384,
+      "grad_norm": 0.3571698135102629,
+      "learning_rate": 4.338184626554043e-06,
+      "loss": 0.448,
+      "step": 4169
+    },
+    {
+      "epoch": 2.8042363621080946,
+      "grad_norm": 1.1019836197426351,
+      "learning_rate": 4.336002524890442e-06,
+      "loss": 0.4559,
+      "step": 4170
+    },
+    {
+      "epoch": 2.8049088005379508,
+      "grad_norm": 0.5615591083474871,
+      "learning_rate": 4.333820551956381e-06,
+      "loss": 0.4432,
+      "step": 4171
+    },
+    {
+      "epoch": 2.805581238967807,
+      "grad_norm": 0.42330074239721943,
+      "learning_rate": 4.331638708174877e-06,
+      "loss": 0.4495,
+      "step": 4172
+    },
+    {
+      "epoch": 2.806253677397663,
+      "grad_norm": 0.41645460574901194,
+      "learning_rate": 4.329456993968926e-06,
+      "loss": 0.431,
+      "step": 4173
+    },
+    {
+      "epoch": 2.8069261158275194,
+      "grad_norm": 0.3709084412027702,
+      "learning_rate": 4.327275409761499e-06,
+      "loss": 0.4378,
+      "step": 4174
+    },
+    {
+      "epoch": 2.8075985542573756,
+      "grad_norm": 0.353736146094937,
+      "learning_rate": 4.325093955975541e-06,
+      "loss": 0.4446,
+      "step": 4175
+    },
+    {
+      "epoch": 2.8082709926872322,
+      "grad_norm": 0.3818149149271061,
+      "learning_rate": 4.322912633033972e-06,
+      "loss": 0.4341,
+      "step": 4176
+    },
+    {
+      "epoch": 2.8089434311170884,
+      "grad_norm": 0.4666823798883692,
+      "learning_rate": 4.320731441359684e-06,
+      "loss": 0.4504,
+      "step": 4177
+    },
+    {
+      "epoch": 2.8096158695469446,
+      "grad_norm": 0.3905860953030715,
+      "learning_rate": 4.318550381375548e-06,
+      "loss": 0.4313,
+      "step": 4178
+    },
+    {
+      "epoch": 2.810288307976801,
+      "grad_norm": 0.4089372449811153,
+      "learning_rate": 4.3163694535044055e-06,
+      "loss": 0.4312,
+      "step": 4179
+    },
+    {
+      "epoch": 2.810960746406657,
+      "grad_norm": 0.37258055042296084,
+      "learning_rate": 4.314188658169078e-06,
+      "loss": 0.4441,
+      "step": 4180
+    },
+    {
+      "epoch": 2.8116331848365133,
+      "grad_norm": 0.35262230137425327,
+      "learning_rate": 4.312007995792353e-06,
+      "loss": 0.4423,
+      "step": 4181
+    },
+    {
+      "epoch": 2.8123056232663695,
+      "grad_norm": 0.4462012393921559,
+      "learning_rate": 4.309827466796996e-06,
+      "loss": 0.4647,
+      "step": 4182
+    },
+    {
+      "epoch": 2.812978061696226,
+      "grad_norm": 0.423231503154175,
+      "learning_rate": 4.307647071605751e-06,
+      "loss": 0.4494,
+      "step": 4183
+    },
+    {
+      "epoch": 2.8136505001260823,
+      "grad_norm": 0.37000476204116944,
+      "learning_rate": 4.30546681064133e-06,
+      "loss": 0.4381,
+      "step": 4184
+    },
+    {
+      "epoch": 2.8143229385559385,
+      "grad_norm": 0.4382581644812144,
+      "learning_rate": 4.303286684326424e-06,
+      "loss": 0.4331,
+      "step": 4185
+    },
+    {
+      "epoch": 2.8149953769857947,
+      "grad_norm": 0.3426587641141498,
+      "learning_rate": 4.301106693083692e-06,
+      "loss": 0.4344,
+      "step": 4186
+    },
+    {
+      "epoch": 2.815667815415651,
+      "grad_norm": 0.4400919829618176,
+      "learning_rate": 4.298926837335772e-06,
+      "loss": 0.4499,
+      "step": 4187
+    },
+    {
+      "epoch": 2.816340253845507,
+      "grad_norm": 0.43886870253233085,
+      "learning_rate": 4.2967471175052725e-06,
+      "loss": 0.4469,
+      "step": 4188
+    },
+    {
+      "epoch": 2.8170126922753633,
+      "grad_norm": 0.35837375545485195,
+      "learning_rate": 4.2945675340147796e-06,
+      "loss": 0.4345,
+      "step": 4189
+    },
+    {
+      "epoch": 2.81768513070522,
+      "grad_norm": 0.3634353514020071,
+      "learning_rate": 4.292388087286846e-06,
+      "loss": 0.431,
+      "step": 4190
+    },
+    {
+      "epoch": 2.818357569135076,
+      "grad_norm": 0.3444790616942859,
+      "learning_rate": 4.290208777744006e-06,
+      "loss": 0.4385,
+      "step": 4191
+    },
+    {
+      "epoch": 2.8190300075649324,
+      "grad_norm": 0.4673667498158592,
+      "learning_rate": 4.288029605808761e-06,
+      "loss": 0.4545,
+      "step": 4192
+    },
+    {
+      "epoch": 2.8197024459947886,
+      "grad_norm": 0.4901550150844472,
+      "learning_rate": 4.28585057190359e-06,
+      "loss": 0.4252,
+      "step": 4193
+    },
+    {
+      "epoch": 2.820374884424645,
+      "grad_norm": 0.34004725640003464,
+      "learning_rate": 4.283671676450941e-06,
+      "loss": 0.4339,
+      "step": 4194
+    },
+    {
+      "epoch": 2.821047322854501,
+      "grad_norm": 0.37589666520860127,
+      "learning_rate": 4.28149291987324e-06,
+      "loss": 0.4454,
+      "step": 4195
+    },
+    {
+      "epoch": 2.821719761284357,
+      "grad_norm": 0.43697203610507124,
+      "learning_rate": 4.2793143025928835e-06,
+      "loss": 0.4246,
+      "step": 4196
+    },
+    {
+      "epoch": 2.822392199714214,
+      "grad_norm": 0.32464672726116545,
+      "learning_rate": 4.27713582503224e-06,
+      "loss": 0.4458,
+      "step": 4197
+    },
+    {
+      "epoch": 2.82306463814407,
+      "grad_norm": 0.362476088230034,
+      "learning_rate": 4.274957487613654e-06,
+      "loss": 0.4426,
+      "step": 4198
+    },
+    {
+      "epoch": 2.8237370765739263,
+      "grad_norm": 0.3578027590759422,
+      "learning_rate": 4.272779290759439e-06,
+      "loss": 0.424,
+      "step": 4199
+    },
+    {
+      "epoch": 2.8244095150037825,
+      "grad_norm": 0.742708640591173,
+      "learning_rate": 4.270601234891883e-06,
+      "loss": 0.4413,
+      "step": 4200
+    },
+    {
+      "epoch": 2.8250819534336387,
+      "grad_norm": 0.3279396144376603,
+      "learning_rate": 4.268423320433247e-06,
+      "loss": 0.4392,
+      "step": 4201
+    },
+    {
+      "epoch": 2.825754391863495,
+      "grad_norm": 0.35208092982367156,
+      "learning_rate": 4.266245547805767e-06,
+      "loss": 0.4416,
+      "step": 4202
+    },
+    {
+      "epoch": 2.826426830293351,
+      "grad_norm": 0.36251601382145465,
+      "learning_rate": 4.264067917431645e-06,
+      "loss": 0.4561,
+      "step": 4203
+    },
+    {
+      "epoch": 2.8270992687232077,
+      "grad_norm": 0.4306499066372681,
+      "learning_rate": 4.261890429733063e-06,
+      "loss": 0.4324,
+      "step": 4204
+    },
+    {
+      "epoch": 2.827771707153064,
+      "grad_norm": 0.35960347296205575,
+      "learning_rate": 4.2597130851321685e-06,
+      "loss": 0.4246,
+      "step": 4205
+    },
+    {
+      "epoch": 2.82844414558292,
+      "grad_norm": 0.3312427958990518,
+      "learning_rate": 4.257535884051086e-06,
+      "loss": 0.4662,
+      "step": 4206
+    },
+    {
+      "epoch": 2.8291165840127763,
+      "grad_norm": 0.3127332077974222,
+      "learning_rate": 4.255358826911912e-06,
+      "loss": 0.4256,
+      "step": 4207
+    },
+    {
+      "epoch": 2.8297890224426325,
+      "grad_norm": 0.36043322840857506,
+      "learning_rate": 4.253181914136711e-06,
+      "loss": 0.4442,
+      "step": 4208
+    },
+    {
+      "epoch": 2.8304614608724887,
+      "grad_norm": 0.3183818541164439,
+      "learning_rate": 4.251005146147522e-06,
+      "loss": 0.4271,
+      "step": 4209
+    },
+    {
+      "epoch": 2.831133899302345,
+      "grad_norm": 0.4745175769316324,
+      "learning_rate": 4.248828523366359e-06,
+      "loss": 0.4312,
+      "step": 4210
+    },
+    {
+      "epoch": 2.8318063377322016,
+      "grad_norm": 0.33201401995383073,
+      "learning_rate": 4.246652046215206e-06,
+      "loss": 0.453,
+      "step": 4211
+    },
+    {
+      "epoch": 2.832478776162058,
+      "grad_norm": 0.324963818023879,
+      "learning_rate": 4.244475715116012e-06,
+      "loss": 0.4339,
+      "step": 4212
+    },
+    {
+      "epoch": 2.833151214591914,
+      "grad_norm": 0.3568436731790368,
+      "learning_rate": 4.242299530490708e-06,
+      "loss": 0.4498,
+      "step": 4213
+    },
+    {
+      "epoch": 2.83382365302177,
+      "grad_norm": 0.4182822403247682,
+      "learning_rate": 4.24012349276119e-06,
+      "loss": 0.4609,
+      "step": 4214
+    },
+    {
+      "epoch": 2.8344960914516264,
+      "grad_norm": 3.131545947607575,
+      "learning_rate": 4.23794760234933e-06,
+      "loss": 0.4352,
+      "step": 4215
+    },
+    {
+      "epoch": 2.8351685298814826,
+      "grad_norm": 0.4583669298743143,
+      "learning_rate": 4.2357718596769655e-06,
+      "loss": 0.4469,
+      "step": 4216
+    },
+    {
+      "epoch": 2.835840968311339,
+      "grad_norm": 0.3794591328890642,
+      "learning_rate": 4.233596265165912e-06,
+      "loss": 0.4319,
+      "step": 4217
+    },
+    {
+      "epoch": 2.8365134067411955,
+      "grad_norm": 0.4474970862074643,
+      "learning_rate": 4.231420819237951e-06,
+      "loss": 0.4468,
+      "step": 4218
+    },
+    {
+      "epoch": 2.8371858451710517,
+      "grad_norm": 0.4264990089485767,
+      "learning_rate": 4.229245522314839e-06,
+      "loss": 0.4457,
+      "step": 4219
+    },
+    {
+      "epoch": 2.837858283600908,
+      "grad_norm": 0.32880018687738627,
+      "learning_rate": 4.2270703748183025e-06,
+      "loss": 0.4394,
+      "step": 4220
+    },
+    {
+      "epoch": 2.838530722030764,
+      "grad_norm": 0.4347059065127276,
+      "learning_rate": 4.2248953771700356e-06,
+      "loss": 0.45,
+      "step": 4221
+    },
+    {
+      "epoch": 2.8392031604606203,
+      "grad_norm": 0.3539200637449874,
+      "learning_rate": 4.222720529791708e-06,
+      "loss": 0.44,
+      "step": 4222
+    },
+    {
+      "epoch": 2.8398755988904765,
+      "grad_norm": 0.31172589745609186,
+      "learning_rate": 4.220545833104958e-06,
+      "loss": 0.4481,
+      "step": 4223
+    },
+    {
+      "epoch": 2.8405480373203327,
+      "grad_norm": 0.35007131216723775,
+      "learning_rate": 4.218371287531397e-06,
+      "loss": 0.4293,
+      "step": 4224
+    },
+    {
+      "epoch": 2.8412204757501893,
+      "grad_norm": 0.449397329929107,
+      "learning_rate": 4.216196893492602e-06,
+      "loss": 0.4361,
+      "step": 4225
+    },
+    {
+      "epoch": 2.8418929141800455,
+      "grad_norm": 0.4494250507992524,
+      "learning_rate": 4.2140226514101265e-06,
+      "loss": 0.4327,
+      "step": 4226
+    },
+    {
+      "epoch": 2.8425653526099017,
+      "grad_norm": 0.328475017212245,
+      "learning_rate": 4.2118485617054916e-06,
+      "loss": 0.4507,
+      "step": 4227
+    },
+    {
+      "epoch": 2.843237791039758,
+      "grad_norm": 0.4306809663698929,
+      "learning_rate": 4.20967462480019e-06,
+      "loss": 0.4374,
+      "step": 4228
+    },
+    {
+      "epoch": 2.843910229469614,
+      "grad_norm": 0.35037605536830907,
+      "learning_rate": 4.207500841115682e-06,
+      "loss": 0.4409,
+      "step": 4229
+    },
+    {
+      "epoch": 2.8445826678994703,
+      "grad_norm": 0.33765625836092356,
+      "learning_rate": 4.205327211073399e-06,
+      "loss": 0.4381,
+      "step": 4230
+    },
+    {
+      "epoch": 2.8452551063293265,
+      "grad_norm": 0.5551967692371169,
+      "learning_rate": 4.203153735094747e-06,
+      "loss": 0.4328,
+      "step": 4231
+    },
+    {
+      "epoch": 2.845927544759183,
+      "grad_norm": 0.40994218394479104,
+      "learning_rate": 4.200980413601097e-06,
+      "loss": 0.4471,
+      "step": 4232
+    },
+    {
+      "epoch": 2.8465999831890394,
+      "grad_norm": 0.37418125193428436,
+      "learning_rate": 4.198807247013794e-06,
+      "loss": 0.4518,
+      "step": 4233
+    },
+    {
+      "epoch": 2.8472724216188956,
+      "grad_norm": 0.41210781906159627,
+      "learning_rate": 4.1966342357541486e-06,
+      "loss": 0.4363,
+      "step": 4234
+    },
+    {
+      "epoch": 2.847944860048752,
+      "grad_norm": 0.4155183764811362,
+      "learning_rate": 4.194461380243445e-06,
+      "loss": 0.4373,
+      "step": 4235
+    },
+    {
+      "epoch": 2.848617298478608,
+      "grad_norm": 0.3669197743717589,
+      "learning_rate": 4.192288680902935e-06,
+      "loss": 0.4453,
+      "step": 4236
+    },
+    {
+      "epoch": 2.849289736908464,
+      "grad_norm": 0.425957665347839,
+      "learning_rate": 4.190116138153844e-06,
+      "loss": 0.4562,
+      "step": 4237
+    },
+    {
+      "epoch": 2.8499621753383204,
+      "grad_norm": 0.3404491078149701,
+      "learning_rate": 4.187943752417359e-06,
+      "loss": 0.4301,
+      "step": 4238
+    },
+    {
+      "epoch": 2.850634613768177,
+      "grad_norm": 0.3677165947890313,
+      "learning_rate": 4.185771524114644e-06,
+      "loss": 0.4263,
+      "step": 4239
+    },
+    {
+      "epoch": 2.8513070521980333,
+      "grad_norm": 0.31412292626872296,
+      "learning_rate": 4.18359945366683e-06,
+      "loss": 0.4299,
+      "step": 4240
+    },
+    {
+      "epoch": 2.8519794906278895,
+      "grad_norm": 0.3716609911042492,
+      "learning_rate": 4.181427541495018e-06,
+      "loss": 0.4427,
+      "step": 4241
+    },
+    {
+      "epoch": 2.8526519290577457,
+      "grad_norm": 0.3518355157188309,
+      "learning_rate": 4.179255788020278e-06,
+      "loss": 0.4342,
+      "step": 4242
+    },
+    {
+      "epoch": 2.853324367487602,
+      "grad_norm": 0.35252230576663934,
+      "learning_rate": 4.177084193663647e-06,
+      "loss": 0.4337,
+      "step": 4243
+    },
+    {
+      "epoch": 2.853996805917458,
+      "grad_norm": 0.3379641016136965,
+      "learning_rate": 4.174912758846134e-06,
+      "loss": 0.4207,
+      "step": 4244
+    },
+    {
+      "epoch": 2.8546692443473143,
+      "grad_norm": 0.3524487510511978,
+      "learning_rate": 4.172741483988717e-06,
+      "loss": 0.4353,
+      "step": 4245
+    },
+    {
+      "epoch": 2.855341682777171,
+      "grad_norm": 0.40793087769440395,
+      "learning_rate": 4.170570369512344e-06,
+      "loss": 0.4253,
+      "step": 4246
+    },
+    {
+      "epoch": 2.856014121207027,
+      "grad_norm": 0.422585208007085,
+      "learning_rate": 4.1683994158379255e-06,
+      "loss": 0.4295,
+      "step": 4247
+    },
+    {
+      "epoch": 2.8566865596368833,
+      "grad_norm": 0.45569305713966374,
