diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,23833 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.21789284798769545,
+  "eval_steps": 500,
+  "global_step": 3400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.40861317610869e-05,
+      "grad_norm": 3.049792280054225,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001281722635221738,
+      "grad_norm": 2.736097146503981,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 2
+    },
+    {
+      "epoch": 0.00019225839528326071,
+      "grad_norm": 2.696671878484295,
+      "learning_rate": 1e-06,
+      "loss": 0.4815,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002563445270443476,
+      "grad_norm": 3.337136551938015,
+      "learning_rate": 1e-06,
+      "loss": 0.4904,
+      "step": 4
+    },
+    {
+      "epoch": 0.0003204306588054345,
+      "grad_norm": 2.6647797201470778,
+      "learning_rate": 1e-06,
+      "loss": 0.4185,
+      "step": 5
+    },
+    {
+      "epoch": 0.00038451679056652143,
+      "grad_norm": 2.5193486744426967,
+      "learning_rate": 1e-06,
+      "loss": 0.4688,
+      "step": 6
+    },
+    {
+      "epoch": 0.0004486029223276083,
+      "grad_norm": 2.66396905228593,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005126890540886952,
+      "grad_norm": 3.1409024702898307,
+      "learning_rate": 1e-06,
+      "loss": 0.4413,
+      "step": 8
+    },
+    {
+      "epoch": 0.0005767751858497821,
+      "grad_norm": 2.6431941146744,
+      "learning_rate": 1e-06,
+      "loss": 0.449,
+      "step": 9
+    },
+    {
+      "epoch": 0.000640861317610869,
+      "grad_norm": 2.924241054090828,
+      "learning_rate": 1e-06,
+      "loss": 0.451,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007049474493719559,
+      "grad_norm": 2.622132191231163,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 11
+    },
+    {
+      "epoch": 0.0007690335811330429,
+      "grad_norm": 2.6442231709945525,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 12
+    },
+    {
+      "epoch": 0.0008331197128941297,
+      "grad_norm": 2.8066741487653886,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 13
+    },
+    {
+      "epoch": 0.0008972058446552166,
+      "grad_norm": 2.4952793737222794,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 14
+    },
+    {
+      "epoch": 0.0009612919764163035,
+      "grad_norm": 2.736382692932069,
+      "learning_rate": 1e-06,
+      "loss": 0.3844,
+      "step": 15
+    },
+    {
+      "epoch": 0.0010253781081773904,
+      "grad_norm": 2.690189041127437,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 16
+    },
+    {
+      "epoch": 0.0010894642399384773,
+      "grad_norm": 2.4809643616320916,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 17
+    },
+    {
+      "epoch": 0.0011535503716995643,
+      "grad_norm": 2.470983130778616,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 18
+    },
+    {
+      "epoch": 0.001217636503460651,
+      "grad_norm": 2.9011921770535833,
+      "learning_rate": 1e-06,
+      "loss": 0.4941,
+      "step": 19
+    },
+    {
+      "epoch": 0.001281722635221738,
+      "grad_norm": 2.592260682165995,
+      "learning_rate": 1e-06,
+      "loss": 0.3991,
+      "step": 20
+    },
+    {
+      "epoch": 0.001345808766982825,
+      "grad_norm": 2.7190365461237502,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 21
+    },
+    {
+      "epoch": 0.0014098948987439118,
+      "grad_norm": 2.623825797975284,
+      "learning_rate": 1e-06,
+      "loss": 0.4459,
+      "step": 22
+    },
+    {
+      "epoch": 0.0014739810305049988,
+      "grad_norm": 4.130313126154051,
+      "learning_rate": 1e-06,
+      "loss": 0.4236,
+      "step": 23
+    },
+    {
+      "epoch": 0.0015380671622660857,
+      "grad_norm": 2.6968346941151387,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 24
+    },
+    {
+      "epoch": 0.0016021532940271724,
+      "grad_norm": 2.478154009816993,
+      "learning_rate": 1e-06,
+      "loss": 0.4064,
+      "step": 25
+    },
+    {
+      "epoch": 0.0016662394257882594,
+      "grad_norm": 2.878743074857694,
+      "learning_rate": 1e-06,
+      "loss": 0.4746,
+      "step": 26
+    },
+    {
+      "epoch": 0.0017303255575493463,
+      "grad_norm": 2.572012522185562,
+      "learning_rate": 1e-06,
+      "loss": 0.4063,
+      "step": 27
+    },
+    {
+      "epoch": 0.0017944116893104333,
+      "grad_norm": 2.6480699294244108,
+      "learning_rate": 1e-06,
+      "loss": 0.4893,
+      "step": 28
+    },
+    {
+      "epoch": 0.0018584978210715202,
+      "grad_norm": 2.5117617032878536,
+      "learning_rate": 1e-06,
+      "loss": 0.4271,
+      "step": 29
+    },
+    {
+      "epoch": 0.001922583952832607,
+      "grad_norm": 2.6045342374650704,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 30
+    },
+    {
+      "epoch": 0.001986670084593694,
+      "grad_norm": 2.598933479100832,
+      "learning_rate": 1e-06,
+      "loss": 0.447,
+      "step": 31
+    },
+    {
+      "epoch": 0.002050756216354781,
+      "grad_norm": 2.6774168288479685,
+      "learning_rate": 1e-06,
+      "loss": 0.4337,
+      "step": 32
+    },
+    {
+      "epoch": 0.0021148423481158678,
+      "grad_norm": 2.559910547832753,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 33
+    },
+    {
+      "epoch": 0.0021789284798769547,
+      "grad_norm": 2.8131005080313427,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 34
+    },
+    {
+      "epoch": 0.0022430146116380416,
+      "grad_norm": 2.6128742530557196,
+      "learning_rate": 1e-06,
+      "loss": 0.4635,
+      "step": 35
+    },
+    {
+      "epoch": 0.0023071007433991286,
+      "grad_norm": 2.790837783591932,
+      "learning_rate": 1e-06,
+      "loss": 0.4692,
+      "step": 36
+    },
+    {
+      "epoch": 0.0023711868751602155,
+      "grad_norm": 2.677228029063294,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 37
+    },
+    {
+      "epoch": 0.002435273006921302,
+      "grad_norm": 2.899817329376979,
+      "learning_rate": 1e-06,
+      "loss": 0.4265,
+      "step": 38
+    },
+    {
+      "epoch": 0.002499359138682389,
+      "grad_norm": 2.5848276627009055,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 39
+    },
+    {
+      "epoch": 0.002563445270443476,
+      "grad_norm": 2.5591688177632674,
+      "learning_rate": 1e-06,
+      "loss": 0.4879,
+      "step": 40
+    },
+    {
+      "epoch": 0.002627531402204563,
+      "grad_norm": 2.628718605922321,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 41
+    },
+    {
+      "epoch": 0.00269161753396565,
+      "grad_norm": 2.5750637451230385,
+      "learning_rate": 1e-06,
+      "loss": 0.3824,
+      "step": 42
+    },
+    {
+      "epoch": 0.0027557036657267367,
+      "grad_norm": 2.583335490883839,
+      "learning_rate": 1e-06,
+      "loss": 0.4289,
+      "step": 43
+    },
+    {
+      "epoch": 0.0028197897974878237,
+      "grad_norm": 2.5301462397652985,
+      "learning_rate": 1e-06,
+      "loss": 0.4284,
+      "step": 44
+    },
+    {
+      "epoch": 0.0028838759292489106,
+      "grad_norm": 2.629254554816327,
+      "learning_rate": 1e-06,
+      "loss": 0.4944,
+      "step": 45
+    },
+    {
+      "epoch": 0.0029479620610099976,
+      "grad_norm": 2.3796843494411215,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 46
+    },
+    {
+      "epoch": 0.0030120481927710845,
+      "grad_norm": 2.337285306067337,
+      "learning_rate": 1e-06,
+      "loss": 0.3691,
+      "step": 47
+    },
+    {
+      "epoch": 0.0030761343245321714,
+      "grad_norm": 2.4625613624689686,
+      "learning_rate": 1e-06,
+      "loss": 0.435,
+      "step": 48
+    },
+    {
+      "epoch": 0.003140220456293258,
+      "grad_norm": 2.6425603802991904,
+      "learning_rate": 1e-06,
+      "loss": 0.4156,
+      "step": 49
+    },
+    {
+      "epoch": 0.003204306588054345,
+      "grad_norm": 2.640849648601871,
+      "learning_rate": 1e-06,
+      "loss": 0.4261,
+      "step": 50
+    },
+    {
+      "epoch": 0.003268392719815432,
+      "grad_norm": 2.5632636645352895,
+      "learning_rate": 1e-06,
+      "loss": 0.4231,
+      "step": 51
+    },
+    {
+      "epoch": 0.0033324788515765188,
+      "grad_norm": 2.576641529620584,
+      "learning_rate": 1e-06,
+      "loss": 0.4591,
+      "step": 52
+    },
+    {
+      "epoch": 0.0033965649833376057,
+      "grad_norm": 2.485785384457921,
+      "learning_rate": 1e-06,
+      "loss": 0.436,
+      "step": 53
+    },
+    {
+      "epoch": 0.0034606511150986926,
+      "grad_norm": 2.546660745565414,
+      "learning_rate": 1e-06,
+      "loss": 0.4197,
+      "step": 54
+    },
+    {
+      "epoch": 0.0035247372468597796,
+      "grad_norm": 2.4421749353029747,
+      "learning_rate": 1e-06,
+      "loss": 0.3905,
+      "step": 55
+    },
+    {
+      "epoch": 0.0035888233786208665,
+      "grad_norm": 2.660810200932209,
+      "learning_rate": 1e-06,
+      "loss": 0.3768,
+      "step": 56
+    },
+    {
+      "epoch": 0.0036529095103819535,
+      "grad_norm": 2.4907286529936625,
+      "learning_rate": 1e-06,
+      "loss": 0.4203,
+      "step": 57
+    },
+    {
+      "epoch": 0.0037169956421430404,
+      "grad_norm": 2.5185462082716965,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 58
+    },
+    {
+      "epoch": 0.0037810817739041274,
+      "grad_norm": 2.5075831979076475,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 59
+    },
+    {
+      "epoch": 0.003845167905665214,
+      "grad_norm": 2.5893984879897434,
+      "learning_rate": 1e-06,
+      "loss": 0.4869,
+      "step": 60
+    },
+    {
+      "epoch": 0.003909254037426301,
+      "grad_norm": 2.5513574941347916,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 61
+    },
+    {
+      "epoch": 0.003973340169187388,
+      "grad_norm": 2.3634405046489824,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 62
+    },
+    {
+      "epoch": 0.004037426300948475,
+      "grad_norm": 2.794196161444718,
+      "learning_rate": 1e-06,
+      "loss": 0.474,
+      "step": 63
+    },
+    {
+      "epoch": 0.004101512432709562,
+      "grad_norm": 2.7786386772732636,
+      "learning_rate": 1e-06,
+      "loss": 0.4546,
+      "step": 64
+    },
+    {
+      "epoch": 0.0041655985644706486,
+      "grad_norm": 2.6427053217076577,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 65
+    },
+    {
+      "epoch": 0.0042296846962317355,
+      "grad_norm": 2.9375865167158963,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 66
+    },
+    {
+      "epoch": 0.0042937708279928224,
+      "grad_norm": 2.3163933176028584,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 67
+    },
+    {
+      "epoch": 0.004357856959753909,
+      "grad_norm": 2.4717258151766455,
+      "learning_rate": 1e-06,
+      "loss": 0.4136,
+      "step": 68
+    },
+    {
+      "epoch": 0.004421943091514996,
+      "grad_norm": 2.6985368472703435,
+      "learning_rate": 1e-06,
+      "loss": 0.4447,
+      "step": 69
+    },
+    {
+      "epoch": 0.004486029223276083,
+      "grad_norm": 2.6854921516916264,
+      "learning_rate": 1e-06,
+      "loss": 0.4055,
+      "step": 70
+    },
+    {
+      "epoch": 0.00455011535503717,
+      "grad_norm": 2.69736848706748,
+      "learning_rate": 1e-06,
+      "loss": 0.4256,
+      "step": 71
+    },
+    {
+      "epoch": 0.004614201486798257,
+      "grad_norm": 2.596490884499458,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 72
+    },
+    {
+      "epoch": 0.004678287618559344,
+      "grad_norm": 2.8156615368629216,
+      "learning_rate": 1e-06,
+      "loss": 0.3991,
+      "step": 73
+    },
+    {
+      "epoch": 0.004742373750320431,
+      "grad_norm": 2.4653779008574017,
+      "learning_rate": 1e-06,
+      "loss": 0.4834,
+      "step": 74
+    },
+    {
+      "epoch": 0.004806459882081518,
+      "grad_norm": 2.619661319531094,
+      "learning_rate": 1e-06,
+      "loss": 0.4507,
+      "step": 75
+    },
+    {
+      "epoch": 0.004870546013842604,
+      "grad_norm": 2.611916458285565,
+      "learning_rate": 1e-06,
+      "loss": 0.4103,
+      "step": 76
+    },
+    {
+      "epoch": 0.004934632145603691,
+      "grad_norm": 2.5636207784022225,
+      "learning_rate": 1e-06,
+      "loss": 0.4467,
+      "step": 77
+    },
+    {
+      "epoch": 0.004998718277364778,
+      "grad_norm": 2.4963493016425935,
+      "learning_rate": 1e-06,
+      "loss": 0.4775,
+      "step": 78
+    },
+    {
+      "epoch": 0.005062804409125865,
+      "grad_norm": 2.8481558198252133,
+      "learning_rate": 1e-06,
+      "loss": 0.4645,
+      "step": 79
+    },
+    {
+      "epoch": 0.005126890540886952,
+      "grad_norm": 2.724709371989913,
+      "learning_rate": 1e-06,
+      "loss": 0.4768,
+      "step": 80
+    },
+    {
+      "epoch": 0.005190976672648039,
+      "grad_norm": 2.474072117134112,
+      "learning_rate": 1e-06,
+      "loss": 0.415,
+      "step": 81
+    },
+    {
+      "epoch": 0.005255062804409126,
+      "grad_norm": 2.6671004940790204,
+      "learning_rate": 1e-06,
+      "loss": 0.4297,
+      "step": 82
+    },
+    {
+      "epoch": 0.005319148936170213,
+      "grad_norm": 2.636356788264366,
+      "learning_rate": 1e-06,
+      "loss": 0.4816,
+      "step": 83
+    },
+    {
+      "epoch": 0.0053832350679313,
+      "grad_norm": 2.5333332281115566,
+      "learning_rate": 1e-06,
+      "loss": 0.4175,
+      "step": 84
+    },
+    {
+      "epoch": 0.0054473211996923865,
+      "grad_norm": 2.5815065252078324,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 85
+    },
+    {
+      "epoch": 0.0055114073314534735,
+      "grad_norm": 2.532968445439128,
+      "learning_rate": 1e-06,
+      "loss": 0.4621,
+      "step": 86
+    },
+    {
+      "epoch": 0.00557549346321456,
+      "grad_norm": 2.5729492311933275,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 87
+    },
+    {
+      "epoch": 0.005639579594975647,
+      "grad_norm": 2.550980847594244,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 88
+    },
+    {
+      "epoch": 0.005703665726736734,
+      "grad_norm": 2.341442489669966,
+      "learning_rate": 1e-06,
+      "loss": 0.3631,
+      "step": 89
+    },
+    {
+      "epoch": 0.005767751858497821,
+      "grad_norm": 2.4357259388266046,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 90
+    },
+    {
+      "epoch": 0.005831837990258908,
+      "grad_norm": 2.8327584495472657,
+      "learning_rate": 1e-06,
+      "loss": 0.4202,
+      "step": 91
+    },
+    {
+      "epoch": 0.005895924122019995,
+      "grad_norm": 2.67552409067869,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 92
+    },
+    {
+      "epoch": 0.005960010253781082,
+      "grad_norm": 2.358660039118358,
+      "learning_rate": 1e-06,
+      "loss": 0.4177,
+      "step": 93
+    },
+    {
+      "epoch": 0.006024096385542169,
+      "grad_norm": 2.56361346512011,
+      "learning_rate": 1e-06,
+      "loss": 0.369,
+      "step": 94
+    },
+    {
+      "epoch": 0.006088182517303256,
+      "grad_norm": 2.5406342261425143,
+      "learning_rate": 1e-06,
+      "loss": 0.4096,
+      "step": 95
+    },
+    {
+      "epoch": 0.006152268649064343,
+      "grad_norm": 2.4792750353456876,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 96
+    },
+    {
+      "epoch": 0.006216354780825429,
+      "grad_norm": 2.632504157019107,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 97
+    },
+    {
+      "epoch": 0.006280440912586516,
+      "grad_norm": 2.6260305723231485,
+      "learning_rate": 1e-06,
+      "loss": 0.4319,
+      "step": 98
+    },
+    {
+      "epoch": 0.006344527044347603,
+      "grad_norm": 2.6485300234091356,
+      "learning_rate": 1e-06,
+      "loss": 0.3941,
+      "step": 99
+    },
+    {
+      "epoch": 0.00640861317610869,
+      "grad_norm": 2.611785234749641,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 100
+    },
+    {
+      "epoch": 0.006472699307869777,
+      "grad_norm": 2.485020096159456,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 101
+    },
+    {
+      "epoch": 0.006536785439630864,
+      "grad_norm": 2.515516066727853,
+      "learning_rate": 1e-06,
+      "loss": 0.4082,
+      "step": 102
+    },
+    {
+      "epoch": 0.006600871571391951,
+      "grad_norm": 2.6842572405670575,
+      "learning_rate": 1e-06,
+      "loss": 0.4795,
+      "step": 103
+    },
+    {
+      "epoch": 0.0066649577031530375,
+      "grad_norm": 2.7120822334389514,
+      "learning_rate": 1e-06,
+      "loss": 0.4101,
+      "step": 104
+    },
+    {
+      "epoch": 0.0067290438349141245,
+      "grad_norm": 2.69713806260704,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 105
+    },
+    {
+      "epoch": 0.006793129966675211,
+      "grad_norm": 2.823995680469878,
+      "learning_rate": 1e-06,
+      "loss": 0.4376,
+      "step": 106
+    },
+    {
+      "epoch": 0.006857216098436298,
+      "grad_norm": 2.971152302019369,
+      "learning_rate": 1e-06,
+      "loss": 0.4351,
+      "step": 107
+    },
+    {
+      "epoch": 0.006921302230197385,
+      "grad_norm": 2.6318324758558522,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 108
+    },
+    {
+      "epoch": 0.006985388361958472,
+      "grad_norm": 2.663343057511306,
+      "learning_rate": 1e-06,
+      "loss": 0.4143,
+      "step": 109
+    },
+    {
+      "epoch": 0.007049474493719559,
+      "grad_norm": 2.5787240892091043,
+      "learning_rate": 1e-06,
+      "loss": 0.4553,
+      "step": 110
+    },
+    {
+      "epoch": 0.007113560625480646,
+      "grad_norm": 2.904813690584719,
+      "learning_rate": 1e-06,
+      "loss": 0.4463,
+      "step": 111
+    },
+    {
+      "epoch": 0.007177646757241733,
+      "grad_norm": 2.6164388520336543,
+      "learning_rate": 1e-06,
+      "loss": 0.4237,
+      "step": 112
+    },
+    {
+      "epoch": 0.00724173288900282,
+      "grad_norm": 3.62468026756726,
+      "learning_rate": 1e-06,
+      "loss": 0.5197,
+      "step": 113
+    },
+    {
+      "epoch": 0.007305819020763907,
+      "grad_norm": 2.6283606516515183,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 114
+    },
+    {
+      "epoch": 0.007369905152524994,
+      "grad_norm": 2.733978140448335,
+      "learning_rate": 1e-06,
+      "loss": 0.4945,
+      "step": 115
+    },
+    {
+      "epoch": 0.007433991284286081,
+      "grad_norm": 2.6142513762887365,
+      "learning_rate": 1e-06,
+      "loss": 0.3966,
+      "step": 116
+    },
+    {
+      "epoch": 0.007498077416047168,
+      "grad_norm": 2.9462705802375964,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 117
+    },
+    {
+      "epoch": 0.007562163547808255,
+      "grad_norm": 2.5077415901431492,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 118
+    },
+    {
+      "epoch": 0.007626249679569341,
+      "grad_norm": 2.484246468433061,
+      "learning_rate": 1e-06,
+      "loss": 0.4683,
+      "step": 119
+    },
+    {
+      "epoch": 0.007690335811330428,
+      "grad_norm": 2.544320892555243,
+      "learning_rate": 1e-06,
+      "loss": 0.4388,
+      "step": 120
+    },
+    {
+      "epoch": 0.007754421943091515,
+      "grad_norm": 2.4444504338812396,
+      "learning_rate": 1e-06,
+      "loss": 0.4482,
+      "step": 121
+    },
+    {
+      "epoch": 0.007818508074852602,
+      "grad_norm": 2.6270003993720485,
+      "learning_rate": 1e-06,
+      "loss": 0.4309,
+      "step": 122
+    },
+    {
+      "epoch": 0.00788259420661369,
+      "grad_norm": 2.5560433655850656,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 123
+    },
+    {
+      "epoch": 0.007946680338374775,
+      "grad_norm": 2.583021760048913,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 124
+    },
+    {
+      "epoch": 0.008010766470135863,
+      "grad_norm": 2.516809486753932,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 125
+    },
+    {
+      "epoch": 0.00807485260189695,
+      "grad_norm": 2.7202512962911167,
+      "learning_rate": 1e-06,
+      "loss": 0.4496,
+      "step": 126
+    },
+    {
+      "epoch": 0.008138938733658037,
+      "grad_norm": 2.561504469620483,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 127
+    },
+    {
+      "epoch": 0.008203024865419123,
+      "grad_norm": 2.70197541717228,
+      "learning_rate": 1e-06,
+      "loss": 0.4708,
+      "step": 128
+    },
+    {
+      "epoch": 0.008267110997180211,
+      "grad_norm": 2.703655222038152,
+      "learning_rate": 1e-06,
+      "loss": 0.4013,
+      "step": 129
+    },
+    {
+      "epoch": 0.008331197128941297,
+      "grad_norm": 2.801075229740231,
+      "learning_rate": 1e-06,
+      "loss": 0.4957,
+      "step": 130
+    },
+    {
+      "epoch": 0.008395283260702383,
+      "grad_norm": 2.7000542176029283,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 131
+    },
+    {
+      "epoch": 0.008459369392463471,
+      "grad_norm": 2.728367510056538,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 132
+    },
+    {
+      "epoch": 0.008523455524224557,
+      "grad_norm": 2.4926115155921744,
+      "learning_rate": 1e-06,
+      "loss": 0.4143,
+      "step": 133
+    },
+    {
+      "epoch": 0.008587541655985645,
+      "grad_norm": 2.5170415635469663,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 134
+    },
+    {
+      "epoch": 0.008651627787746731,
+      "grad_norm": 2.766298946998558,
+      "learning_rate": 1e-06,
+      "loss": 0.3751,
+      "step": 135
+    },
+    {
+      "epoch": 0.008715713919507819,
+      "grad_norm": 2.725850116300731,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 136
+    },
+    {
+      "epoch": 0.008779800051268905,
+      "grad_norm": 2.53900556364546,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 137
+    },
+    {
+      "epoch": 0.008843886183029993,
+      "grad_norm": 2.477924972118881,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 138
+    },
+    {
+      "epoch": 0.008907972314791079,
+      "grad_norm": 2.826510511701279,
+      "learning_rate": 1e-06,
+      "loss": 0.4638,
+      "step": 139
+    },
+    {
+      "epoch": 0.008972058446552167,
+      "grad_norm": 2.6376643369143764,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 140
+    },
+    {
+      "epoch": 0.009036144578313253,
+      "grad_norm": 2.554145291137539,
+      "learning_rate": 1e-06,
+      "loss": 0.4451,
+      "step": 141
+    },
+    {
+      "epoch": 0.00910023071007434,
+      "grad_norm": 2.6743052984416154,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 142
+    },
+    {
+      "epoch": 0.009164316841835426,
+      "grad_norm": 2.56483434081286,
+      "learning_rate": 1e-06,
+      "loss": 0.4436,
+      "step": 143
+    },
+    {
+      "epoch": 0.009228402973596514,
+      "grad_norm": 2.4254491231280455,
+      "learning_rate": 1e-06,
+      "loss": 0.42,
+      "step": 144
+    },
+    {
+      "epoch": 0.0092924891053576,
+      "grad_norm": 2.558112650901851,
+      "learning_rate": 1e-06,
+      "loss": 0.4531,
+      "step": 145
+    },
+    {
+      "epoch": 0.009356575237118688,
+      "grad_norm": 2.405260114602089,
+      "learning_rate": 1e-06,
+      "loss": 0.4223,
+      "step": 146
+    },
+    {
+      "epoch": 0.009420661368879774,
+      "grad_norm": 2.60820574354277,
+      "learning_rate": 1e-06,
+      "loss": 0.3979,
+      "step": 147
+    },
+    {
+      "epoch": 0.009484747500640862,
+      "grad_norm": 2.6897694601294533,
+      "learning_rate": 1e-06,
+      "loss": 0.478,
+      "step": 148
+    },
+    {
+      "epoch": 0.009548833632401948,
+      "grad_norm": 2.579453394198776,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 149
+    },
+    {
+      "epoch": 0.009612919764163036,
+      "grad_norm": 2.7424643174105716,
+      "learning_rate": 1e-06,
+      "loss": 0.3937,
+      "step": 150
+    },
+    {
+      "epoch": 0.009677005895924122,
+      "grad_norm": 2.7023142583494875,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 151
+    },
+    {
+      "epoch": 0.009741092027685208,
+      "grad_norm": 2.506484730613173,
+      "learning_rate": 1e-06,
+      "loss": 0.3936,
+      "step": 152
+    },
+    {
+      "epoch": 0.009805178159446296,
+      "grad_norm": 2.4749366097633874,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 153
+    },
+    {
+      "epoch": 0.009869264291207382,
+      "grad_norm": 2.3269287496593414,
+      "learning_rate": 1e-06,
+      "loss": 0.3664,
+      "step": 154
+    },
+    {
+      "epoch": 0.00993335042296847,
+      "grad_norm": 2.4212911977056604,
+      "learning_rate": 1e-06,
+      "loss": 0.4353,
+      "step": 155
+    },
+    {
+      "epoch": 0.009997436554729556,
+      "grad_norm": 2.4468505337849593,
+      "learning_rate": 1e-06,
+      "loss": 0.4622,
+      "step": 156
+    },
+    {
+      "epoch": 0.010061522686490644,
+      "grad_norm": 2.5645734340411206,
+      "learning_rate": 1e-06,
+      "loss": 0.3703,
+      "step": 157
+    },
+    {
+      "epoch": 0.01012560881825173,
+      "grad_norm": 2.6534084292524764,
+      "learning_rate": 1e-06,