+      "learning_rate": 4.166228623386347e-06,
+      "loss": 0.4325,
+      "step": 4248
+    },
+    {
+      "epoch": 2.8573589980667395,
+      "grad_norm": 0.5739734252923965,
+      "learning_rate": 4.164057992578463e-06,
+      "loss": 0.4316,
+      "step": 4249
+    },
+    {
+      "epoch": 2.8580314364965957,
+      "grad_norm": 0.38539807899996864,
+      "learning_rate": 4.1618875238350945e-06,
+      "loss": 0.4518,
+      "step": 4250
+    },
+    {
+      "epoch": 2.858703874926452,
+      "grad_norm": 0.4344583117890214,
+      "learning_rate": 4.15971721757703e-06,
+      "loss": 0.435,
+      "step": 4251
+    },
+    {
+      "epoch": 2.859376313356308,
+      "grad_norm": 0.37028539101648733,
+      "learning_rate": 4.157547074225028e-06,
+      "loss": 0.4475,
+      "step": 4252
+    },
+    {
+      "epoch": 2.860048751786165,
+      "grad_norm": 0.3612453578542226,
+      "learning_rate": 4.155377094199816e-06,
+      "loss": 0.4343,
+      "step": 4253
+    },
+    {
+      "epoch": 2.860721190216021,
+      "grad_norm": 0.46991632221235885,
+      "learning_rate": 4.153207277922089e-06,
+      "loss": 0.433,
+      "step": 4254
+    },
+    {
+      "epoch": 2.861393628645877,
+      "grad_norm": 0.34980489848240764,
+      "learning_rate": 4.151037625812513e-06,
+      "loss": 0.4365,
+      "step": 4255
+    },
+    {
+      "epoch": 2.8620660670757334,
+      "grad_norm": 0.3750905114580679,
+      "learning_rate": 4.148868138291714e-06,
+      "loss": 0.4365,
+      "step": 4256
+    },
+    {
+      "epoch": 2.8627385055055896,
+      "grad_norm": 0.38187841266966155,
+      "learning_rate": 4.146698815780295e-06,
+      "loss": 0.4443,
+      "step": 4257
+    },
+    {
+      "epoch": 2.863410943935446,
+      "grad_norm": 0.33677571728502814,
+      "learning_rate": 4.144529658698824e-06,
+      "loss": 0.4423,
+      "step": 4258
+    },
+    {
+      "epoch": 2.864083382365302,
+      "grad_norm": 0.402470212628385,
+      "learning_rate": 4.142360667467838e-06,
+      "loss": 0.4259,
+      "step": 4259
+    },
+    {
+      "epoch": 2.8647558207951587,
+      "grad_norm": 0.6996119095805915,
+      "learning_rate": 4.1401918425078365e-06,
+      "loss": 0.4397,
+      "step": 4260
+    },
+    {
+      "epoch": 2.865428259225015,
+      "grad_norm": 0.32779799605137916,
+      "learning_rate": 4.138023184239294e-06,
+      "loss": 0.4488,
+      "step": 4261
+    },
+    {
+      "epoch": 2.866100697654871,
+      "grad_norm": 0.31941981810706765,
+      "learning_rate": 4.135854693082646e-06,
+      "loss": 0.4242,
+      "step": 4262
+    },
+    {
+      "epoch": 2.8667731360847273,
+      "grad_norm": 0.4026405587453875,
+      "learning_rate": 4.133686369458303e-06,
+      "loss": 0.437,
+      "step": 4263
+    },
+    {
+      "epoch": 2.8674455745145835,
+      "grad_norm": 0.3102554702479224,
+      "learning_rate": 4.131518213786638e-06,
+      "loss": 0.453,
+      "step": 4264
+    },
+    {
+      "epoch": 2.8681180129444397,
+      "grad_norm": 0.4258431129398313,
+      "learning_rate": 4.1293502264879895e-06,
+      "loss": 0.4534,
+      "step": 4265
+    },
+    {
+      "epoch": 2.868790451374296,
+      "grad_norm": 0.6528324085825112,
+      "learning_rate": 4.1271824079826695e-06,
+      "loss": 0.4371,
+      "step": 4266
+    },
+    {
+      "epoch": 2.8694628898041525,
+      "grad_norm": 0.376635634239984,
+      "learning_rate": 4.1250147586909534e-06,
+      "loss": 0.4323,
+      "step": 4267
+    },
+    {
+      "epoch": 2.8701353282340083,
+      "grad_norm": 0.45667317364587295,
+      "learning_rate": 4.122847279033086e-06,
+      "loss": 0.4647,
+      "step": 4268
+    },
+    {
+      "epoch": 2.870807766663865,
+      "grad_norm": 0.40142667751818195,
+      "learning_rate": 4.120679969429274e-06,
+      "loss": 0.4471,
+      "step": 4269
+    },
+    {
+      "epoch": 2.871480205093721,
+      "grad_norm": 0.39789184379035186,
+      "learning_rate": 4.118512830299696e-06,
+      "loss": 0.4405,
+      "step": 4270
+    },
+    {
+      "epoch": 2.8721526435235774,
+      "grad_norm": 0.34470586336361553,
+      "learning_rate": 4.116345862064498e-06,
+      "loss": 0.4551,
+      "step": 4271
+    },
+    {
+      "epoch": 2.8728250819534336,
+      "grad_norm": 0.3288657110198987,
+      "learning_rate": 4.114179065143792e-06,
+      "loss": 0.4506,
+      "step": 4272
+    },
+    {
+      "epoch": 2.8734975203832898,
+      "grad_norm": 0.316861492206836,
+      "learning_rate": 4.112012439957653e-06,
+      "loss": 0.4336,
+      "step": 4273
+    },
+    {
+      "epoch": 2.8741699588131464,
+      "grad_norm": 0.3757760463793481,
+      "learning_rate": 4.109845986926127e-06,
+      "loss": 0.4187,
+      "step": 4274
+    },
+    {
+      "epoch": 2.874842397243002,
+      "grad_norm": 0.36283664838320717,
+      "learning_rate": 4.107679706469226e-06,
+      "loss": 0.4286,
+      "step": 4275
+    },
+    {
+      "epoch": 2.875514835672859,
+      "grad_norm": 0.3972832536818271,
+      "learning_rate": 4.105513599006927e-06,
+      "loss": 0.4417,
+      "step": 4276
+    },
+    {
+      "epoch": 2.876187274102715,
+      "grad_norm": 0.3395551214278122,
+      "learning_rate": 4.103347664959179e-06,
+      "loss": 0.447,
+      "step": 4277
+    },
+    {
+      "epoch": 2.8768597125325712,
+      "grad_norm": 0.3844461209754712,
+      "learning_rate": 4.101181904745885e-06,
+      "loss": 0.4556,
+      "step": 4278
+    },
+    {
+      "epoch": 2.8775321509624274,
+      "grad_norm": 0.3931268402696742,
+      "learning_rate": 4.099016318786926e-06,
+      "loss": 0.4535,
+      "step": 4279
+    },
+    {
+      "epoch": 2.8782045893922836,
+      "grad_norm": 0.3069355737860178,
+      "learning_rate": 4.096850907502145e-06,
+      "loss": 0.4391,
+      "step": 4280
+    },
+    {
+      "epoch": 2.8788770278221403,
+      "grad_norm": 1.1404325595603302,
+      "learning_rate": 4.094685671311353e-06,
+      "loss": 0.4477,
+      "step": 4281
+    },
+    {
+      "epoch": 2.879549466251996,
+      "grad_norm": 0.3256061095754914,
+      "learning_rate": 4.092520610634323e-06,
+      "loss": 0.435,
+      "step": 4282
+    },
+    {
+      "epoch": 2.8802219046818527,
+      "grad_norm": 0.3556430518345657,
+      "learning_rate": 4.0903557258907975e-06,
+      "loss": 0.4357,
+      "step": 4283
+    },
+    {
+      "epoch": 2.880894343111709,
+      "grad_norm": 0.4232484835520356,
+      "learning_rate": 4.088191017500484e-06,
+      "loss": 0.4407,
+      "step": 4284
+    },
+    {
+      "epoch": 2.881566781541565,
+      "grad_norm": 0.5641353968555147,
+      "learning_rate": 4.0860264858830545e-06,
+      "loss": 0.4343,
+      "step": 4285
+    },
+    {
+      "epoch": 2.8822392199714213,
+      "grad_norm": 0.35668894593439726,
+      "learning_rate": 4.083862131458152e-06,
+      "loss": 0.4384,
+      "step": 4286
+    },
+    {
+      "epoch": 2.8829116584012775,
+      "grad_norm": 0.5268429700767152,
+      "learning_rate": 4.081697954645375e-06,
+      "loss": 0.4454,
+      "step": 4287
+    },
+    {
+      "epoch": 2.883584096831134,
+      "grad_norm": 0.3592019898868432,
+      "learning_rate": 4.0795339558642964e-06,
+      "loss": 0.43,
+      "step": 4288
+    },
+    {
+      "epoch": 2.88425653526099,
+      "grad_norm": 0.3355202443504104,
+      "learning_rate": 4.077370135534452e-06,
+      "loss": 0.4256,
+      "step": 4289
+    },
+    {
+      "epoch": 2.8849289736908466,
+      "grad_norm": 0.35761333702572906,
+      "learning_rate": 4.075206494075344e-06,
+      "loss": 0.4093,
+      "step": 4290
+    },
+    {
+      "epoch": 2.8856014121207028,
+      "grad_norm": 0.36261425680219084,
+      "learning_rate": 4.073043031906437e-06,
+      "loss": 0.434,
+      "step": 4291
+    },
+    {
+      "epoch": 2.886273850550559,
+      "grad_norm": 0.48381493475754367,
+      "learning_rate": 4.070879749447162e-06,
+      "loss": 0.4384,
+      "step": 4292
+    },
+    {
+      "epoch": 2.886946288980415,
+      "grad_norm": 0.3368170714364876,
+      "learning_rate": 4.0687166471169156e-06,
+      "loss": 0.4339,
+      "step": 4293
+    },
+    {
+      "epoch": 2.8876187274102714,
+      "grad_norm": 0.37122650595240925,
+      "learning_rate": 4.066553725335064e-06,
+      "loss": 0.4456,
+      "step": 4294
+    },
+    {
+      "epoch": 2.888291165840128,
+      "grad_norm": 0.472617643332299,
+      "learning_rate": 4.0643909845209286e-06,
+      "loss": 0.4395,
+      "step": 4295
+    },
+    {
+      "epoch": 2.8889636042699838,
+      "grad_norm": 0.4182466531187184,
+      "learning_rate": 4.062228425093802e-06,
+      "loss": 0.4311,
+      "step": 4296
+    },
+    {
+      "epoch": 2.8896360426998404,
+      "grad_norm": 0.4305460750947574,
+      "learning_rate": 4.060066047472942e-06,
+      "loss": 0.4396,
+      "step": 4297
+    },
+    {
+      "epoch": 2.8903084811296966,
+      "grad_norm": 0.33529669856356104,
+      "learning_rate": 4.05790385207757e-06,
+      "loss": 0.4318,
+      "step": 4298
+    },
+    {
+      "epoch": 2.890980919559553,
+      "grad_norm": 0.31791553757939606,
+      "learning_rate": 4.055741839326872e-06,
+      "loss": 0.4511,
+      "step": 4299
+    },
+    {
+      "epoch": 2.891653357989409,
+      "grad_norm": 0.4054930402209033,
+      "learning_rate": 4.053580009639998e-06,
+      "loss": 0.442,
+      "step": 4300
+    },
+    {
+      "epoch": 2.8923257964192652,
+      "grad_norm": 0.4424163335137873,
+      "learning_rate": 4.051418363436062e-06,
+      "loss": 0.4409,
+      "step": 4301
+    },
+    {
+      "epoch": 2.892998234849122,
+      "grad_norm": 0.3115659861446721,
+      "learning_rate": 4.049256901134146e-06,
+      "loss": 0.4344,
+      "step": 4302
+    },
+    {
+      "epoch": 2.8936706732789776,
+      "grad_norm": 0.35125570354866636,
+      "learning_rate": 4.047095623153294e-06,
+      "loss": 0.4417,
+      "step": 4303
+    },
+    {
+      "epoch": 2.8943431117088343,
+      "grad_norm": 0.5242489123159184,
+      "learning_rate": 4.0449345299125105e-06,
+      "loss": 0.4424,
+      "step": 4304
+    },
+    {
+      "epoch": 2.8950155501386905,
+      "grad_norm": 0.39148138912349173,
+      "learning_rate": 4.042773621830769e-06,
+      "loss": 0.4341,
+      "step": 4305
+    },
+    {
+      "epoch": 2.8956879885685467,
+      "grad_norm": 0.4032256382439055,
+      "learning_rate": 4.040612899327007e-06,
+      "loss": 0.4374,
+      "step": 4306
+    },
+    {
+      "epoch": 2.896360426998403,
+      "grad_norm": 0.45494911382000336,
+      "learning_rate": 4.0384523628201246e-06,
+      "loss": 0.4405,
+      "step": 4307
+    },
+    {
+      "epoch": 2.897032865428259,
+      "grad_norm": 0.5259742575105177,
+      "learning_rate": 4.03629201272899e-06,
+      "loss": 0.4419,
+      "step": 4308
+    },
+    {
+      "epoch": 2.8977053038581158,
+      "grad_norm": 0.3713800624548828,
+      "learning_rate": 4.034131849472423e-06,
+      "loss": 0.4203,
+      "step": 4309
+    },
+    {
+      "epoch": 2.8983777422879715,
+      "grad_norm": 0.35812631028349245,
+      "learning_rate": 4.031971873469222e-06,
+      "loss": 0.4432,
+      "step": 4310
+    },
+    {
+      "epoch": 2.899050180717828,
+      "grad_norm": 0.3508903930067445,
+      "learning_rate": 4.0298120851381405e-06,
+      "loss": 0.4601,
+      "step": 4311
+    },
+    {
+      "epoch": 2.8997226191476844,
+      "grad_norm": 0.36295881177394324,
+      "learning_rate": 4.0276524848978985e-06,
+      "loss": 0.4389,
+      "step": 4312
+    },
+    {
+      "epoch": 2.9003950575775406,
+      "grad_norm": 0.31969764112024057,
+      "learning_rate": 4.0254930731671785e-06,
+      "loss": 0.4254,
+      "step": 4313
+    },
+    {
+      "epoch": 2.9010674960073968,
+      "grad_norm": 0.3590482150560216,
+      "learning_rate": 4.0233338503646255e-06,
+      "loss": 0.4522,
+      "step": 4314
+    },
+    {
+      "epoch": 2.901739934437253,
+      "grad_norm": 0.42649867839913475,
+      "learning_rate": 4.021174816908852e-06,
+      "loss": 0.44,
+      "step": 4315
+    },
+    {
+      "epoch": 2.9024123728671096,
+      "grad_norm": 0.35706898525852293,
+      "learning_rate": 4.0190159732184305e-06,
+      "loss": 0.4449,
+      "step": 4316
+    },
+    {
+      "epoch": 2.9030848112969654,
+      "grad_norm": 0.3270427305336125,
+      "learning_rate": 4.016857319711893e-06,
+      "loss": 0.4353,
+      "step": 4317
+    },
+    {
+      "epoch": 2.903757249726822,
+      "grad_norm": 0.3247344819275317,
+      "learning_rate": 4.014698856807744e-06,
+      "loss": 0.4326,
+      "step": 4318
+    },
+    {
+      "epoch": 2.9044296881566782,
+      "grad_norm": 0.36525032765947724,
+      "learning_rate": 4.012540584924442e-06,
+      "loss": 0.4276,
+      "step": 4319
+    },
+    {
+      "epoch": 2.9051021265865344,
+      "grad_norm": 0.4000180432810828,
+      "learning_rate": 4.010382504480415e-06,
+      "loss": 0.4422,
+      "step": 4320
+    },
+    {
+      "epoch": 2.9057745650163906,
+      "grad_norm": 0.3840704237209392,
+      "learning_rate": 4.00822461589405e-06,
+      "loss": 0.435,
+      "step": 4321
+    },
+    {
+      "epoch": 2.906447003446247,
+      "grad_norm": 0.3185174756415056,
+      "learning_rate": 4.006066919583698e-06,
+      "loss": 0.4412,
+      "step": 4322
+    },
+    {
+      "epoch": 2.9071194418761035,
+      "grad_norm": 0.35060641499913836,
+      "learning_rate": 4.003909415967672e-06,
+      "loss": 0.446,
+      "step": 4323
+    },
+    {
+      "epoch": 2.9077918803059593,
+      "grad_norm": 0.3229457585115087,
+      "learning_rate": 4.001752105464249e-06,
+      "loss": 0.4338,
+      "step": 4324
+    },
+    {
+      "epoch": 2.908464318735816,
+      "grad_norm": 0.35988129806189556,
+      "learning_rate": 3.999594988491672e-06,
+      "loss": 0.4393,
+      "step": 4325
+    },
+    {
+      "epoch": 2.909136757165672,
+      "grad_norm": 0.5260762120795631,
+      "learning_rate": 3.997438065468135e-06,
+      "loss": 0.4366,
+      "step": 4326
+    },
+    {
+      "epoch": 2.9098091955955283,
+      "grad_norm": 0.33383556534355013,
+      "learning_rate": 3.9952813368118054e-06,