+      "loss": 0.4307,
+      "step": 158
+    },
+    {
+      "epoch": 0.010189694950012818,
+      "grad_norm": 2.510221138642233,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 159
+    },
+    {
+      "epoch": 0.010253781081773904,
+      "grad_norm": 2.6653843643185047,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 160
+    },
+    {
+      "epoch": 0.010317867213534991,
+      "grad_norm": 2.647156935213225,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 161
+    },
+    {
+      "epoch": 0.010381953345296078,
+      "grad_norm": 2.4645904023031364,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 162
+    },
+    {
+      "epoch": 0.010446039477057165,
+      "grad_norm": 2.5273636798670447,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 163
+    },
+    {
+      "epoch": 0.010510125608818251,
+      "grad_norm": 2.5759837139960355,
+      "learning_rate": 1e-06,
+      "loss": 0.4219,
+      "step": 164
+    },
+    {
+      "epoch": 0.01057421174057934,
+      "grad_norm": 2.7530163451644483,
+      "learning_rate": 1e-06,
+      "loss": 0.4473,
+      "step": 165
+    },
+    {
+      "epoch": 0.010638297872340425,
+      "grad_norm": 2.771144588040565,
+      "learning_rate": 1e-06,
+      "loss": 0.3832,
+      "step": 166
+    },
+    {
+      "epoch": 0.010702384004101513,
+      "grad_norm": 2.6609737067836194,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 167
+    },
+    {
+      "epoch": 0.0107664701358626,
+      "grad_norm": 2.76214436031771,
+      "learning_rate": 1e-06,
+      "loss": 0.3864,
+      "step": 168
+    },
+    {
+      "epoch": 0.010830556267623687,
+      "grad_norm": 2.722402456149691,
+      "learning_rate": 1e-06,
+      "loss": 0.404,
+      "step": 169
+    },
+    {
+      "epoch": 0.010894642399384773,
+      "grad_norm": 2.5998788986620696,
+      "learning_rate": 1e-06,
+      "loss": 0.4311,
+      "step": 170
+    },
+    {
+      "epoch": 0.01095872853114586,
+      "grad_norm": 2.7559088943343335,
+      "learning_rate": 1e-06,
+      "loss": 0.4491,
+      "step": 171
+    },
+    {
+      "epoch": 0.011022814662906947,
+      "grad_norm": 2.7409712519131686,
+      "learning_rate": 1e-06,
+      "loss": 0.4435,
+      "step": 172
+    },
+    {
+      "epoch": 0.011086900794668033,
+      "grad_norm": 2.664956072118899,
+      "learning_rate": 1e-06,
+      "loss": 0.4193,
+      "step": 173
+    },
+    {
+      "epoch": 0.01115098692642912,
+      "grad_norm": 3.1192465909037774,
+      "learning_rate": 1e-06,
+      "loss": 0.4669,
+      "step": 174
+    },
+    {
+      "epoch": 0.011215073058190207,
+      "grad_norm": 2.563621451938816,
+      "learning_rate": 1e-06,
+      "loss": 0.3767,
+      "step": 175
+    },
+    {
+      "epoch": 0.011279159189951295,
+      "grad_norm": 2.570045233440838,
+      "learning_rate": 1e-06,
+      "loss": 0.4765,
+      "step": 176
+    },
+    {
+      "epoch": 0.01134324532171238,
+      "grad_norm": 2.4916231423662043,
+      "learning_rate": 1e-06,
+      "loss": 0.4392,
+      "step": 177
+    },
+    {
+      "epoch": 0.011407331453473469,
+      "grad_norm": 2.7651096653059195,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 178
+    },
+    {
+      "epoch": 0.011471417585234555,
+      "grad_norm": 2.562958698429324,
+      "learning_rate": 1e-06,
+      "loss": 0.4092,
+      "step": 179
+    },
+    {
+      "epoch": 0.011535503716995642,
+      "grad_norm": 2.464035920659328,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 180
+    },
+    {
+      "epoch": 0.011599589848756729,
+      "grad_norm": 2.751881851503854,
+      "learning_rate": 1e-06,
+      "loss": 0.3687,
+      "step": 181
+    },
+    {
+      "epoch": 0.011663675980517816,
+      "grad_norm": 2.559719631734224,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 182
+    },
+    {
+      "epoch": 0.011727762112278902,
+      "grad_norm": 2.4418892922224074,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 183
+    },
+    {
+      "epoch": 0.01179184824403999,
+      "grad_norm": 2.6724559730934545,
+      "learning_rate": 1e-06,
+      "loss": 0.4036,
+      "step": 184
+    },
+    {
+      "epoch": 0.011855934375801076,
+      "grad_norm": 2.6322280599536936,
+      "learning_rate": 1e-06,
+      "loss": 0.4517,
+      "step": 185
+    },
+    {
+      "epoch": 0.011920020507562164,
+      "grad_norm": 2.605353550494746,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 186
+    },
+    {
+      "epoch": 0.01198410663932325,
+      "grad_norm": 2.564073940099639,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 187
+    },
+    {
+      "epoch": 0.012048192771084338,
+      "grad_norm": 2.535479880548571,
+      "learning_rate": 1e-06,
+      "loss": 0.4454,
+      "step": 188
+    },
+    {
+      "epoch": 0.012112278902845424,
+      "grad_norm": 2.566704466588977,
+      "learning_rate": 1e-06,
+      "loss": 0.4025,
+      "step": 189
+    },
+    {
+      "epoch": 0.012176365034606512,
+      "grad_norm": 2.5806716178681293,
+      "learning_rate": 1e-06,
+      "loss": 0.3537,
+      "step": 190
+    },
+    {
+      "epoch": 0.012240451166367598,
+      "grad_norm": 2.825882756789249,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 191
+    },
+    {
+      "epoch": 0.012304537298128686,
+      "grad_norm": 2.578508321046613,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 192
+    },
+    {
+      "epoch": 0.012368623429889772,
+      "grad_norm": 2.460610078013274,
+      "learning_rate": 1e-06,
+      "loss": 0.3796,
+      "step": 193
+    },
+    {
+      "epoch": 0.012432709561650858,
+      "grad_norm": 2.52758953552192,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 194
+    },
+    {
+      "epoch": 0.012496795693411946,
+      "grad_norm": 2.680670213978359,
+      "learning_rate": 1e-06,
+      "loss": 0.4494,
+      "step": 195
+    },
+    {
+      "epoch": 0.012560881825173032,
+      "grad_norm": 2.5727240318324345,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 196
+    },
+    {
+      "epoch": 0.01262496795693412,
+      "grad_norm": 2.5479336539567035,
+      "learning_rate": 1e-06,
+      "loss": 0.4283,
+      "step": 197
+    },
+    {
+      "epoch": 0.012689054088695206,
+      "grad_norm": 2.5115294399067687,
+      "learning_rate": 1e-06,
+      "loss": 0.4641,
+      "step": 198
+    },
+    {
+      "epoch": 0.012753140220456293,
+      "grad_norm": 2.78330697315584,
+      "learning_rate": 1e-06,
+      "loss": 0.4351,
+      "step": 199
+    },
+    {
+      "epoch": 0.01281722635221738,
+      "grad_norm": 2.45488310212574,
+      "learning_rate": 1e-06,
+      "loss": 0.414,
+      "step": 200
+    },
+    {
+      "epoch": 0.012881312483978467,
+      "grad_norm": 2.3852827309025404,
+      "learning_rate": 1e-06,
+      "loss": 0.3923,
+      "step": 201
+    },
+    {
+      "epoch": 0.012945398615739553,
+      "grad_norm": 2.5864768221972354,
+      "learning_rate": 1e-06,
+      "loss": 0.4649,
+      "step": 202
+    },
+    {
+      "epoch": 0.013009484747500641,
+      "grad_norm": 2.7974137736304803,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 203
+    },
+    {
+      "epoch": 0.013073570879261727,
+      "grad_norm": 2.406651303043376,
+      "learning_rate": 1e-06,
+      "loss": 0.409,
+      "step": 204
+    },
+    {
+      "epoch": 0.013137657011022815,
+      "grad_norm": 2.5980042928944735,
+      "learning_rate": 1e-06,
+      "loss": 0.4234,
+      "step": 205
+    },
+    {
+      "epoch": 0.013201743142783901,
+      "grad_norm": 2.5707835336578237,
+      "learning_rate": 1e-06,
+      "loss": 0.4188,
+      "step": 206
+    },
+    {
+      "epoch": 0.013265829274544989,
+      "grad_norm": 2.5628325989007665,
+      "learning_rate": 1e-06,
+      "loss": 0.4019,
+      "step": 207
+    },
+    {
+      "epoch": 0.013329915406306075,
+      "grad_norm": 2.3178028538939057,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 208
+    },
+    {
+      "epoch": 0.013394001538067163,
+      "grad_norm": 2.4943832148029372,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 209
+    },
+    {
+      "epoch": 0.013458087669828249,
+      "grad_norm": 2.634069670687289,
+      "learning_rate": 1e-06,
+      "loss": 0.4756,
+      "step": 210
+    },
+    {
+      "epoch": 0.013522173801589337,
+      "grad_norm": 2.557527545449268,
+      "learning_rate": 1e-06,
+      "loss": 0.4653,
+      "step": 211
+    },
+    {
+      "epoch": 0.013586259933350423,
+      "grad_norm": 2.463861613281935,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 212
+    },
+    {
+      "epoch": 0.01365034606511151,
+      "grad_norm": 2.3854742264069397,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 213
+    },
+    {
+      "epoch": 0.013714432196872597,
+      "grad_norm": 2.4260505846618186,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 214
+    },
+    {
+      "epoch": 0.013778518328633685,
+      "grad_norm": 2.5804340331860343,
+      "learning_rate": 1e-06,
+      "loss": 0.4042,
+      "step": 215
+    },
+    {
+      "epoch": 0.01384260446039477,
+      "grad_norm": 2.6587083377401455,
+      "learning_rate": 1e-06,
+      "loss": 0.4072,
+      "step": 216
+    },
+    {
+      "epoch": 0.013906690592155857,
+      "grad_norm": 2.617929149718754,
+      "learning_rate": 1e-06,
+      "loss": 0.4605,
+      "step": 217
+    },
+    {
+      "epoch": 0.013970776723916944,
+      "grad_norm": 2.5803501828834143,
+      "learning_rate": 1e-06,
+      "loss": 0.454,
+      "step": 218
+    },
+    {
+      "epoch": 0.01403486285567803,
+      "grad_norm": 2.788641206889281,
+      "learning_rate": 1e-06,
+      "loss": 0.404,
+      "step": 219
+    },
+    {
+      "epoch": 0.014098948987439118,
+      "grad_norm": 2.506065493967879,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 220
+    },
+    {
+      "epoch": 0.014163035119200204,
+      "grad_norm": 2.62297260217538,
+      "learning_rate": 1e-06,
+      "loss": 0.4524,
+      "step": 221
+    },
+    {
+      "epoch": 0.014227121250961292,
+      "grad_norm": 2.396202673688117,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 222
+    },
+    {
+      "epoch": 0.014291207382722378,
+      "grad_norm": 2.5608484595985788,
+      "learning_rate": 1e-06,
+      "loss": 0.394,
+      "step": 223
+    },
+    {
+      "epoch": 0.014355293514483466,
+      "grad_norm": 2.7259635787986722,
+      "learning_rate": 1e-06,
+      "loss": 0.4533,
+      "step": 224
+    },
+    {
+      "epoch": 0.014419379646244552,
+      "grad_norm": 2.7655406885266967,
+      "learning_rate": 1e-06,
+      "loss": 0.4553,
+      "step": 225
+    },
+    {
+      "epoch": 0.01448346577800564,
+      "grad_norm": 2.548446603379672,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 226
+    },
+    {
+      "epoch": 0.014547551909766726,
+      "grad_norm": 2.4670836949298565,
+      "learning_rate": 1e-06,
+      "loss": 0.4037,
+      "step": 227
+    },
+    {
+      "epoch": 0.014611638041527814,
+      "grad_norm": 2.4857360759200557,
+      "learning_rate": 1e-06,
+      "loss": 0.3866,
+      "step": 228
+    },
+    {
+      "epoch": 0.0146757241732889,
+      "grad_norm": 2.6690124449544754,
+      "learning_rate": 1e-06,
+      "loss": 0.4127,
+      "step": 229
+    },
+    {
+      "epoch": 0.014739810305049988,
+      "grad_norm": 2.6132105365625464,
+      "learning_rate": 1e-06,
+      "loss": 0.4843,
+      "step": 230
+    },
+    {
+      "epoch": 0.014803896436811074,
+      "grad_norm": 2.7432970666186347,
+      "learning_rate": 1e-06,
+      "loss": 0.4536,
+      "step": 231
+    },
+    {
+      "epoch": 0.014867982568572162,
+      "grad_norm": 2.7658731495653672,
+      "learning_rate": 1e-06,
+      "loss": 0.3939,
+      "step": 232
+    },
+    {
+      "epoch": 0.014932068700333248,
+      "grad_norm": 2.580362454305093,
+      "learning_rate": 1e-06,
+      "loss": 0.3823,
+      "step": 233
+    },
+    {
+      "epoch": 0.014996154832094336,
+      "grad_norm": 2.6052893301698745,
+      "learning_rate": 1e-06,
+      "loss": 0.4735,
+      "step": 234
+    },
+    {
+      "epoch": 0.015060240963855422,
+      "grad_norm": 2.48609104419947,
+      "learning_rate": 1e-06,
+      "loss": 0.4368,
+      "step": 235
+    },
+    {
+      "epoch": 0.01512432709561651,
+      "grad_norm": 2.4309934849231962,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 236
+    },
+    {
+      "epoch": 0.015188413227377595,
+      "grad_norm": 2.437877921605441,
+      "learning_rate": 1e-06,
+      "loss": 0.4504,
+      "step": 237
+    },
+    {
+      "epoch": 0.015252499359138682,
+      "grad_norm": 2.445601223786844,
+      "learning_rate": 1e-06,
+      "loss": 0.3719,
+      "step": 238
+    },
+    {
+      "epoch": 0.01531658549089977,
+      "grad_norm": 2.566610647012455,
+      "learning_rate": 1e-06,
+      "loss": 0.4522,
+      "step": 239
+    },
+    {
+      "epoch": 0.015380671622660855,
+      "grad_norm": 2.3629967174097883,
+      "learning_rate": 1e-06,
+      "loss": 0.3975,
+      "step": 240
+    },
+    {
+      "epoch": 0.015444757754421943,
+      "grad_norm": 2.7502843636448358,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 241
+    },
+    {
+      "epoch": 0.01550884388618303,
+      "grad_norm": 2.708324699769681,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 242
+    },
+    {
+      "epoch": 0.015572930017944117,
+      "grad_norm": 2.444268394765568,
+      "learning_rate": 1e-06,
+      "loss": 0.4329,
+      "step": 243
+    },
+    {
+      "epoch": 0.015637016149705203,
+      "grad_norm": 2.3920689319195025,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 244
+    },
+    {
+      "epoch": 0.01570110228146629,
+      "grad_norm": 2.4825639040106156,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 245
+    },
+    {
+      "epoch": 0.01576518841322738,
+      "grad_norm": 2.4814384615199123,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 246
+    },
+    {
+      "epoch": 0.015829274544988465,
+      "grad_norm": 2.724229129070956,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 247
+    },
+    {
+      "epoch": 0.01589336067674955,
+      "grad_norm": 2.7175053474836766,
+      "learning_rate": 1e-06,
+      "loss": 0.3619,
+      "step": 248
+    },
+    {
+      "epoch": 0.015957446808510637,
+      "grad_norm": 2.4697050879589275,
+      "learning_rate": 1e-06,
+      "loss": 0.3635,
+      "step": 249
+    },
+    {
+      "epoch": 0.016021532940271727,
+      "grad_norm": 2.4595339460868573,
+      "learning_rate": 1e-06,
+      "loss": 0.476,
+      "step": 250
+    },
+    {
+      "epoch": 0.016085619072032813,
+      "grad_norm": 2.6445467786638814,
+      "learning_rate": 1e-06,
+      "loss": 0.4999,
+      "step": 251
+    },
+    {
+      "epoch": 0.0161497052037939,
+      "grad_norm": 2.4693317351326414,
+      "learning_rate": 1e-06,
+      "loss": 0.4237,
+      "step": 252
+    },
+    {
+      "epoch": 0.016213791335554985,
+      "grad_norm": 2.4702693466347125,
+      "learning_rate": 1e-06,
+      "loss": 0.4443,
+      "step": 253
+    },
+    {
+      "epoch": 0.016277877467316074,
+      "grad_norm": 2.717654695399045,
+      "learning_rate": 1e-06,
+      "loss": 0.4019,
+      "step": 254
+    },
+    {
+      "epoch": 0.01634196359907716,
+      "grad_norm": 2.603236954441173,
+      "learning_rate": 1e-06,
+      "loss": 0.469,
+      "step": 255
+    },
+    {
+      "epoch": 0.016406049730838246,
+      "grad_norm": 2.7996949150270867,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 256
+    },
+    {
+      "epoch": 0.016470135862599333,
+      "grad_norm": 2.6193082828705125,
+      "learning_rate": 1e-06,
+      "loss": 0.4333,
+      "step": 257
+    },
+    {
+      "epoch": 0.016534221994360422,
+      "grad_norm": 2.473437533047886,
+      "learning_rate": 1e-06,
+      "loss": 0.4774,
+      "step": 258
+    },
+    {
+      "epoch": 0.016598308126121508,
+      "grad_norm": 2.564354049838647,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 259
+    },
+    {
+      "epoch": 0.016662394257882594,
+      "grad_norm": 2.547984763105764,
+      "learning_rate": 1e-06,
+      "loss": 0.4982,
+      "step": 260
+    },
+    {
+      "epoch": 0.01672648038964368,
+      "grad_norm": 2.693431115018885,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 261
+    },
+    {
+      "epoch": 0.016790566521404766,
+      "grad_norm": 2.5918028116724146,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 262
+    },
+    {
+      "epoch": 0.016854652653165856,
+      "grad_norm": 2.586212163796859,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 263
+    },
+    {
+      "epoch": 0.016918738784926942,
+      "grad_norm": 2.5085076025134962,
+      "learning_rate": 1e-06,
+      "loss": 0.4387,
+      "step": 264
+    },
+    {
+      "epoch": 0.016982824916688028,
+      "grad_norm": 2.7284858517103294,
+      "learning_rate": 1e-06,
+      "loss": 0.4369,
+      "step": 265
+    },
+    {
+      "epoch": 0.017046911048449114,
+      "grad_norm": 2.698912647175665,
+      "learning_rate": 1e-06,
+      "loss": 0.3981,
+      "step": 266
+    },
+    {
+      "epoch": 0.017110997180210204,
+      "grad_norm": 2.433393766997923,
+      "learning_rate": 1e-06,
+      "loss": 0.3888,
+      "step": 267
+    },
+    {
+      "epoch": 0.01717508331197129,
+      "grad_norm": 2.6338501838202606,
+      "learning_rate": 1e-06,
+      "loss": 0.4465,
+      "step": 268
+    },
+    {
+      "epoch": 0.017239169443732376,
+      "grad_norm": 2.728265238380636,
+      "learning_rate": 1e-06,
+      "loss": 0.471,
+      "step": 269
+    },
+    {
+      "epoch": 0.017303255575493462,
+      "grad_norm": 2.588489728565312,
+      "learning_rate": 1e-06,
+      "loss": 0.4759,
+      "step": 270
+    },
+    {
+      "epoch": 0.01736734170725455,
+      "grad_norm": 2.611963629217684,
+      "learning_rate": 1e-06,
+      "loss": 0.4743,
+      "step": 271
+    },
+    {
+      "epoch": 0.017431427839015638,
+      "grad_norm": 2.633648442017364,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 272
+    },
+    {
+      "epoch": 0.017495513970776724,
+      "grad_norm": 2.9530713121350414,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 273
+    },
+    {
+      "epoch": 0.01755960010253781,
+      "grad_norm": 2.499549993514414,
+      "learning_rate": 1e-06,
+      "loss": 0.4312,
+      "step": 274
+    },
+    {
+      "epoch": 0.0176236862342989,
+      "grad_norm": 2.6602932818625926,
+      "learning_rate": 1e-06,
+      "loss": 0.4594,
+      "step": 275
+    },
+    {
+      "epoch": 0.017687772366059985,
+      "grad_norm": 2.4729820946013943,
+      "learning_rate": 1e-06,
+      "loss": 0.3503,
+      "step": 276
+    },
+    {
+      "epoch": 0.01775185849782107,
+      "grad_norm": 2.568481943130322,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 277
+    },
+    {
+      "epoch": 0.017815944629582157,
+      "grad_norm": 2.5410630389732356,
+      "learning_rate": 1e-06,
+      "loss": 0.4558,
+      "step": 278
+    },
+    {
+      "epoch": 0.017880030761343247,
+      "grad_norm": 2.8834562082794633,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 279
+    },
+    {
+      "epoch": 0.017944116893104333,
+      "grad_norm": 2.9142632748077593,
+      "learning_rate": 1e-06,
+      "loss": 0.3867,
+      "step": 280
+    },
+    {
+      "epoch": 0.01800820302486542,
+      "grad_norm": 2.685459029147027,
+      "learning_rate": 1e-06,
+      "loss": 0.4265,
+      "step": 281
+    },
+    {
+      "epoch": 0.018072289156626505,
+      "grad_norm": 2.8592794071139687,
+      "learning_rate": 1e-06,
+      "loss": 0.4207,
+      "step": 282
+    },
+    {
+      "epoch": 0.01813637528838759,
+      "grad_norm": 2.577623379545379,
+      "learning_rate": 1e-06,
+      "loss": 0.3723,
+      "step": 283
+    },
+    {
+      "epoch": 0.01820046142014868,
+      "grad_norm": 2.639894455655153,
+      "learning_rate": 1e-06,
+      "loss": 0.4035,
+      "step": 284
+    },
+    {
+      "epoch": 0.018264547551909767,
+      "grad_norm": 2.744866333497609,
+      "learning_rate": 1e-06,
+      "loss": 0.4558,
+      "step": 285
+    },
+    {
+      "epoch": 0.018328633683670853,
+      "grad_norm": 2.7701088109503855,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 286
+    },
+    {
+      "epoch": 0.01839271981543194,
+      "grad_norm": 2.7164498040000233,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 287
+    },
+    {
+      "epoch": 0.01845680594719303,
+      "grad_norm": 2.8216916740476417,
+      "learning_rate": 1e-06,
+      "loss": 0.4087,
+      "step": 288
+    },
+    {
+      "epoch": 0.018520892078954115,
+      "grad_norm": 2.570695596833045,
+      "learning_rate": 1e-06,
+      "loss": 0.3929,
+      "step": 289
+    },
+    {
+      "epoch": 0.0185849782107152,
+      "grad_norm": 2.5436410160123093,
+      "learning_rate": 1e-06,
+      "loss": 0.4209,
+      "step": 290
+    },
+    {
+      "epoch": 0.018649064342476287,
+      "grad_norm": 2.464564508450358,
+      "learning_rate": 1e-06,
+      "loss": 0.4682,
+      "step": 291
+    },
+    {
+      "epoch": 0.018713150474237376,
+      "grad_norm": 2.6414925530399103,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 292
+    },
+    {
+      "epoch": 0.018777236605998462,
+      "grad_norm": 2.612675091607748,
+      "learning_rate": 1e-06,
+      "loss": 0.4331,
+      "step": 293
+    },
+    {
+      "epoch": 0.01884132273775955,
+      "grad_norm": 2.443575754509268,
+      "learning_rate": 1e-06,
+      "loss": 0.3982,
+      "step": 294
+    },
+    {
+      "epoch": 0.018905408869520635,
+      "grad_norm": 2.580805303286195,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 295
+    },
+    {
+      "epoch": 0.018969495001281724,
+      "grad_norm": 2.5273581888412453,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 296
+    },
+    {
+      "epoch": 0.01903358113304281,
+      "grad_norm": 2.489408203399578,
+      "learning_rate": 1e-06,
+      "loss": 0.4248,
+      "step": 297
+    },
+    {
+      "epoch": 0.019097667264803896,
+      "grad_norm": 2.649191810622033,
+      "learning_rate": 1e-06,
+      "loss": 0.4376,
+      "step": 298
+    },
+    {
+      "epoch": 0.019161753396564982,
+      "grad_norm": 2.531864638870876,
+      "learning_rate": 1e-06,
+      "loss": 0.4435,
+      "step": 299
+    },
+    {
+      "epoch": 0.019225839528326072,
+      "grad_norm": 2.405100794839331,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 300
+    },
+    {
+      "epoch": 0.019289925660087158,
+      "grad_norm": 2.8102035159421064,
+      "learning_rate": 1e-06,
+      "loss": 0.4716,
+      "step": 301
+    },
+    {
+      "epoch": 0.019354011791848244,
+      "grad_norm": 2.502247457851022,
+      "learning_rate": 1e-06,
+      "loss": 0.4135,
+      "step": 302
+    },
+    {
+      "epoch": 0.01941809792360933,
+      "grad_norm": 2.4034822507856375,
+      "learning_rate": 1e-06,
+      "loss": 0.375,
+      "step": 303
+    },
+    {
+      "epoch": 0.019482184055370416,
+      "grad_norm": 2.6665336665510355,
+      "learning_rate": 1e-06,
+      "loss": 0.4282,
+      "step": 304
+    },
+    {
+      "epoch": 0.019546270187131506,
+      "grad_norm": 2.6706007873681754,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 305
+    },
+    {
+      "epoch": 0.019610356318892592,
+      "grad_norm": 2.9943586682230596,
+      "learning_rate": 1e-06,
+      "loss": 0.4563,
+      "step": 306
+    },
+    {
+      "epoch": 0.019674442450653678,
+      "grad_norm": 2.529442058426431,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 307
+    },
+    {
+      "epoch": 0.019738528582414764,
+      "grad_norm": 2.6846719971532287,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 308
+    },
+    {
+      "epoch": 0.019802614714175854,
+      "grad_norm": 2.656046389414779,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 309
+    },
+    {
+      "epoch": 0.01986670084593694,
+      "grad_norm": 2.3228700793409316,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 310
+    },
+    {
+      "epoch": 0.019930786977698026,
+      "grad_norm": 2.3570092668653313,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 311
+    },
+    {
+      "epoch": 0.01999487310945911,
+      "grad_norm": 2.6047654223851087,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 312
+    },
+    {
+      "epoch": 0.0200589592412202,
+      "grad_norm": 2.762796362913787,
+      "learning_rate": 1e-06,
+      "loss": 0.3973,
+      "step": 313
+    },
+    {
+      "epoch": 0.020123045372981287,
+      "grad_norm": 2.684177046272746,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 314
+    },
+    {
+      "epoch": 0.020187131504742373,
+      "grad_norm": 2.7783820778054875,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 315
+    },
+    {
+      "epoch": 0.02025121763650346,
+      "grad_norm": 2.6167852095843864,
+      "learning_rate": 1e-06,
+      "loss": 0.4247,
+      "step": 316