+      "loss": 0.4236,
+      "step": 4327
+    },
+    {
+      "epoch": 2.9104816340253845,
+      "grad_norm": 0.3549944344912472,
+      "learning_rate": 3.9931248029408096e-06,
+      "loss": 0.4194,
+      "step": 4328
+    },
+    {
+      "epoch": 2.9111540724552407,
+      "grad_norm": 0.340081867293386,
+      "learning_rate": 3.9909684642732346e-06,
+      "loss": 0.438,
+      "step": 4329
+    },
+    {
+      "epoch": 2.911826510885097,
+      "grad_norm": 0.41570221401733837,
+      "learning_rate": 3.988812321227134e-06,
+      "loss": 0.4427,
+      "step": 4330
+    },
+    {
+      "epoch": 2.912498949314953,
+      "grad_norm": 0.39149912394492825,
+      "learning_rate": 3.986656374220516e-06,
+      "loss": 0.4401,
+      "step": 4331
+    },
+    {
+      "epoch": 2.9131713877448098,
+      "grad_norm": 0.5597983452383962,
+      "learning_rate": 3.984500623671359e-06,
+      "loss": 0.4343,
+      "step": 4332
+    },
+    {
+      "epoch": 2.913843826174666,
+      "grad_norm": 0.46909294423105663,
+      "learning_rate": 3.982345069997597e-06,
+      "loss": 0.4362,
+      "step": 4333
+    },
+    {
+      "epoch": 2.914516264604522,
+      "grad_norm": 0.4440529949481681,
+      "learning_rate": 3.980189713617132e-06,
+      "loss": 0.447,
+      "step": 4334
+    },
+    {
+      "epoch": 2.9151887030343784,
+      "grad_norm": 0.3846651417453312,
+      "learning_rate": 3.9780345549478185e-06,
+      "loss": 0.4319,
+      "step": 4335
+    },
+    {
+      "epoch": 2.9158611414642346,
+      "grad_norm": 0.3907841131878748,
+      "learning_rate": 3.975879594407481e-06,
+      "loss": 0.4302,
+      "step": 4336
+    },
+    {
+      "epoch": 2.916533579894091,
+      "grad_norm": 0.34113166552315544,
+      "learning_rate": 3.9737248324139035e-06,
+      "loss": 0.4581,
+      "step": 4337
+    },
+    {
+      "epoch": 2.917206018323947,
+      "grad_norm": 0.4001528551480132,
+      "learning_rate": 3.971570269384831e-06,
+      "loss": 0.4256,
+      "step": 4338
+    },
+    {
+      "epoch": 2.9178784567538036,
+      "grad_norm": 0.3707242145332742,
+      "learning_rate": 3.969415905737967e-06,
+      "loss": 0.4358,
+      "step": 4339
+    },
+    {
+      "epoch": 2.91855089518366,
+      "grad_norm": 0.3583102598519517,
+      "learning_rate": 3.967261741890982e-06,
+      "loss": 0.4364,
+      "step": 4340
+    },
+    {
+      "epoch": 2.919223333613516,
+      "grad_norm": 0.3255702338682238,
+      "learning_rate": 3.965107778261504e-06,
+      "loss": 0.4398,
+      "step": 4341
+    },
+    {
+      "epoch": 2.9198957720433723,
+      "grad_norm": 0.34596311967659127,
+      "learning_rate": 3.962954015267123e-06,
+      "loss": 0.4166,
+      "step": 4342
+    },
+    {
+      "epoch": 2.9205682104732285,
+      "grad_norm": 0.3569938436608133,
+      "learning_rate": 3.960800453325392e-06,
+      "loss": 0.4401,
+      "step": 4343
+    },
+    {
+      "epoch": 2.9212406489030847,
+      "grad_norm": 0.4270059848200292,
+      "learning_rate": 3.958647092853819e-06,
+      "loss": 0.4399,
+      "step": 4344
+    },
+    {
+      "epoch": 2.921913087332941,
+      "grad_norm": 0.45722521864625526,
+      "learning_rate": 3.95649393426988e-06,
+      "loss": 0.4399,
+      "step": 4345
+    },
+    {
+      "epoch": 2.9225855257627975,
+      "grad_norm": 0.326045276220502,
+      "learning_rate": 3.954340977991008e-06,
+      "loss": 0.4371,
+      "step": 4346
+    },
+    {
+      "epoch": 2.9232579641926537,
+      "grad_norm": 0.5610976445828544,
+      "learning_rate": 3.9521882244345996e-06,
+      "loss": 0.443,
+      "step": 4347
+    },
+    {
+      "epoch": 2.92393040262251,
+      "grad_norm": 0.37308037941952066,
+      "learning_rate": 3.950035674018008e-06,
+      "loss": 0.4418,
+      "step": 4348
+    },
+    {
+      "epoch": 2.924602841052366,
+      "grad_norm": 0.2977451673809849,
+      "learning_rate": 3.9478833271585494e-06,
+      "loss": 0.4423,
+      "step": 4349
+    },
+    {
+      "epoch": 2.9252752794822223,
+      "grad_norm": 0.3645061290192601,
+      "learning_rate": 3.945731184273502e-06,
+      "loss": 0.4206,
+      "step": 4350
+    },
+    {
+      "epoch": 2.9259477179120785,
+      "grad_norm": 0.37103446938698675,
+      "learning_rate": 3.943579245780101e-06,
+      "loss": 0.4396,
+      "step": 4351
+    },
+    {
+      "epoch": 2.9266201563419347,
+      "grad_norm": 0.33635430725600896,
+      "learning_rate": 3.941427512095548e-06,
+      "loss": 0.4409,
+      "step": 4352
+    },
+    {
+      "epoch": 2.9272925947717914,
+      "grad_norm": 0.44784358321516776,
+      "learning_rate": 3.939275983636995e-06,
+      "loss": 0.443,
+      "step": 4353
+    },
+    {
+      "epoch": 2.9279650332016476,
+      "grad_norm": 0.32436690281265607,
+      "learning_rate": 3.937124660821562e-06,
+      "loss": 0.4371,
+      "step": 4354
+    },
+    {
+      "epoch": 2.928637471631504,
+      "grad_norm": 0.33151113447889863,
+      "learning_rate": 3.934973544066328e-06,
+      "loss": 0.4312,
+      "step": 4355
+    },
+    {
+      "epoch": 2.92930991006136,
+      "grad_norm": 0.39549657605539945,
+      "learning_rate": 3.932822633788334e-06,
+      "loss": 0.4484,
+      "step": 4356
+    },
+    {
+      "epoch": 2.929982348491216,
+      "grad_norm": 0.3303839780249808,
+      "learning_rate": 3.930671930404572e-06,
+      "loss": 0.4392,
+      "step": 4357
+    },
+    {
+      "epoch": 2.9306547869210724,
+      "grad_norm": 0.45877058828119155,
+      "learning_rate": 3.928521434332002e-06,
+      "loss": 0.4452,
+      "step": 4358
+    },
+    {
+      "epoch": 2.9313272253509286,
+      "grad_norm": 0.3174532018147674,
+      "learning_rate": 3.926371145987544e-06,
+      "loss": 0.4355,
+      "step": 4359
+    },
+    {
+      "epoch": 2.9319996637807852,
+      "grad_norm": 0.3864330463989357,
+      "learning_rate": 3.9242210657880745e-06,
+      "loss": 0.4528,
+      "step": 4360
+    },
+    {
+      "epoch": 2.9326721022106415,
+      "grad_norm": 0.376800759867609,
+      "learning_rate": 3.92207119415043e-06,
+      "loss": 0.438,
+      "step": 4361
+    },
+    {
+      "epoch": 2.9333445406404977,
+      "grad_norm": 0.32073634660336636,
+      "learning_rate": 3.919921531491407e-06,
+      "loss": 0.4443,
+      "step": 4362
+    },
+    {
+      "epoch": 2.934016979070354,
+      "grad_norm": 0.3172303339676808,
+      "learning_rate": 3.917772078227763e-06,
+      "loss": 0.4454,
+      "step": 4363
+    },
+    {
+      "epoch": 2.93468941750021,
+      "grad_norm": 0.3850781141799853,
+      "learning_rate": 3.9156228347762115e-06,
+      "loss": 0.4228,
+      "step": 4364
+    },
+    {
+      "epoch": 2.9353618559300663,
+      "grad_norm": 0.2930249871479131,
+      "learning_rate": 3.913473801553433e-06,
+      "loss": 0.4452,
+      "step": 4365
+    },
+    {
+      "epoch": 2.9360342943599225,
+      "grad_norm": 0.3473829718278103,
+      "learning_rate": 3.911324978976054e-06,
+      "loss": 0.4289,
+      "step": 4366
+    },
+    {
+      "epoch": 2.936706732789779,
+      "grad_norm": 0.3217487492540504,
+      "learning_rate": 3.909176367460672e-06,
+      "loss": 0.4403,
+      "step": 4367
+    },
+    {
+      "epoch": 2.9373791712196353,
+      "grad_norm": 0.4522829535415397,
+      "learning_rate": 3.907027967423839e-06,
+      "loss": 0.4355,
+      "step": 4368
+    },
+    {
+      "epoch": 2.9380516096494915,
+      "grad_norm": 0.3485041332119282,
+      "learning_rate": 3.904879779282067e-06,
+      "loss": 0.4336,
+      "step": 4369
+    },
+    {
+      "epoch": 2.9387240480793477,
+      "grad_norm": 0.4578987985171744,
+      "learning_rate": 3.9027318034518245e-06,
+      "loss": 0.4625,
+      "step": 4370
+    },
+    {
+      "epoch": 2.939396486509204,
+      "grad_norm": 0.4543264370604624,
+      "learning_rate": 3.900584040349543e-06,
+      "loss": 0.4496,
+      "step": 4371
+    },
+    {
+      "epoch": 2.94006892493906,
+      "grad_norm": 0.3338600468951495,
+      "learning_rate": 3.8984364903916086e-06,
+      "loss": 0.4394,
+      "step": 4372
+    },
+    {
+      "epoch": 2.9407413633689163,
+      "grad_norm": 0.3022669546557657,
+      "learning_rate": 3.896289153994369e-06,
+      "loss": 0.4248,
+      "step": 4373
+    },
+    {
+      "epoch": 2.941413801798773,
+      "grad_norm": 0.36615243953150717,
+      "learning_rate": 3.894142031574131e-06,
+      "loss": 0.4413,
+      "step": 4374
+    },
+    {
+      "epoch": 2.942086240228629,
+      "grad_norm": 0.8356858213298126,
+      "learning_rate": 3.891995123547154e-06,
+      "loss": 0.437,
+      "step": 4375
+    },
+    {
+      "epoch": 2.9427586786584854,
+      "grad_norm": 0.3351703752251006,
+      "learning_rate": 3.889848430329664e-06,
+      "loss": 0.4423,
+      "step": 4376
+    },
+    {
+      "epoch": 2.9434311170883416,
+      "grad_norm": 0.29626714008291044,
+      "learning_rate": 3.887701952337839e-06,
+      "loss": 0.4284,
+      "step": 4377
+    },
+    {
+      "epoch": 2.944103555518198,
+      "grad_norm": 0.28788910704281123,
+      "learning_rate": 3.88555568998782e-06,
+      "loss": 0.4332,
+      "step": 4378
+    },
+    {
+      "epoch": 2.944775993948054,
+      "grad_norm": 0.3341832475121459,
+      "learning_rate": 3.883409643695702e-06,
+      "loss": 0.4526,
+      "step": 4379
+    },
+    {
+      "epoch": 2.94544843237791,
+      "grad_norm": 0.3747649741539901,
+      "learning_rate": 3.881263813877542e-06,
+      "loss": 0.4448,
+      "step": 4380
+    },
+    {
+      "epoch": 2.946120870807767,
+      "grad_norm": 0.34952520777191076,
+      "learning_rate": 3.879118200949352e-06,
+      "loss": 0.4326,
+      "step": 4381
+    },
+    {
+      "epoch": 2.946793309237623,
+      "grad_norm": 0.4237596153227539,
+      "learning_rate": 3.876972805327105e-06,
+      "loss": 0.4422,
+      "step": 4382
+    },
+    {
+      "epoch": 2.9474657476674793,
+      "grad_norm": 0.3219451157593167,
+      "learning_rate": 3.874827627426727e-06,
+      "loss": 0.4405,
+      "step": 4383
+    },
+    {
+      "epoch": 2.9481381860973355,
+      "grad_norm": 0.44034715574883926,
+      "learning_rate": 3.872682667664105e-06,
+      "loss": 0.4244,
+      "step": 4384
+    },
+    {
+      "epoch": 2.9488106245271917,
+      "grad_norm": 0.3202514401249179,
+      "learning_rate": 3.870537926455086e-06,
+      "loss": 0.4474,
+      "step": 4385
+    },
+    {
+      "epoch": 2.949483062957048,
+      "grad_norm": 0.3718417568910564,
+      "learning_rate": 3.868393404215469e-06,
+      "loss": 0.4301,
+      "step": 4386
+    },
+    {
+      "epoch": 2.950155501386904,
+      "grad_norm": 0.35650965227510006,
+      "learning_rate": 3.866249101361018e-06,
+      "loss": 0.4379,
+      "step": 4387
+    },
+    {
+      "epoch": 2.9508279398167607,
+      "grad_norm": 0.4331958919885307,
+      "learning_rate": 3.864105018307446e-06,
+      "loss": 0.4652,
+      "step": 4388
+    },
+    {
+      "epoch": 2.951500378246617,
+      "grad_norm": 0.36349900094158416,
+      "learning_rate": 3.861961155470428e-06,
+      "loss": 0.4429,
+      "step": 4389
+    },
+    {
+      "epoch": 2.952172816676473,
+      "grad_norm": 0.34935962224647293,
+      "learning_rate": 3.8598175132655975e-06,
+      "loss": 0.4501,
+      "step": 4390
+    },
+    {
+      "epoch": 2.9528452551063293,
+      "grad_norm": 0.36038031146784655,
+      "learning_rate": 3.8576740921085455e-06,
+      "loss": 0.4352,
+      "step": 4391
+    },
+    {
+      "epoch": 2.9535176935361855,
+      "grad_norm": 0.39743775222788014,
+      "learning_rate": 3.855530892414813e-06,
+      "loss": 0.4352,
+      "step": 4392
+    },
+    {
+      "epoch": 2.9541901319660417,
+      "grad_norm": 0.32815871640241545,
+      "learning_rate": 3.853387914599905e-06,
+      "loss": 0.4474,
+      "step": 4393
+    },
+    {
+      "epoch": 2.954862570395898,
+      "grad_norm": 0.3279877663767734,
+      "learning_rate": 3.851245159079283e-06,
+      "loss": 0.4388,
+      "step": 4394
+    },
+    {
+      "epoch": 2.9555350088257546,
+      "grad_norm": 0.37024229207542203,
+      "learning_rate": 3.849102626268364e-06,
+      "loss": 0.4348,
+      "step": 4395
+    },
+    {
+      "epoch": 2.956207447255611,
+      "grad_norm": 0.3295138624701636,
+      "learning_rate": 3.8469603165825226e-06,
+      "loss": 0.4402,
+      "step": 4396
+    },
+    {
+      "epoch": 2.956879885685467,
+      "grad_norm": 0.36905622085010975,
+      "learning_rate": 3.844818230437087e-06,
+      "loss": 0.4338,
+      "step": 4397
+    },
+    {
+      "epoch": 2.957552324115323,
+      "grad_norm": 0.32310577320388223,
+      "learning_rate": 3.842676368247347e-06,
+      "loss": 0.436,
+      "step": 4398
+    },
+    {
+      "epoch": 2.9582247625451794,
+      "grad_norm": 0.3368728139063044,
+      "learning_rate": 3.840534730428545e-06,
+      "loss": 0.4124,
+      "step": 4399
+    },
+    {
+      "epoch": 2.9588972009750356,
+      "grad_norm": 0.356940651326783,
+      "learning_rate": 3.8383933173958835e-06,
+      "loss": 0.4377,
+      "step": 4400
+    },
+    {
+      "epoch": 2.959569639404892,
+      "grad_norm": 0.4407601298417127,
+      "learning_rate": 3.8362521295645164e-06,
+      "loss": 0.4388,
+      "step": 4401
+    },
+    {
+      "epoch": 2.9602420778347485,
+      "grad_norm": 0.3699052538556628,
+      "learning_rate": 3.834111167349558e-06,
+      "loss": 0.4566,
+      "step": 4402
+    },
+    {
+      "epoch": 2.9609145162646047,
+      "grad_norm": 0.41370129265662003,
+      "learning_rate": 3.8319704311660785e-06,
+      "loss": 0.4443,
+      "step": 4403
+    },
+    {
+      "epoch": 2.961586954694461,
+      "grad_norm": 0.3563826234769003,
+      "learning_rate": 3.829829921429103e-06,
+      "loss": 0.4551,
+      "step": 4404
+    },
+    {
+      "epoch": 2.962259393124317,
+      "grad_norm": 0.5052402730139329,
+      "learning_rate": 3.8276896385536145e-06,
+      "loss": 0.4235,
+      "step": 4405
+    },
+    {
+      "epoch": 2.9629318315541733,
+      "grad_norm": 0.3742072239640784,
+      "learning_rate": 3.825549582954548e-06,
+      "loss": 0.425,