+    },
+    {
+      "epoch": 0.02031530376826455,
+      "grad_norm": 2.6845860637575756,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 317
+    },
+    {
+      "epoch": 0.020379389900025635,
+      "grad_norm": 2.602696481988515,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 318
+    },
+    {
+      "epoch": 0.02044347603178672,
+      "grad_norm": 2.5342579150615006,
+      "learning_rate": 1e-06,
+      "loss": 0.4055,
+      "step": 319
+    },
+    {
+      "epoch": 0.020507562163547807,
+      "grad_norm": 2.5854239688820377,
+      "learning_rate": 1e-06,
+      "loss": 0.465,
+      "step": 320
+    },
+    {
+      "epoch": 0.020571648295308897,
+      "grad_norm": 2.792877790090399,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 321
+    },
+    {
+      "epoch": 0.020635734427069983,
+      "grad_norm": 2.6879878145297913,
+      "learning_rate": 1e-06,
+      "loss": 0.3912,
+      "step": 322
+    },
+    {
+      "epoch": 0.02069982055883107,
+      "grad_norm": 2.5764792723219703,
+      "learning_rate": 1e-06,
+      "loss": 0.4293,
+      "step": 323
+    },
+    {
+      "epoch": 0.020763906690592155,
+      "grad_norm": 2.584799346003916,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 324
+    },
+    {
+      "epoch": 0.02082799282235324,
+      "grad_norm": 2.52310368174129,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 325
+    },
+    {
+      "epoch": 0.02089207895411433,
+      "grad_norm": 2.631813901191913,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 326
+    },
+    {
+      "epoch": 0.020956165085875417,
+      "grad_norm": 2.640467081871137,
+      "learning_rate": 1e-06,
+      "loss": 0.4428,
+      "step": 327
+    },
+    {
+      "epoch": 0.021020251217636503,
+      "grad_norm": 2.9492291344272874,
+      "learning_rate": 1e-06,
+      "loss": 0.4127,
+      "step": 328
+    },
+    {
+      "epoch": 0.02108433734939759,
+      "grad_norm": 2.5242390714272114,
+      "learning_rate": 1e-06,
+      "loss": 0.5015,
+      "step": 329
+    },
+    {
+      "epoch": 0.02114842348115868,
+      "grad_norm": 2.6479950311376954,
+      "learning_rate": 1e-06,
+      "loss": 0.4574,
+      "step": 330
+    },
+    {
+      "epoch": 0.021212509612919764,
+      "grad_norm": 2.5907035344735116,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 331
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 2.4904752366096203,
+      "learning_rate": 1e-06,
+      "loss": 0.4231,
+      "step": 332
+    },
+    {
+      "epoch": 0.021340681876441937,
+      "grad_norm": 2.762482929644451,
+      "learning_rate": 1e-06,
+      "loss": 0.4319,
+      "step": 333
+    },
+    {
+      "epoch": 0.021404768008203026,
+      "grad_norm": 2.257106119148726,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 334
+    },
+    {
+      "epoch": 0.021468854139964112,
+      "grad_norm": 2.687648586593707,
+      "learning_rate": 1e-06,
+      "loss": 0.4059,
+      "step": 335
+    },
+    {
+      "epoch": 0.0215329402717252,
+      "grad_norm": 2.6657368936212715,
+      "learning_rate": 1e-06,
+      "loss": 0.3806,
+      "step": 336
+    },
+    {
+      "epoch": 0.021597026403486284,
+      "grad_norm": 2.6482827833947997,
+      "learning_rate": 1e-06,
+      "loss": 0.4097,
+      "step": 337
+    },
+    {
+      "epoch": 0.021661112535247374,
+      "grad_norm": 2.7093340741541425,
+      "learning_rate": 1e-06,
+      "loss": 0.4485,
+      "step": 338
+    },
+    {
+      "epoch": 0.02172519866700846,
+      "grad_norm": 2.3762505723651057,
+      "learning_rate": 1e-06,
+      "loss": 0.428,
+      "step": 339
+    },
+    {
+      "epoch": 0.021789284798769546,
+      "grad_norm": 2.423159148830173,
+      "learning_rate": 1e-06,
+      "loss": 0.3828,
+      "step": 340
+    },
+    {
+      "epoch": 0.021853370930530632,
+      "grad_norm": 2.716362646966935,
+      "learning_rate": 1e-06,
+      "loss": 0.4466,
+      "step": 341
+    },
+    {
+      "epoch": 0.02191745706229172,
+      "grad_norm": 2.6724315541317494,
+      "learning_rate": 1e-06,
+      "loss": 0.4175,
+      "step": 342
+    },
+    {
+      "epoch": 0.021981543194052808,
+      "grad_norm": 2.622452376572196,
+      "learning_rate": 1e-06,
+      "loss": 0.4326,
+      "step": 343
+    },
+    {
+      "epoch": 0.022045629325813894,
+      "grad_norm": 2.6641408148600485,
+      "learning_rate": 1e-06,
+      "loss": 0.4213,
+      "step": 344
+    },
+    {
+      "epoch": 0.02210971545757498,
+      "grad_norm": 2.587246078788654,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 345
+    },
+    {
+      "epoch": 0.022173801589336066,
+      "grad_norm": 2.684412921798805,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 346
+    },
+    {
+      "epoch": 0.022237887721097156,
+      "grad_norm": 2.4607146739904007,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 347
+    },
+    {
+      "epoch": 0.02230197385285824,
+      "grad_norm": 2.7249177701726817,
+      "learning_rate": 1e-06,
+      "loss": 0.4347,
+      "step": 348
+    },
+    {
+      "epoch": 0.022366059984619328,
+      "grad_norm": 2.7097559627929053,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 349
+    },
+    {
+      "epoch": 0.022430146116380414,
+      "grad_norm": 2.6075455207333387,
+      "learning_rate": 1e-06,
+      "loss": 0.3605,
+      "step": 350
+    },
+    {
+      "epoch": 0.022494232248141503,
+      "grad_norm": 2.526788415971298,
+      "learning_rate": 1e-06,
+      "loss": 0.4505,
+      "step": 351
+    },
+    {
+      "epoch": 0.02255831837990259,
+      "grad_norm": 2.6078284647598204,
+      "learning_rate": 1e-06,
+      "loss": 0.4309,
+      "step": 352
+    },
+    {
+      "epoch": 0.022622404511663675,
+      "grad_norm": 2.443089947367771,
+      "learning_rate": 1e-06,
+      "loss": 0.394,
+      "step": 353
+    },
+    {
+      "epoch": 0.02268649064342476,
+      "grad_norm": 2.694646337485449,
+      "learning_rate": 1e-06,
+      "loss": 0.4094,
+      "step": 354
+    },
+    {
+      "epoch": 0.02275057677518585,
+      "grad_norm": 2.539210140474977,
+      "learning_rate": 1e-06,
+      "loss": 0.4183,
+      "step": 355
+    },
+    {
+      "epoch": 0.022814662906946937,
+      "grad_norm": 2.5494243341233602,
+      "learning_rate": 1e-06,
+      "loss": 0.3934,
+      "step": 356
+    },
+    {
+      "epoch": 0.022878749038708023,
+      "grad_norm": 2.518198059218748,
+      "learning_rate": 1e-06,
+      "loss": 0.3786,
+      "step": 357
+    },
+    {
+      "epoch": 0.02294283517046911,
+      "grad_norm": 2.61819759245241,
+      "learning_rate": 1e-06,
+      "loss": 0.4574,
+      "step": 358
+    },
+    {
+      "epoch": 0.0230069213022302,
+      "grad_norm": 2.5581859112297005,
+      "learning_rate": 1e-06,
+      "loss": 0.3627,
+      "step": 359
+    },
+    {
+      "epoch": 0.023071007433991285,
+      "grad_norm": 2.587465107279955,
+      "learning_rate": 1e-06,
+      "loss": 0.3798,
+      "step": 360
+    },
+    {
+      "epoch": 0.02313509356575237,
+      "grad_norm": 2.648677113740983,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 361
+    },
+    {
+      "epoch": 0.023199179697513457,
+      "grad_norm": 2.7540181767323664,
+      "learning_rate": 1e-06,
+      "loss": 0.3658,
+      "step": 362
+    },
+    {
+      "epoch": 0.023263265829274547,
+      "grad_norm": 2.755584999096141,
+      "learning_rate": 1e-06,
+      "loss": 0.4769,
+      "step": 363
+    },
+    {
+      "epoch": 0.023327351961035633,
+      "grad_norm": 2.6396695649199176,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 364
+    },
+    {
+      "epoch": 0.02339143809279672,
+      "grad_norm": 2.6761169006385845,
+      "learning_rate": 1e-06,
+      "loss": 0.4838,
+      "step": 365
+    },
+    {
+      "epoch": 0.023455524224557805,
+      "grad_norm": 2.6889950878228004,
+      "learning_rate": 1e-06,
+      "loss": 0.4501,
+      "step": 366
+    },
+    {
+      "epoch": 0.02351961035631889,
+      "grad_norm": 2.7548073384891003,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 367
+    },
+    {
+      "epoch": 0.02358369648807998,
+      "grad_norm": 2.542233508453795,
+      "learning_rate": 1e-06,
+      "loss": 0.4433,
+      "step": 368
+    },
+    {
+      "epoch": 0.023647782619841066,
+      "grad_norm": 2.5380547510477487,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 369
+    },
+    {
+      "epoch": 0.023711868751602153,
+      "grad_norm": 2.5495340569743252,
+      "learning_rate": 1e-06,
+      "loss": 0.3672,
+      "step": 370
+    },
+    {
+      "epoch": 0.02377595488336324,
+      "grad_norm": 2.775600728413592,
+      "learning_rate": 1e-06,
+      "loss": 0.3908,
+      "step": 371
+    },
+    {
+      "epoch": 0.023840041015124328,
+      "grad_norm": 2.60490757100387,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 372
+    },
+    {
+      "epoch": 0.023904127146885414,
+      "grad_norm": 2.6266435496694682,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 373
+    },
+    {
+      "epoch": 0.0239682132786465,
+      "grad_norm": 2.3238570132425744,
+      "learning_rate": 1e-06,
+      "loss": 0.4079,
+      "step": 374
+    },
+    {
+      "epoch": 0.024032299410407586,
+      "grad_norm": 2.4584692833391917,
+      "learning_rate": 1e-06,
+      "loss": 0.4871,
+      "step": 375
+    },
+    {
+      "epoch": 0.024096385542168676,
+      "grad_norm": 2.6874135978719065,
+      "learning_rate": 1e-06,
+      "loss": 0.3946,
+      "step": 376
+    },
+    {
+      "epoch": 0.024160471673929762,
+      "grad_norm": 2.5538834092647704,
+      "learning_rate": 1e-06,
+      "loss": 0.3631,
+      "step": 377
+    },
+    {
+      "epoch": 0.024224557805690848,
+      "grad_norm": 2.8327606606394724,
+      "learning_rate": 1e-06,
+      "loss": 0.3849,
+      "step": 378
+    },
+    {
+      "epoch": 0.024288643937451934,
+      "grad_norm": 2.514808684223404,
+      "learning_rate": 1e-06,
+      "loss": 0.4206,
+      "step": 379
+    },
+    {
+      "epoch": 0.024352730069213024,
+      "grad_norm": 2.6324540798174185,
+      "learning_rate": 1e-06,
+      "loss": 0.4487,
+      "step": 380
+    },
+    {
+      "epoch": 0.02441681620097411,
+      "grad_norm": 2.6056342440608526,
+      "learning_rate": 1e-06,
+      "loss": 0.3951,
+      "step": 381
+    },
+    {
+      "epoch": 0.024480902332735196,
+      "grad_norm": 2.675148296855455,
+      "learning_rate": 1e-06,
+      "loss": 0.4448,
+      "step": 382
+    },
+    {
+      "epoch": 0.024544988464496282,
+      "grad_norm": 2.5831013701900947,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 383
+    },
+    {
+      "epoch": 0.02460907459625737,
+      "grad_norm": 2.5554733350090575,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 384
+    },
+    {
+      "epoch": 0.024673160728018458,
+      "grad_norm": 2.825553455546965,
+      "learning_rate": 1e-06,
+      "loss": 0.4427,
+      "step": 385
+    },
+    {
+      "epoch": 0.024737246859779544,
+      "grad_norm": 2.598203447299543,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 386
+    },
+    {
+      "epoch": 0.02480133299154063,
+      "grad_norm": 2.7532512685061796,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 387
+    },
+    {
+      "epoch": 0.024865419123301716,
+      "grad_norm": 2.7759105997753264,
+      "learning_rate": 1e-06,
+      "loss": 0.4646,
+      "step": 388
+    },
+    {
+      "epoch": 0.024929505255062805,
+      "grad_norm": 2.486755526528995,
+      "learning_rate": 1e-06,
+      "loss": 0.4424,
+      "step": 389
+    },
+    {
+      "epoch": 0.02499359138682389,
+      "grad_norm": 2.5947708705458394,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 390
+    },
+    {
+      "epoch": 0.025057677518584977,
+      "grad_norm": 2.4297539299884683,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 391
+    },
+    {
+      "epoch": 0.025121763650346064,
+      "grad_norm": 2.434233908402395,
+      "learning_rate": 1e-06,
+      "loss": 0.4528,
+      "step": 392
+    },
+    {
+      "epoch": 0.025185849782107153,
+      "grad_norm": 2.757448942774155,
+      "learning_rate": 1e-06,
+      "loss": 0.4467,
+      "step": 393
+    },
+    {
+      "epoch": 0.02524993591386824,
+      "grad_norm": 2.4998807548143014,
+      "learning_rate": 1e-06,
+      "loss": 0.4334,
+      "step": 394
+    },
+    {
+      "epoch": 0.025314022045629325,
+      "grad_norm": 2.419465872584267,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 395
+    },
+    {
+      "epoch": 0.02537810817739041,
+      "grad_norm": 2.679576216317354,
+      "learning_rate": 1e-06,
+      "loss": 0.4507,
+      "step": 396
+    },
+    {
+      "epoch": 0.0254421943091515,
+      "grad_norm": 2.5869533374512335,
+      "learning_rate": 1e-06,
+      "loss": 0.464,
+      "step": 397
+    },
+    {
+      "epoch": 0.025506280440912587,
+      "grad_norm": 2.8162022718685678,
+      "learning_rate": 1e-06,
+      "loss": 0.4536,
+      "step": 398
+    },
+    {
+      "epoch": 0.025570366572673673,
+      "grad_norm": 2.49440453274208,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 399
+    },
+    {
+      "epoch": 0.02563445270443476,
+      "grad_norm": 2.7267037535153813,
+      "learning_rate": 1e-06,
+      "loss": 0.3788,
+      "step": 400
+    },
+    {
+      "epoch": 0.02569853883619585,
+      "grad_norm": 2.6275422608283123,
+      "learning_rate": 1e-06,
+      "loss": 0.455,
+      "step": 401
+    },
+    {
+      "epoch": 0.025762624967956935,
+      "grad_norm": 2.468485363505419,
+      "learning_rate": 1e-06,
+      "loss": 0.4348,
+      "step": 402
+    },
+    {
+      "epoch": 0.02582671109971802,
+      "grad_norm": 2.6322373705642015,
+      "learning_rate": 1e-06,
+      "loss": 0.4474,
+      "step": 403
+    },
+    {
+      "epoch": 0.025890797231479107,
+      "grad_norm": 2.535796436272985,
+      "learning_rate": 1e-06,
+      "loss": 0.4392,
+      "step": 404
+    },
+    {
+      "epoch": 0.025954883363240196,
+      "grad_norm": 2.5836972859728915,
+      "learning_rate": 1e-06,
+      "loss": 0.429,
+      "step": 405
+    },
+    {
+      "epoch": 0.026018969495001282,
+      "grad_norm": 2.7316052338513432,
+      "learning_rate": 1e-06,
+      "loss": 0.396,
+      "step": 406
+    },
+    {
+      "epoch": 0.02608305562676237,
+      "grad_norm": 2.736398307464729,
+      "learning_rate": 1e-06,
+      "loss": 0.4456,
+      "step": 407
+    },
+    {
+      "epoch": 0.026147141758523455,
+      "grad_norm": 2.7440451094469394,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 408
+    },
+    {
+      "epoch": 0.026211227890284544,
+      "grad_norm": 2.4414167181146635,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 409
+    },
+    {
+      "epoch": 0.02627531402204563,
+      "grad_norm": 2.7337025666681787,
+      "learning_rate": 1e-06,
+      "loss": 0.4728,
+      "step": 410
+    },
+    {
+      "epoch": 0.026339400153806716,
+      "grad_norm": 2.6555366248003076,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 411
+    },
+    {
+      "epoch": 0.026403486285567802,
+      "grad_norm": 2.5538593566825405,
+      "learning_rate": 1e-06,
+      "loss": 0.4807,
+      "step": 412
+    },
+    {
+      "epoch": 0.02646757241732889,
+      "grad_norm": 2.5362557233108833,
+      "learning_rate": 1e-06,
+      "loss": 0.3934,
+      "step": 413
+    },
+    {
+      "epoch": 0.026531658549089978,
+      "grad_norm": 2.8275479404432775,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 414
+    },
+    {
+      "epoch": 0.026595744680851064,
+      "grad_norm": 2.8543582707538655,
+      "learning_rate": 1e-06,
+      "loss": 0.4315,
+      "step": 415
+    },
+    {
+      "epoch": 0.02665983081261215,
+      "grad_norm": 2.950759907375976,
+      "learning_rate": 1e-06,
+      "loss": 0.5034,
+      "step": 416
+    },
+    {
+      "epoch": 0.026723916944373236,
+      "grad_norm": 2.6064488094160705,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 417
+    },
+    {
+      "epoch": 0.026788003076134326,
+      "grad_norm": 2.7096530877343534,
+      "learning_rate": 1e-06,
+      "loss": 0.4349,
+      "step": 418
+    },
+    {
+      "epoch": 0.026852089207895412,
+      "grad_norm": 2.563785025605755,
+      "learning_rate": 1e-06,
+      "loss": 0.4283,
+      "step": 419
+    },
+    {
+      "epoch": 0.026916175339656498,
+      "grad_norm": 2.5884296422116213,
+      "learning_rate": 1e-06,
+      "loss": 0.3884,
+      "step": 420
+    },
+    {
+      "epoch": 0.026980261471417584,
+      "grad_norm": 2.652177567014019,
+      "learning_rate": 1e-06,
+      "loss": 0.3815,
+      "step": 421
+    },
+    {
+      "epoch": 0.027044347603178674,
+      "grad_norm": 2.8145378730635477,
+      "learning_rate": 1e-06,
+      "loss": 0.469,
+      "step": 422
+    },
+    {
+      "epoch": 0.02710843373493976,
+      "grad_norm": 2.6413005099038918,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 423
+    },
+    {
+      "epoch": 0.027172519866700846,
+      "grad_norm": 2.6259835527792803,
+      "learning_rate": 1e-06,
+      "loss": 0.4584,
+      "step": 424
+    },
+    {
+      "epoch": 0.02723660599846193,
+      "grad_norm": 2.59772191403426,
+      "learning_rate": 1e-06,
+      "loss": 0.4332,
+      "step": 425
+    },
+    {
+      "epoch": 0.02730069213022302,
+      "grad_norm": 2.6447993885707106,
+      "learning_rate": 1e-06,
+      "loss": 0.4098,
+      "step": 426
+    },
+    {
+      "epoch": 0.027364778261984107,
+      "grad_norm": 2.8823068371516865,
+      "learning_rate": 1e-06,
+      "loss": 0.4728,
+      "step": 427
+    },
+    {
+      "epoch": 0.027428864393745193,
+      "grad_norm": 2.6895802156764996,
+      "learning_rate": 1e-06,
+      "loss": 0.4462,
+      "step": 428
+    },
+    {
+      "epoch": 0.02749295052550628,
+      "grad_norm": 2.775049320708902,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 429
+    },
+    {
+      "epoch": 0.02755703665726737,
+      "grad_norm": 2.4854036703312783,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 430
+    },
+    {
+      "epoch": 0.027621122789028455,
+      "grad_norm": 2.628854745968563,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 431
+    },
+    {
+      "epoch": 0.02768520892078954,
+      "grad_norm": 2.543581473195899,
+      "learning_rate": 1e-06,
+      "loss": 0.4403,
+      "step": 432
+    },
+    {
+      "epoch": 0.027749295052550627,
+      "grad_norm": 2.5870776263158115,
+      "learning_rate": 1e-06,
+      "loss": 0.4503,
+      "step": 433
+    },
+    {
+      "epoch": 0.027813381184311713,
+      "grad_norm": 2.5550230493211266,
+      "learning_rate": 1e-06,
+      "loss": 0.4454,
+      "step": 434
+    },
+    {
+      "epoch": 0.027877467316072803,
+      "grad_norm": 2.517836595844116,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 435
+    },
+    {
+      "epoch": 0.02794155344783389,
+      "grad_norm": 2.532564498900615,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 436
+    },
+    {
+      "epoch": 0.028005639579594975,
+      "grad_norm": 2.617217239557506,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 437
+    },
+    {
+      "epoch": 0.02806972571135606,
+      "grad_norm": 2.4932234326126985,
+      "learning_rate": 1e-06,
+      "loss": 0.4112,
+      "step": 438
+    },
+    {
+      "epoch": 0.02813381184311715,
+      "grad_norm": 2.632681267604381,
+      "learning_rate": 1e-06,
+      "loss": 0.4631,
+      "step": 439
+    },
+    {
+      "epoch": 0.028197897974878237,
+      "grad_norm": 2.700129131791681,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 440
+    },
+    {
+      "epoch": 0.028261984106639323,
+      "grad_norm": 2.3125223839321034,
+      "learning_rate": 1e-06,
+      "loss": 0.4322,
+      "step": 441
+    },
+    {
+      "epoch": 0.02832607023840041,
+      "grad_norm": 2.6480507139738187,
+      "learning_rate": 1e-06,
+      "loss": 0.4028,
+      "step": 442
+    },
+    {
+      "epoch": 0.0283901563701615,
+      "grad_norm": 2.5860040775843087,
+      "learning_rate": 1e-06,
+      "loss": 0.441,
+      "step": 443
+    },
+    {
+      "epoch": 0.028454242501922584,
+      "grad_norm": 2.623689520293853,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 444
+    },
+    {
+      "epoch": 0.02851832863368367,
+      "grad_norm": 2.558893067356012,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 445
+    },
+    {
+      "epoch": 0.028582414765444757,
+      "grad_norm": 2.827213781887415,
+      "learning_rate": 1e-06,
+      "loss": 0.4805,
+      "step": 446
+    },
+    {
+      "epoch": 0.028646500897205846,
+      "grad_norm": 2.626705801630092,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 447
+    },
+    {
+      "epoch": 0.028710587028966932,
+      "grad_norm": 2.3921237896610315,
+      "learning_rate": 1e-06,
+      "loss": 0.4341,
+      "step": 448
+    },
+    {
+      "epoch": 0.02877467316072802,
+      "grad_norm": 2.594901028723936,
+      "learning_rate": 1e-06,
+      "loss": 0.4329,
+      "step": 449
+    },
+    {
+      "epoch": 0.028838759292489104,
+      "grad_norm": 2.6928586153085465,
+      "learning_rate": 1e-06,
+      "loss": 0.4559,
+      "step": 450
+    },
+    {
+      "epoch": 0.028902845424250194,
+      "grad_norm": 2.5095686668184713,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 451
+    },
+    {
+      "epoch": 0.02896693155601128,
+      "grad_norm": 2.7755150350769773,
+      "learning_rate": 1e-06,
+      "loss": 0.443,
+      "step": 452
+    },
+    {
+      "epoch": 0.029031017687772366,
+      "grad_norm": 2.3618305061019202,
+      "learning_rate": 1e-06,
+      "loss": 0.4055,
+      "step": 453
+    },
+    {
+      "epoch": 0.029095103819533452,
+      "grad_norm": 2.5548014934465026,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 454
+    },
+    {
+      "epoch": 0.029159189951294538,
+      "grad_norm": 2.565871109083848,
+      "learning_rate": 1e-06,
+      "loss": 0.4101,
+      "step": 455
+    },
+    {
+      "epoch": 0.029223276083055628,
+      "grad_norm": 2.653111195337013,
+      "learning_rate": 1e-06,
+      "loss": 0.4435,
+      "step": 456
+    },
+    {
+      "epoch": 0.029287362214816714,
+      "grad_norm": 2.4368979780190245,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 457
+    },
+    {
+      "epoch": 0.0293514483465778,
+      "grad_norm": 2.7911192897610047,
+      "learning_rate": 1e-06,
+      "loss": 0.3813,
+      "step": 458
+    },
+    {
+      "epoch": 0.029415534478338886,
+      "grad_norm": 2.4492006758826106,
+      "learning_rate": 1e-06,
+      "loss": 0.4112,
+      "step": 459
+    },
+    {
+      "epoch": 0.029479620610099976,
+      "grad_norm": 2.6617405331446555,
+      "learning_rate": 1e-06,
+      "loss": 0.4823,
+      "step": 460
+    },
+    {
+      "epoch": 0.02954370674186106,
+      "grad_norm": 2.43862787048992,
+      "learning_rate": 1e-06,
+      "loss": 0.4443,
+      "step": 461
+    },
+    {
+      "epoch": 0.029607792873622148,
+      "grad_norm": 2.4792268303893783,
+      "learning_rate": 1e-06,
+      "loss": 0.4563,
+      "step": 462
+    },
+    {
+      "epoch": 0.029671879005383234,
+      "grad_norm": 2.4850255441260267,
+      "learning_rate": 1e-06,
+      "loss": 0.4743,
+      "step": 463
+    },
+    {
+      "epoch": 0.029735965137144323,
+      "grad_norm": 2.5140153994819427,
+      "learning_rate": 1e-06,
+      "loss": 0.4696,
+      "step": 464
+    },
+    {
+      "epoch": 0.02980005126890541,
+      "grad_norm": 2.6261266820771585,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 465
+    },
+    {
+      "epoch": 0.029864137400666495,
+      "grad_norm": 2.6310050025757508,
+      "learning_rate": 1e-06,
+      "loss": 0.3847,
+      "step": 466
+    },
+    {
+      "epoch": 0.02992822353242758,
+      "grad_norm": 2.7082985977439176,
+      "learning_rate": 1e-06,
+      "loss": 0.3926,
+      "step": 467
+    },
+    {
+      "epoch": 0.02999230966418867,
+      "grad_norm": 2.553112208833126,
+      "learning_rate": 1e-06,
+      "loss": 0.396,
+      "step": 468
+    },
+    {
+      "epoch": 0.030056395795949757,
+      "grad_norm": 2.612291745367972,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 469
+    },
+    {
+      "epoch": 0.030120481927710843,
+      "grad_norm": 2.621969421745666,
+      "learning_rate": 1e-06,
+      "loss": 0.4045,
+      "step": 470
+    },
+    {
+      "epoch": 0.03018456805947193,
+      "grad_norm": 2.5381238717470183,
+      "learning_rate": 1e-06,
+      "loss": 0.4543,
+      "step": 471
+    },
+    {
+      "epoch": 0.03024865419123302,
+      "grad_norm": 2.665916627779648,
+      "learning_rate": 1e-06,
+      "loss": 0.4564,
+      "step": 472
+    },
+    {
+      "epoch": 0.030312740322994105,
+      "grad_norm": 2.4444266781182664,
+      "learning_rate": 1e-06,
+      "loss": 0.4394,
+      "step": 473
+    },
+    {
+      "epoch": 0.03037682645475519,
+      "grad_norm": 2.669252093062087,