+      "step": 4406
+    },
+    {
+      "epoch": 2.9636042699840295,
+      "grad_norm": 0.484152781124374,
+      "learning_rate": 3.8234097550468e-06,
+      "loss": 0.4319,
+      "step": 4407
+    },
+    {
+      "epoch": 2.9642767084138857,
+      "grad_norm": 0.41888443325550634,
+      "learning_rate": 3.821270155245219e-06,
+      "loss": 0.4564,
+      "step": 4408
+    },
+    {
+      "epoch": 2.9649491468437423,
+      "grad_norm": 0.3124106039017705,
+      "learning_rate": 3.81913078396461e-06,
+      "loss": 0.4434,
+      "step": 4409
+    },
+    {
+      "epoch": 2.9656215852735985,
+      "grad_norm": 0.3310828872080542,
+      "learning_rate": 3.816991641619736e-06,
+      "loss": 0.449,
+      "step": 4410
+    },
+    {
+      "epoch": 2.9662940237034547,
+      "grad_norm": 0.3283652682944646,
+      "learning_rate": 3.81485272862531e-06,
+      "loss": 0.4258,
+      "step": 4411
+    },
+    {
+      "epoch": 2.966966462133311,
+      "grad_norm": 0.3375407699567821,
+      "learning_rate": 3.8127140453960065e-06,
+      "loss": 0.4444,
+      "step": 4412
+    },
+    {
+      "epoch": 2.967638900563167,
+      "grad_norm": 0.3991924856965535,
+      "learning_rate": 3.810575592346455e-06,
+      "loss": 0.4355,
+      "step": 4413
+    },
+    {
+      "epoch": 2.9683113389930234,
+      "grad_norm": 0.31755611053343813,
+      "learning_rate": 3.8084373698912334e-06,
+      "loss": 0.4395,
+      "step": 4414
+    },
+    {
+      "epoch": 2.9689837774228796,
+      "grad_norm": 0.3789286228953469,
+      "learning_rate": 3.806299378444884e-06,
+      "loss": 0.4415,
+      "step": 4415
+    },
+    {
+      "epoch": 2.969656215852736,
+      "grad_norm": 0.3697470094575646,
+      "learning_rate": 3.804161618421899e-06,
+      "loss": 0.4376,
+      "step": 4416
+    },
+    {
+      "epoch": 2.9703286542825924,
+      "grad_norm": 0.3486963933380699,
+      "learning_rate": 3.802024090236727e-06,
+      "loss": 0.4323,
+      "step": 4417
+    },
+    {
+      "epoch": 2.9710010927124486,
+      "grad_norm": 0.3036241491697172,
+      "learning_rate": 3.799886794303773e-06,
+      "loss": 0.4355,
+      "step": 4418
+    },
+    {
+      "epoch": 2.971673531142305,
+      "grad_norm": 0.3629807388059824,
+      "learning_rate": 3.7977497310373946e-06,
+      "loss": 0.4411,
+      "step": 4419
+    },
+    {
+      "epoch": 2.972345969572161,
+      "grad_norm": 0.33530428915772964,
+      "learning_rate": 3.7956129008519046e-06,
+      "loss": 0.4275,
+      "step": 4420
+    },
+    {
+      "epoch": 2.973018408002017,
+      "grad_norm": 0.32335921903193104,
+      "learning_rate": 3.7934763041615717e-06,
+      "loss": 0.432,
+      "step": 4421
+    },
+    {
+      "epoch": 2.9736908464318734,
+      "grad_norm": 0.5711599036875669,
+      "learning_rate": 3.7913399413806227e-06,
+      "loss": 0.4504,
+      "step": 4422
+    },
+    {
+      "epoch": 2.97436328486173,
+      "grad_norm": 0.32928993885534397,
+      "learning_rate": 3.78920381292323e-06,
+      "loss": 0.428,
+      "step": 4423
+    },
+    {
+      "epoch": 2.9750357232915863,
+      "grad_norm": 0.3275175794711282,
+      "learning_rate": 3.787067919203528e-06,
+      "loss": 0.4517,
+      "step": 4424
+    },
+    {
+      "epoch": 2.9757081617214425,
+      "grad_norm": 0.4021150224715988,
+      "learning_rate": 3.784932260635604e-06,
+      "loss": 0.4355,
+      "step": 4425
+    },
+    {
+      "epoch": 2.9763806001512987,
+      "grad_norm": 0.4063896786721347,
+      "learning_rate": 3.7827968376334996e-06,
+      "loss": 0.432,
+      "step": 4426
+    },
+    {
+      "epoch": 2.977053038581155,
+      "grad_norm": 0.45938151658635823,
+      "learning_rate": 3.780661650611209e-06,
+      "loss": 0.4371,
+      "step": 4427
+    },
+    {
+      "epoch": 2.977725477011011,
+      "grad_norm": 0.3171925401712923,
+      "learning_rate": 3.7785266999826826e-06,
+      "loss": 0.449,
+      "step": 4428
+    },
+    {
+      "epoch": 2.9783979154408673,
+      "grad_norm": 0.3228017458469971,
+      "learning_rate": 3.7763919861618247e-06,
+      "loss": 0.4408,
+      "step": 4429
+    },
+    {
+      "epoch": 2.979070353870724,
+      "grad_norm": 0.4159499590049895,
+      "learning_rate": 3.774257509562493e-06,
+      "loss": 0.4389,
+      "step": 4430
+    },
+    {
+      "epoch": 2.9797427923005797,
+      "grad_norm": 0.3722689785310936,
+      "learning_rate": 3.7721232705985022e-06,
+      "loss": 0.4492,
+      "step": 4431
+    },
+    {
+      "epoch": 2.9804152307304363,
+      "grad_norm": 0.4000912931409732,
+      "learning_rate": 3.769989269683614e-06,
+      "loss": 0.4395,
+      "step": 4432
+    },
+    {
+      "epoch": 2.9810876691602926,
+      "grad_norm": 0.35943578317497027,
+      "learning_rate": 3.7678555072315496e-06,
+      "loss": 0.4355,
+      "step": 4433
+    },
+    {
+      "epoch": 2.9817601075901488,
+      "grad_norm": 0.31215073655493997,
+      "learning_rate": 3.765721983655984e-06,
+      "loss": 0.432,
+      "step": 4434
+    },
+    {
+      "epoch": 2.982432546020005,
+      "grad_norm": 0.32125224482976794,
+      "learning_rate": 3.7635886993705443e-06,
+      "loss": 0.4465,
+      "step": 4435
+    },
+    {
+      "epoch": 2.983104984449861,
+      "grad_norm": 0.35945927270651473,
+      "learning_rate": 3.76145565478881e-06,
+      "loss": 0.4427,
+      "step": 4436
+    },
+    {
+      "epoch": 2.983777422879718,
+      "grad_norm": 0.3154339889699578,
+      "learning_rate": 3.7593228503243173e-06,
+      "loss": 0.4375,
+      "step": 4437
+    },
+    {
+      "epoch": 2.9844498613095736,
+      "grad_norm": 0.3717430845968554,
+      "learning_rate": 3.7571902863905524e-06,
+      "loss": 0.432,
+      "step": 4438
+    },
+    {
+      "epoch": 2.98512229973943,
+      "grad_norm": 0.3863997070542928,
+      "learning_rate": 3.7550579634009582e-06,
+      "loss": 0.4455,
+      "step": 4439
+    },
+    {
+      "epoch": 2.9857947381692864,
+      "grad_norm": 0.2909428722373954,
+      "learning_rate": 3.752925881768931e-06,
+      "loss": 0.4362,
+      "step": 4440
+    },
+    {
+      "epoch": 2.9864671765991426,
+      "grad_norm": 0.34718339418997324,
+      "learning_rate": 3.7507940419078127e-06,
+      "loss": 0.447,
+      "step": 4441
+    },
+    {
+      "epoch": 2.987139615028999,
+      "grad_norm": 0.40098733777174367,
+      "learning_rate": 3.7486624442309087e-06,
+      "loss": 0.4382,
+      "step": 4442
+    },
+    {
+      "epoch": 2.987812053458855,
+      "grad_norm": 0.3695593860512126,
+      "learning_rate": 3.7465310891514716e-06,
+      "loss": 0.4378,
+      "step": 4443
+    },
+    {
+      "epoch": 2.9884844918887117,
+      "grad_norm": 0.3398051541434089,
+      "learning_rate": 3.74439997708271e-06,
+      "loss": 0.4374,
+      "step": 4444
+    },
+    {
+      "epoch": 2.9891569303185674,
+      "grad_norm": 0.34199486719995503,
+      "learning_rate": 3.7422691084377817e-06,
+      "loss": 0.4422,
+      "step": 4445
+    },
+    {
+      "epoch": 2.989829368748424,
+      "grad_norm": 0.35147755808673253,
+      "learning_rate": 3.7401384836297994e-06,
+      "loss": 0.4316,
+      "step": 4446
+    },
+    {
+      "epoch": 2.9905018071782803,
+      "grad_norm": 0.32203234479982756,
+      "learning_rate": 3.7380081030718296e-06,
+      "loss": 0.4385,
+      "step": 4447
+    },
+    {
+      "epoch": 2.9911742456081365,
+      "grad_norm": 0.3200714589833798,
+      "learning_rate": 3.7358779671768917e-06,
+      "loss": 0.448,
+      "step": 4448
+    },
+    {
+      "epoch": 2.9918466840379927,
+      "grad_norm": 0.3481011119507389,
+      "learning_rate": 3.733748076357952e-06,
+      "loss": 0.4416,
+      "step": 4449
+    },
+    {
+      "epoch": 2.992519122467849,
+      "grad_norm": 0.32614372611646925,
+      "learning_rate": 3.7316184310279356e-06,
+      "loss": 0.434,
+      "step": 4450
+    },
+    {
+      "epoch": 2.9931915608977055,
+      "grad_norm": 1.4320532111873607,
+      "learning_rate": 3.729489031599719e-06,
+      "loss": 0.4359,
+      "step": 4451
+    },
+    {
+      "epoch": 2.9938639993275613,
+      "grad_norm": 0.34225309102035456,
+      "learning_rate": 3.7273598784861288e-06,
+      "loss": 0.4575,
+      "step": 4452
+    },
+    {
+      "epoch": 2.994536437757418,
+      "grad_norm": 0.3854212013922402,
+      "learning_rate": 3.725230972099947e-06,
+      "loss": 0.438,
+      "step": 4453
+    },
+    {
+      "epoch": 2.995208876187274,
+      "grad_norm": 0.49023520566482387,
+      "learning_rate": 3.7231023128539023e-06,
+      "loss": 0.4372,
+      "step": 4454
+    },
+    {
+      "epoch": 2.9958813146171304,
+      "grad_norm": 0.343889522075791,
+      "learning_rate": 3.7209739011606814e-06,
+      "loss": 0.4305,
+      "step": 4455
+    },
+    {
+      "epoch": 2.9965537530469866,
+      "grad_norm": 0.3953877234343036,
+      "learning_rate": 3.71884573743292e-06,
+      "loss": 0.4341,
+      "step": 4456
+    },
+    {
+      "epoch": 2.9972261914768428,
+      "grad_norm": 0.3238747253395985,
+      "learning_rate": 3.716717822083209e-06,
+      "loss": 0.4341,
+      "step": 4457
+    },
+    {
+      "epoch": 2.9978986299066994,
+      "grad_norm": 0.3730996167012987,
+      "learning_rate": 3.714590155524084e-06,
+      "loss": 0.4407,
+      "step": 4458
+    },
+    {
+      "epoch": 2.998571068336555,
+      "grad_norm": 0.37637480012700475,
+      "learning_rate": 3.7124627381680367e-06,
+      "loss": 0.4266,
+      "step": 4459
+    },
+    {
+      "epoch": 2.999243506766412,
+      "grad_norm": 0.38744039917159373,
+      "learning_rate": 3.7103355704275136e-06,
+      "loss": 0.4359,
+      "step": 4460
+    },
+    {
+      "epoch": 2.999915945196268,
+      "grad_norm": 0.32943006457297874,
+      "learning_rate": 3.708208652714912e-06,
+      "loss": 0.4349,
+      "step": 4461
+    },
+    {
+      "epoch": 3.0005883836261242,
+      "grad_norm": 0.6176820385280498,
+      "learning_rate": 3.7060819854425723e-06,
+      "loss": 0.4499,
+      "step": 4462
+    },
+    {
+      "epoch": 3.0012608220559804,
+      "grad_norm": 0.33412830767673857,
+      "learning_rate": 3.7039555690227963e-06,
+      "loss": 0.442,
+      "step": 4463
+    },
+    {
+      "epoch": 3.0019332604858366,
+      "grad_norm": 0.40413572576992607,
+      "learning_rate": 3.7018294038678326e-06,
+      "loss": 0.4357,
+      "step": 4464
+    },
+    {
+      "epoch": 3.002605698915693,
+      "grad_norm": 0.3732340624532239,
+      "learning_rate": 3.6997034903898826e-06,
+      "loss": 0.4394,
+      "step": 4465
+    },
+    {
+      "epoch": 3.0032781373455495,
+      "grad_norm": 0.3435046155970762,
+      "learning_rate": 3.697577829001099e-06,
+      "loss": 0.4296,
+      "step": 4466
+    },
+    {
+      "epoch": 3.0039505757754057,
+      "grad_norm": 0.2881616624216002,
+      "learning_rate": 3.6954524201135823e-06,
+      "loss": 0.4372,
+      "step": 4467
+    },
+    {
+      "epoch": 3.004623014205262,
+      "grad_norm": 0.35940667588121067,
+      "learning_rate": 3.693327264139388e-06,
+      "loss": 0.4407,
+      "step": 4468
+    },
+    {
+      "epoch": 3.005295452635118,
+      "grad_norm": 0.3429020968951988,
+      "learning_rate": 3.6912023614905218e-06,
+      "loss": 0.4262,
+      "step": 4469
+    },
+    {
+      "epoch": 3.0059678910649743,
+      "grad_norm": 0.31452139822598385,
+      "learning_rate": 3.689077712578941e-06,
+      "loss": 0.4474,
+      "step": 4470
+    },
+    {
+      "epoch": 3.0066403294948305,
+      "grad_norm": 0.35211524591414006,
+      "learning_rate": 3.686953317816548e-06,
+      "loss": 0.4178,
+      "step": 4471
+    },
+    {
+      "epoch": 3.0073127679246867,
+      "grad_norm": 0.38554047368580263,
+      "learning_rate": 3.6848291776152044e-06,
+      "loss": 0.4281,
+      "step": 4472
+    },
+    {
+      "epoch": 3.0079852063545434,
+      "grad_norm": 0.37807020266172264,
+      "learning_rate": 3.682705292386716e-06,
+      "loss": 0.4405,
+      "step": 4473
+    },
+    {
+      "epoch": 3.0086576447843996,
+      "grad_norm": 0.5783834918565653,
+      "learning_rate": 3.6805816625428424e-06,
+      "loss": 0.4303,
+      "step": 4474
+    },
+    {
+      "epoch": 3.0093300832142558,
+      "grad_norm": 0.3027429867873164,
+      "learning_rate": 3.6784582884952935e-06,
+      "loss": 0.4284,
+      "step": 4475
+    },
+    {
+      "epoch": 3.010002521644112,
+      "grad_norm": 0.3573843602129143,
+      "learning_rate": 3.676335170655728e-06,
+      "loss": 0.4257,
+      "step": 4476
+    },
+    {
+      "epoch": 3.010674960073968,
+      "grad_norm": 0.5170737055546644,
+      "learning_rate": 3.674212309435756e-06,
+      "loss": 0.4343,
+      "step": 4477
+    },
+    {
+      "epoch": 3.0113473985038244,
+      "grad_norm": 0.39620461894404074,
+      "learning_rate": 3.6720897052469374e-06,
+      "loss": 0.4352,
+      "step": 4478
+    },
+    {
+      "epoch": 3.0120198369336806,
+      "grad_norm": 0.34278914287919265,
+      "learning_rate": 3.6699673585007858e-06,
+      "loss": 0.4303,
+      "step": 4479
+    },
+    {
+      "epoch": 3.0126922753635372,
+      "grad_norm": 0.3574083848789049,
+      "learning_rate": 3.6678452696087563e-06,
+      "loss": 0.4235,
+      "step": 4480
+    },
+    {
+      "epoch": 3.0133647137933934,
+      "grad_norm": 0.4886644604880471,
+      "learning_rate": 3.6657234389822626e-06,
+      "loss": 0.4326,
+      "step": 4481
+    },
+    {
+      "epoch": 3.0140371522232496,
+      "grad_norm": 0.4046875113455944,
+      "learning_rate": 3.6636018670326646e-06,
+      "loss": 0.4461,
+      "step": 4482
+    },
+    {
+      "epoch": 3.014709590653106,
+      "grad_norm": 0.3058637048547598,
+      "learning_rate": 3.6614805541712726e-06,
+      "loss": 0.4428,
+      "step": 4483
+    },
+    {
+      "epoch": 3.015382029082962,
+      "grad_norm": 0.3532756588998309,
+      "learning_rate": 3.6593595008093464e-06,
+      "loss": 0.419,
+      "step": 4484
+    },
+    {
+      "epoch": 3.0160544675128182,
+      "grad_norm": 0.4818989394385238,
+      "learning_rate": 3.657238707358096e-06,
+      "loss": 0.4308,