+      "learning_rate": 1e-06,
+      "loss": 0.4722,
+      "step": 474
+    },
+    {
+      "epoch": 0.030440912586516277,
+      "grad_norm": 2.5986644470297136,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 475
+    },
+    {
+      "epoch": 0.030504998718277363,
+      "grad_norm": 2.516452232958566,
+      "learning_rate": 1e-06,
+      "loss": 0.453,
+      "step": 476
+    },
+    {
+      "epoch": 0.030569084850038453,
+      "grad_norm": 2.4799852114064214,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 477
+    },
+    {
+      "epoch": 0.03063317098179954,
+      "grad_norm": 2.691271105524432,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 478
+    },
+    {
+      "epoch": 0.030697257113560625,
+      "grad_norm": 2.6177226276670886,
+      "learning_rate": 1e-06,
+      "loss": 0.3892,
+      "step": 479
+    },
+    {
+      "epoch": 0.03076134324532171,
+      "grad_norm": 2.6112474010731055,
+      "learning_rate": 1e-06,
+      "loss": 0.4304,
+      "step": 480
+    },
+    {
+      "epoch": 0.0308254293770828,
+      "grad_norm": 2.43910372990363,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 481
+    },
+    {
+      "epoch": 0.030889515508843886,
+      "grad_norm": 2.560767071357324,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 482
+    },
+    {
+      "epoch": 0.030953601640604973,
+      "grad_norm": 2.545682557199924,
+      "learning_rate": 1e-06,
+      "loss": 0.4691,
+      "step": 483
+    },
+    {
+      "epoch": 0.03101768777236606,
+      "grad_norm": 2.575507981220968,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 484
+    },
+    {
+      "epoch": 0.031081773904127148,
+      "grad_norm": 2.398778610972209,
+      "learning_rate": 1e-06,
+      "loss": 0.4322,
+      "step": 485
+    },
+    {
+      "epoch": 0.031145860035888234,
+      "grad_norm": 2.468832999943894,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 486
+    },
+    {
+      "epoch": 0.03120994616764932,
+      "grad_norm": 2.6723453958820045,
+      "learning_rate": 1e-06,
+      "loss": 0.4476,
+      "step": 487
+    },
+    {
+      "epoch": 0.031274032299410406,
+      "grad_norm": 3.007937745124945,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 488
+    },
+    {
+      "epoch": 0.03133811843117149,
+      "grad_norm": 2.5372976061475625,
+      "learning_rate": 1e-06,
+      "loss": 0.4811,
+      "step": 489
+    },
+    {
+      "epoch": 0.03140220456293258,
+      "grad_norm": 2.4099062889186245,
+      "learning_rate": 1e-06,
+      "loss": 0.4239,
+      "step": 490
+    },
+    {
+      "epoch": 0.03146629069469367,
+      "grad_norm": 2.561389579372859,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 491
+    },
+    {
+      "epoch": 0.03153037682645476,
+      "grad_norm": 2.7289712094507492,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 492
+    },
+    {
+      "epoch": 0.031594462958215844,
+      "grad_norm": 2.8885100719999253,
+      "learning_rate": 1e-06,
+      "loss": 0.4456,
+      "step": 493
+    },
+    {
+      "epoch": 0.03165854908997693,
+      "grad_norm": 2.5064397467754573,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 494
+    },
+    {
+      "epoch": 0.031722635221738016,
+      "grad_norm": 2.412203696247204,
+      "learning_rate": 1e-06,
+      "loss": 0.3731,
+      "step": 495
+    },
+    {
+      "epoch": 0.0317867213534991,
+      "grad_norm": 2.629095888173398,
+      "learning_rate": 1e-06,
+      "loss": 0.3977,
+      "step": 496
+    },
+    {
+      "epoch": 0.03185080748526019,
+      "grad_norm": 2.4280514452263406,
+      "learning_rate": 1e-06,
+      "loss": 0.4447,
+      "step": 497
+    },
+    {
+      "epoch": 0.031914893617021274,
+      "grad_norm": 2.5575542980194443,
+      "learning_rate": 1e-06,
+      "loss": 0.3789,
+      "step": 498
+    },
+    {
+      "epoch": 0.03197897974878236,
+      "grad_norm": 2.5872529006915412,
+      "learning_rate": 1e-06,
+      "loss": 0.4307,
+      "step": 499
+    },
+    {
+      "epoch": 0.03204306588054345,
+      "grad_norm": 2.5440251604452992,
+      "learning_rate": 1e-06,
+      "loss": 0.3903,
+      "step": 500
+    },
+    {
+      "epoch": 0.03210715201230454,
+      "grad_norm": 2.7137592808907884,
+      "learning_rate": 1e-06,
+      "loss": 0.4266,
+      "step": 501
+    },
+    {
+      "epoch": 0.032171238144065625,
+      "grad_norm": 2.6186486459780167,
+      "learning_rate": 1e-06,
+      "loss": 0.434,
+      "step": 502
+    },
+    {
+      "epoch": 0.03223532427582671,
+      "grad_norm": 2.570526490487541,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 503
+    },
+    {
+      "epoch": 0.0322994104075878,
+      "grad_norm": 2.60892271445792,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 504
+    },
+    {
+      "epoch": 0.032363496539348884,
+      "grad_norm": 2.8058791368375835,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 505
+    },
+    {
+      "epoch": 0.03242758267110997,
+      "grad_norm": 2.3428317112470345,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 506
+    },
+    {
+      "epoch": 0.032491668802871056,
+      "grad_norm": 2.532222954783407,
+      "learning_rate": 1e-06,
+      "loss": 0.4619,
+      "step": 507
+    },
+    {
+      "epoch": 0.03255575493463215,
+      "grad_norm": 2.7351971936980672,
+      "learning_rate": 1e-06,
+      "loss": 0.3857,
+      "step": 508
+    },
+    {
+      "epoch": 0.032619841066393235,
+      "grad_norm": 2.4604615768666416,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 509
+    },
+    {
+      "epoch": 0.03268392719815432,
+      "grad_norm": 2.831748431791256,
+      "learning_rate": 1e-06,
+      "loss": 0.4475,
+      "step": 510
+    },
+    {
+      "epoch": 0.03274801332991541,
+      "grad_norm": 2.441688126278881,
+      "learning_rate": 1e-06,
+      "loss": 0.4578,
+      "step": 511
+    },
+    {
+      "epoch": 0.03281209946167649,
+      "grad_norm": 2.5457154513847606,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 512
+    },
+    {
+      "epoch": 0.03287618559343758,
+      "grad_norm": 2.6596542207786333,
+      "learning_rate": 1e-06,
+      "loss": 0.4234,
+      "step": 513
+    },
+    {
+      "epoch": 0.032940271725198665,
+      "grad_norm": 2.577103093592158,
+      "learning_rate": 1e-06,
+      "loss": 0.4362,
+      "step": 514
+    },
+    {
+      "epoch": 0.03300435785695975,
+      "grad_norm": 2.6865205065165476,
+      "learning_rate": 1e-06,
+      "loss": 0.457,
+      "step": 515
+    },
+    {
+      "epoch": 0.033068443988720844,
+      "grad_norm": 2.7935603445979593,
+      "learning_rate": 1e-06,
+      "loss": 0.3538,
+      "step": 516
+    },
+    {
+      "epoch": 0.03313253012048193,
+      "grad_norm": 2.5709612528638646,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 517
+    },
+    {
+      "epoch": 0.033196616252243016,
+      "grad_norm": 2.679906178087004,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 518
+    },
+    {
+      "epoch": 0.0332607023840041,
+      "grad_norm": 2.631332537538208,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 519
+    },
+    {
+      "epoch": 0.03332478851576519,
+      "grad_norm": 2.742087588652849,
+      "learning_rate": 1e-06,
+      "loss": 0.3902,
+      "step": 520
+    },
+    {
+      "epoch": 0.033388874647526275,
+      "grad_norm": 2.607913480915616,
+      "learning_rate": 1e-06,
+      "loss": 0.3595,
+      "step": 521
+    },
+    {
+      "epoch": 0.03345296077928736,
+      "grad_norm": 2.6446962621851564,
+      "learning_rate": 1e-06,
+      "loss": 0.3932,
+      "step": 522
+    },
+    {
+      "epoch": 0.03351704691104845,
+      "grad_norm": 2.7401141685722337,
+      "learning_rate": 1e-06,
+      "loss": 0.4399,
+      "step": 523
+    },
+    {
+      "epoch": 0.03358113304280953,
+      "grad_norm": 2.505928983276003,
+      "learning_rate": 1e-06,
+      "loss": 0.501,
+      "step": 524
+    },
+    {
+      "epoch": 0.033645219174570626,
+      "grad_norm": 2.7600474954065444,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 525
+    },
+    {
+      "epoch": 0.03370930530633171,
+      "grad_norm": 2.7364210341460997,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 526
+    },
+    {
+      "epoch": 0.0337733914380928,
+      "grad_norm": 2.750607243162528,
+      "learning_rate": 1e-06,
+      "loss": 0.4495,
+      "step": 527
+    },
+    {
+      "epoch": 0.033837477569853884,
+      "grad_norm": 2.8190619687628273,
+      "learning_rate": 1e-06,
+      "loss": 0.3647,
+      "step": 528
+    },
+    {
+      "epoch": 0.03390156370161497,
+      "grad_norm": 2.7488605163437243,
+      "learning_rate": 1e-06,
+      "loss": 0.3953,
+      "step": 529
+    },
+    {
+      "epoch": 0.033965649833376056,
+      "grad_norm": 2.4591430760247475,
+      "learning_rate": 1e-06,
+      "loss": 0.4333,
+      "step": 530
+    },
+    {
+      "epoch": 0.03402973596513714,
+      "grad_norm": 2.456421101011619,
+      "learning_rate": 1e-06,
+      "loss": 0.3921,
+      "step": 531
+    },
+    {
+      "epoch": 0.03409382209689823,
+      "grad_norm": 2.6154427035143994,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 532
+    },
+    {
+      "epoch": 0.03415790822865932,
+      "grad_norm": 2.6735778265914716,
+      "learning_rate": 1e-06,
+      "loss": 0.3936,
+      "step": 533
+    },
+    {
+      "epoch": 0.03422199436042041,
+      "grad_norm": 2.890439024855884,
+      "learning_rate": 1e-06,
+      "loss": 0.4686,
+      "step": 534
+    },
+    {
+      "epoch": 0.034286080492181494,
+      "grad_norm": 2.4630446860653006,
+      "learning_rate": 1e-06,
+      "loss": 0.427,
+      "step": 535
+    },
+    {
+      "epoch": 0.03435016662394258,
+      "grad_norm": 2.6268371157846073,
+      "learning_rate": 1e-06,
+      "loss": 0.4102,
+      "step": 536
+    },
+    {
+      "epoch": 0.034414252755703666,
+      "grad_norm": 2.610285178020935,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 537
+    },
+    {
+      "epoch": 0.03447833888746475,
+      "grad_norm": 2.6839716680166577,
+      "learning_rate": 1e-06,
+      "loss": 0.463,
+      "step": 538
+    },
+    {
+      "epoch": 0.03454242501922584,
+      "grad_norm": 2.749663653273078,
+      "learning_rate": 1e-06,
+      "loss": 0.4618,
+      "step": 539
+    },
+    {
+      "epoch": 0.034606511150986924,
+      "grad_norm": 2.7732262749372962,
+      "learning_rate": 1e-06,
+      "loss": 0.4329,
+      "step": 540
+    },
+    {
+      "epoch": 0.03467059728274801,
+      "grad_norm": 2.4709968867303482,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 541
+    },
+    {
+      "epoch": 0.0347346834145091,
+      "grad_norm": 2.636652286257393,
+      "learning_rate": 1e-06,
+      "loss": 0.4409,
+      "step": 542
+    },
+    {
+      "epoch": 0.03479876954627019,
+      "grad_norm": 2.6364517480767207,
+      "learning_rate": 1e-06,
+      "loss": 0.3699,
+      "step": 543
+    },
+    {
+      "epoch": 0.034862855678031275,
+      "grad_norm": 2.602465187056353,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 544
+    },
+    {
+      "epoch": 0.03492694180979236,
+      "grad_norm": 2.5206229660480135,
+      "learning_rate": 1e-06,
+      "loss": 0.4329,
+      "step": 545
+    },
+    {
+      "epoch": 0.03499102794155345,
+      "grad_norm": 2.404100546779136,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 546
+    },
+    {
+      "epoch": 0.03505511407331453,
+      "grad_norm": 2.5357402525324373,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 547
+    },
+    {
+      "epoch": 0.03511920020507562,
+      "grad_norm": 2.759792501919909,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 548
+    },
+    {
+      "epoch": 0.035183286336836705,
+      "grad_norm": 2.6433173572474207,
+      "learning_rate": 1e-06,
+      "loss": 0.4271,
+      "step": 549
+    },
+    {
+      "epoch": 0.0352473724685978,
+      "grad_norm": 2.4955771953873986,
+      "learning_rate": 1e-06,
+      "loss": 0.4189,
+      "step": 550
+    },
+    {
+      "epoch": 0.035311458600358885,
+      "grad_norm": 2.6502029354718526,
+      "learning_rate": 1e-06,
+      "loss": 0.4253,
+      "step": 551
+    },
+    {
+      "epoch": 0.03537554473211997,
+      "grad_norm": 2.5023269519541556,
+      "learning_rate": 1e-06,
+      "loss": 0.4171,
+      "step": 552
+    },
+    {
+      "epoch": 0.03543963086388106,
+      "grad_norm": 2.427919137521944,
+      "learning_rate": 1e-06,
+      "loss": 0.3839,
+      "step": 553
+    },
+    {
+      "epoch": 0.03550371699564214,
+      "grad_norm": 2.547181435232563,
+      "learning_rate": 1e-06,
+      "loss": 0.4268,
+      "step": 554
+    },
+    {
+      "epoch": 0.03556780312740323,
+      "grad_norm": 2.8027164655802412,
+      "learning_rate": 1e-06,
+      "loss": 0.4133,
+      "step": 555
+    },
+    {
+      "epoch": 0.035631889259164315,
+      "grad_norm": 2.6901481129495024,
+      "learning_rate": 1e-06,
+      "loss": 0.4793,
+      "step": 556
+    },
+    {
+      "epoch": 0.0356959753909254,
+      "grad_norm": 2.5079034785333008,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 557
+    },
+    {
+      "epoch": 0.035760061522686494,
+      "grad_norm": 2.403696079239236,
+      "learning_rate": 1e-06,
+      "loss": 0.444,
+      "step": 558
+    },
+    {
+      "epoch": 0.03582414765444758,
+      "grad_norm": 2.569305452424107,
+      "learning_rate": 1e-06,
+      "loss": 0.3475,
+      "step": 559
+    },
+    {
+      "epoch": 0.035888233786208666,
+      "grad_norm": 2.6545444286825983,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 560
+    },
+    {
+      "epoch": 0.03595231991796975,
+      "grad_norm": 2.440955279891903,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 561
+    },
+    {
+      "epoch": 0.03601640604973084,
+      "grad_norm": 2.847512451653497,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 562
+    },
+    {
+      "epoch": 0.036080492181491924,
+      "grad_norm": 2.6466369305096666,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 563
+    },
+    {
+      "epoch": 0.03614457831325301,
+      "grad_norm": 2.6832876331361475,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 564
+    },
+    {
+      "epoch": 0.0362086644450141,
+      "grad_norm": 2.5616868912458157,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 565
+    },
+    {
+      "epoch": 0.03627275057677518,
+      "grad_norm": 2.4878609095378854,
+      "learning_rate": 1e-06,
+      "loss": 0.455,
+      "step": 566
+    },
+    {
+      "epoch": 0.036336836708536276,
+      "grad_norm": 2.46531966648379,
+      "learning_rate": 1e-06,
+      "loss": 0.4627,
+      "step": 567
+    },
+    {
+      "epoch": 0.03640092284029736,
+      "grad_norm": 2.516195232483696,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 568
+    },
+    {
+      "epoch": 0.03646500897205845,
+      "grad_norm": 2.6044742204218174,
+      "learning_rate": 1e-06,
+      "loss": 0.4174,
+      "step": 569
+    },
+    {
+      "epoch": 0.036529095103819534,
+      "grad_norm": 2.5714531773465312,
+      "learning_rate": 1e-06,
+      "loss": 0.4185,
+      "step": 570
+    },
+    {
+      "epoch": 0.03659318123558062,
+      "grad_norm": 2.5553692530565004,
+      "learning_rate": 1e-06,
+      "loss": 0.4518,
+      "step": 571
+    },
+    {
+      "epoch": 0.036657267367341706,
+      "grad_norm": 2.5353351960631363,
+      "learning_rate": 1e-06,
+      "loss": 0.4417,
+      "step": 572
+    },
+    {
+      "epoch": 0.03672135349910279,
+      "grad_norm": 2.690225025998638,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 573
+    },
+    {
+      "epoch": 0.03678543963086388,
+      "grad_norm": 2.709566881886002,
+      "learning_rate": 1e-06,
+      "loss": 0.4066,
+      "step": 574
+    },
+    {
+      "epoch": 0.03684952576262497,
+      "grad_norm": 2.755681344115124,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 575
+    },
+    {
+      "epoch": 0.03691361189438606,
+      "grad_norm": 2.5693249686778685,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 576
+    },
+    {
+      "epoch": 0.03697769802614714,
+      "grad_norm": 2.4624250441741364,
+      "learning_rate": 1e-06,
+      "loss": 0.3905,
+      "step": 577
+    },
+    {
+      "epoch": 0.03704178415790823,
+      "grad_norm": 2.73004795352225,
+      "learning_rate": 1e-06,
+      "loss": 0.456,
+      "step": 578
+    },
+    {
+      "epoch": 0.037105870289669315,
+      "grad_norm": 2.7527083634276592,
+      "learning_rate": 1e-06,
+      "loss": 0.3965,
+      "step": 579
+    },
+    {
+      "epoch": 0.0371699564214304,
+      "grad_norm": 2.8028124217258004,
+      "learning_rate": 1e-06,
+      "loss": 0.471,
+      "step": 580
+    },
+    {
+      "epoch": 0.03723404255319149,
+      "grad_norm": 2.5880344409030807,
+      "learning_rate": 1e-06,
+      "loss": 0.3902,
+      "step": 581
+    },
+    {
+      "epoch": 0.037298128684952574,
+      "grad_norm": 2.4467098820143263,
+      "learning_rate": 1e-06,
+      "loss": 0.4309,
+      "step": 582
+    },
+    {
+      "epoch": 0.03736221481671366,
+      "grad_norm": 2.6909263668214414,
+      "learning_rate": 1e-06,
+      "loss": 0.4424,
+      "step": 583
+    },
+    {
+      "epoch": 0.03742630094847475,
+      "grad_norm": 2.5028705246054064,
+      "learning_rate": 1e-06,
+      "loss": 0.4534,
+      "step": 584
+    },
+    {
+      "epoch": 0.03749038708023584,
+      "grad_norm": 2.632149240813846,
+      "learning_rate": 1e-06,
+      "loss": 0.4429,
+      "step": 585
+    },
+    {
+      "epoch": 0.037554473211996925,
+      "grad_norm": 2.676547497619598,
+      "learning_rate": 1e-06,
+      "loss": 0.4092,
+      "step": 586
+    },
+    {
+      "epoch": 0.03761855934375801,
+      "grad_norm": 2.593412911659521,
+      "learning_rate": 1e-06,
+      "loss": 0.3827,
+      "step": 587
+    },
+    {
+      "epoch": 0.0376826454755191,
+      "grad_norm": 2.5646155672189868,
+      "learning_rate": 1e-06,
+      "loss": 0.4445,
+      "step": 588
+    },
+    {
+      "epoch": 0.03774673160728018,
+      "grad_norm": 2.6958374982600857,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 589
+    },
+    {
+      "epoch": 0.03781081773904127,
+      "grad_norm": 2.944875349581748,
+      "learning_rate": 1e-06,
+      "loss": 0.4577,
+      "step": 590
+    },
+    {
+      "epoch": 0.037874903870802355,
+      "grad_norm": 2.4909466851587285,
+      "learning_rate": 1e-06,
+      "loss": 0.3851,
+      "step": 591
+    },
+    {
+      "epoch": 0.03793899000256345,
+      "grad_norm": 2.6354951431246474,
+      "learning_rate": 1e-06,
+      "loss": 0.4192,
+      "step": 592
+    },
+    {
+      "epoch": 0.038003076134324534,
+      "grad_norm": 2.5659179494331514,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 593
+    },
+    {
+      "epoch": 0.03806716226608562,
+      "grad_norm": 2.7030130238756387,
+      "learning_rate": 1e-06,
+      "loss": 0.4631,
+      "step": 594
+    },
+    {
+      "epoch": 0.038131248397846706,
+      "grad_norm": 2.5036363687595413,
+      "learning_rate": 1e-06,
+      "loss": 0.3573,
+      "step": 595
+    },
+    {
+      "epoch": 0.03819533452960779,
+      "grad_norm": 2.568713565345006,
+      "learning_rate": 1e-06,
+      "loss": 0.4051,
+      "step": 596
+    },
+    {
+      "epoch": 0.03825942066136888,
+      "grad_norm": 2.7356219550249543,
+      "learning_rate": 1e-06,
+      "loss": 0.3943,
+      "step": 597
+    },
+    {
+      "epoch": 0.038323506793129965,
+      "grad_norm": 2.589753760976193,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 598
+    },
+    {
+      "epoch": 0.03838759292489105,
+      "grad_norm": 2.5004031202976598,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 599
+    },
+    {
+      "epoch": 0.038451679056652144,
+      "grad_norm": 2.640214926702102,
+      "learning_rate": 1e-06,
+      "loss": 0.4696,
+      "step": 600
+    },
+    {
+      "epoch": 0.03851576518841323,
+      "grad_norm": 2.4476421037580227,
+      "learning_rate": 1e-06,
+      "loss": 0.3422,
+      "step": 601
+    },
+    {
+      "epoch": 0.038579851320174316,
+      "grad_norm": 2.6129179653473673,
+      "learning_rate": 1e-06,
+      "loss": 0.385,
+      "step": 602
+    },
+    {
+      "epoch": 0.0386439374519354,
+      "grad_norm": 2.8065814805198643,
+      "learning_rate": 1e-06,
+      "loss": 0.4985,
+      "step": 603
+    },
+    {
+      "epoch": 0.03870802358369649,
+      "grad_norm": 2.7506065468121665,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 604
+    },
+    {
+      "epoch": 0.038772109715457574,
+      "grad_norm": 2.743913537942316,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 605
+    },
+    {
+      "epoch": 0.03883619584721866,
+      "grad_norm": 2.7130495638219982,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 606
+    },
+    {
+      "epoch": 0.038900281978979746,
+      "grad_norm": 2.5914175128432153,
+      "learning_rate": 1e-06,
+      "loss": 0.4005,
+      "step": 607
+    },
+    {
+      "epoch": 0.03896436811074083,
+      "grad_norm": 2.5482424585290366,
+      "learning_rate": 1e-06,
+      "loss": 0.4562,
+      "step": 608
+    },
+    {
+      "epoch": 0.039028454242501925,
+      "grad_norm": 2.7290427011123395,
+      "learning_rate": 1e-06,
+      "loss": 0.4013,
+      "step": 609
+    },
+    {
+      "epoch": 0.03909254037426301,
+      "grad_norm": 2.5427635742774117,
+      "learning_rate": 1e-06,
+      "loss": 0.3788,
+      "step": 610
+    },
+    {
+      "epoch": 0.0391566265060241,
+      "grad_norm": 2.561653737428255,
+      "learning_rate": 1e-06,
+      "loss": 0.4245,
+      "step": 611
+    },
+    {
+      "epoch": 0.039220712637785184,
+      "grad_norm": 2.5955070887416323,
+      "learning_rate": 1e-06,
+      "loss": 0.4266,
+      "step": 612
+    },
+    {
+      "epoch": 0.03928479876954627,
+      "grad_norm": 2.54265029522293,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 613
+    },
+    {
+      "epoch": 0.039348884901307356,
+      "grad_norm": 2.606212536478632,
+      "learning_rate": 1e-06,
+      "loss": 0.43,
+      "step": 614
+    },
+    {
+      "epoch": 0.03941297103306844,
+      "grad_norm": 2.6840178730682682,
+      "learning_rate": 1e-06,
+      "loss": 0.4871,
+      "step": 615
+    },
+    {
+      "epoch": 0.03947705716482953,
+      "grad_norm": 2.650414922571063,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 616
+    },
+    {
+      "epoch": 0.03954114329659062,
+      "grad_norm": 2.673526302945202,
+      "learning_rate": 1e-06,
+      "loss": 0.5218,
+      "step": 617
+    },
+    {
+      "epoch": 0.03960522942835171,
+      "grad_norm": 2.553206108448424,
+      "learning_rate": 1e-06,
+      "loss": 0.4068,
+      "step": 618
+    },
+    {
+      "epoch": 0.03966931556011279,
+      "grad_norm": 3.8623068130329914,
+      "learning_rate": 1e-06,
+      "loss": 0.4685,
+      "step": 619
+    },
+    {
+      "epoch": 0.03973340169187388,
+      "grad_norm": 2.611209530336014,
+      "learning_rate": 1e-06,
+      "loss": 0.4692,
+      "step": 620
+    },
+    {
+      "epoch": 0.039797487823634965,
+      "grad_norm": 2.612983189021595,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 621
+    },
+    {
+      "epoch": 0.03986157395539605,
+      "grad_norm": 2.642479051481998,
+      "learning_rate": 1e-06,
+      "loss": 0.4703,
+      "step": 622
+    },
+    {
+      "epoch": 0.03992566008715714,
+      "grad_norm": 2.7796089353924867,
+      "learning_rate": 1e-06,
+      "loss": 0.4151,
+      "step": 623
+    },
+    {
+      "epoch": 0.03998974621891822,
+      "grad_norm": 2.58513886175909,
+      "learning_rate": 1e-06,
+      "loss": 0.3998,
+      "step": 624
+    },
+    {
+      "epoch": 0.040053832350679316,
+      "grad_norm": 2.5219815352139916,
+      "learning_rate": 1e-06,
+      "loss": 0.4732,
+      "step": 625
+    },
+    {
+      "epoch": 0.0401179184824404,
+      "grad_norm": 2.427279598308452,
+      "learning_rate": 1e-06,
+      "loss": 0.4312,
+      "step": 626
+    },
+    {
+      "epoch": 0.04018200461420149,
+      "grad_norm": 2.7451148950328763,
+      "learning_rate": 1e-06,
+      "loss": 0.451,
+      "step": 627
+    },
+    {
+      "epoch": 0.040246090745962575,
+      "grad_norm": 2.935612939952673,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 628
+    },
+    {
+      "epoch": 0.04031017687772366,
+      "grad_norm": 2.315351642452188,
+      "learning_rate": 1e-06,
+      "loss": 0.42,
+      "step": 629
+    },
+    {
+      "epoch": 0.04037426300948475,
+      "grad_norm": 2.616335080998465,
+      "learning_rate": 1e-06,
+      "loss": 0.459,
+      "step": 630
+    },
+    {
+      "epoch": 0.04043834914124583,
+      "grad_norm": 2.7994650340660527,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 631
+    },
+    {
+      "epoch": 0.04050243527300692,
+      "grad_norm": 2.7990589665063315,
+      "learning_rate": 1e-06,
+      "loss": 0.433,