+      "step": 4485
+    },
+    {
+      "epoch": 3.0167269059426745,
+      "grad_norm": 0.38255122084895843,
+      "learning_rate": 3.6551181742286803e-06,
+      "loss": 0.4438,
+      "step": 4486
+    },
+    {
+      "epoch": 3.017399344372531,
+      "grad_norm": 0.3754752510920237,
+      "learning_rate": 3.652997901832208e-06,
+      "loss": 0.4444,
+      "step": 4487
+    },
+    {
+      "epoch": 3.0180717828023873,
+      "grad_norm": 0.36663369477192764,
+      "learning_rate": 3.6508778905797404e-06,
+      "loss": 0.4342,
+      "step": 4488
+    },
+    {
+      "epoch": 3.0187442212322435,
+      "grad_norm": 0.31769006125646865,
+      "learning_rate": 3.648758140882281e-06,
+      "loss": 0.4397,
+      "step": 4489
+    },
+    {
+      "epoch": 3.0194166596620997,
+      "grad_norm": 0.37354204632756116,
+      "learning_rate": 3.646638653150788e-06,
+      "loss": 0.4223,
+      "step": 4490
+    },
+    {
+      "epoch": 3.020089098091956,
+      "grad_norm": 0.3219142891575576,
+      "learning_rate": 3.644519427796168e-06,
+      "loss": 0.4477,
+      "step": 4491
+    },
+    {
+      "epoch": 3.020761536521812,
+      "grad_norm": 0.347480754410478,
+      "learning_rate": 3.6424004652292778e-06,
+      "loss": 0.4461,
+      "step": 4492
+    },
+    {
+      "epoch": 3.0214339749516683,
+      "grad_norm": 0.30296795201848176,
+      "learning_rate": 3.640281765860919e-06,
+      "loss": 0.4481,
+      "step": 4493
+    },
+    {
+      "epoch": 3.022106413381525,
+      "grad_norm": 0.39347590564289325,
+      "learning_rate": 3.638163330101847e-06,
+      "loss": 0.4327,
+      "step": 4494
+    },
+    {
+      "epoch": 3.022778851811381,
+      "grad_norm": 0.32219672839312713,
+      "learning_rate": 3.636045158362763e-06,
+      "loss": 0.451,
+      "step": 4495
+    },
+    {
+      "epoch": 3.0234512902412374,
+      "grad_norm": 0.34578066800769225,
+      "learning_rate": 3.6339272510543184e-06,
+      "loss": 0.4284,
+      "step": 4496
+    },
+    {
+      "epoch": 3.0241237286710936,
+      "grad_norm": 0.5388113685987523,
+      "learning_rate": 3.6318096085871148e-06,
+      "loss": 0.4295,
+      "step": 4497
+    },
+    {
+      "epoch": 3.02479616710095,
+      "grad_norm": 0.38941952487264886,
+      "learning_rate": 3.6296922313716976e-06,
+      "loss": 0.4391,
+      "step": 4498
+    },
+    {
+      "epoch": 3.025468605530806,
+      "grad_norm": 0.3382159734636173,
+      "learning_rate": 3.6275751198185644e-06,
+      "loss": 0.4447,
+      "step": 4499
+    },
+    {
+      "epoch": 3.026141043960662,
+      "grad_norm": 0.3602686931463578,
+      "learning_rate": 3.6254582743381617e-06,
+      "loss": 0.4295,
+      "step": 4500
+    },
+    {
+      "epoch": 3.0268134823905184,
+      "grad_norm": 0.44577634985746306,
+      "learning_rate": 3.6233416953408844e-06,
+      "loss": 0.4402,
+      "step": 4501
+    },
+    {
+      "epoch": 3.027485920820375,
+      "grad_norm": 0.35712384379947537,
+      "learning_rate": 3.6212253832370727e-06,
+      "loss": 0.4294,
+      "step": 4502
+    },
+    {
+      "epoch": 3.0281583592502312,
+      "grad_norm": 0.3874521973589743,
+      "learning_rate": 3.6191093384370173e-06,
+      "loss": 0.4288,
+      "step": 4503
+    },
+    {
+      "epoch": 3.0288307976800874,
+      "grad_norm": 0.43785321982355346,
+      "learning_rate": 3.6169935613509585e-06,
+      "loss": 0.4489,
+      "step": 4504
+    },
+    {
+      "epoch": 3.0295032361099437,
+      "grad_norm": 0.35880506553273245,
+      "learning_rate": 3.6148780523890836e-06,
+      "loss": 0.4387,
+      "step": 4505
+    },
+    {
+      "epoch": 3.0301756745398,
+      "grad_norm": 0.31171991877169425,
+      "learning_rate": 3.6127628119615245e-06,
+      "loss": 0.4415,
+      "step": 4506
+    },
+    {
+      "epoch": 3.030848112969656,
+      "grad_norm": 0.5210472160505484,
+      "learning_rate": 3.610647840478365e-06,
+      "loss": 0.4404,
+      "step": 4507
+    },
+    {
+      "epoch": 3.0315205513995123,
+      "grad_norm": 0.3050302944530371,
+      "learning_rate": 3.6085331383496357e-06,
+      "loss": 0.4542,
+      "step": 4508
+    },
+    {
+      "epoch": 3.032192989829369,
+      "grad_norm": 0.3415842205365484,
+      "learning_rate": 3.6064187059853173e-06,
+      "loss": 0.4304,
+      "step": 4509
+    },
+    {
+      "epoch": 3.032865428259225,
+      "grad_norm": 0.3195233141623651,
+      "learning_rate": 3.604304543795335e-06,
+      "loss": 0.424,
+      "step": 4510
+    },
+    {
+      "epoch": 3.0335378666890813,
+      "grad_norm": 0.3758174522413181,
+      "learning_rate": 3.60219065218956e-06,
+      "loss": 0.4417,
+      "step": 4511
+    },
+    {
+      "epoch": 3.0342103051189375,
+      "grad_norm": 0.3266941100665439,
+      "learning_rate": 3.6000770315778157e-06,
+      "loss": 0.432,
+      "step": 4512
+    },
+    {
+      "epoch": 3.0348827435487937,
+      "grad_norm": 0.3321494460098997,
+      "learning_rate": 3.5979636823698704e-06,
+      "loss": 0.4364,
+      "step": 4513
+    },
+    {
+      "epoch": 3.03555518197865,
+      "grad_norm": 0.390727902210108,
+      "learning_rate": 3.595850604975441e-06,
+      "loss": 0.4429,
+      "step": 4514
+    },
+    {
+      "epoch": 3.036227620408506,
+      "grad_norm": 0.3715492533589567,
+      "learning_rate": 3.5937377998041888e-06,
+      "loss": 0.4391,
+      "step": 4515
+    },
+    {
+      "epoch": 3.036900058838363,
+      "grad_norm": 0.4056712467446377,
+      "learning_rate": 3.591625267265727e-06,
+      "loss": 0.4296,
+      "step": 4516
+    },
+    {
+      "epoch": 3.037572497268219,
+      "grad_norm": 0.9494418041097662,
+      "learning_rate": 3.589513007769611e-06,
+      "loss": 0.418,
+      "step": 4517
+    },
+    {
+      "epoch": 3.038244935698075,
+      "grad_norm": 0.3523303983594626,
+      "learning_rate": 3.5874010217253473e-06,
+      "loss": 0.4373,
+      "step": 4518
+    },
+    {
+      "epoch": 3.0389173741279314,
+      "grad_norm": 0.3689719434747409,
+      "learning_rate": 3.58528930954239e-06,
+      "loss": 0.425,
+      "step": 4519
+    },
+    {
+      "epoch": 3.0395898125577876,
+      "grad_norm": 0.3906852991531941,
+      "learning_rate": 3.5831778716301325e-06,
+      "loss": 0.4267,
+      "step": 4520
+    },
+    {
+      "epoch": 3.040262250987644,
+      "grad_norm": 0.5857781559126515,
+      "learning_rate": 3.5810667083979228e-06,
+      "loss": 0.4309,
+      "step": 4521
+    },
+    {
+      "epoch": 3.0409346894175,
+      "grad_norm": 0.4321142668318707,
+      "learning_rate": 3.5789558202550533e-06,
+      "loss": 0.4344,
+      "step": 4522
+    },
+    {
+      "epoch": 3.0416071278473567,
+      "grad_norm": 0.3307722284146976,
+      "learning_rate": 3.576845207610765e-06,
+      "loss": 0.448,
+      "step": 4523
+    },
+    {
+      "epoch": 3.042279566277213,
+      "grad_norm": 0.4335724324922873,
+      "learning_rate": 3.5747348708742404e-06,
+      "loss": 0.4443,
+      "step": 4524
+    },
+    {
+      "epoch": 3.042952004707069,
+      "grad_norm": 0.4024654859333773,
+      "learning_rate": 3.572624810454612e-06,
+      "loss": 0.4296,
+      "step": 4525
+    },
+    {
+      "epoch": 3.0436244431369253,
+      "grad_norm": 0.34310394163753305,
+      "learning_rate": 3.5705150267609596e-06,
+      "loss": 0.4468,
+      "step": 4526
+    },
+    {
+      "epoch": 3.0442968815667815,
+      "grad_norm": 0.3634416843777775,
+      "learning_rate": 3.5684055202023093e-06,
+      "loss": 0.4484,
+      "step": 4527
+    },
+    {
+      "epoch": 3.0449693199966377,
+      "grad_norm": 0.3003360998378971,
+      "learning_rate": 3.566296291187629e-06,
+      "loss": 0.4466,
+      "step": 4528
+    },
+    {
+      "epoch": 3.045641758426494,
+      "grad_norm": 0.32281364279096514,
+      "learning_rate": 3.5641873401258377e-06,
+      "loss": 0.4367,
+      "step": 4529
+    },
+    {
+      "epoch": 3.0463141968563505,
+      "grad_norm": 0.3521914518525131,
+      "learning_rate": 3.5620786674257983e-06,
+      "loss": 0.4342,
+      "step": 4530
+    },
+    {
+      "epoch": 3.0469866352862067,
+      "grad_norm": 0.3169201581905446,
+      "learning_rate": 3.559970273496321e-06,
+      "loss": 0.4329,
+      "step": 4531
+    },
+    {
+      "epoch": 3.047659073716063,
+      "grad_norm": 0.3902227518137739,
+      "learning_rate": 3.5578621587461615e-06,
+      "loss": 0.4413,
+      "step": 4532
+    },
+    {
+      "epoch": 3.048331512145919,
+      "grad_norm": 0.4085131117285104,
+      "learning_rate": 3.55575432358402e-06,
+      "loss": 0.4384,
+      "step": 4533
+    },
+    {
+      "epoch": 3.0490039505757753,
+      "grad_norm": 0.3264488825660449,
+      "learning_rate": 3.553646768418544e-06,
+      "loss": 0.4228,
+      "step": 4534
+    },
+    {
+      "epoch": 3.0496763890056315,
+      "grad_norm": 0.4476131212157327,
+      "learning_rate": 3.5515394936583265e-06,
+      "loss": 0.4418,
+      "step": 4535
+    },
+    {
+      "epoch": 3.0503488274354877,
+      "grad_norm": 0.5043788346159002,
+      "learning_rate": 3.549432499711908e-06,
+      "loss": 0.4466,
+      "step": 4536
+    },
+    {
+      "epoch": 3.0510212658653444,
+      "grad_norm": 0.3253028263392405,
+      "learning_rate": 3.547325786987768e-06,
+      "loss": 0.431,
+      "step": 4537
+    },
+    {
+      "epoch": 3.0516937042952006,
+      "grad_norm": 0.34369162109844337,
+      "learning_rate": 3.545219355894339e-06,
+      "loss": 0.4341,
+      "step": 4538
+    },
+    {
+      "epoch": 3.052366142725057,
+      "grad_norm": 0.33327263778378285,
+      "learning_rate": 3.543113206839995e-06,
+      "loss": 0.4323,
+      "step": 4539
+    },
+    {
+      "epoch": 3.053038581154913,
+      "grad_norm": 0.3719428111807546,
+      "learning_rate": 3.5410073402330565e-06,
+      "loss": 0.4514,
+      "step": 4540
+    },
+    {
+      "epoch": 3.053711019584769,
+      "grad_norm": 0.37129283477446157,
+      "learning_rate": 3.53890175648179e-06,
+      "loss": 0.4186,
+      "step": 4541
+    },
+    {
+      "epoch": 3.0543834580146254,
+      "grad_norm": 0.37228958234421883,
+      "learning_rate": 3.5367964559944045e-06,
+      "loss": 0.4449,
+      "step": 4542
+    },
+    {
+      "epoch": 3.0550558964444816,
+      "grad_norm": 0.32190002037023063,
+      "learning_rate": 3.5346914391790566e-06,
+      "loss": 0.4429,
+      "step": 4543
+    },
+    {
+      "epoch": 3.0557283348743383,
+      "grad_norm": 0.31796748081394494,
+      "learning_rate": 3.532586706443846e-06,
+      "loss": 0.4451,
+      "step": 4544
+    },
+    {
+      "epoch": 3.0564007733041945,
+      "grad_norm": 0.31043316844484037,
+      "learning_rate": 3.5304822581968214e-06,
+      "loss": 0.4321,
+      "step": 4545
+    },
+    {
+      "epoch": 3.0570732117340507,
+      "grad_norm": 0.44760175036381356,
+      "learning_rate": 3.528378094845969e-06,
+      "loss": 0.4357,
+      "step": 4546
+    },
+    {
+      "epoch": 3.057745650163907,
+      "grad_norm": 0.3789634341908856,
+      "learning_rate": 3.5262742167992265e-06,
+      "loss": 0.4279,
+      "step": 4547
+    },
+    {
+      "epoch": 3.058418088593763,
+      "grad_norm": 0.30327069801108236,
+      "learning_rate": 3.5241706244644724e-06,
+      "loss": 0.434,
+      "step": 4548
+    },
+    {
+      "epoch": 3.0590905270236193,
+      "grad_norm": 0.3517385667754944,
+      "learning_rate": 3.5220673182495346e-06,
+      "loss": 0.4369,
+      "step": 4549
+    },
+    {
+      "epoch": 3.0597629654534755,
+      "grad_norm": 0.3115038237779994,
+      "learning_rate": 3.5199642985621775e-06,
+      "loss": 0.4481,
+      "step": 4550
+    },
+    {
+      "epoch": 3.060435403883332,
+      "grad_norm": 0.4064540175949745,
+      "learning_rate": 3.517861565810118e-06,
+      "loss": 0.4277,
+      "step": 4551
+    },
+    {
+      "epoch": 3.0611078423131883,
+      "grad_norm": 0.40234286779242157,
+      "learning_rate": 3.5157591204010123e-06,
+      "loss": 0.4371,
+      "step": 4552
+    },
+    {
+      "epoch": 3.0617802807430445,
+      "grad_norm": 0.4273834795088497,
+      "learning_rate": 3.513656962742463e-06,
+      "loss": 0.4359,
+      "step": 4553
+    },
+    {
+      "epoch": 3.0624527191729007,
+      "grad_norm": 0.4395651494398156,
+      "learning_rate": 3.5115550932420194e-06,
+      "loss": 0.4392,
+      "step": 4554
+    },
+    {
+      "epoch": 3.063125157602757,
+      "grad_norm": 0.47998674278000614,
+      "learning_rate": 3.509453512307167e-06,
+      "loss": 0.4382,
+      "step": 4555
+    },
+    {
+      "epoch": 3.063797596032613,
+      "grad_norm": 0.2963015918779441,
+      "learning_rate": 3.507352220345343e-06,
+      "loss": 0.4365,
+      "step": 4556
+    },
+    {
+      "epoch": 3.0644700344624693,
+      "grad_norm": 0.3417701165760403,
+      "learning_rate": 3.505251217763925e-06,
+      "loss": 0.433,
+      "step": 4557
+    },
+    {
+      "epoch": 3.065142472892326,
+      "grad_norm": 0.5514900169517631,
+      "learning_rate": 3.5031505049702388e-06,
+      "loss": 0.4288,
+      "step": 4558
+    },
+    {
+      "epoch": 3.065814911322182,
+      "grad_norm": 0.3463449584856086,
+      "learning_rate": 3.5010500823715453e-06,
+      "loss": 0.4267,
+      "step": 4559
+    },
+    {
+      "epoch": 3.0664873497520384,
+      "grad_norm": 0.3585885707085387,
+      "learning_rate": 3.498949950375059e-06,
+      "loss": 0.4235,
+      "step": 4560
+    },
+    {
+      "epoch": 3.0671597881818946,
+      "grad_norm": 0.3581184175510112,
+      "learning_rate": 3.496850109387931e-06,
+      "loss": 0.4363,
+      "step": 4561
+    },
+    {
+      "epoch": 3.067832226611751,
+      "grad_norm": 0.3978062858624349,
+      "learning_rate": 3.494750559817259e-06,
+      "loss": 0.4301,
+      "step": 4562
+    },
+    {
+      "epoch": 3.068504665041607,
+      "grad_norm": 0.3134466285837552,
+      "learning_rate": 3.4926513020700862e-06,
+      "loss": 0.4457,
+      "step": 4563
+    },
+    {
+      "epoch": 3.069177103471463,
+      "grad_norm": 0.43058514574800544,
+      "learning_rate": 3.4905523365533933e-06,
+      "loss": 0.4355,
+      "step": 4564
+    },
+    {
+      "epoch": 3.06984954190132,
+      "grad_norm": 0.34477356186422137,
+      "learning_rate": 3.48845366367411e-06,
+      "loss": 0.4391,
+      "step": 4565
+    },
+    {
+      "epoch": 3.070521980331176,
+      "grad_norm": 0.3388934365750416,
+      "learning_rate": 3.4863552838391063e-06,
+      "loss": 0.4331,
+      "step": 4566
+    },
+    {
+      "epoch": 3.0711944187610323,
+      "grad_norm": 0.4684156526808039,
+      "learning_rate": 3.4842571974551988e-06,
+      "loss": 0.4234,
+      "step": 4567
+    },
+    {
+      "epoch": 3.0718668571908885,
+      "grad_norm": 0.3313737505540861,
+      "learning_rate": 3.4821594049291397e-06,
+      "loss": 0.4276,
+      "step": 4568
+    },
+    {
+      "epoch": 3.0725392956207447,
+      "grad_norm": 0.36942260054250975,
+      "learning_rate": 3.480061906667631e-06,
+      "loss": 0.4314,
+      "step": 4569
+    },
+    {
+      "epoch": 3.073211734050601,
+      "grad_norm": 0.32521269600746494,
+      "learning_rate": 3.477964703077318e-06,
+      "loss": 0.4495,
+      "step": 4570
+    },
+    {
+      "epoch": 3.073884172480457,
+      "grad_norm": 0.3608809742566614,
+      "learning_rate": 3.4758677945647845e-06,
+      "loss": 0.4246,
+      "step": 4571
+    },
+    {
+      "epoch": 3.0745566109103137,
+      "grad_norm": 0.3344989550477347,
+      "learning_rate": 3.47377118153656e-06,
+      "loss": 0.4497,
+      "step": 4572
+    },
+    {
+      "epoch": 3.07522904934017,
+      "grad_norm": 0.3572676656882878,
+      "learning_rate": 3.4716748643991156e-06,
+      "loss": 0.431,
+      "step": 4573
+    },
+    {
+      "epoch": 3.075901487770026,
+      "grad_norm": 0.553248539316614,
+      "learning_rate": 3.469578843558865e-06,
+      "loss": 0.4304,
+      "step": 4574
+    },
+    {
+      "epoch": 3.0765739261998823,
+      "grad_norm": 0.35331663767167226,
+      "learning_rate": 3.4674831194221664e-06,
+      "loss": 0.4319,
+      "step": 4575
+    },
+    {
+      "epoch": 3.0772463646297386,
+      "grad_norm": 0.42518322373957207,
+      "learning_rate": 3.46538769239532e-06,
+      "loss": 0.4308,
+      "step": 4576
+    },
+    {
+      "epoch": 3.0779188030595948,
+      "grad_norm": 0.3338628786350596,
+      "learning_rate": 3.4632925628845627e-06,
+      "loss": 0.4416,
+      "step": 4577
+    },
+    {
+      "epoch": 3.078591241489451,
+      "grad_norm": 0.32805731158939805,
+      "learning_rate": 3.461197731296081e-06,
+      "loss": 0.4427,
+      "step": 4578
+    },
+    {
+      "epoch": 3.0792636799193076,
+      "grad_norm": 0.36026355171320523,
+      "learning_rate": 3.4591031980360014e-06,
+      "loss": 0.4369,
+      "step": 4579
+    },
+    {
+      "epoch": 3.079936118349164,
+      "grad_norm": 0.30893179819019195,
+      "learning_rate": 3.4570089635103934e-06,
+      "loss": 0.4347,
+      "step": 4580
+    },
+    {
+      "epoch": 3.08060855677902,
+      "grad_norm": 0.4345687627764726,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.4427,
+      "step": 4581
+    },
+    {
+      "epoch": 3.081280995208876,
+      "grad_norm": 0.40103013471154253,
+      "learning_rate": 3.452821392286567e-06,
+      "loss": 0.4304,
+      "step": 4582
+    },
+    {
+      "epoch": 3.0819534336387324,
+      "grad_norm": 0.30194327621664857,
+      "learning_rate": 3.4507280564001968e-06,
+      "loss": 0.4434,
+      "step": 4583
+    },
+    {
+      "epoch": 3.0826258720685886,
+      "grad_norm": 0.332364236177735,
+      "learning_rate": 3.44863502087199e-06,
+      "loss": 0.4478,
+      "step": 4584
+    },
+    {
+      "epoch": 3.083298310498445,
+      "grad_norm": 0.33594200809082103,
+      "learning_rate": 3.4465422861077267e-06,
+      "loss": 0.4263,
+      "step": 4585
+    },
+    {
+      "epoch": 3.0839707489283015,
+      "grad_norm": 0.3689223762605296,
+      "learning_rate": 3.4444498525131215e-06,
+      "loss": 0.446,
+      "step": 4586
+    },
+    {
+      "epoch": 3.0846431873581577,
+      "grad_norm": 0.7450107564695669,
+      "learning_rate": 3.442357720493838e-06,
+      "loss": 0.4424,
+      "step": 4587
+    },
+    {
+      "epoch": 3.085315625788014,
+      "grad_norm": 0.3466911797566331,
+      "learning_rate": 3.4402658904554785e-06,
+      "loss": 0.4394,
+      "step": 4588
+    },
+    {
+      "epoch": 3.08598806421787,
+      "grad_norm": 0.34958332313672374,
+      "learning_rate": 3.43817436280359e-06,
+      "loss": 0.4432,
+      "step": 4589
+    },
+    {
+      "epoch": 3.0866605026477263,
+      "grad_norm": 0.42559182941801493,
+      "learning_rate": 3.4360831379436533e-06,
+      "loss": 0.4315,
+      "step": 4590
+    },
+    {
+      "epoch": 3.0873329410775825,
+      "grad_norm": 0.33547003791092794,
+      "learning_rate": 3.4339922162810983e-06,
+      "loss": 0.4357,
+      "step": 4591
+    },
+    {
+      "epoch": 3.0880053795074387,
+      "grad_norm": 0.38646265904258303,
+      "learning_rate": 3.4319015982212922e-06,
+      "loss": 0.4434,
+      "step": 4592
+    },
+    {
+      "epoch": 3.0886778179372953,
+      "grad_norm": 0.35206075599972964,
+      "learning_rate": 3.4298112841695477e-06,
+      "loss": 0.4416,
+      "step": 4593
+    },
+    {
+      "epoch": 3.0893502563671515,
+      "grad_norm": 0.34848617744389093,
+      "learning_rate": 3.4277212745311084e-06,
+      "loss": 0.4366,
+      "step": 4594
+    },
+    {
+      "epoch": 3.0900226947970078,
+      "grad_norm": 0.3533361661758744,
+      "learning_rate": 3.425631569711169e-06,
+      "loss": 0.4401,
+      "step": 4595
+    },
+    {
+      "epoch": 3.090695133226864,
+      "grad_norm": 0.36406170069571986,
+      "learning_rate": 3.4235421701148625e-06,
+      "loss": 0.4273,
+      "step": 4596
+    },
+    {
+      "epoch": 3.09136757165672,
+      "grad_norm": 1.4039836290122003,
+      "learning_rate": 3.4214530761472607e-06,
+      "loss": 0.4357,
+      "step": 4597
+    },
+    {
+      "epoch": 3.0920400100865764,
+      "grad_norm": 0.3308515481068115,
+      "learning_rate": 3.419364288213379e-06,
+      "loss": 0.4441,
+      "step": 4598
+    },
+    {
+      "epoch": 3.0927124485164326,
+      "grad_norm": 0.3119654188294005,
+      "learning_rate": 3.4172758067181687e-06,
+      "loss": 0.4208,
+      "step": 4599
+    },
+    {
+      "epoch": 3.0933848869462888,
+      "grad_norm": 0.2867191112981178,
+      "learning_rate": 3.4151876320665276e-06,
+      "loss": 0.4422,
+      "step": 4600
+    },
+    {
+      "epoch": 3.0940573253761454,
+      "grad_norm": 0.35474956609831393,
+      "learning_rate": 3.4130997646632895e-06,
+      "loss": 0.4314,
+      "step": 4601
+    },
+    {
+      "epoch": 3.0947297638060016,
+      "grad_norm": 0.3159002316785881,
+      "learning_rate": 3.411012204913233e-06,
+      "loss": 0.4459,
+      "step": 4602
+    },
+    {
+      "epoch": 3.095402202235858,
+      "grad_norm": 0.3194766649563873,
+      "learning_rate": 3.4089249532210717e-06,
+      "loss": 0.4496,
+      "step": 4603
+    },
+    {
+      "epoch": 3.096074640665714,
+      "grad_norm": 0.6258388924333276,
+      "learning_rate": 3.406838009991462e-06,
+      "loss": 0.4433,
+      "step": 4604
+    },
+    {
+      "epoch": 3.0967470790955702,
+      "grad_norm": 0.3838652445382094,
+      "learning_rate": 3.404751375629002e-06,
+      "loss": 0.4416,
+      "step": 4605
+    },
+    {
+      "epoch": 3.0974195175254264,
+      "grad_norm": 0.32257798582969266,
+      "learning_rate": 3.402665050538228e-06,
+      "loss": 0.4256,
+      "step": 4606
+    },
+    {
+      "epoch": 3.0980919559552826,
+      "grad_norm": 0.31236500506479076,
+      "learning_rate": 3.4005790351236185e-06,
+      "loss": 0.4342,
+      "step": 4607
+    },
+    {
+      "epoch": 3.0987643943851393,
+      "grad_norm": 0.6161557458503163,
+      "learning_rate": 3.3984933297895876e-06,
+      "loss": 0.4219,
+      "step": 4608
+    },
+    {
+      "epoch": 3.0994368328149955,
+      "grad_norm": 0.31925387633231034,
+      "learning_rate": 3.3964079349404937e-06,
+      "loss": 0.4293,
+      "step": 4609
+    },
+    {
+      "epoch": 3.1001092712448517,
+      "grad_norm": 0.43522890072392556,
+      "learning_rate": 3.3943228509806337e-06,
+      "loss": 0.4343,
+      "step": 4610
+    },
+    {
+      "epoch": 3.100781709674708,
+      "grad_norm": 0.345784014529943,
+      "learning_rate": 3.392238078314245e-06,
+      "loss": 0.4335,
+      "step": 4611
+    },
+    {
+      "epoch": 3.101454148104564,
+      "grad_norm": 0.42119438239765744,
+      "learning_rate": 3.390153617345501e-06,
+      "loss": 0.4222,
+      "step": 4612
+    },
+    {
+      "epoch": 3.1021265865344203,
+      "grad_norm": 0.32674762135053875,
+      "learning_rate": 3.388069468478516e-06,
+      "loss": 0.4317,
+      "step": 4613
+    },
+    {
+      "epoch": 3.1027990249642765,
+      "grad_norm": 0.4307593993758202,
+      "learning_rate": 3.385985632117349e-06,
+      "loss": 0.4522,
+      "step": 4614
+    },
+    {
+      "epoch": 3.103471463394133,
+      "grad_norm": 0.432891284456854,
+      "learning_rate": 3.3839021086659944e-06,
+      "loss": 0.4315,
+      "step": 4615
+    },
+    {
+      "epoch": 3.1041439018239894,
+      "grad_norm": 0.48510915643605573,
+      "learning_rate": 3.3818188985283827e-06,
+      "loss": 0.4209,
+      "step": 4616
+    },
+    {
+      "epoch": 3.1048163402538456,
+      "grad_norm": 0.5217158789525718,
+      "learning_rate": 3.379736002108388e-06,
+      "loss": 0.4352,
+      "step": 4617
+    },
+    {
+      "epoch": 3.1054887786837018,
+      "grad_norm": 0.38286786080522917,
+      "learning_rate": 3.3776534198098245e-06,
+      "loss": 0.4428,
+      "step": 4618
+    },
+    {
+      "epoch": 3.106161217113558,
+      "grad_norm": 0.39261664609612695,
+      "learning_rate": 3.375571152036441e-06,
+      "loss": 0.4394,
+      "step": 4619
+    },
+    {
+      "epoch": 3.106833655543414,
+      "grad_norm": 0.9636478000474139,
+      "learning_rate": 3.37348919919193e-06,
+      "loss": 0.4277,
+      "step": 4620
+    },
+    {
+      "epoch": 3.1075060939732704,
+      "grad_norm": 0.5307538668903624,
+      "learning_rate": 3.371407561679919e-06,
+      "loss": 0.4312,
+      "step": 4621
+    },
+    {
+      "epoch": 3.108178532403127,
+      "grad_norm": 0.39308857470248204,
+      "learning_rate": 3.3693262399039765e-06,
+      "loss": 0.422,
+      "step": 4622
+    },
+    {
+      "epoch": 3.1088509708329832,
+      "grad_norm": 0.42116258159068504,
+      "learning_rate": 3.3672452342676104e-06,
+      "loss": 0.4508,
+      "step": 4623
+    },
+    {
+      "epoch": 3.1095234092628394,
+      "grad_norm": 0.34333009940872494,
+      "learning_rate": 3.3651645451742677e-06,
+      "loss": 0.4381,
+      "step": 4624
+    },
+    {
+      "epoch": 3.1101958476926956,
+      "grad_norm": 0.32574292490361406,
+      "learning_rate": 3.363084173027329e-06,
+      "loss": 0.4373,
+      "step": 4625
+    },
+    {
+      "epoch": 3.110868286122552,
+      "grad_norm": 0.47585240478550994,
+      "learning_rate": 3.3610041182301185e-06,
+      "loss": 0.4407,
+      "step": 4626
+    },
+    {
+      "epoch": 3.111540724552408,
+      "grad_norm": 0.30577248963550213,
+      "learning_rate": 3.358924381185898e-06,
+      "loss": 0.4431,
+      "step": 4627
+    },
+    {
+      "epoch": 3.1122131629822642,
+      "grad_norm": 0.34454421287433445,
+      "learning_rate": 3.3568449622978672e-06,
+      "loss": 0.4249,
+      "step": 4628
+    },
+    {
+      "epoch": 3.112885601412121,
+      "grad_norm": 0.3124320595245157,
+      "learning_rate": 3.354765861969165e-06,
+      "loss": 0.428,
+      "step": 4629
+    },
+    {
+      "epoch": 3.113558039841977,
+      "grad_norm": 0.33255926005497843,
+      "learning_rate": 3.352687080602866e-06,
+      "loss": 0.4339,
+      "step": 4630
+    },
+    {
+      "epoch": 3.1142304782718333,
+      "grad_norm": 0.3263425191480317,
+      "learning_rate": 3.3506086186019853e-06,
+      "loss": 0.4396,
+      "step": 4631
+    },
+    {
+      "epoch": 3.1149029167016895,
+      "grad_norm": 0.3311160032209623,
+      "learning_rate": 3.348530476369476e-06,
+      "loss": 0.4413,
+      "step": 4632
+    },
+    {
+      "epoch": 3.1155753551315457,
+      "grad_norm": 0.3542262654933314,
+      "learning_rate": 3.3464526543082294e-06,
+      "loss": 0.4459,
+      "step": 4633
+    },
+    {
+      "epoch": 3.116247793561402,
+      "grad_norm": 0.325571930550553,
+      "learning_rate": 3.344375152821071e-06,
+      "loss": 0.4362,
+      "step": 4634
+    },
+    {
+      "epoch": 3.116920231991258,
+      "grad_norm": 0.3358903620675028,
+      "learning_rate": 3.3422979723107684e-06,
+      "loss": 0.4424,
+      "step": 4635
+    },
+    {
+      "epoch": 3.1175926704211148,
+      "grad_norm": 0.305451735662324,
+      "learning_rate": 3.3402211131800267e-06,
+      "loss": 0.4259,
+      "step": 4636
+    },
+    {
+      "epoch": 3.118265108850971,
+      "grad_norm": 0.42671107335206243,
+      "learning_rate": 3.338144575831488e-06,
+      "loss": 0.4215,
+      "step": 4637
+    },
+    {
+      "epoch": 3.118937547280827,
+      "grad_norm": 0.43720263887286137,
+      "learning_rate": 3.336068360667729e-06,
+      "loss": 0.4331,
+      "step": 4638
+    },
+    {
+      "epoch": 3.1196099857106834,
+      "grad_norm": 0.32109468966367455,
+      "learning_rate": 3.333992468091268e-06,
+      "loss": 0.4357,
+      "step": 4639
+    },
+    {
+      "epoch": 3.1202824241405396,
+      "grad_norm": 0.3038497432055437,
+      "learning_rate": 3.3319168985045613e-06,
+      "loss": 0.4396,
+      "step": 4640
+    },
+    {
+      "epoch": 3.120954862570396,
+      "grad_norm": 0.34822304470317245,
+      "learning_rate": 3.329841652309998e-06,
+      "loss": 0.4324,
+      "step": 4641
+    },
+    {
+      "epoch": 3.121627301000252,
+      "grad_norm": 0.33377784623106194,
+      "learning_rate": 3.3277667299099104e-06,
+      "loss": 0.4339,
+      "step": 4642
+    },
+    {
+      "epoch": 3.1222997394301086,
+      "grad_norm": 0.3474041561897187,
+      "learning_rate": 3.3256921317065603e-06,
+      "loss": 0.4531,
+      "step": 4643