+      "step": 632
+    },
+    {
+      "epoch": 0.040566521404768005,
+      "grad_norm": 2.726625812194557,
+      "learning_rate": 1e-06,
+      "loss": 0.4934,
+      "step": 633
+    },
+    {
+      "epoch": 0.0406306075365291,
+      "grad_norm": 2.6595974411965004,
+      "learning_rate": 1e-06,
+      "loss": 0.3912,
+      "step": 634
+    },
+    {
+      "epoch": 0.040694693668290184,
+      "grad_norm": 2.6317170132064054,
+      "learning_rate": 1e-06,
+      "loss": 0.4233,
+      "step": 635
+    },
+    {
+      "epoch": 0.04075877980005127,
+      "grad_norm": 2.7113262836328507,
+      "learning_rate": 1e-06,
+      "loss": 0.3809,
+      "step": 636
+    },
+    {
+      "epoch": 0.040822865931812356,
+      "grad_norm": 2.7276916836582963,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 637
+    },
+    {
+      "epoch": 0.04088695206357344,
+      "grad_norm": 2.511155076141722,
+      "learning_rate": 1e-06,
+      "loss": 0.3905,
+      "step": 638
+    },
+    {
+      "epoch": 0.04095103819533453,
+      "grad_norm": 2.6231514302817867,
+      "learning_rate": 1e-06,
+      "loss": 0.4993,
+      "step": 639
+    },
+    {
+      "epoch": 0.041015124327095615,
+      "grad_norm": 2.8738210559312836,
+      "learning_rate": 1e-06,
+      "loss": 0.4589,
+      "step": 640
+    },
+    {
+      "epoch": 0.0410792104588567,
+      "grad_norm": 2.630954950788824,
+      "learning_rate": 1e-06,
+      "loss": 0.4045,
+      "step": 641
+    },
+    {
+      "epoch": 0.041143296590617794,
+      "grad_norm": 2.3813698177760885,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 642
+    },
+    {
+      "epoch": 0.04120738272237888,
+      "grad_norm": 2.559089835922891,
+      "learning_rate": 1e-06,
+      "loss": 0.4153,
+      "step": 643
+    },
+    {
+      "epoch": 0.041271468854139966,
+      "grad_norm": 2.6301614669286764,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 644
+    },
+    {
+      "epoch": 0.04133555498590105,
+      "grad_norm": 2.826724753336134,
+      "learning_rate": 1e-06,
+      "loss": 0.4785,
+      "step": 645
+    },
+    {
+      "epoch": 0.04139964111766214,
+      "grad_norm": 2.5852268129152094,
+      "learning_rate": 1e-06,
+      "loss": 0.4127,
+      "step": 646
+    },
+    {
+      "epoch": 0.041463727249423224,
+      "grad_norm": 2.5517838435167337,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 647
+    },
+    {
+      "epoch": 0.04152781338118431,
+      "grad_norm": 2.5003343630514485,
+      "learning_rate": 1e-06,
+      "loss": 0.4664,
+      "step": 648
+    },
+    {
+      "epoch": 0.041591899512945396,
+      "grad_norm": 2.660162354251772,
+      "learning_rate": 1e-06,
+      "loss": 0.4587,
+      "step": 649
+    },
+    {
+      "epoch": 0.04165598564470648,
+      "grad_norm": 2.519788649872128,
+      "learning_rate": 1e-06,
+      "loss": 0.4124,
+      "step": 650
+    },
+    {
+      "epoch": 0.041720071776467575,
+      "grad_norm": 2.6811482996986684,
+      "learning_rate": 1e-06,
+      "loss": 0.4979,
+      "step": 651
+    },
+    {
+      "epoch": 0.04178415790822866,
+      "grad_norm": 2.617520724416409,
+      "learning_rate": 1e-06,
+      "loss": 0.4485,
+      "step": 652
+    },
+    {
+      "epoch": 0.04184824403998975,
+      "grad_norm": 2.488909344232931,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 653
+    },
+    {
+      "epoch": 0.04191233017175083,
+      "grad_norm": 2.6149410343930644,
+      "learning_rate": 1e-06,
+      "loss": 0.3825,
+      "step": 654
+    },
+    {
+      "epoch": 0.04197641630351192,
+      "grad_norm": 2.5128514869435525,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 655
+    },
+    {
+      "epoch": 0.042040502435273006,
+      "grad_norm": 2.677894898570438,
+      "learning_rate": 1e-06,
+      "loss": 0.3868,
+      "step": 656
+    },
+    {
+      "epoch": 0.04210458856703409,
+      "grad_norm": 2.5073846226183494,
+      "learning_rate": 1e-06,
+      "loss": 0.4001,
+      "step": 657
+    },
+    {
+      "epoch": 0.04216867469879518,
+      "grad_norm": 2.5959443672213887,
+      "learning_rate": 1e-06,
+      "loss": 0.3518,
+      "step": 658
+    },
+    {
+      "epoch": 0.04223276083055627,
+      "grad_norm": 2.6503459543932113,
+      "learning_rate": 1e-06,
+      "loss": 0.4283,
+      "step": 659
+    },
+    {
+      "epoch": 0.04229684696231736,
+      "grad_norm": 2.43063879482731,
+      "learning_rate": 1e-06,
+      "loss": 0.4251,
+      "step": 660
+    },
+    {
+      "epoch": 0.04236093309407844,
+      "grad_norm": 2.349576722965158,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 661
+    },
+    {
+      "epoch": 0.04242501922583953,
+      "grad_norm": 2.537089027277767,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 662
+    },
+    {
+      "epoch": 0.042489105357600615,
+      "grad_norm": 2.7224838638581756,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 663
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 2.7403703661711303,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 664
+    },
+    {
+      "epoch": 0.04261727762112279,
+      "grad_norm": 2.481829066148711,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 665
+    },
+    {
+      "epoch": 0.04268136375288387,
+      "grad_norm": 2.7058959104350286,
+      "learning_rate": 1e-06,
+      "loss": 0.3839,
+      "step": 666
+    },
+    {
+      "epoch": 0.042745449884644966,
+      "grad_norm": 2.664470985436932,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 667
+    },
+    {
+      "epoch": 0.04280953601640605,
+      "grad_norm": 2.989381219322504,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 668
+    },
+    {
+      "epoch": 0.04287362214816714,
+      "grad_norm": 2.456391983663052,
+      "learning_rate": 1e-06,
+      "loss": 0.4497,
+      "step": 669
+    },
+    {
+      "epoch": 0.042937708279928224,
+      "grad_norm": 2.5686791944245355,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 670
+    },
+    {
+      "epoch": 0.04300179441168931,
+      "grad_norm": 2.5171257630542496,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 671
+    },
+    {
+      "epoch": 0.0430658805434504,
+      "grad_norm": 2.7514044064489958,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 672
+    },
+    {
+      "epoch": 0.04312996667521148,
+      "grad_norm": 2.588593452159588,
+      "learning_rate": 1e-06,
+      "loss": 0.4361,
+      "step": 673
+    },
+    {
+      "epoch": 0.04319405280697257,
+      "grad_norm": 2.5337204798422133,
+      "learning_rate": 1e-06,
+      "loss": 0.4651,
+      "step": 674
+    },
+    {
+      "epoch": 0.043258138938733655,
+      "grad_norm": 2.80848924235811,
+      "learning_rate": 1e-06,
+      "loss": 0.3884,
+      "step": 675
+    },
+    {
+      "epoch": 0.04332222507049475,
+      "grad_norm": 2.4781318443632703,
+      "learning_rate": 1e-06,
+      "loss": 0.4071,
+      "step": 676
+    },
+    {
+      "epoch": 0.043386311202255834,
+      "grad_norm": 2.6636080398686017,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 677
+    },
+    {
+      "epoch": 0.04345039733401692,
+      "grad_norm": 2.6875653127583643,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 678
+    },
+    {
+      "epoch": 0.043514483465778006,
+      "grad_norm": 2.593545387753299,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 679
+    },
+    {
+      "epoch": 0.04357856959753909,
+      "grad_norm": 2.6927729543543393,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 680
+    },
+    {
+      "epoch": 0.04364265572930018,
+      "grad_norm": 2.683321727952175,
+      "learning_rate": 1e-06,
+      "loss": 0.4527,
+      "step": 681
+    },
+    {
+      "epoch": 0.043706741861061264,
+      "grad_norm": 2.4442159701665838,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 682
+    },
+    {
+      "epoch": 0.04377082799282235,
+      "grad_norm": 2.6443057440007722,
+      "learning_rate": 1e-06,
+      "loss": 0.4226,
+      "step": 683
+    },
+    {
+      "epoch": 0.04383491412458344,
+      "grad_norm": 2.771577021353034,
+      "learning_rate": 1e-06,
+      "loss": 0.4209,
+      "step": 684
+    },
+    {
+      "epoch": 0.04389900025634453,
+      "grad_norm": 2.5830604191286577,
+      "learning_rate": 1e-06,
+      "loss": 0.4699,
+      "step": 685
+    },
+    {
+      "epoch": 0.043963086388105616,
+      "grad_norm": 2.7145392457704367,
+      "learning_rate": 1e-06,
+      "loss": 0.3852,
+      "step": 686
+    },
+    {
+      "epoch": 0.0440271725198667,
+      "grad_norm": 2.6292966393223773,
+      "learning_rate": 1e-06,
+      "loss": 0.4188,
+      "step": 687
+    },
+    {
+      "epoch": 0.04409125865162779,
+      "grad_norm": 2.549423161227808,
+      "learning_rate": 1e-06,
+      "loss": 0.4607,
+      "step": 688
+    },
+    {
+      "epoch": 0.044155344783388874,
+      "grad_norm": 2.8191967007014416,
+      "learning_rate": 1e-06,
+      "loss": 0.4822,
+      "step": 689
+    },
+    {
+      "epoch": 0.04421943091514996,
+      "grad_norm": 2.7238873369280596,
+      "learning_rate": 1e-06,
+      "loss": 0.4639,
+      "step": 690
+    },
+    {
+      "epoch": 0.044283517046911046,
+      "grad_norm": 2.633151381982928,
+      "learning_rate": 1e-06,
+      "loss": 0.4557,
+      "step": 691
+    },
+    {
+      "epoch": 0.04434760317867213,
+      "grad_norm": 2.7228749845037745,
+      "learning_rate": 1e-06,
+      "loss": 0.4778,
+      "step": 692
+    },
+    {
+      "epoch": 0.044411689310433225,
+      "grad_norm": 2.7679898994678944,
+      "learning_rate": 1e-06,
+      "loss": 0.4181,
+      "step": 693
+    },
+    {
+      "epoch": 0.04447577544219431,
+      "grad_norm": 2.589789446094907,
+      "learning_rate": 1e-06,
+      "loss": 0.4236,
+      "step": 694
+    },
+    {
+      "epoch": 0.0445398615739554,
+      "grad_norm": 2.511018179407139,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 695
+    },
+    {
+      "epoch": 0.04460394770571648,
+      "grad_norm": 2.7885072030309206,
+      "learning_rate": 1e-06,
+      "loss": 0.4553,
+      "step": 696
+    },
+    {
+      "epoch": 0.04466803383747757,
+      "grad_norm": 2.7021041650758506,
+      "learning_rate": 1e-06,
+      "loss": 0.3864,
+      "step": 697
+    },
+    {
+      "epoch": 0.044732119969238655,
+      "grad_norm": 2.788293327898669,
+      "learning_rate": 1e-06,
+      "loss": 0.4676,
+      "step": 698
+    },
+    {
+      "epoch": 0.04479620610099974,
+      "grad_norm": 2.5530815918507215,
+      "learning_rate": 1e-06,
+      "loss": 0.4304,
+      "step": 699
+    },
+    {
+      "epoch": 0.04486029223276083,
+      "grad_norm": 2.3701294923623166,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 700
+    },
+    {
+      "epoch": 0.04492437836452192,
+      "grad_norm": 2.521705987774555,
+      "learning_rate": 1e-06,
+      "loss": 0.4147,
+      "step": 701
+    },
+    {
+      "epoch": 0.04498846449628301,
+      "grad_norm": 2.7337919356294273,
+      "learning_rate": 1e-06,
+      "loss": 0.4036,
+      "step": 702
+    },
+    {
+      "epoch": 0.04505255062804409,
+      "grad_norm": 2.7334466833635713,
+      "learning_rate": 1e-06,
+      "loss": 0.4737,
+      "step": 703
+    },
+    {
+      "epoch": 0.04511663675980518,
+      "grad_norm": 2.412574306067936,
+      "learning_rate": 1e-06,
+      "loss": 0.4092,
+      "step": 704
+    },
+    {
+      "epoch": 0.045180722891566265,
+      "grad_norm": 2.6647181788543675,
+      "learning_rate": 1e-06,
+      "loss": 0.4644,
+      "step": 705
+    },
+    {
+      "epoch": 0.04524480902332735,
+      "grad_norm": 2.8176686628649867,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 706
+    },
+    {
+      "epoch": 0.04530889515508844,
+      "grad_norm": 3.5376406758684555,
+      "learning_rate": 1e-06,
+      "loss": 0.4637,
+      "step": 707
+    },
+    {
+      "epoch": 0.04537298128684952,
+      "grad_norm": 2.648046979088884,
+      "learning_rate": 1e-06,
+      "loss": 0.436,
+      "step": 708
+    },
+    {
+      "epoch": 0.045437067418610616,
+      "grad_norm": 2.714391240179539,
+      "learning_rate": 1e-06,
+      "loss": 0.4007,
+      "step": 709
+    },
+    {
+      "epoch": 0.0455011535503717,
+      "grad_norm": 2.4256539477484993,
+      "learning_rate": 1e-06,
+      "loss": 0.4697,
+      "step": 710
+    },
+    {
+      "epoch": 0.04556523968213279,
+      "grad_norm": 2.53217824501919,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 711
+    },
+    {
+      "epoch": 0.045629325813893874,
+      "grad_norm": 2.627890030554265,
+      "learning_rate": 1e-06,
+      "loss": 0.3706,
+      "step": 712
+    },
+    {
+      "epoch": 0.04569341194565496,
+      "grad_norm": 2.518196492721495,
+      "learning_rate": 1e-06,
+      "loss": 0.4087,
+      "step": 713
+    },
+    {
+      "epoch": 0.045757498077416046,
+      "grad_norm": 2.496806126024874,
+      "learning_rate": 1e-06,
+      "loss": 0.4553,
+      "step": 714
+    },
+    {
+      "epoch": 0.04582158420917713,
+      "grad_norm": 2.5758802594573242,
+      "learning_rate": 1e-06,
+      "loss": 0.3946,
+      "step": 715
+    },
+    {
+      "epoch": 0.04588567034093822,
+      "grad_norm": 2.6327983887455875,
+      "learning_rate": 1e-06,
+      "loss": 0.4266,
+      "step": 716
+    },
+    {
+      "epoch": 0.045949756472699305,
+      "grad_norm": 2.5940640401988104,
+      "learning_rate": 1e-06,
+      "loss": 0.4717,
+      "step": 717
+    },
+    {
+      "epoch": 0.0460138426044604,
+      "grad_norm": 2.6292851190752056,
+      "learning_rate": 1e-06,
+      "loss": 0.4645,
+      "step": 718
+    },
+    {
+      "epoch": 0.046077928736221484,
+      "grad_norm": 2.5407643752820706,
+      "learning_rate": 1e-06,
+      "loss": 0.4496,
+      "step": 719
+    },
+    {
+      "epoch": 0.04614201486798257,
+      "grad_norm": 2.7968598288395667,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 720
+    },
+    {
+      "epoch": 0.046206100999743656,
+      "grad_norm": 2.5396197101295246,
+      "learning_rate": 1e-06,
+      "loss": 0.4702,
+      "step": 721
+    },
+    {
+      "epoch": 0.04627018713150474,
+      "grad_norm": 2.6880034453423662,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 722
+    },
+    {
+      "epoch": 0.04633427326326583,
+      "grad_norm": 2.534639229971012,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 723
+    },
+    {
+      "epoch": 0.046398359395026914,
+      "grad_norm": 2.7035380317990705,
+      "learning_rate": 1e-06,
+      "loss": 0.4194,
+      "step": 724
+    },
+    {
+      "epoch": 0.046462445526788,
+      "grad_norm": 2.5100605403625913,
+      "learning_rate": 1e-06,
+      "loss": 0.4387,
+      "step": 725
+    },
+    {
+      "epoch": 0.04652653165854909,
+      "grad_norm": 2.787377804865803,
+      "learning_rate": 1e-06,
+      "loss": 0.4391,
+      "step": 726
+    },
+    {
+      "epoch": 0.04659061779031018,
+      "grad_norm": 2.758137307249647,
+      "learning_rate": 1e-06,
+      "loss": 0.4297,
+      "step": 727
+    },
+    {
+      "epoch": 0.046654703922071265,
+      "grad_norm": 2.6893335848609987,
+      "learning_rate": 1e-06,
+      "loss": 0.4442,
+      "step": 728
+    },
+    {
+      "epoch": 0.04671879005383235,
+      "grad_norm": 2.596806970800409,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 729
+    },
+    {
+      "epoch": 0.04678287618559344,
+      "grad_norm": 2.607569512934613,
+      "learning_rate": 1e-06,
+      "loss": 0.3982,
+      "step": 730
+    },
+    {
+      "epoch": 0.046846962317354524,
+      "grad_norm": 2.5297392747277994,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 731
+    },
+    {
+      "epoch": 0.04691104844911561,
+      "grad_norm": 2.6521936744466195,
+      "learning_rate": 1e-06,
+      "loss": 0.4147,
+      "step": 732
+    },
+    {
+      "epoch": 0.046975134580876696,
+      "grad_norm": 2.876213805294052,
+      "learning_rate": 1e-06,
+      "loss": 0.395,
+      "step": 733
+    },
+    {
+      "epoch": 0.04703922071263778,
+      "grad_norm": 2.4043885041918243,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 734
+    },
+    {
+      "epoch": 0.047103306844398875,
+      "grad_norm": 2.5867612136864055,
+      "learning_rate": 1e-06,
+      "loss": 0.4444,
+      "step": 735
+    },
+    {
+      "epoch": 0.04716739297615996,
+      "grad_norm": 2.5203294696988134,
+      "learning_rate": 1e-06,
+      "loss": 0.437,
+      "step": 736
+    },
+    {
+      "epoch": 0.04723147910792105,
+      "grad_norm": 2.6451605123098587,
+      "learning_rate": 1e-06,
+      "loss": 0.4449,
+      "step": 737
+    },
+    {
+      "epoch": 0.04729556523968213,
+      "grad_norm": 2.6941293120299137,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 738
+    },
+    {
+      "epoch": 0.04735965137144322,
+      "grad_norm": 2.6302676678079546,
+      "learning_rate": 1e-06,
+      "loss": 0.4505,
+      "step": 739
+    },
+    {
+      "epoch": 0.047423737503204305,
+      "grad_norm": 2.592261969149947,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 740
+    },
+    {
+      "epoch": 0.04748782363496539,
+      "grad_norm": 2.682832075298025,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 741
+    },
+    {
+      "epoch": 0.04755190976672648,
+      "grad_norm": 2.5907747282023887,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 742
+    },
+    {
+      "epoch": 0.04761599589848757,
+      "grad_norm": 2.6833032775341943,
+      "learning_rate": 1e-06,
+      "loss": 0.399,
+      "step": 743
+    },
+    {
+      "epoch": 0.047680082030248656,
+      "grad_norm": 2.518606400690256,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 744
+    },
+    {
+      "epoch": 0.04774416816200974,
+      "grad_norm": 2.479270755919287,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 745
+    },
+    {
+      "epoch": 0.04780825429377083,
+      "grad_norm": 2.72515135116294,
+      "learning_rate": 1e-06,
+      "loss": 0.4247,
+      "step": 746
+    },
+    {
+      "epoch": 0.047872340425531915,
+      "grad_norm": 2.4749722434919716,
+      "learning_rate": 1e-06,
+      "loss": 0.4729,
+      "step": 747
+    },
+    {
+      "epoch": 0.047936426557293,
+      "grad_norm": 2.5106869846945323,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 748
+    },
+    {
+      "epoch": 0.04800051268905409,
+      "grad_norm": 2.665593207338953,
+      "learning_rate": 1e-06,
+      "loss": 0.4037,
+      "step": 749
+    },
+    {
+      "epoch": 0.04806459882081517,
+      "grad_norm": 2.7037283823905405,
+      "learning_rate": 1e-06,
+      "loss": 0.451,
+      "step": 750
+    },
+    {
+      "epoch": 0.048128684952576266,
+      "grad_norm": 2.433991765145583,
+      "learning_rate": 1e-06,
+      "loss": 0.4518,
+      "step": 751
+    },
+    {
+      "epoch": 0.04819277108433735,
+      "grad_norm": 2.6227479993797833,
+      "learning_rate": 1e-06,
+      "loss": 0.327,
+      "step": 752
+    },
+    {
+      "epoch": 0.04825685721609844,
+      "grad_norm": 2.5651412483654323,
+      "learning_rate": 1e-06,
+      "loss": 0.4565,
+      "step": 753
+    },
+    {
+      "epoch": 0.048320943347859524,
+      "grad_norm": 2.4244933108312683,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 754
+    },
+    {
+      "epoch": 0.04838502947962061,
+      "grad_norm": 2.41146156297117,
+      "learning_rate": 1e-06,
+      "loss": 0.4178,
+      "step": 755
+    },
+    {
+      "epoch": 0.048449115611381696,
+      "grad_norm": 2.8872723239228097,
+      "learning_rate": 1e-06,
+      "loss": 0.4369,
+      "step": 756
+    },
+    {
+      "epoch": 0.04851320174314278,
+      "grad_norm": 2.762725037885391,
+      "learning_rate": 1e-06,
+      "loss": 0.3781,
+      "step": 757
+    },
+    {
+      "epoch": 0.04857728787490387,
+      "grad_norm": 3.836723679916666,
+      "learning_rate": 1e-06,
+      "loss": 0.4037,
+      "step": 758
+    },
+    {
+      "epoch": 0.048641374006664954,
+      "grad_norm": 2.8109148560486727,
+      "learning_rate": 1e-06,
+      "loss": 0.4655,
+      "step": 759
+    },
+    {
+      "epoch": 0.04870546013842605,
+      "grad_norm": 2.577080298605142,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 760
+    },
+    {
+      "epoch": 0.048769546270187134,
+      "grad_norm": 2.65526854422781,
+      "learning_rate": 1e-06,
+      "loss": 0.5098,
+      "step": 761
+    },
+    {
+      "epoch": 0.04883363240194822,
+      "grad_norm": 2.525793932433678,
+      "learning_rate": 1e-06,
+      "loss": 0.3975,
+      "step": 762
+    },
+    {
+      "epoch": 0.048897718533709306,
+      "grad_norm": 2.7031533973927915,
+      "learning_rate": 1e-06,
+      "loss": 0.4672,
+      "step": 763
+    },
+    {
+      "epoch": 0.04896180466547039,
+      "grad_norm": 2.772546979327886,
+      "learning_rate": 1e-06,
+      "loss": 0.4592,
+      "step": 764
+    },
+    {
+      "epoch": 0.04902589079723148,
+      "grad_norm": 2.7047272161441316,
+      "learning_rate": 1e-06,
+      "loss": 0.4611,
+      "step": 765
+    },
+    {
+      "epoch": 0.049089976928992564,
+      "grad_norm": 2.6979587125081963,
+      "learning_rate": 1e-06,
+      "loss": 0.4094,
+      "step": 766
+    },
+    {
+      "epoch": 0.04915406306075365,
+      "grad_norm": 2.525668570583122,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 767
+    },
+    {
+      "epoch": 0.04921814919251474,
+      "grad_norm": 2.538200424294564,
+      "learning_rate": 1e-06,
+      "loss": 0.3992,
+      "step": 768
+    },
+    {
+      "epoch": 0.04928223532427583,
+      "grad_norm": 2.4589512816952372,
+      "learning_rate": 1e-06,
+      "loss": 0.3886,
+      "step": 769
+    },
+    {
+      "epoch": 0.049346321456036915,
+      "grad_norm": 2.74365547311966,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 770
+    },
+    {
+      "epoch": 0.049410407587798,
+      "grad_norm": 2.8613608191226456,
+      "learning_rate": 1e-06,
+      "loss": 0.417,
+      "step": 771
+    },
+    {
+      "epoch": 0.04947449371955909,
+      "grad_norm": 2.4818805818190293,
+      "learning_rate": 1e-06,
+      "loss": 0.4414,
+      "step": 772
+    },
+    {
+      "epoch": 0.04953857985132017,
+      "grad_norm": 3.073554752078375,
+      "learning_rate": 1e-06,
+      "loss": 0.3623,
+      "step": 773
+    },
+    {
+      "epoch": 0.04960266598308126,
+      "grad_norm": 2.6000204539317986,
+      "learning_rate": 1e-06,
+      "loss": 0.4448,
+      "step": 774
+    },
+    {
+      "epoch": 0.049666752114842345,
+      "grad_norm": 2.619867444073265,
+      "learning_rate": 1e-06,
+      "loss": 0.4846,
+      "step": 775
+    },
+    {
+      "epoch": 0.04973083824660343,
+      "grad_norm": 2.670953914332981,
+      "learning_rate": 1e-06,
+      "loss": 0.4124,
+      "step": 776
+    },
+    {
+      "epoch": 0.049794924378364525,
+      "grad_norm": 2.6431772566079537,
+      "learning_rate": 1e-06,
+      "loss": 0.4111,
+      "step": 777
+    },
+    {
+      "epoch": 0.04985901051012561,
+      "grad_norm": 2.379782074934585,
+      "learning_rate": 1e-06,
+      "loss": 0.4399,
+      "step": 778
+    },
+    {
+      "epoch": 0.0499230966418867,
+      "grad_norm": 2.5242383740225645,
+      "learning_rate": 1e-06,
+      "loss": 0.3977,
+      "step": 779
+    },
+    {
+      "epoch": 0.04998718277364778,
+      "grad_norm": 2.4585612977491214,
+      "learning_rate": 1e-06,
+      "loss": 0.4342,
+      "step": 780
+    },
+    {
+      "epoch": 0.05005126890540887,
+      "grad_norm": 2.5353463721233833,
+      "learning_rate": 1e-06,
+      "loss": 0.409,
+      "step": 781
+    },
+    {
+      "epoch": 0.050115355037169955,
+      "grad_norm": 2.771798641700129,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 782
+    },
+    {
+      "epoch": 0.05017944116893104,
+      "grad_norm": 2.4567227275972785,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 783
+    },
+    {
+      "epoch": 0.05024352730069213,
+      "grad_norm": 2.776241471884275,
+      "learning_rate": 1e-06,
+      "loss": 0.3772,
+      "step": 784
+    },
+    {
+      "epoch": 0.05030761343245322,
+      "grad_norm": 2.73917716665947,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 785
+    },
+    {
+      "epoch": 0.050371699564214306,
+      "grad_norm": 2.4608060384184696,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 786
+    },
+    {
+      "epoch": 0.05043578569597539,
+      "grad_norm": 2.7782795289847333,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 787
+    },
+    {
+      "epoch": 0.05049987182773648,
+      "grad_norm": 2.4978989115524106,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 788
+    },
+    {
+      "epoch": 0.050563957959497564,
+      "grad_norm": 2.5533215903850977,
+      "learning_rate": 1e-06,
+      "loss": 0.3687,
+      "step": 789
+    },
+    {
+      "epoch": 0.05062804409125865,
+      "grad_norm": 2.7587787615557113,
+      "learning_rate": 1e-06,
+      "loss": 0.427,
+      "step": 790
+    },
+    {
+      "epoch": 0.05069213022301974,
+      "grad_norm": 2.6589908486758125,
+      "learning_rate": 1e-06,
+      "loss": 0.4661,
+      "step": 791
+    },
+    {
+      "epoch": 0.05075621635478082,
+      "grad_norm": 2.574238405368089,
+      "learning_rate": 1e-06,
+      "loss": 0.4779,
+      "step": 792
+    },
+    {
+      "epoch": 0.050820302486541916,
+      "grad_norm": 2.6274966108889255,
+      "learning_rate": 1e-06,
+      "loss": 0.4217,
+      "step": 793
+    },
+    {
+      "epoch": 0.050884388618303,
+      "grad_norm": 3.2801557437132556,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 794
+    },
+    {
+      "epoch": 0.05094847475006409,
+      "grad_norm": 2.6941030070682874,
+      "learning_rate": 1e-06,
+      "loss": 0.466,
+      "step": 795
+    },
+    {
+      "epoch": 0.051012560881825174,
+      "grad_norm": 2.5176002839833918,
+      "learning_rate": 1e-06,
+      "loss": 0.4492,
+      "step": 796
+    },
+    {
+      "epoch": 0.05107664701358626,
+      "grad_norm": 2.7530193209358567,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 797
+    },
+    {
+      "epoch": 0.051140733145347346,
+      "grad_norm": 2.565968695255342,
+      "learning_rate": 1e-06,
+      "loss": 0.3498,
+      "step": 798
+    },
+    {
+      "epoch": 0.05120481927710843,
+      "grad_norm": 2.6259720811960445,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 799
+    },
+    {
+      "epoch": 0.05126890540886952,
+      "grad_norm": 2.6154860713109773,
+      "learning_rate": 1e-06,
+      "loss": 0.4164,
+      "step": 800
+    },
+    {
+      "epoch": 0.051332991540630604,
+      "grad_norm": 2.7563490613331054,
+      "learning_rate": 1e-06,
+      "loss": 0.4331,
+      "step": 801
+    },
+    {
+      "epoch": 0.0513970776723917,
+      "grad_norm": 2.588060868381382,
+      "learning_rate": 1e-06,
+      "loss": 0.4109,
+      "step": 802
+    },
+    {
+      "epoch": 0.05146116380415278,
+      "grad_norm": 2.4579680369870234,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 803
+    },
+    {
+      "epoch": 0.05152524993591387,
+      "grad_norm": 2.4116736631822464,
+      "learning_rate": 1e-06,
+      "loss": 0.4438,
+      "step": 804
+    },
+    {
+      "epoch": 0.051589336067674955,
+      "grad_norm": 2.695563644208034,
+      "learning_rate": 1e-06,
+      "loss": 0.4688,
+      "step": 805
+    },
+    {
+      "epoch": 0.05165342219943604,
+      "grad_norm": 2.675047438176051,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 806
+    },
+    {
+      "epoch": 0.05171750833119713,
+      "grad_norm": 2.748181109672208,
+      "learning_rate": 1e-06,
+      "loss": 0.4171,
+      "step": 807
+    },
+    {
+      "epoch": 0.051781594462958214,
+      "grad_norm": 2.3059289383795263,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 808
+    },
+    {
+      "epoch": 0.0518456805947193,
+      "grad_norm": 2.5384132760352434,
+      "learning_rate": 1e-06,
+      "loss": 0.4316,
+      "step": 809
+    },
+    {
+      "epoch": 0.05190976672648039,
+      "grad_norm": 2.647356877858427,
+      "learning_rate": 1e-06,
+      "loss": 0.4583,
+      "step": 810
+    },
+    {
+      "epoch": 0.05197385285824148,
+      "grad_norm": 2.6468297347585636,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 811
+    },
+    {
+      "epoch": 0.052037938990002565,
+      "grad_norm": 2.694772877787142,
+      "learning_rate": 1e-06,
+      "loss": 0.4127,
+      "step": 812
+    },
+    {
+      "epoch": 0.05210202512176365,
+      "grad_norm": 2.7739794514088914,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 813
+    },
+    {
+      "epoch": 0.05216611125352474,
+      "grad_norm": 2.7432106258958444,
+      "learning_rate": 1e-06,
+      "loss": 0.4478,
+      "step": 814
+    },
+    {
+      "epoch": 0.05223019738528582,
+      "grad_norm": 2.59447934250964,
+      "learning_rate": 1e-06,
+      "loss": 0.4175,
+      "step": 815
+    },
+    {
+      "epoch": 0.05229428351704691,
+      "grad_norm": 2.482449882094235,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 816
+    },
+    {
+      "epoch": 0.052358369648807995,
+      "grad_norm": 2.6200358287917687,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 817
+    },
+    {
+      "epoch": 0.05242245578056909,
+      "grad_norm": 2.6787235723782645,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 818
+    },
+    {
+      "epoch": 0.052486541912330174,
+      "grad_norm": 2.857976316647533,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 819
+    },
+    {
+      "epoch": 0.05255062804409126,
+      "grad_norm": 2.4674037209798287,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 820
+    },
+    {
+      "epoch": 0.052614714175852347,
+      "grad_norm": 2.581328829253118,
+      "learning_rate": 1e-06,
+      "loss": 0.4306,
+      "step": 821
+    },
+    {
+      "epoch": 0.05267880030761343,
+      "grad_norm": 2.5713295173776984,
+      "learning_rate": 1e-06,
+      "loss": 0.4197,
+      "step": 822
+    },
+    {
+      "epoch": 0.05274288643937452,
+      "grad_norm": 2.6617855766435903,
+      "learning_rate": 1e-06,
+      "loss": 0.453,
+      "step": 823
+    },
+    {
+      "epoch": 0.052806972571135605,
+      "grad_norm": 2.5223890207481925,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 824
+    },
+    {
+      "epoch": 0.05287105870289669,
+      "grad_norm": 2.532919291158444,
+      "learning_rate": 1e-06,
+      "loss": 0.4401,
+      "step": 825
+    },
+    {
+      "epoch": 0.05293514483465778,
+      "grad_norm": 2.6565628872172256,
+      "learning_rate": 1e-06,
+      "loss": 0.4112,
+      "step": 826
+    },
+    {
+      "epoch": 0.05299923096641887,
+      "grad_norm": 2.593028943009522,
+      "learning_rate": 1e-06,
+      "loss": 0.429,
+      "step": 827
+    },
+    {
+      "epoch": 0.053063317098179956,
+      "grad_norm": 2.713800694208457,
+      "learning_rate": 1e-06,
+      "loss": 0.4659,
+      "step": 828
+    },
+    {
+      "epoch": 0.05312740322994104,
+      "grad_norm": 2.4576869289326706,
+      "learning_rate": 1e-06,
+      "loss": 0.4199,
+      "step": 829
+    },
+    {
+      "epoch": 0.05319148936170213,
+      "grad_norm": 2.4801827666076552,
+      "learning_rate": 1e-06,
+      "loss": 0.4272,
+      "step": 830
+    },
+    {
+      "epoch": 0.053255575493463214,
+      "grad_norm": 2.582789118414582,
+      "learning_rate": 1e-06,
+      "loss": 0.4289,
+      "step": 831
+    },
+    {
+      "epoch": 0.0533196616252243,
+      "grad_norm": 2.48821093605307,
+      "learning_rate": 1e-06,
+      "loss": 0.4121,
+      "step": 832
+    },
+    {
+      "epoch": 0.053383747756985386,
+      "grad_norm": 2.6186323943996435,
+      "learning_rate": 1e-06,
+      "loss": 0.441,
+      "step": 833
+    },
+    {
+      "epoch": 0.05344783388874647,
+      "grad_norm": 2.5517032828506334,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 834
+    },
+    {
+      "epoch": 0.053511920020507565,
+      "grad_norm": 2.5356349325994647,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 835
+    },
+    {
+      "epoch": 0.05357600615226865,
+      "grad_norm": 2.428198485601343,
+      "learning_rate": 1e-06,
+      "loss": 0.4706,
+      "step": 836
+    },
+    {
+      "epoch": 0.05364009228402974,
+      "grad_norm": 2.587909226071667,
+      "learning_rate": 1e-06,
+      "loss": 0.4062,
+      "step": 837
+    },
+    {
+      "epoch": 0.053704178415790824,
+      "grad_norm": 2.309073113032567,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 838
+    },
+    {
+      "epoch": 0.05376826454755191,
+      "grad_norm": 2.6703159072154077,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 839
+    },
+    {
+      "epoch": 0.053832350679312996,
+      "grad_norm": 2.71350739697573,
+      "learning_rate": 1e-06,
+      "loss": 0.4495,
+      "step": 840
+    },
+    {
+      "epoch": 0.05389643681107408,
+      "grad_norm": 2.434892539079085,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 841
+    },
+    {
+      "epoch": 0.05396052294283517,
+      "grad_norm": 2.5963354526268425,
+      "learning_rate": 1e-06,
+      "loss": 0.425,
+      "step": 842
+    },
+    {
+      "epoch": 0.054024609074596254,
+      "grad_norm": 2.483702809890466,
+      "learning_rate": 1e-06,
+      "loss": 0.4452,
+      "step": 843
+    },
+    {
+      "epoch": 0.05408869520635735,
+      "grad_norm": 2.5440750960584957,
+      "learning_rate": 1e-06,
+      "loss": 0.3545,
+      "step": 844
+    },
+    {
+      "epoch": 0.05415278133811843,
+      "grad_norm": 2.483060350154876,
+      "learning_rate": 1e-06,
+      "loss": 0.4014,
+      "step": 845
+    },
+    {
+      "epoch": 0.05421686746987952,
+      "grad_norm": 2.65872885985031,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 846
+    },
+    {
+      "epoch": 0.054280953601640605,
+      "grad_norm": 2.5818989112102173,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 847
+    },
+    {
+      "epoch": 0.05434503973340169,
+      "grad_norm": 2.4076980150335734,
+      "learning_rate": 1e-06,
+      "loss": 0.3354,
+      "step": 848
+    },
+    {
+      "epoch": 0.05440912586516278,
+      "grad_norm": 2.698007185792458,
+      "learning_rate": 1e-06,
+      "loss": 0.4506,
+      "step": 849
+    },
+    {
+      "epoch": 0.05447321199692386,
+      "grad_norm": 2.519075483872261,
+      "learning_rate": 1e-06,
+      "loss": 0.4133,
+      "step": 850
+    },
+    {
+      "epoch": 0.05453729812868495,
+      "grad_norm": 2.717885074101232,
+      "learning_rate": 1e-06,
+      "loss": 0.4175,
+      "step": 851
+    },
+    {
+      "epoch": 0.05460138426044604,
+      "grad_norm": 2.5856521915485646,
+      "learning_rate": 1e-06,
+      "loss": 0.4253,
+      "step": 852
+    },
+    {
+      "epoch": 0.05466547039220713,
+      "grad_norm": 2.4556651119087127,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 853
+    },
+    {
+      "epoch": 0.054729556523968215,
+      "grad_norm": 2.7721277540478146,
+      "learning_rate": 1e-06,
+      "loss": 0.3671,
+      "step": 854
+    },
+    {
+      "epoch": 0.0547936426557293,
+      "grad_norm": 2.657352492338648,
+      "learning_rate": 1e-06,
+      "loss": 0.4178,
+      "step": 855
+    },
+    {
+      "epoch": 0.05485772878749039,
+      "grad_norm": 2.7006089581535075,
+      "learning_rate": 1e-06,
+      "loss": 0.4183,
+      "step": 856
+    },
+    {
+      "epoch": 0.05492181491925147,
+      "grad_norm": 2.5670709358295647,
+      "learning_rate": 1e-06,
+      "loss": 0.4558,
+      "step": 857
+    },
+    {
+      "epoch": 0.05498590105101256,
+      "grad_norm": 2.628637826655462,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 858
+    },
+    {
+      "epoch": 0.055049987182773645,
+      "grad_norm": 2.6148005379775783,
+      "learning_rate": 1e-06,
+      "loss": 0.4136,
+      "step": 859
+    },
+    {
+      "epoch": 0.05511407331453474,
+      "grad_norm": 2.591000403813115,
+      "learning_rate": 1e-06,
+      "loss": 0.3866,
+      "step": 860
+    },
+    {
+      "epoch": 0.055178159446295824,
+      "grad_norm": 2.507601296484238,
+      "learning_rate": 1e-06,
+      "loss": 0.3998,
+      "step": 861
+    },
+    {
+      "epoch": 0.05524224557805691,
+      "grad_norm": 2.787771022744685,
+      "learning_rate": 1e-06,
+      "loss": 0.4026,
+      "step": 862
+    },
+    {
+      "epoch": 0.055306331709817996,
+      "grad_norm": 2.518082388993556,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 863
+    },
+    {
+      "epoch": 0.05537041784157908,
+      "grad_norm": 2.4587915178077777,
+      "learning_rate": 1e-06,
+      "loss": 0.4266,
+      "step": 864
+    },
+    {
+      "epoch": 0.05543450397334017,
+      "grad_norm": 2.615686357034752,
+      "learning_rate": 1e-06,
+      "loss": 0.368,
+      "step": 865
+    },
+    {
+      "epoch": 0.055498590105101255,
+      "grad_norm": 2.761059961241006,
+      "learning_rate": 1e-06,
+      "loss": 0.5015,
+      "step": 866
+    },
+    {
+      "epoch": 0.05556267623686234,
+      "grad_norm": 2.5272432083229535,
+      "learning_rate": 1e-06,
+      "loss": 0.4705,
+      "step": 867
+    },
+    {
+      "epoch": 0.05562676236862343,
+      "grad_norm": 2.59284819194525,
+      "learning_rate": 1e-06,
+      "loss": 0.403,
+      "step": 868
+    },
+    {
+      "epoch": 0.05569084850038452,
+      "grad_norm": 2.533049808944894,
+      "learning_rate": 1e-06,
+      "loss": 0.394,
+      "step": 869
+    },
+    {
+      "epoch": 0.055754934632145606,
+      "grad_norm": 2.504421255278961,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 870
+    },
+    {
+      "epoch": 0.05581902076390669,
+      "grad_norm": 2.771424162674612,
+      "learning_rate": 1e-06,
+      "loss": 0.4824,
+      "step": 871
+    },
+    {
+      "epoch": 0.05588310689566778,
+      "grad_norm": 2.5494936421830285,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 872
+    },
+    {
+      "epoch": 0.055947193027428864,
+      "grad_norm": 2.626180064034078,
+      "learning_rate": 1e-06,
+      "loss": 0.4615,
+      "step": 873
+    },
+    {
+      "epoch": 0.05601127915918995,
+      "grad_norm": 2.845992108008082,
+      "learning_rate": 1e-06,
+      "loss": 0.4442,
+      "step": 874
+    },
+    {
+      "epoch": 0.056075365290951036,
+      "grad_norm": 2.741059913668609,
+      "learning_rate": 1e-06,
+      "loss": 0.4047,
+      "step": 875
+    },
+    {
+      "epoch": 0.05613945142271212,
+      "grad_norm": 2.870070922473511,
+      "learning_rate": 1e-06,
+      "loss": 0.4648,
+      "step": 876
+    },
+    {
+      "epoch": 0.056203537554473215,
+      "grad_norm": 2.5058746300769434,
+      "learning_rate": 1e-06,
+      "loss": 0.3809,
+      "step": 877
+    },
+    {
+      "epoch": 0.0562676236862343,
+      "grad_norm": 2.560616702011279,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 878
+    },
+    {
+      "epoch": 0.05633170981799539,
+      "grad_norm": 2.6594905071835955,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 879
+    },
+    {
+      "epoch": 0.05639579594975647,
+      "grad_norm": 2.5214292041437827,
+      "learning_rate": 1e-06,
+      "loss": 0.4427,
+      "step": 880
+    },
+    {
+      "epoch": 0.05645988208151756,
+      "grad_norm": 2.5077576684919407,
+      "learning_rate": 1e-06,
+      "loss": 0.4656,
+      "step": 881
+    },
+    {
+      "epoch": 0.056523968213278646,
+      "grad_norm": 2.5092725959002546,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 882
+    },
+    {
+      "epoch": 0.05658805434503973,
+      "grad_norm": 2.580445517000057,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 883
+    },
+    {
+      "epoch": 0.05665214047680082,
+      "grad_norm": 2.5735197772804193,
+      "learning_rate": 1e-06,
+      "loss": 0.4221,
+      "step": 884
+    },
+    {
+      "epoch": 0.056716226608561904,
+      "grad_norm": 2.574408375918761,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 885
+    },
+    {
+      "epoch": 0.056780312740323,
+      "grad_norm": 2.780282548148286,
+      "learning_rate": 1e-06,
+      "loss": 0.4947,
+      "step": 886
+    },
+    {
+      "epoch": 0.05684439887208408,
+      "grad_norm": 2.4928626518173274,
+      "learning_rate": 1e-06,
+      "loss": 0.3504,
+      "step": 887
+    },
+    {
+      "epoch": 0.05690848500384517,
+      "grad_norm": 2.553051950988801,
+      "learning_rate": 1e-06,
+      "loss": 0.3975,
+      "step": 888
+    },
+    {
+      "epoch": 0.056972571135606255,
+      "grad_norm": 2.658416930515607,
+      "learning_rate": 1e-06,
+      "loss": 0.4394,
+      "step": 889
+    },
+    {
+      "epoch": 0.05703665726736734,
+      "grad_norm": 2.485961423782349,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 890
+    },
+    {
+      "epoch": 0.05710074339912843,
+      "grad_norm": 2.667713489712823,
+      "learning_rate": 1e-06,
+      "loss": 0.4705,
+      "step": 891
+    },
+    {
+      "epoch": 0.05716482953088951,
+      "grad_norm": 2.534564282196064,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 892
+    },
+    {
+      "epoch": 0.0572289156626506,
+      "grad_norm": 2.4356599669017496,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 893
+    },
+    {
+      "epoch": 0.05729300179441169,
+      "grad_norm": 2.7300363917836066,
+      "learning_rate": 1e-06,
+      "loss": 0.4208,
+      "step": 894
+    },
+    {
+      "epoch": 0.05735708792617278,
+      "grad_norm": 2.4426263028281503,
+      "learning_rate": 1e-06,
+      "loss": 0.4138,
+      "step": 895
+    },
+    {
+      "epoch": 0.057421174057933864,
+      "grad_norm": 2.4273359063929116,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 896
+    },
+    {
+      "epoch": 0.05748526018969495,
+      "grad_norm": 2.683961557261951,
+      "learning_rate": 1e-06,
+      "loss": 0.4253,
+      "step": 897
+    },
+    {
+      "epoch": 0.05754934632145604,
+      "grad_norm": 2.716449807419154,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 898
+    },
+    {
+      "epoch": 0.05761343245321712,
+      "grad_norm": 2.489474578355706,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 899
+    },
+    {
+      "epoch": 0.05767751858497821,
+      "grad_norm": 2.621299699094432,
+      "learning_rate": 1e-06,
+      "loss": 0.3961,
+      "step": 900
+    },
+    {
+      "epoch": 0.057741604716739295,
+      "grad_norm": 2.7042519805441545,
+      "learning_rate": 1e-06,
+      "loss": 0.4085,
+      "step": 901
+    },
+    {
+      "epoch": 0.05780569084850039,
+      "grad_norm": 2.8709190595594793,
+      "learning_rate": 1e-06,
+      "loss": 0.4351,
+      "step": 902
+    },
+    {
+      "epoch": 0.057869776980261474,
+      "grad_norm": 2.7239595338843543,
+      "learning_rate": 1e-06,
+      "loss": 0.4457,
+      "step": 903
+    },
+    {
+      "epoch": 0.05793386311202256,
+      "grad_norm": 2.6942314635143743,
+      "learning_rate": 1e-06,
+      "loss": 0.437,
+      "step": 904
+    },
+    {
+      "epoch": 0.057997949243783646,
+      "grad_norm": 2.561892748218017,
+      "learning_rate": 1e-06,
+      "loss": 0.4445,
+      "step": 905
+    },
+    {
+      "epoch": 0.05806203537554473,
+      "grad_norm": 2.6940166517257977,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 906
+    },
+    {
+      "epoch": 0.05812612150730582,
+      "grad_norm": 2.8794761385227083,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 907
+    },
+    {
+      "epoch": 0.058190207639066904,
+      "grad_norm": 2.6035218967013662,
+      "learning_rate": 1e-06,
+      "loss": 0.4566,
+      "step": 908
+    },
+    {
+      "epoch": 0.05825429377082799,
+      "grad_norm": 2.658548866501046,
+      "learning_rate": 1e-06,
+      "loss": 0.4506,
+      "step": 909
+    },
+    {
+      "epoch": 0.058318379902589076,
+      "grad_norm": 2.7266152577743723,
+      "learning_rate": 1e-06,
+      "loss": 0.502,
+      "step": 910
+    },
+    {
+      "epoch": 0.05838246603435017,
+      "grad_norm": 2.480477229574691,
+      "learning_rate": 1e-06,
+      "loss": 0.4381,
+      "step": 911
+    },
+    {
+      "epoch": 0.058446552166111256,
+      "grad_norm": 2.571579618617921,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 912
+    },
+    {
+      "epoch": 0.05851063829787234,
+      "grad_norm": 2.673325790509034,
+      "learning_rate": 1e-06,
+      "loss": 0.4085,
+      "step": 913
+    },
+    {
+      "epoch": 0.05857472442963343,
+      "grad_norm": 2.585655375519541,
+      "learning_rate": 1e-06,
+      "loss": 0.4323,
+      "step": 914
+    },
+    {
+      "epoch": 0.058638810561394514,
+      "grad_norm": 2.5704842362508846,
+      "learning_rate": 1e-06,
+      "loss": 0.4449,
+      "step": 915
+    },
+    {
+      "epoch": 0.0587028966931556,
+      "grad_norm": 2.6248859605653276,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 916
+    },
+    {
+      "epoch": 0.058766982824916686,
+      "grad_norm": 2.5772275270301703,
+      "learning_rate": 1e-06,
+      "loss": 0.4135,
+      "step": 917
+    },
+    {
+      "epoch": 0.05883106895667777,
+      "grad_norm": 2.524083865479341,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 918
+    },
+    {
+      "epoch": 0.058895155088438865,
+      "grad_norm": 2.5691457666125697,
+      "learning_rate": 1e-06,
+      "loss": 0.4609,
+      "step": 919
+    },
+    {
+      "epoch": 0.05895924122019995,
+      "grad_norm": 2.613129783593128,
+      "learning_rate": 1e-06,
+      "loss": 0.4304,
+      "step": 920
+    },
+    {
+      "epoch": 0.05902332735196104,
+      "grad_norm": 2.645453363847247,
+      "learning_rate": 1e-06,
+      "loss": 0.4506,
+      "step": 921
+    },
+    {
+      "epoch": 0.05908741348372212,
+      "grad_norm": 2.5056096531333516,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 922
+    },
+    {
+      "epoch": 0.05915149961548321,
+      "grad_norm": 2.8282030249135,
+      "learning_rate": 1e-06,
+      "loss": 0.4979,
+      "step": 923
+    },
+    {
+      "epoch": 0.059215585747244295,
+      "grad_norm": 2.545655906976193,
+      "learning_rate": 1e-06,
+      "loss": 0.4205,
+      "step": 924
+    },
+    {
+      "epoch": 0.05927967187900538,
+      "grad_norm": 2.4742283942662486,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 925
+    },
+    {
+      "epoch": 0.05934375801076647,
+      "grad_norm": 2.609486423567062,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 926
+    },
+    {
+      "epoch": 0.059407844142527554,
+      "grad_norm": 2.433383851079032,
+      "learning_rate": 1e-06,
+      "loss": 0.465,
+      "step": 927
+    },
+    {
+      "epoch": 0.05947193027428865,
+      "grad_norm": 2.4780835152841405,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 928
+    },
+    {
+      "epoch": 0.05953601640604973,
+      "grad_norm": 2.5995419876722883,
+      "learning_rate": 1e-06,
+      "loss": 0.4444,
+      "step": 929
+    },
+    {
+      "epoch": 0.05960010253781082,
+      "grad_norm": 2.802476257110543,
+      "learning_rate": 1e-06,
+      "loss": 0.4203,
+      "step": 930
+    },
+    {
+      "epoch": 0.059664188669571905,
+      "grad_norm": 2.696579952644756,
+      "learning_rate": 1e-06,
+      "loss": 0.369,
+      "step": 931
+    },
+    {
+      "epoch": 0.05972827480133299,
+      "grad_norm": 2.764140417821217,
+      "learning_rate": 1e-06,
+      "loss": 0.4162,
+      "step": 932
+    },
+    {
+      "epoch": 0.05979236093309408,
+      "grad_norm": 2.7153933671973727,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 933
+    },
+    {
+      "epoch": 0.05985644706485516,
+      "grad_norm": 2.5802340494659806,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 934
+    },
+    {
+      "epoch": 0.05992053319661625,
+      "grad_norm": 2.8596854328037984,
+      "learning_rate": 1e-06,
+      "loss": 0.4285,
+      "step": 935
+    },
+    {
+      "epoch": 0.05998461932837734,
+      "grad_norm": 2.4558109539962976,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 936
+    },
+    {
+      "epoch": 0.06004870546013843,
+      "grad_norm": 2.565179500088762,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 937
+    },
+    {
+      "epoch": 0.060112791591899514,
+      "grad_norm": 2.8418067743587625,
+      "learning_rate": 1e-06,
+      "loss": 0.3897,
+      "step": 938
+    },
+    {
+      "epoch": 0.0601768777236606,
+      "grad_norm": 2.664804134655138,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 939
+    },
+    {
+      "epoch": 0.060240963855421686,
+      "grad_norm": 2.4571750352285804,
+      "learning_rate": 1e-06,
+      "loss": 0.3878,
+      "step": 940
+    },
+    {
+      "epoch": 0.06030504998718277,
+      "grad_norm": 2.61630303904412,
+      "learning_rate": 1e-06,
+      "loss": 0.4347,
+      "step": 941
+    },
+    {
+      "epoch": 0.06036913611894386,
+      "grad_norm": 2.7027575415190896,
+      "learning_rate": 1e-06,
+      "loss": 0.3921,
+      "step": 942
+    },
+    {
+      "epoch": 0.060433222250704945,
+      "grad_norm": 3.010427817021502,
+      "learning_rate": 1e-06,
+      "loss": 0.3863,
+      "step": 943
+    },
+    {
+      "epoch": 0.06049730838246604,
+      "grad_norm": 2.7087312174774065,
+      "learning_rate": 1e-06,
+      "loss": 0.3995,
+      "step": 944
+    },
+    {
+      "epoch": 0.060561394514227124,
+      "grad_norm": 2.4384517149103226,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 945
+    },
+    {
+      "epoch": 0.06062548064598821,
+      "grad_norm": 2.3482061954090616,
+      "learning_rate": 1e-06,
+      "loss": 0.3846,
+      "step": 946
+    },
+    {
+      "epoch": 0.060689566777749296,
+      "grad_norm": 2.5136646304971033,
+      "learning_rate": 1e-06,
+      "loss": 0.4334,
+      "step": 947
+    },
+    {
+      "epoch": 0.06075365290951038,
+      "grad_norm": 2.5457038683619015,
+      "learning_rate": 1e-06,
+      "loss": 0.3798,
+      "step": 948
+    },
+    {
+      "epoch": 0.06081773904127147,