+    },
+    {
+      "epoch": 3.122972177859965,
+      "grad_norm": 0.35275400166282933,
+      "learning_rate": 3.3236178581021543e-06,
+      "loss": 0.4435,
+      "step": 4644
+    },
+    {
+      "epoch": 3.123644616289821,
+      "grad_norm": 0.3766660078631167,
+      "learning_rate": 3.3215439094988315e-06,
+      "loss": 0.4408,
+      "step": 4645
+    },
+    {
+      "epoch": 3.1243170547196772,
+      "grad_norm": 0.34077492250456387,
+      "learning_rate": 3.3194702862986716e-06,
+      "loss": 0.4237,
+      "step": 4646
+    },
+    {
+      "epoch": 3.1249894931495334,
+      "grad_norm": 0.3990472748064627,
+      "learning_rate": 3.3173969889036858e-06,
+      "loss": 0.4252,
+      "step": 4647
+    },
+    {
+      "epoch": 3.1256619315793897,
+      "grad_norm": 0.3324141727579477,
+      "learning_rate": 3.3153240177158267e-06,
+      "loss": 0.4459,
+      "step": 4648
+    },
+    {
+      "epoch": 3.126334370009246,
+      "grad_norm": 0.3740768618563349,
+      "learning_rate": 3.3132513731369832e-06,
+      "loss": 0.4497,
+      "step": 4649
+    },
+    {
+      "epoch": 3.127006808439102,
+      "grad_norm": 0.5408116642166466,
+      "learning_rate": 3.311179055568979e-06,
+      "loss": 0.433,
+      "step": 4650
+    },
+    {
+      "epoch": 3.1276792468689587,
+      "grad_norm": 0.29868137271676387,
+      "learning_rate": 3.3091070654135777e-06,
+      "loss": 0.4381,
+      "step": 4651
+    },
+    {
+      "epoch": 3.128351685298815,
+      "grad_norm": 0.31638763072719994,
+      "learning_rate": 3.3070354030724735e-06,
+      "loss": 0.4414,
+      "step": 4652
+    },
+    {
+      "epoch": 3.129024123728671,
+      "grad_norm": 0.3291056686507435,
+      "learning_rate": 3.3049640689473015e-06,
+      "loss": 0.4389,
+      "step": 4653
+    },
+    {
+      "epoch": 3.1296965621585273,
+      "grad_norm": 0.3468954593487525,
+      "learning_rate": 3.302893063439634e-06,
+      "loss": 0.4297,
+      "step": 4654
+    },
+    {
+      "epoch": 3.1303690005883835,
+      "grad_norm": 0.3827760044042586,
+      "learning_rate": 3.300822386950978e-06,
+      "loss": 0.4352,
+      "step": 4655
+    },
+    {
+      "epoch": 3.1310414390182397,
+      "grad_norm": 0.35013542676495674,
+      "learning_rate": 3.298752039882774e-06,
+      "loss": 0.4232,
+      "step": 4656
+    },
+    {
+      "epoch": 3.131713877448096,
+      "grad_norm": 0.3148957999424402,
+      "learning_rate": 3.2966820226364037e-06,
+      "loss": 0.4345,
+      "step": 4657
+    },
+    {
+      "epoch": 3.1323863158779526,
+      "grad_norm": 0.31776486418488736,
+      "learning_rate": 3.2946123356131817e-06,
+      "loss": 0.4474,
+      "step": 4658
+    },
+    {
+      "epoch": 3.133058754307809,
+      "grad_norm": 0.336646129660973,
+      "learning_rate": 3.292542979214361e-06,
+      "loss": 0.4159,
+      "step": 4659
+    },
+    {
+      "epoch": 3.133731192737665,
+      "grad_norm": 0.3276513923565536,
+      "learning_rate": 3.2904739538411253e-06,
+      "loss": 0.4296,
+      "step": 4660
+    },
+    {
+      "epoch": 3.134403631167521,
+      "grad_norm": 0.4116312387173316,
+      "learning_rate": 3.288405259894599e-06,
+      "loss": 0.4351,
+      "step": 4661
+    },
+    {
+      "epoch": 3.1350760695973774,
+      "grad_norm": 0.3403757223273257,
+      "learning_rate": 3.2863368977758408e-06,
+      "loss": 0.4371,
+      "step": 4662
+    },
+    {
+      "epoch": 3.1357485080272336,
+      "grad_norm": 0.3659654866043058,
+      "learning_rate": 3.284268867885847e-06,
+      "loss": 0.4295,
+      "step": 4663
+    },
+    {
+      "epoch": 3.13642094645709,
+      "grad_norm": 0.4652447138237276,
+      "learning_rate": 3.2822011706255473e-06,
+      "loss": 0.4324,
+      "step": 4664
+    },
+    {
+      "epoch": 3.1370933848869464,
+      "grad_norm": 0.33165855878747724,
+      "learning_rate": 3.2801338063958055e-06,
+      "loss": 0.4229,
+      "step": 4665
+    },
+    {
+      "epoch": 3.1377658233168026,
+      "grad_norm": 0.3838062127845923,
+      "learning_rate": 3.278066775597423e-06,
+      "loss": 0.4432,
+      "step": 4666
+    },
+    {
+      "epoch": 3.138438261746659,
+      "grad_norm": 0.3439324081941863,
+      "learning_rate": 3.276000078631136e-06,
+      "loss": 0.4308,
+      "step": 4667
+    },
+    {
+      "epoch": 3.139110700176515,
+      "grad_norm": 0.3777137266840129,
+      "learning_rate": 3.2739337158976183e-06,
+      "loss": 0.438,
+      "step": 4668
+    },
+    {
+      "epoch": 3.1397831386063713,
+      "grad_norm": 0.3770860508041263,
+      "learning_rate": 3.2718676877974732e-06,
+      "loss": 0.4284,
+      "step": 4669
+    },
+    {
+      "epoch": 3.1404555770362275,
+      "grad_norm": 0.4052584725799753,
+      "learning_rate": 3.2698019947312447e-06,
+      "loss": 0.419,
+      "step": 4670
+    },
+    {
+      "epoch": 3.1411280154660837,
+      "grad_norm": 0.33877127331626405,
+      "learning_rate": 3.26773663709941e-06,
+      "loss": 0.4375,
+      "step": 4671
+    },
+    {
+      "epoch": 3.1418004538959403,
+      "grad_norm": 0.38442473662802157,
+      "learning_rate": 3.2656716153023806e-06,
+      "loss": 0.4204,
+      "step": 4672
+    },
+    {
+      "epoch": 3.1424728923257965,
+      "grad_norm": 0.38480592954673515,
+      "learning_rate": 3.263606929740505e-06,
+      "loss": 0.4339,
+      "step": 4673
+    },
+    {
+      "epoch": 3.1431453307556527,
+      "grad_norm": 0.4169876926992838,
+      "learning_rate": 3.2615425808140617e-06,
+      "loss": 0.4431,
+      "step": 4674
+    },
+    {
+      "epoch": 3.143817769185509,
+      "grad_norm": 0.3583867437549412,
+      "learning_rate": 3.259478568923269e-06,
+      "loss": 0.4353,
+      "step": 4675
+    },
+    {
+      "epoch": 3.144490207615365,
+      "grad_norm": 0.36524560376873244,
+      "learning_rate": 3.257414894468277e-06,
+      "loss": 0.4236,
+      "step": 4676
+    },
+    {
+      "epoch": 3.1451626460452213,
+      "grad_norm": 0.4022822932069097,
+      "learning_rate": 3.2553515578491727e-06,
+      "loss": 0.4367,
+      "step": 4677
+    },
+    {
+      "epoch": 3.1458350844750775,
+      "grad_norm": 0.34481558379751426,
+      "learning_rate": 3.2532885594659756e-06,
+      "loss": 0.4174,
+      "step": 4678
+    },
+    {
+      "epoch": 3.146507522904934,
+      "grad_norm": 0.3678747660739935,
+      "learning_rate": 3.2512258997186396e-06,
+      "loss": 0.4448,
+      "step": 4679
+    },
+    {
+      "epoch": 3.1471799613347904,
+      "grad_norm": 0.36854708191939056,
+      "learning_rate": 3.249163579007054e-06,
+      "loss": 0.4303,
+      "step": 4680
+    },
+    {
+      "epoch": 3.1478523997646466,
+      "grad_norm": 0.33787152789012104,
+      "learning_rate": 3.247101597731045e-06,
+      "loss": 0.4429,
+      "step": 4681
+    },
+    {
+      "epoch": 3.148524838194503,
+      "grad_norm": 0.3562201680510042,
+      "learning_rate": 3.245039956290365e-06,
+      "loss": 0.4181,
+      "step": 4682
+    },
+    {
+      "epoch": 3.149197276624359,
+      "grad_norm": 0.3401836758777454,
+      "learning_rate": 3.2429786550847074e-06,
+      "loss": 0.4351,
+      "step": 4683
+    },
+    {
+      "epoch": 3.149869715054215,
+      "grad_norm": 0.4177430294864894,
+      "learning_rate": 3.240917694513699e-06,
+      "loss": 0.4335,
+      "step": 4684
+    },
+    {
+      "epoch": 3.1505421534840714,
+      "grad_norm": 0.4272694717550732,
+      "learning_rate": 3.238857074976898e-06,
+      "loss": 0.4403,
+      "step": 4685
+    },
+    {
+      "epoch": 3.151214591913928,
+      "grad_norm": 0.5249306400567656,
+      "learning_rate": 3.2367967968737982e-06,
+      "loss": 0.4442,
+      "step": 4686
+    },
+    {
+      "epoch": 3.1518870303437843,
+      "grad_norm": 0.29953421863302154,
+      "learning_rate": 3.2347368606038265e-06,
+      "loss": 0.4291,
+      "step": 4687
+    },
+    {
+      "epoch": 3.1525594687736405,
+      "grad_norm": 0.37863611117307133,
+      "learning_rate": 3.2326772665663443e-06,
+      "loss": 0.4394,
+      "step": 4688
+    },
+    {
+      "epoch": 3.1532319072034967,
+      "grad_norm": 0.3344756042740373,
+      "learning_rate": 3.230618015160646e-06,
+      "loss": 0.4205,
+      "step": 4689
+    },
+    {
+      "epoch": 3.153904345633353,
+      "grad_norm": 0.3554230676368455,
+      "learning_rate": 3.228559106785961e-06,
+      "loss": 0.4313,
+      "step": 4690
+    },
+    {
+      "epoch": 3.154576784063209,
+      "grad_norm": 0.36182306292558464,
+      "learning_rate": 3.2265005418414486e-06,
+      "loss": 0.4412,
+      "step": 4691
+    },
+    {
+      "epoch": 3.1552492224930653,
+      "grad_norm": 0.3154854946548312,
+      "learning_rate": 3.2244423207262047e-06,
+      "loss": 0.4303,
+      "step": 4692
+    },
+    {
+      "epoch": 3.155921660922922,
+      "grad_norm": 0.42983282243562276,
+      "learning_rate": 3.2223844438392583e-06,
+      "loss": 0.4446,
+      "step": 4693
+    },
+    {
+      "epoch": 3.156594099352778,
+      "grad_norm": 0.30269250195619035,
+      "learning_rate": 3.220326911579571e-06,
+      "loss": 0.432,
+      "step": 4694
+    },
+    {
+      "epoch": 3.1572665377826343,
+      "grad_norm": 0.424479235474068,
+      "learning_rate": 3.218269724346037e-06,
+      "loss": 0.4561,
+      "step": 4695
+    },
+    {
+      "epoch": 3.1579389762124905,
+      "grad_norm": 0.33574116245111074,
+      "learning_rate": 3.216212882537484e-06,
+      "loss": 0.4264,
+      "step": 4696
+    },
+    {
+      "epoch": 3.1586114146423467,
+      "grad_norm": 0.3498294739538529,
+      "learning_rate": 3.214156386552674e-06,
+      "loss": 0.4368,
+      "step": 4697
+    },
+    {
+      "epoch": 3.159283853072203,
+      "grad_norm": 0.37375534804380817,
+      "learning_rate": 3.2121002367903005e-06,
+      "loss": 0.4257,
+      "step": 4698
+    },
+    {
+      "epoch": 3.159956291502059,
+      "grad_norm": 0.3090641749528765,
+      "learning_rate": 3.2100444336489923e-06,
+      "loss": 0.4339,
+      "step": 4699
+    },
+    {
+      "epoch": 3.160628729931916,
+      "grad_norm": 0.9572396401829548,
+      "learning_rate": 3.2079889775273053e-06,
+      "loss": 0.4307,
+      "step": 4700
+    },
+    {
+      "epoch": 3.161301168361772,
+      "grad_norm": 0.45192994900712563,
+      "learning_rate": 3.205933868823734e-06,
+      "loss": 0.4384,
+      "step": 4701
+    },
+    {
+      "epoch": 3.161973606791628,
+      "grad_norm": 0.36941191503540555,
+      "learning_rate": 3.2038791079367025e-06,
+      "loss": 0.4322,
+      "step": 4702
+    },
+    {
+      "epoch": 3.1626460452214844,
+      "grad_norm": 4.38470789270265,
+      "learning_rate": 3.20182469526457e-06,
+      "loss": 0.4199,
+      "step": 4703
+    },
+    {
+      "epoch": 3.1633184836513406,
+      "grad_norm": 0.34675448007122195,
+      "learning_rate": 3.1997706312056254e-06,
+      "loss": 0.4402,
+      "step": 4704
+    },
+    {
+      "epoch": 3.163990922081197,
+      "grad_norm": 0.40977690757777413,
+      "learning_rate": 3.1977169161580905e-06,
+      "loss": 0.4452,
+      "step": 4705
+    },
+    {
+      "epoch": 3.164663360511053,
+      "grad_norm": 0.4220151900113677,
+      "learning_rate": 3.1956635505201217e-06,
+      "loss": 0.4345,
+      "step": 4706
+    },
+    {
+      "epoch": 3.1653357989409097,
+      "grad_norm": 0.478450870024991,
+      "learning_rate": 3.193610534689805e-06,
+      "loss": 0.4352,
+      "step": 4707
+    },
+    {
+      "epoch": 3.166008237370766,
+      "grad_norm": 0.3188804493495357,
+      "learning_rate": 3.1915578690651614e-06,
+      "loss": 0.4395,
+      "step": 4708
+    },
+    {
+      "epoch": 3.166680675800622,
+      "grad_norm": 0.3183761201737437,
+      "learning_rate": 3.189505554044139e-06,
+      "loss": 0.4423,
+      "step": 4709
+    },
+    {
+      "epoch": 3.1673531142304783,
+      "grad_norm": 0.30600258196367147,
+      "learning_rate": 3.1874535900246232e-06,
+      "loss": 0.4316,
+      "step": 4710
+    },
+    {
+      "epoch": 3.1680255526603345,
+      "grad_norm": 0.3562444091319478,
+      "learning_rate": 3.1854019774044293e-06,
+      "loss": 0.4238,
+      "step": 4711
+    },
+    {
+      "epoch": 3.1686979910901907,
+      "grad_norm": 0.3773529419732106,
+      "learning_rate": 3.183350716581305e-06,
+      "loss": 0.4271,
+      "step": 4712
+    },
+    {
+      "epoch": 3.169370429520047,
+      "grad_norm": 0.4118248790288195,
+      "learning_rate": 3.181299807952928e-06,
+      "loss": 0.4305,
+      "step": 4713
+    },
+    {
+      "epoch": 3.1700428679499035,
+      "grad_norm": 0.4317878872718693,
+      "learning_rate": 3.1792492519169094e-06,
+      "loss": 0.4404,
+      "step": 4714
+    },
+    {
+      "epoch": 3.1707153063797597,
+      "grad_norm": 0.3565325816330687,
+      "learning_rate": 3.177199048870792e-06,
+      "loss": 0.4442,
+      "step": 4715
+    },
+    {
+      "epoch": 3.171387744809616,
+      "grad_norm": 0.5234862327575597,
+      "learning_rate": 3.175149199212052e-06,
+      "loss": 0.428,
+      "step": 4716
+    },
+    {
+      "epoch": 3.172060183239472,
+      "grad_norm": 0.38414172568977173,
+      "learning_rate": 3.173099703338089e-06,
+      "loss": 0.4614,
+      "step": 4717
+    },
+    {
+      "epoch": 3.1727326216693283,
+      "grad_norm": 0.43272101090757875,
+      "learning_rate": 3.171050561646244e-06,
+      "loss": 0.4336,
+      "step": 4718
+    },
+    {
+      "epoch": 3.1734050600991845,
+      "grad_norm": 0.3643160602771628,
+      "learning_rate": 3.169001774533785e-06,
+      "loss": 0.4373,
+      "step": 4719
+    },
+    {
+      "epoch": 3.1740774985290408,
+      "grad_norm": 0.388990393612831,
+      "learning_rate": 3.1669533423979105e-06,
+      "loss": 0.4231,
+      "step": 4720
+    },
+    {
+      "epoch": 3.1747499369588974,
+      "grad_norm": 0.38034827680550565,
+      "learning_rate": 3.1649052656357536e-06,
+      "loss": 0.4433,
+      "step": 4721
+    },
+    {
+      "epoch": 3.1754223753887536,
+      "grad_norm": 0.3678383641889589,
+      "learning_rate": 3.1628575446443722e-06,
+      "loss": 0.4442,
+      "step": 4722