+      "grad_norm": 2.420555660582623,
+      "learning_rate": 1e-06,
+      "loss": 0.3932,
+      "step": 949
+    },
+    {
+      "epoch": 0.060881825173032554,
+      "grad_norm": 2.6412791172215524,
+      "learning_rate": 1e-06,
+      "loss": 0.4662,
+      "step": 950
+    },
+    {
+      "epoch": 0.06094591130479364,
+      "grad_norm": 2.9291322019172688,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 951
+    },
+    {
+      "epoch": 0.061009997436554726,
+      "grad_norm": 2.7145651389710177,
+      "learning_rate": 1e-06,
+      "loss": 0.4171,
+      "step": 952
+    },
+    {
+      "epoch": 0.06107408356831582,
+      "grad_norm": 2.6986578195358666,
+      "learning_rate": 1e-06,
+      "loss": 0.3915,
+      "step": 953
+    },
+    {
+      "epoch": 0.061138169700076905,
+      "grad_norm": 2.554566833482712,
+      "learning_rate": 1e-06,
+      "loss": 0.4024,
+      "step": 954
+    },
+    {
+      "epoch": 0.06120225583183799,
+      "grad_norm": 2.476875377384491,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 955
+    },
+    {
+      "epoch": 0.06126634196359908,
+      "grad_norm": 2.587063107108279,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 956
+    },
+    {
+      "epoch": 0.061330428095360164,
+      "grad_norm": 2.5604522601800404,
+      "learning_rate": 1e-06,
+      "loss": 0.4748,
+      "step": 957
+    },
+    {
+      "epoch": 0.06139451422712125,
+      "grad_norm": 2.5258207322773614,
+      "learning_rate": 1e-06,
+      "loss": 0.44,
+      "step": 958
+    },
+    {
+      "epoch": 0.061458600358882336,
+      "grad_norm": 2.6373864349591107,
+      "learning_rate": 1e-06,
+      "loss": 0.4521,
+      "step": 959
+    },
+    {
+      "epoch": 0.06152268649064342,
+      "grad_norm": 2.702180991893202,
+      "learning_rate": 1e-06,
+      "loss": 0.424,
+      "step": 960
+    },
+    {
+      "epoch": 0.061586772622404515,
+      "grad_norm": 2.4921759193243074,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 961
+    },
+    {
+      "epoch": 0.0616508587541656,
+      "grad_norm": 2.450635064392266,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 962
+    },
+    {
+      "epoch": 0.06171494488592669,
+      "grad_norm": 2.489839839860473,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 963
+    },
+    {
+      "epoch": 0.06177903101768777,
+      "grad_norm": 2.563181821436447,
+      "learning_rate": 1e-06,
+      "loss": 0.4138,
+      "step": 964
+    },
+    {
+      "epoch": 0.06184311714944886,
+      "grad_norm": 2.7120831137952637,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 965
+    },
+    {
+      "epoch": 0.061907203281209945,
+      "grad_norm": 2.623528912470007,
+      "learning_rate": 1e-06,
+      "loss": 0.4902,
+      "step": 966
+    },
+    {
+      "epoch": 0.06197128941297103,
+      "grad_norm": 2.70653288950328,
+      "learning_rate": 1e-06,
+      "loss": 0.3872,
+      "step": 967
+    },
+    {
+      "epoch": 0.06203537554473212,
+      "grad_norm": 2.589210582203363,
+      "learning_rate": 1e-06,
+      "loss": 0.4239,
+      "step": 968
+    },
+    {
+      "epoch": 0.06209946167649321,
+      "grad_norm": 2.5398221503803544,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 969
+    },
+    {
+      "epoch": 0.062163547808254296,
+      "grad_norm": 2.454628695755026,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 970
+    },
+    {
+      "epoch": 0.06222763394001538,
+      "grad_norm": 2.848700306726633,
+      "learning_rate": 1e-06,
+      "loss": 0.4546,
+      "step": 971
+    },
+    {
+      "epoch": 0.06229172007177647,
+      "grad_norm": 2.6158789298430816,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 972
+    },
+    {
+      "epoch": 0.062355806203537555,
+      "grad_norm": 2.4948748651670978,
+      "learning_rate": 1e-06,
+      "loss": 0.4011,
+      "step": 973
+    },
+    {
+      "epoch": 0.06241989233529864,
+      "grad_norm": 2.559300985421126,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 974
+    },
+    {
+      "epoch": 0.06248397846705973,
+      "grad_norm": 2.506236058687631,
+      "learning_rate": 1e-06,
+      "loss": 0.3754,
+      "step": 975
+    },
+    {
+      "epoch": 0.06254806459882081,
+      "grad_norm": 2.802408713917196,
+      "learning_rate": 1e-06,
+      "loss": 0.4177,
+      "step": 976
+    },
+    {
+      "epoch": 0.0626121507305819,
+      "grad_norm": 2.5662771461384297,
+      "learning_rate": 1e-06,
+      "loss": 0.4466,
+      "step": 977
+    },
+    {
+      "epoch": 0.06267623686234298,
+      "grad_norm": 2.4273852123129736,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 978
+    },
+    {
+      "epoch": 0.06274032299410408,
+      "grad_norm": 2.695506665685708,
+      "learning_rate": 1e-06,
+      "loss": 0.4042,
+      "step": 979
+    },
+    {
+      "epoch": 0.06280440912586516,
+      "grad_norm": 2.5089248203469117,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 980
+    },
+    {
+      "epoch": 0.06286849525762625,
+      "grad_norm": 2.610225075497294,
+      "learning_rate": 1e-06,
+      "loss": 0.425,
+      "step": 981
+    },
+    {
+      "epoch": 0.06293258138938734,
+      "grad_norm": 2.6515675578565365,
+      "learning_rate": 1e-06,
+      "loss": 0.4858,
+      "step": 982
+    },
+    {
+      "epoch": 0.06299666752114842,
+      "grad_norm": 2.662942034843004,
+      "learning_rate": 1e-06,
+      "loss": 0.4146,
+      "step": 983
+    },
+    {
+      "epoch": 0.06306075365290952,
+      "grad_norm": 2.483294863504189,
+      "learning_rate": 1e-06,
+      "loss": 0.4229,
+      "step": 984
+    },
+    {
+      "epoch": 0.0631248397846706,
+      "grad_norm": 2.774408912609533,
+      "learning_rate": 1e-06,
+      "loss": 0.4477,
+      "step": 985
+    },
+    {
+      "epoch": 0.06318892591643169,
+      "grad_norm": 2.6992735362714746,
+      "learning_rate": 1e-06,
+      "loss": 0.3877,
+      "step": 986
+    },
+    {
+      "epoch": 0.06325301204819277,
+      "grad_norm": 2.574806326857814,
+      "learning_rate": 1e-06,
+      "loss": 0.4137,
+      "step": 987
+    },
+    {
+      "epoch": 0.06331709817995386,
+      "grad_norm": 2.6815055430989316,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 988
+    },
+    {
+      "epoch": 0.06338118431171494,
+      "grad_norm": 2.519521321534601,
+      "learning_rate": 1e-06,
+      "loss": 0.389,
+      "step": 989
+    },
+    {
+      "epoch": 0.06344527044347603,
+      "grad_norm": 2.33602785029866,
+      "learning_rate": 1e-06,
+      "loss": 0.4372,
+      "step": 990
+    },
+    {
+      "epoch": 0.06350935657523712,
+      "grad_norm": 2.71707391218128,
+      "learning_rate": 1e-06,
+      "loss": 0.4465,
+      "step": 991
+    },
+    {
+      "epoch": 0.0635734427069982,
+      "grad_norm": 2.636719039925872,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 992
+    },
+    {
+      "epoch": 0.0636375288387593,
+      "grad_norm": 2.4058618312611983,
+      "learning_rate": 1e-06,
+      "loss": 0.3711,
+      "step": 993
+    },
+    {
+      "epoch": 0.06370161497052038,
+      "grad_norm": 2.6341360576740214,
+      "learning_rate": 1e-06,
+      "loss": 0.4046,
+      "step": 994
+    },
+    {
+      "epoch": 0.06376570110228147,
+      "grad_norm": 2.6734271373551226,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 995
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 2.448615128413396,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 996
+    },
+    {
+      "epoch": 0.06389387336580364,
+      "grad_norm": 2.465073956868142,
+      "learning_rate": 1e-06,
+      "loss": 0.4479,
+      "step": 997
+    },
+    {
+      "epoch": 0.06395795949756472,
+      "grad_norm": 2.740786732279467,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 998
+    },
+    {
+      "epoch": 0.06402204562932581,
+      "grad_norm": 2.9571363244427626,
+      "learning_rate": 1e-06,
+      "loss": 0.4127,
+      "step": 999
+    },
+    {
+      "epoch": 0.0640861317610869,
+      "grad_norm": 2.6266858312799912,
+      "learning_rate": 1e-06,
+      "loss": 0.4022,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06415021789284799,
+      "grad_norm": 2.566332751578703,
+      "learning_rate": 1e-06,
+      "loss": 0.4415,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06421430402460908,
+      "grad_norm": 2.8348967920960946,
+      "learning_rate": 1e-06,
+      "loss": 0.4631,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06427839015637016,
+      "grad_norm": 2.783573396035872,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06434247628813125,
+      "grad_norm": 2.573686094710799,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06440656241989233,
+      "grad_norm": 2.4367691385400554,
+      "learning_rate": 1e-06,
+      "loss": 0.4291,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06447064855165342,
+      "grad_norm": 2.5403039658130417,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06453473468341452,
+      "grad_norm": 2.6028748942628352,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0645988208151756,
+      "grad_norm": 2.4662519444773148,
+      "learning_rate": 1e-06,
+      "loss": 0.399,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06466290694693669,
+      "grad_norm": 2.51916140654751,
+      "learning_rate": 1e-06,
+      "loss": 0.4602,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06472699307869777,
+      "grad_norm": 2.5823634639004185,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06479107921045886,
+      "grad_norm": 2.778381438547188,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06485516534221994,
+      "grad_norm": 2.556163096331897,
+      "learning_rate": 1e-06,
+      "loss": 0.3915,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06491925147398103,
+      "grad_norm": 2.759435920952366,
+      "learning_rate": 1e-06,
+      "loss": 0.4893,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06498333760574211,
+      "grad_norm": 2.6059925655979033,
+      "learning_rate": 1e-06,
+      "loss": 0.4138,
+      "step": 1014
+    },
+    {
+      "epoch": 0.0650474237375032,
+      "grad_norm": 2.6958934341111997,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0651115098692643,
+      "grad_norm": 2.5129883965336983,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06517559600102538,
+      "grad_norm": 2.617160055089755,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06523968213278647,
+      "grad_norm": 2.596654082237727,
+      "learning_rate": 1e-06,
+      "loss": 0.3705,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06530376826454755,
+      "grad_norm": 2.6853485295779964,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06536785439630864,
+      "grad_norm": 2.731951426496971,
+      "learning_rate": 1e-06,
+      "loss": 0.4325,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06543194052806972,
+      "grad_norm": 2.615306745896242,
+      "learning_rate": 1e-06,
+      "loss": 0.4863,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06549602665983081,
+      "grad_norm": 2.7173205806268603,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06556011279159189,
+      "grad_norm": 2.6805449715216567,
+      "learning_rate": 1e-06,
+      "loss": 0.394,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06562419892335299,
+      "grad_norm": 2.580930131918419,
+      "learning_rate": 1e-06,
+      "loss": 0.4065,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06568828505511408,
+      "grad_norm": 2.597849794416895,
+      "learning_rate": 1e-06,
+      "loss": 0.4519,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06575237118687516,
+      "grad_norm": 2.599092164234108,
+      "learning_rate": 1e-06,
+      "loss": 0.4076,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06581645731863625,
+      "grad_norm": 2.496092273953171,
+      "learning_rate": 1e-06,
+      "loss": 0.3984,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06588054345039733,
+      "grad_norm": 2.5031105951259547,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06594462958215842,
+      "grad_norm": 2.580869432971234,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0660087157139195,
+      "grad_norm": 2.3408972954509957,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0660728018456806,
+      "grad_norm": 2.404906585724594,
+      "learning_rate": 1e-06,
+      "loss": 0.4575,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06613688797744169,
+      "grad_norm": 2.635559029994182,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06620097410920277,
+      "grad_norm": 2.7759398385370195,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06626506024096386,
+      "grad_norm": 2.7080679996277444,
+      "learning_rate": 1e-06,
+      "loss": 0.4226,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06632914637272494,
+      "grad_norm": 2.5479868488751345,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06639323250448603,
+      "grad_norm": 2.5285529800347284,
+      "learning_rate": 1e-06,
+      "loss": 0.4695,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06645731863624711,
+      "grad_norm": 2.6519770537147407,
+      "learning_rate": 1e-06,
+      "loss": 0.4483,
+      "step": 1037
+    },
+    {
+      "epoch": 0.0665214047680082,
+      "grad_norm": 2.533724568813007,
+      "learning_rate": 1e-06,
+      "loss": 0.4122,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06658549089976928,
+      "grad_norm": 2.521873472809296,
+      "learning_rate": 1e-06,
+      "loss": 0.3602,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06664957703153038,
+      "grad_norm": 2.8267483275910443,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06671366316329147,
+      "grad_norm": 2.467528084282326,
+      "learning_rate": 1e-06,
+      "loss": 0.4311,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06677774929505255,
+      "grad_norm": 2.561755755376321,
+      "learning_rate": 1e-06,
+      "loss": 0.4501,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06684183542681364,
+      "grad_norm": 2.6711046370183373,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06690592155857472,
+      "grad_norm": 2.7762462243085735,
+      "learning_rate": 1e-06,
+      "loss": 0.434,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06697000769033581,
+      "grad_norm": 2.5653129293074213,
+      "learning_rate": 1e-06,
+      "loss": 0.372,
+      "step": 1045
+    },
+    {
+      "epoch": 0.0670340938220969,
+      "grad_norm": 2.706282450560898,
+      "learning_rate": 1e-06,
+      "loss": 0.444,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06709817995385799,
+      "grad_norm": 2.622565934269703,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06716226608561907,
+      "grad_norm": 2.781376720798491,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06722635221738016,
+      "grad_norm": 2.8092295996044228,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06729043834914125,
+      "grad_norm": 2.4394636221770556,
+      "learning_rate": 1e-06,
+      "loss": 0.5048,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06735452448090233,
+      "grad_norm": 2.452372374028152,
+      "learning_rate": 1e-06,
+      "loss": 0.4492,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06741861061266342,
+      "grad_norm": 2.3619178075691307,
+      "learning_rate": 1e-06,
+      "loss": 0.3886,
+      "step": 1052
+    },
+    {
+      "epoch": 0.0674826967444245,
+      "grad_norm": 2.800387735902425,
+      "learning_rate": 1e-06,
+      "loss": 0.4693,
+      "step": 1053
+    },
+    {
+      "epoch": 0.0675467828761856,
+      "grad_norm": 2.6335374299550316,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06761086900794668,
+      "grad_norm": 2.6774069184367177,
+      "learning_rate": 1e-06,
+      "loss": 0.3891,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06767495513970777,
+      "grad_norm": 2.7446624322855886,
+      "learning_rate": 1e-06,
+      "loss": 0.4485,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06773904127146885,
+      "grad_norm": 2.5979310178492585,
+      "learning_rate": 1e-06,
+      "loss": 0.3936,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06780312740322994,
+      "grad_norm": 2.565876902003627,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06786721353499103,
+      "grad_norm": 2.670634426279264,
+      "learning_rate": 1e-06,
+      "loss": 0.4304,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06793129966675211,
+      "grad_norm": 2.6113141898048293,
+      "learning_rate": 1e-06,
+      "loss": 0.3744,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0679953857985132,
+      "grad_norm": 2.7067225739696212,
+      "learning_rate": 1e-06,
+      "loss": 0.4545,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06805947193027428,
+      "grad_norm": 2.793453856124571,
+      "learning_rate": 1e-06,
+      "loss": 0.4007,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06812355806203538,
+      "grad_norm": 2.756931561264504,
+      "learning_rate": 1e-06,
+      "loss": 0.4316,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06818764419379646,
+      "grad_norm": 2.597642965053177,
+      "learning_rate": 1e-06,
+      "loss": 0.3865,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06825173032555755,
+      "grad_norm": 2.6663999203185385,
+      "learning_rate": 1e-06,
+      "loss": 0.4348,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06831581645731864,
+      "grad_norm": 2.8609809542202487,
+      "learning_rate": 1e-06,
+      "loss": 0.3609,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06837990258907972,
+      "grad_norm": 2.652883468791596,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06844398872084081,
+      "grad_norm": 2.9494281680723646,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 1068
+    },
+    {
+      "epoch": 0.0685080748526019,
+      "grad_norm": 2.6494659758233983,
+      "learning_rate": 1e-06,
+      "loss": 0.4532,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06857216098436299,
+      "grad_norm": 2.687343190573424,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06863624711612407,
+      "grad_norm": 2.522789534611682,
+      "learning_rate": 1e-06,
+      "loss": 0.4329,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06870033324788516,
+      "grad_norm": 2.668827458426054,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06876441937964624,
+      "grad_norm": 2.5380058649439676,
+      "learning_rate": 1e-06,
+      "loss": 0.3881,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06882850551140733,
+      "grad_norm": 2.5193129973843287,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06889259164316842,
+      "grad_norm": 2.6544689506543473,
+      "learning_rate": 1e-06,
+      "loss": 0.4046,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0689566777749295,
+      "grad_norm": 2.752467894455635,
+      "learning_rate": 1e-06,
+      "loss": 0.4065,
+      "step": 1076
+    },
+    {
+      "epoch": 0.0690207639066906,
+      "grad_norm": 2.6254575219907803,
+      "learning_rate": 1e-06,
+      "loss": 0.3959,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06908485003845168,
+      "grad_norm": 2.5639116654438254,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06914893617021277,
+      "grad_norm": 2.816930112267722,
+      "learning_rate": 1e-06,
+      "loss": 0.4839,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06921302230197385,
+      "grad_norm": 2.6424585821266313,
+      "learning_rate": 1e-06,
+      "loss": 0.4152,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06927710843373494,
+      "grad_norm": 2.598199833909251,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06934119456549602,
+      "grad_norm": 2.6818415968968385,
+      "learning_rate": 1e-06,
+      "loss": 0.3833,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06940528069725711,
+      "grad_norm": 2.7587247856867463,
+      "learning_rate": 1e-06,
+      "loss": 0.4621,
+      "step": 1083
+    },
+    {
+      "epoch": 0.0694693668290182,
+      "grad_norm": 2.669134371598925,
+      "learning_rate": 1e-06,
+      "loss": 0.4269,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06953345296077929,
+      "grad_norm": 2.506076752346152,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06959753909254038,
+      "grad_norm": 2.5858512706812475,
+      "learning_rate": 1e-06,
+      "loss": 0.4303,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06966162522430146,
+      "grad_norm": 2.661423397910268,
+      "learning_rate": 1e-06,
+      "loss": 0.4027,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06972571135606255,
+      "grad_norm": 2.5027730869065947,
+      "learning_rate": 1e-06,
+      "loss": 0.369,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06978979748782363,
+      "grad_norm": 2.5945275163137262,
+      "learning_rate": 1e-06,
+      "loss": 0.4553,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06985388361958472,
+      "grad_norm": 2.847369043442018,
+      "learning_rate": 1e-06,
+      "loss": 0.4604,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06991796975134582,
+      "grad_norm": 2.513424892957305,
+      "learning_rate": 1e-06,
+      "loss": 0.3673,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0699820558831069,
+      "grad_norm": 2.7317286397718914,
+      "learning_rate": 1e-06,
+      "loss": 0.4119,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07004614201486799,
+      "grad_norm": 2.4113150816480986,
+      "learning_rate": 1e-06,
+      "loss": 0.415,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07011022814662907,
+      "grad_norm": 2.6688521108098375,
+      "learning_rate": 1e-06,
+      "loss": 0.4192,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07017431427839016,
+      "grad_norm": 2.5120122115920633,
+      "learning_rate": 1e-06,
+      "loss": 0.4205,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07023840041015124,
+      "grad_norm": 2.65132626428691,
+      "learning_rate": 1e-06,
+      "loss": 0.4336,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07030248654191233,
+      "grad_norm": 2.6400689871000056,
+      "learning_rate": 1e-06,
+      "loss": 0.4401,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07036657267367341,
+      "grad_norm": 2.467297818144308,
+      "learning_rate": 1e-06,
+      "loss": 0.4335,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0704306588054345,
+      "grad_norm": 2.478710251122422,
+      "learning_rate": 1e-06,
+      "loss": 0.4625,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0704947449371956,
+      "grad_norm": 2.506389963438959,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07055883106895668,
+      "grad_norm": 2.571801825624722,
+      "learning_rate": 1e-06,
+      "loss": 0.3919,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07062291720071777,
+      "grad_norm": 2.419710867732795,
+      "learning_rate": 1e-06,
+      "loss": 0.3828,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07068700333247885,
+      "grad_norm": 2.599190425523394,
+      "learning_rate": 1e-06,
+      "loss": 0.3828,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07075108946423994,
+      "grad_norm": 2.607254163899496,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07081517559600102,
+      "grad_norm": 2.358621192505561,
+      "learning_rate": 1e-06,
+      "loss": 0.3716,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07087926172776211,
+      "grad_norm": 2.539685863939085,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07094334785952319,
+      "grad_norm": 2.4938855723828772,
+      "learning_rate": 1e-06,
+      "loss": 0.3537,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07100743399128429,
+      "grad_norm": 2.5238087008493304,
+      "learning_rate": 1e-06,
+      "loss": 0.45,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07107152012304538,
+      "grad_norm": 2.7442192669650094,
+      "learning_rate": 1e-06,
+      "loss": 0.3776,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07113560625480646,
+      "grad_norm": 2.5839741345460627,
+      "learning_rate": 1e-06,
+      "loss": 0.5042,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07119969238656755,
+      "grad_norm": 2.6663297246069355,
+      "learning_rate": 1e-06,
+      "loss": 0.3484,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07126377851832863,
+      "grad_norm": 2.598906923733249,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07132786465008972,
+      "grad_norm": 2.6202715077828085,
+      "learning_rate": 1e-06,
+      "loss": 0.388,
+      "step": 1113
+    },
+    {
+      "epoch": 0.0713919507818508,
+      "grad_norm": 2.4825489453871974,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 1114
+    },
+    {
+      "epoch": 0.0714560369136119,
+      "grad_norm": 3.6753116517960454,
+      "learning_rate": 1e-06,
+      "loss": 0.4002,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07152012304537299,