+    },
+    {
+      "epoch": 3.17609481381861,
+      "grad_norm": 0.3981919440636663,
+      "learning_rate": 3.16081017982076e-06,
+      "loss": 0.4163,
+      "step": 4723
+    },
+    {
+      "epoch": 3.176767252248466,
+      "grad_norm": 0.38403716118646863,
+      "learning_rate": 3.158763171561842e-06,
+      "loss": 0.4341,
+      "step": 4724
+    },
+    {
+      "epoch": 3.177439690678322,
+      "grad_norm": 0.3583236464931998,
+      "learning_rate": 3.1567165202644733e-06,
+      "loss": 0.441,
+      "step": 4725
+    },
+    {
+      "epoch": 3.1781121291081784,
+      "grad_norm": 0.36116918921683977,
+      "learning_rate": 3.154670226325437e-06,
+      "loss": 0.4566,
+      "step": 4726
+    },
+    {
+      "epoch": 3.1787845675380346,
+      "grad_norm": 0.34424155111991867,
+      "learning_rate": 3.15262429014145e-06,
+      "loss": 0.4348,
+      "step": 4727
+    },
+    {
+      "epoch": 3.1794570059678913,
+      "grad_norm": 0.36997385256532983,
+      "learning_rate": 3.1505787121091595e-06,
+      "loss": 0.4195,
+      "step": 4728
+    },
+    {
+      "epoch": 3.1801294443977475,
+      "grad_norm": 0.3275366678701998,
+      "learning_rate": 3.1485334926251433e-06,
+      "loss": 0.4444,
+      "step": 4729
+    },
+    {
+      "epoch": 3.1808018828276037,
+      "grad_norm": 0.37805491290009186,
+      "learning_rate": 3.1464886320859096e-06,
+      "loss": 0.4262,
+      "step": 4730
+    },
+    {
+      "epoch": 3.18147432125746,
+      "grad_norm": 0.30860103148911594,
+      "learning_rate": 3.1444441308878935e-06,
+      "loss": 0.4292,
+      "step": 4731
+    },
+    {
+      "epoch": 3.182146759687316,
+      "grad_norm": 0.32504355552537767,
+      "learning_rate": 3.142399989427466e-06,
+      "loss": 0.4366,
+      "step": 4732
+    },
+    {
+      "epoch": 3.1828191981171723,
+      "grad_norm": 0.3226580089600235,
+      "learning_rate": 3.1403562081009252e-06,
+      "loss": 0.4199,
+      "step": 4733
+    },
+    {
+      "epoch": 3.1834916365470285,
+      "grad_norm": 0.4895344773329832,
+      "learning_rate": 3.1383127873045016e-06,
+      "loss": 0.4197,
+      "step": 4734
+    },
+    {
+      "epoch": 3.184164074976885,
+      "grad_norm": 0.3564439182384866,
+      "learning_rate": 3.1362697274343512e-06,
+      "loss": 0.4396,
+      "step": 4735
+    },
+    {
+      "epoch": 3.1848365134067413,
+      "grad_norm": 0.37691746552945693,
+      "learning_rate": 3.1342270288865655e-06,
+      "loss": 0.431,
+      "step": 4736
+    },
+    {
+      "epoch": 3.1855089518365975,
+      "grad_norm": 0.4162265918187403,
+      "learning_rate": 3.1321846920571627e-06,
+      "loss": 0.4268,
+      "step": 4737
+    },
+    {
+      "epoch": 3.1861813902664538,
+      "grad_norm": 0.39649204601421295,
+      "learning_rate": 3.1301427173420935e-06,
+      "loss": 0.4447,
+      "step": 4738
+    },
+    {
+      "epoch": 3.18685382869631,
+      "grad_norm": 0.49702336290092375,
+      "learning_rate": 3.1281011051372327e-06,
+      "loss": 0.4637,
+      "step": 4739
+    },
+    {
+      "epoch": 3.187526267126166,
+      "grad_norm": 0.3006983383812492,
+      "learning_rate": 3.1260598558383913e-06,
+      "loss": 0.4224,
+      "step": 4740
+    },
+    {
+      "epoch": 3.1881987055560224,
+      "grad_norm": 0.40406229463083043,
+      "learning_rate": 3.124018969841307e-06,
+      "loss": 0.4494,
+      "step": 4741
+    },
+    {
+      "epoch": 3.188871143985879,
+      "grad_norm": 0.3584595005860944,
+      "learning_rate": 3.121978447541648e-06,
+      "loss": 0.432,
+      "step": 4742
+    },
+    {
+      "epoch": 3.189543582415735,
+      "grad_norm": 0.3686743773892617,
+      "learning_rate": 3.1199382893350115e-06,
+      "loss": 0.4333,
+      "step": 4743
+    },
+    {
+      "epoch": 3.1902160208455914,
+      "grad_norm": 0.3359050197895751,
+      "learning_rate": 3.1178984956169225e-06,
+      "loss": 0.4175,
+      "step": 4744
+    },
+    {
+      "epoch": 3.1908884592754476,
+      "grad_norm": 0.31833285001091,
+      "learning_rate": 3.1158590667828376e-06,
+      "loss": 0.4212,
+      "step": 4745
+    },
+    {
+      "epoch": 3.191560897705304,
+      "grad_norm": 0.4891485403436755,
+      "learning_rate": 3.113820003228142e-06,
+      "loss": 0.4414,
+      "step": 4746
+    },
+    {
+      "epoch": 3.19223333613516,
+      "grad_norm": 0.4269682563934399,
+      "learning_rate": 3.111781305348153e-06,
+      "loss": 0.4449,
+      "step": 4747
+    },
+    {
+      "epoch": 3.1929057745650162,
+      "grad_norm": 0.3584339588087221,
+      "learning_rate": 3.109742973538108e-06,
+      "loss": 0.4255,
+      "step": 4748
+    },
+    {
+      "epoch": 3.193578212994873,
+      "grad_norm": 0.40390970804608317,
+      "learning_rate": 3.1077050081931835e-06,
+      "loss": 0.4405,
+      "step": 4749
+    },
+    {
+      "epoch": 3.194250651424729,
+      "grad_norm": 0.40806604809088715,
+      "learning_rate": 3.10566740970848e-06,
+      "loss": 0.4223,
+      "step": 4750
+    },
+    {
+      "epoch": 3.1949230898545853,
+      "grad_norm": 0.3056085110647207,
+      "learning_rate": 3.103630178479028e-06,
+      "loss": 0.4376,
+      "step": 4751
+    },
+    {
+      "epoch": 3.1955955282844415,
+      "grad_norm": 0.34709925682469184,
+      "learning_rate": 3.1015933148997868e-06,
+      "loss": 0.4374,
+      "step": 4752
+    },
+    {
+      "epoch": 3.1962679667142977,
+      "grad_norm": 0.30168107092410623,
+      "learning_rate": 3.0995568193656435e-06,
+      "loss": 0.4348,
+      "step": 4753
+    },
+    {
+      "epoch": 3.196940405144154,
+      "grad_norm": 0.45845343718385345,
+      "learning_rate": 3.097520692271414e-06,
+      "loss": 0.4438,
+      "step": 4754
+    },
+    {
+      "epoch": 3.19761284357401,
+      "grad_norm": 0.33802532153175735,
+      "learning_rate": 3.0954849340118454e-06,
+      "loss": 0.4197,
+      "step": 4755
+    },
+    {
+      "epoch": 3.1982852820038667,
+      "grad_norm": 0.4163329714506242,
+      "learning_rate": 3.0934495449816117e-06,
+      "loss": 0.439,
+      "step": 4756
+    },
+    {
+      "epoch": 3.198957720433723,
+      "grad_norm": 0.3638694248396867,
+      "learning_rate": 3.09141452557531e-06,
+      "loss": 0.4208,
+      "step": 4757
+    },
+    {
+      "epoch": 3.199630158863579,
+      "grad_norm": 0.36166532189611367,
+      "learning_rate": 3.0893798761874754e-06,
+      "loss": 0.4257,
+      "step": 4758
+    },
+    {
+      "epoch": 3.2003025972934354,
+      "grad_norm": 0.35809717619963644,
+      "learning_rate": 3.0873455972125644e-06,
+      "loss": 0.4317,
+      "step": 4759
+    },
+    {
+      "epoch": 3.2009750357232916,
+      "grad_norm": 0.35426721678748335,
+      "learning_rate": 3.0853116890449646e-06,
+      "loss": 0.4367,
+      "step": 4760
+    },
+    {
+      "epoch": 3.2016474741531478,
+      "grad_norm": 0.3126887790627608,
+      "learning_rate": 3.08327815207899e-06,
+      "loss": 0.4274,
+      "step": 4761
+    },
+    {
+      "epoch": 3.202319912583004,
+      "grad_norm": 0.39450856417405994,
+      "learning_rate": 3.0812449867088833e-06,
+      "loss": 0.4224,
+      "step": 4762
+    },
+    {
+      "epoch": 3.2029923510128606,
+      "grad_norm": 0.32818810265606196,
+      "learning_rate": 3.0792121933288162e-06,
+      "loss": 0.4395,
+      "step": 4763
+    },
+    {
+      "epoch": 3.203664789442717,
+      "grad_norm": 0.35599959312630275,
+      "learning_rate": 3.0771797723328868e-06,
+      "loss": 0.4312,
+      "step": 4764
+    },
+    {
+      "epoch": 3.204337227872573,
+      "grad_norm": 0.4090587251102353,
+      "learning_rate": 3.075147724115124e-06,
+      "loss": 0.4372,
+      "step": 4765
+    },
+    {
+      "epoch": 3.2050096663024292,
+      "grad_norm": 0.468838029036378,
+      "learning_rate": 3.073116049069478e-06,
+      "loss": 0.4244,
+      "step": 4766
+    },
+    {
+      "epoch": 3.2056821047322854,
+      "grad_norm": 0.37953456910976113,
+      "learning_rate": 3.071084747589832e-06,
+      "loss": 0.4327,
+      "step": 4767
+    },
+    {
+      "epoch": 3.2063545431621416,
+      "grad_norm": 0.3686688100535242,
+      "learning_rate": 3.0690538200699973e-06,
+      "loss": 0.4229,
+      "step": 4768
+    },
+    {
+      "epoch": 3.207026981591998,
+      "grad_norm": 0.31807479799014965,
+      "learning_rate": 3.0670232669037112e-06,
+      "loss": 0.4385,
+      "step": 4769
+    },
+    {
+      "epoch": 3.207699420021854,
+      "grad_norm": 0.3307981654169214,
+      "learning_rate": 3.0649930884846348e-06,
+      "loss": 0.4304,
+      "step": 4770
+    },
+    {
+      "epoch": 3.2083718584517107,
+      "grad_norm": 0.4029112853490653,
+      "learning_rate": 3.0629632852063616e-06,
+      "loss": 0.4358,
+      "step": 4771
+    },
+    {
+      "epoch": 3.209044296881567,
+      "grad_norm": 0.3732050035030929,
+      "learning_rate": 3.060933857462411e-06,
+      "loss": 0.421,
+      "step": 4772
+    },
+    {
+      "epoch": 3.209716735311423,
+      "grad_norm": 0.3742535869075959,
+      "learning_rate": 3.058904805646229e-06,
+      "loss": 0.4622,
+      "step": 4773
+    },
+    {
+      "epoch": 3.2103891737412793,
+      "grad_norm": 0.3659673176100718,
+      "learning_rate": 3.0568761301511894e-06,
+      "loss": 0.4347,
+      "step": 4774
+    },
+    {
+      "epoch": 3.2110616121711355,
+      "grad_norm": 0.3358417737771693,
+      "learning_rate": 3.0548478313705917e-06,
+      "loss": 0.4521,
+      "step": 4775
+    },
+    {
+      "epoch": 3.2117340506009917,
+      "grad_norm": 0.4766779055355382,
+      "learning_rate": 3.052819909697663e-06,
+      "loss": 0.4451,
+      "step": 4776
+    },
+    {
+      "epoch": 3.212406489030848,
+      "grad_norm": 0.3932548693099987,
+      "learning_rate": 3.0507923655255588e-06,
+      "loss": 0.4352,
+      "step": 4777
+    },
+    {
+      "epoch": 3.2130789274607046,
+      "grad_norm": 0.3385982906612951,
+      "learning_rate": 3.048765199247361e-06,
+      "loss": 0.431,
+      "step": 4778
+    },
+    {
+      "epoch": 3.2137513658905608,
+      "grad_norm": 0.5054161825659945,
+      "learning_rate": 3.046738411256074e-06,
+      "loss": 0.4342,
+      "step": 4779
+    },
+    {
+      "epoch": 3.214423804320417,
+      "grad_norm": 0.43428867622105366,
+      "learning_rate": 3.044712001944634e-06,
+      "loss": 0.4428,
+      "step": 4780
+    },
+    {
+      "epoch": 3.215096242750273,
+      "grad_norm": 0.5603923248105427,
+      "learning_rate": 3.042685971705903e-06,
+      "loss": 0.4547,
+      "step": 4781
+    },
+    {
+      "epoch": 3.2157686811801294,
+      "grad_norm": 0.35307316001332334,
+      "learning_rate": 3.040660320932668e-06,
+      "loss": 0.4331,
+      "step": 4782
+    },
+    {
+      "epoch": 3.2164411196099856,
+      "grad_norm": 0.3614379883327843,
+      "learning_rate": 3.038635050017642e-06,
+      "loss": 0.4428,
+      "step": 4783
+    },
+    {
+      "epoch": 3.217113558039842,
+      "grad_norm": 0.3314145010300879,
+      "learning_rate": 3.036610159353466e-06,
+      "loss": 0.4218,
+      "step": 4784
+    },
+    {
+      "epoch": 3.2177859964696984,
+      "grad_norm": 0.31371322318633665,
+      "learning_rate": 3.0345856493327066e-06,
+      "loss": 0.4411,
+      "step": 4785
+    },
+    {
+      "epoch": 3.2184584348995546,
+      "grad_norm": 0.3213531388938575,
+      "learning_rate": 3.0325615203478563e-06,
+      "loss": 0.4182,
+      "step": 4786
+    },
+    {
+      "epoch": 3.219130873329411,
+      "grad_norm": 0.41555390009757326,
+      "learning_rate": 3.0305377727913366e-06,
+      "loss": 0.4341,
+      "step": 4787
+    },
+    {
+      "epoch": 3.219803311759267,
+      "grad_norm": 0.5289558142311216,
+      "learning_rate": 3.0285144070554884e-06,
+      "loss": 0.4428,
+      "step": 4788
+    },
+    {
+      "epoch": 3.2204757501891232,
+      "grad_norm": 0.39461029318470636,
+      "learning_rate": 3.0264914235325847e-06,
+      "loss": 0.4287,
+      "step": 4789
+    },
+    {
+      "epoch": 3.2211481886189794,
+      "grad_norm": 0.4359852144425983,
+      "learning_rate": 3.024468822614822e-06,
+      "loss": 0.4289,
+      "step": 4790
+    },
+    {
+      "epoch": 3.2218206270488357,
+      "grad_norm": 0.3223360264707932,
+      "learning_rate": 3.0224466046943245e-06,
+      "loss": 0.4259,
+      "step": 4791
+    },
+    {
+      "epoch": 3.2224930654786923,
+      "grad_norm": 0.3298543675630738,
+      "learning_rate": 3.020424770163138e-06,
+      "loss": 0.4435,
+      "step": 4792
+    },
+    {
+      "epoch": 3.2231655039085485,
+      "grad_norm": 0.3712648442771377,
+      "learning_rate": 3.018403319413238e-06,
+      "loss": 0.4399,
+      "step": 4793
+    },
+    {
+      "epoch": 3.2238379423384047,
+      "grad_norm": 0.35017531257266965,
+      "learning_rate": 3.016382252836525e-06,
+      "loss": 0.4117,
+      "step": 4794
+    },
+    {
+      "epoch": 3.224510380768261,
+      "grad_norm": 0.3906473263678258,
+      "learning_rate": 3.014361570824823e-06,
+      "loss": 0.4473,
+      "step": 4795
+    },
+    {
+      "epoch": 3.225182819198117,
+      "grad_norm": 0.31963992523794227,
+      "learning_rate": 3.012341273769885e-06,
+      "loss": 0.4388,
+      "step": 4796
+    },
+    {
+      "epoch": 3.2258552576279733,
+      "grad_norm": 0.3652391401530828,
+      "learning_rate": 3.010321362063383e-06,
+      "loss": 0.4234,
+      "step": 4797
+    },
+    {
+      "epoch": 3.2265276960578295,
+      "grad_norm": 0.40334376458618504,
+      "learning_rate": 3.0083018360969213e-06,
+      "loss": 0.4531,
+      "step": 4798
+    },
+    {
+      "epoch": 3.227200134487686,
+      "grad_norm": 0.3849107781967533,
+      "learning_rate": 3.0062826962620252e-06,
+      "loss": 0.4355,
+      "step": 4799
+    },
+    {
+      "epoch": 3.2278725729175424,
+      "grad_norm": 0.41994305213491073,
+      "learning_rate": 3.004263942950148e-06,
+      "loss": 0.4298,
+      "step": 4800
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7435,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.434020777725788e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}