+      "grad_norm": 2.6693428456376393,
+      "learning_rate": 1e-06,
+      "loss": 0.4982,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07158420917713407,
+      "grad_norm": 2.3868609496718136,
+      "learning_rate": 1e-06,
+      "loss": 0.462,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07164829530889516,
+      "grad_norm": 2.7746202720611635,
+      "learning_rate": 1e-06,
+      "loss": 0.3875,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07171238144065624,
+      "grad_norm": 2.4389606051674395,
+      "learning_rate": 1e-06,
+      "loss": 0.427,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07177646757241733,
+      "grad_norm": 2.743220624439705,
+      "learning_rate": 1e-06,
+      "loss": 0.4451,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07184055370417841,
+      "grad_norm": 2.430870254174703,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 1121
+    },
+    {
+      "epoch": 0.0719046398359395,
+      "grad_norm": 2.534393202382834,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07196872596770058,
+      "grad_norm": 2.7817519211332393,
+      "learning_rate": 1e-06,
+      "loss": 0.4502,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07203281209946168,
+      "grad_norm": 2.9100386240706517,
+      "learning_rate": 1e-06,
+      "loss": 0.3884,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07209689823122277,
+      "grad_norm": 2.826915075639032,
+      "learning_rate": 1e-06,
+      "loss": 0.4718,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07216098436298385,
+      "grad_norm": 2.6944257999670826,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07222507049474494,
+      "grad_norm": 2.4687223760052435,
+      "learning_rate": 1e-06,
+      "loss": 0.4122,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07228915662650602,
+      "grad_norm": 2.4178966947676646,
+      "learning_rate": 1e-06,
+      "loss": 0.3825,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07235324275826711,
+      "grad_norm": 2.752834047078145,
+      "learning_rate": 1e-06,
+      "loss": 0.4372,
+      "step": 1129
+    },
+    {
+      "epoch": 0.0724173288900282,
+      "grad_norm": 2.6488729911222206,
+      "learning_rate": 1e-06,
+      "loss": 0.42,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07248141502178929,
+      "grad_norm": 2.501959418170512,
+      "learning_rate": 1e-06,
+      "loss": 0.3996,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07254550115355037,
+      "grad_norm": 2.596725048264129,
+      "learning_rate": 1e-06,
+      "loss": 0.401,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07260958728531146,
+      "grad_norm": 2.787403116588525,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07267367341707255,
+      "grad_norm": 2.658465773541618,
+      "learning_rate": 1e-06,
+      "loss": 0.4132,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07273775954883363,
+      "grad_norm": 2.482592801042877,
+      "learning_rate": 1e-06,
+      "loss": 0.4216,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07280184568059472,
+      "grad_norm": 2.6974112554570926,
+      "learning_rate": 1e-06,
+      "loss": 0.4661,
+      "step": 1136
+    },
+    {
+      "epoch": 0.0728659318123558,
+      "grad_norm": 2.7703393520834614,
+      "learning_rate": 1e-06,
+      "loss": 0.4758,
+      "step": 1137
+    },
+    {
+      "epoch": 0.0729300179441169,
+      "grad_norm": 2.52696180165321,
+      "learning_rate": 1e-06,
+      "loss": 0.4251,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07299410407587797,
+      "grad_norm": 2.543291457166555,
+      "learning_rate": 1e-06,
+      "loss": 0.412,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07305819020763907,
+      "grad_norm": 2.518067877477988,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07312227633940016,
+      "grad_norm": 2.649865343001732,
+      "learning_rate": 1e-06,
+      "loss": 0.4216,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07318636247116124,
+      "grad_norm": 2.6317013476687454,
+      "learning_rate": 1e-06,
+      "loss": 0.4166,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07325044860292233,
+      "grad_norm": 2.5713731021703294,
+      "learning_rate": 1e-06,
+      "loss": 0.3974,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07331453473468341,
+      "grad_norm": 2.7931203073361788,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 1144
+    },
+    {
+      "epoch": 0.0733786208664445,
+      "grad_norm": 2.599003814107677,
+      "learning_rate": 1e-06,
+      "loss": 0.4507,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07344270699820558,
+      "grad_norm": 2.7684915380392354,
+      "learning_rate": 1e-06,
+      "loss": 0.4471,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07350679312996668,
+      "grad_norm": 2.4796130862109735,
+      "learning_rate": 1e-06,
+      "loss": 0.4646,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07357087926172776,
+      "grad_norm": 2.691420190189633,
+      "learning_rate": 1e-06,
+      "loss": 0.4762,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07363496539348885,
+      "grad_norm": 2.7664381558741065,
+      "learning_rate": 1e-06,
+      "loss": 0.448,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07369905152524994,
+      "grad_norm": 2.5162422680138814,
+      "learning_rate": 1e-06,
+      "loss": 0.4012,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07376313765701102,
+      "grad_norm": 2.5781952395234633,
+      "learning_rate": 1e-06,
+      "loss": 0.3675,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07382722378877211,
+      "grad_norm": 2.6625898353464157,
+      "learning_rate": 1e-06,
+      "loss": 0.379,
+      "step": 1152
+    },
+    {
+      "epoch": 0.0738913099205332,
+      "grad_norm": 2.5901429074988553,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07395539605229429,
+      "grad_norm": 2.7942392683449846,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07401948218405537,
+      "grad_norm": 2.6654460748878286,
+      "learning_rate": 1e-06,
+      "loss": 0.409,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07408356831581646,
+      "grad_norm": 2.5476306648846294,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07414765444757754,
+      "grad_norm": 2.591611279936759,
+      "learning_rate": 1e-06,
+      "loss": 0.4648,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07421174057933863,
+      "grad_norm": 2.6704534006186087,
+      "learning_rate": 1e-06,
+      "loss": 0.4079,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07427582671109972,
+      "grad_norm": 2.671768868673718,
+      "learning_rate": 1e-06,
+      "loss": 0.4416,
+      "step": 1159
+    },
+    {
+      "epoch": 0.0743399128428608,
+      "grad_norm": 2.4873744200996404,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0744039989746219,
+      "grad_norm": 2.4953535418400232,
+      "learning_rate": 1e-06,
+      "loss": 0.4331,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07446808510638298,
+      "grad_norm": 2.526056293976957,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07453217123814407,
+      "grad_norm": 2.588818835559356,
+      "learning_rate": 1e-06,
+      "loss": 0.4776,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07459625736990515,
+      "grad_norm": 2.666576479172023,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07466034350166624,
+      "grad_norm": 2.961651092072556,
+      "learning_rate": 1e-06,
+      "loss": 0.5216,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07472442963342732,
+      "grad_norm": 2.520589368360706,
+      "learning_rate": 1e-06,
+      "loss": 0.4646,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07478851576518841,
+      "grad_norm": 2.5775375985686435,
+      "learning_rate": 1e-06,
+      "loss": 0.4166,
+      "step": 1167
+    },
+    {
+      "epoch": 0.0748526018969495,
+      "grad_norm": 2.618379504403705,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07491668802871058,
+      "grad_norm": 2.701174665426732,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07498077416047168,
+      "grad_norm": 2.879711459534598,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07504486029223276,
+      "grad_norm": 2.437675394493719,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07510894642399385,
+      "grad_norm": 2.378217037869276,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07517303255575493,
+      "grad_norm": 2.6864533867239104,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07523711868751602,
+      "grad_norm": 2.9070178703797755,
+      "learning_rate": 1e-06,
+      "loss": 0.4432,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07530120481927711,
+      "grad_norm": 2.394244267169891,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0753652909510382,
+      "grad_norm": 2.613237904551317,
+      "learning_rate": 1e-06,
+      "loss": 0.3646,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07542937708279929,
+      "grad_norm": 2.5704274998773995,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07549346321456037,
+      "grad_norm": 2.647086455310922,
+      "learning_rate": 1e-06,
+      "loss": 0.4598,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07555754934632146,
+      "grad_norm": 2.4851338144046866,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07562163547808254,
+      "grad_norm": 2.6992112402066373,
+      "learning_rate": 1e-06,
+      "loss": 0.4165,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07568572160984363,
+      "grad_norm": 2.443187282946552,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07574980774160471,
+      "grad_norm": 2.5382215760682336,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 1182
+    },
+    {
+      "epoch": 0.0758138938733658,
+      "grad_norm": 2.6296100830741005,
+      "learning_rate": 1e-06,
+      "loss": 0.427,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0758779800051269,
+      "grad_norm": 2.437492691585318,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07594206613688798,
+      "grad_norm": 2.5072485842248082,
+      "learning_rate": 1e-06,
+      "loss": 0.3903,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07600615226864907,
+      "grad_norm": 2.6751785189294455,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07607023840041015,
+      "grad_norm": 2.603241674792852,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07613432453217124,
+      "grad_norm": 2.4459489689703724,
+      "learning_rate": 1e-06,
+      "loss": 0.4164,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07619841066393232,
+      "grad_norm": 2.936377164469067,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07626249679569341,
+      "grad_norm": 2.581801369589373,
+      "learning_rate": 1e-06,
+      "loss": 0.4646,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07632658292745449,
+      "grad_norm": 2.6802303045344513,
+      "learning_rate": 1e-06,
+      "loss": 0.3503,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07639066905921559,
+      "grad_norm": 2.6291026196838994,
+      "learning_rate": 1e-06,
+      "loss": 0.3919,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07645475519097668,
+      "grad_norm": 2.5858158792263684,
+      "learning_rate": 1e-06,
+      "loss": 0.4262,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07651884132273776,
+      "grad_norm": 2.658961870365221,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07658292745449885,
+      "grad_norm": 2.6170446639780445,
+      "learning_rate": 1e-06,
+      "loss": 0.3768,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07664701358625993,
+      "grad_norm": 2.449050801281128,
+      "learning_rate": 1e-06,
+      "loss": 0.4166,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07671109971802102,
+      "grad_norm": 2.510377162453613,
+      "learning_rate": 1e-06,
+      "loss": 0.4748,
+      "step": 1197
+    },
+    {
+      "epoch": 0.0767751858497821,
+      "grad_norm": 2.408176741926398,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0768392719815432,
+      "grad_norm": 2.510120365800847,
+      "learning_rate": 1e-06,
+      "loss": 0.4427,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07690335811330429,
+      "grad_norm": 2.570673514283029,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07696744424506537,
+      "grad_norm": 2.560138302433304,
+      "learning_rate": 1e-06,
+      "loss": 0.4463,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07703153037682646,
+      "grad_norm": 2.6974205413109953,
+      "learning_rate": 1e-06,
+      "loss": 0.4467,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07709561650858754,
+      "grad_norm": 2.6092456235211183,
+      "learning_rate": 1e-06,
+      "loss": 0.3956,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07715970264034863,
+      "grad_norm": 2.64955554722162,
+      "learning_rate": 1e-06,
+      "loss": 0.4856,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07722378877210971,
+      "grad_norm": 2.8128688811623337,
+      "learning_rate": 1e-06,
+      "loss": 0.4009,
+      "step": 1205
+    },
+    {
+      "epoch": 0.0772878749038708,
+      "grad_norm": 2.5192936294423074,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07735196103563188,
+      "grad_norm": 2.68278659957716,
+      "learning_rate": 1e-06,
+      "loss": 0.4704,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07741604716739298,
+      "grad_norm": 2.6168633523606064,
+      "learning_rate": 1e-06,
+      "loss": 0.3387,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07748013329915407,
+      "grad_norm": 2.6114164138499887,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07754421943091515,
+      "grad_norm": 2.6161792038015053,
+      "learning_rate": 1e-06,
+      "loss": 0.406,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07760830556267624,
+      "grad_norm": 2.7520314159563033,
+      "learning_rate": 1e-06,
+      "loss": 0.3833,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07767239169443732,
+      "grad_norm": 2.774588610799305,
+      "learning_rate": 1e-06,
+      "loss": 0.5036,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07773647782619841,
+      "grad_norm": 2.5033422537991994,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07780056395795949,
+      "grad_norm": 2.73481240056627,
+      "learning_rate": 1e-06,
+      "loss": 0.4345,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07786465008972059,
+      "grad_norm": 2.6851998419349026,
+      "learning_rate": 1e-06,
+      "loss": 0.4647,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07792873622148166,
+      "grad_norm": 2.7237868971195596,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07799282235324276,
+      "grad_norm": 2.541489190154758,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07805690848500385,
+      "grad_norm": 2.464482316323183,
+      "learning_rate": 1e-06,
+      "loss": 0.4373,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07812099461676493,
+      "grad_norm": 2.572853991725539,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07818508074852602,
+      "grad_norm": 2.6668459095154216,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0782491668802871,
+      "grad_norm": 2.782296312797821,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 1221
+    },
+    {
+      "epoch": 0.0783132530120482,
+      "grad_norm": 2.6235991335311937,
+      "learning_rate": 1e-06,
+      "loss": 0.4165,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07837733914380927,
+      "grad_norm": 2.7714978455242028,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07844142527557037,
+      "grad_norm": 2.3752856980488266,
+      "learning_rate": 1e-06,
+      "loss": 0.3848,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07850551140733146,
+      "grad_norm": 2.7992009765527515,
+      "learning_rate": 1e-06,
+      "loss": 0.3977,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07856959753909254,
+      "grad_norm": 2.633436222450441,
+      "learning_rate": 1e-06,
+      "loss": 0.4539,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07863368367085363,
+      "grad_norm": 2.577156182418178,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07869776980261471,
+      "grad_norm": 2.720243320628882,
+      "learning_rate": 1e-06,
+      "loss": 0.4416,
+      "step": 1228
+    },
+    {
+      "epoch": 0.0787618559343758,
+      "grad_norm": 2.7155698373395354,
+      "learning_rate": 1e-06,
+      "loss": 0.4508,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07882594206613688,
+      "grad_norm": 2.343026422126699,
+      "learning_rate": 1e-06,
+      "loss": 0.4059,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07889002819789798,
+      "grad_norm": 2.5559831262150574,
+      "learning_rate": 1e-06,
+      "loss": 0.4686,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07895411432965906,
+      "grad_norm": 2.493013880324132,
+      "learning_rate": 1e-06,
+      "loss": 0.4157,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07901820046142015,
+      "grad_norm": 2.5095068573703463,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07908228659318124,
+      "grad_norm": 2.6581528818814975,
+      "learning_rate": 1e-06,
+      "loss": 0.4285,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07914637272494232,
+      "grad_norm": 2.7654271167249793,
+      "learning_rate": 1e-06,
+      "loss": 0.47,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07921045885670341,
+      "grad_norm": 2.6893412954625853,
+      "learning_rate": 1e-06,
+      "loss": 0.4038,
+      "step": 1236
+    },
+    {
+      "epoch": 0.0792745449884645,
+      "grad_norm": 2.456340889110768,
+      "learning_rate": 1e-06,
+      "loss": 0.4103,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07933863112022559,
+      "grad_norm": 2.556312174382247,
+      "learning_rate": 1e-06,
+      "loss": 0.4516,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07940271725198667,
+      "grad_norm": 2.6753669316919417,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07946680338374776,
+      "grad_norm": 2.6035452011928903,
+      "learning_rate": 1e-06,
+      "loss": 0.3405,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07953088951550884,
+      "grad_norm": 2.4434288383339946,
+      "learning_rate": 1e-06,
+      "loss": 0.4881,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07959497564726993,
+      "grad_norm": 2.662120824414917,
+      "learning_rate": 1e-06,
+      "loss": 0.382,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07965906177903102,
+      "grad_norm": 2.6177518200932,
+      "learning_rate": 1e-06,
+      "loss": 0.4306,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0797231479107921,
+      "grad_norm": 2.4114846966775767,
+      "learning_rate": 1e-06,
+      "loss": 0.3994,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0797872340425532,
+      "grad_norm": 2.6835316926141863,
+      "learning_rate": 1e-06,
+      "loss": 0.4706,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07985132017431427,
+      "grad_norm": 2.5651819030486496,
+      "learning_rate": 1e-06,
+      "loss": 0.4262,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07991540630607537,
+      "grad_norm": 2.3934181889049126,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07997949243783645,
+      "grad_norm": 2.643579877871764,
+      "learning_rate": 1e-06,
+      "loss": 0.4524,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08004357856959754,
+      "grad_norm": 2.5313599900999337,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08010766470135863,
+      "grad_norm": 2.619558211523681,
+      "learning_rate": 1e-06,
+      "loss": 0.3862,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08017175083311971,
+      "grad_norm": 2.5146257629070097,
+      "learning_rate": 1e-06,
+      "loss": 0.4005,
+      "step": 1251
+    },
+    {
+      "epoch": 0.0802358369648808,
+      "grad_norm": 2.591349223638374,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08029992309664188,
+      "grad_norm": 2.59587465957244,
+      "learning_rate": 1e-06,
+      "loss": 0.437,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08036400922840298,
+      "grad_norm": 2.9832966117856667,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08042809536016406,
+      "grad_norm": 2.6040546777081657,
+      "learning_rate": 1e-06,
+      "loss": 0.3766,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08049218149192515,
+      "grad_norm": 2.496979332905621,
+      "learning_rate": 1e-06,
+      "loss": 0.4219,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08055626762368623,
+      "grad_norm": 2.629851175282855,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08062035375544732,
+      "grad_norm": 2.5674825630257896,
+      "learning_rate": 1e-06,
+      "loss": 0.4325,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08068443988720841,
+      "grad_norm": 2.940799171634024,
+      "learning_rate": 1e-06,
+      "loss": 0.468,
+      "step": 1259
+    },
+    {
+      "epoch": 0.0807485260189695,
+      "grad_norm": 2.650948845848417,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08081261215073059,
+      "grad_norm": 2.679500087359039,
+      "learning_rate": 1e-06,
+      "loss": 0.4538,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08087669828249167,
+      "grad_norm": 2.390614137526815,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08094078441425276,
+      "grad_norm": 2.6276093482133867,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08100487054601384,
+      "grad_norm": 2.61781590271726,
+      "learning_rate": 1e-06,
+      "loss": 0.4178,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08106895667777493,
+      "grad_norm": 2.779503761768041,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08113304280953601,
+      "grad_norm": 2.561366232789244,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 1266
+    },
+    {
+      "epoch": 0.0811971289412971,
+      "grad_norm": 2.8717807574636174,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 1267
+    },
+    {
+      "epoch": 0.0812612150730582,
+      "grad_norm": 2.7921707046886977,
+      "learning_rate": 1e-06,
+      "loss": 0.4527,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08132530120481928,
+      "grad_norm": 2.5928313310418565,
+      "learning_rate": 1e-06,
+      "loss": 0.4103,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08138938733658037,
+      "grad_norm": 2.9418199515370085,
+      "learning_rate": 1e-06,
+      "loss": 0.4114,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08145347346834145,
+      "grad_norm": 2.474771537185245,
+      "learning_rate": 1e-06,
+      "loss": 0.4398,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08151755960010254,
+      "grad_norm": 2.4288233134888104,
+      "learning_rate": 1e-06,
+      "loss": 0.4071,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08158164573186362,
+      "grad_norm": 2.728783148369517,
+      "learning_rate": 1e-06,
+      "loss": 0.3944,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08164573186362471,
+      "grad_norm": 2.595463299855329,
+      "learning_rate": 1e-06,
+      "loss": 0.3961,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08170981799538579,
+      "grad_norm": 2.4186223539565472,
+      "learning_rate": 1e-06,
+      "loss": 0.3815,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08177390412714688,
+      "grad_norm": 2.5691674213334412,
+      "learning_rate": 1e-06,
+      "loss": 0.4347,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08183799025890798,
+      "grad_norm": 2.772988911794295,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08190207639066906,
+      "grad_norm": 2.5770033358021176,
+      "learning_rate": 1e-06,
+      "loss": 0.4306,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08196616252243015,
+      "grad_norm": 2.7479896736468534,
+      "learning_rate": 1e-06,
+      "loss": 0.4102,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08203024865419123,
+      "grad_norm": 2.463361879732375,
+      "learning_rate": 1e-06,
+      "loss": 0.4234,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08209433478595232,
+      "grad_norm": 2.5773689957088974,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 1281
+    },
+    {
+      "epoch": 0.0821584209177134,
+      "grad_norm": 2.7031134347973205,
+      "learning_rate": 1e-06,
+      "loss": 0.3718,
+      "step": 1282
+    },
+    {
+      "epoch": 0.0822225070494745,
+      "grad_norm": 2.3685374130275436,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08228659318123559,
+      "grad_norm": 2.5898467699040904,
+      "learning_rate": 1e-06,
+      "loss": 0.5035,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08235067931299667,
+      "grad_norm": 2.566905337207484,
+      "learning_rate": 1e-06,
+      "loss": 0.4253,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08241476544475776,
+      "grad_norm": 2.7563322922109172,
+      "learning_rate": 1e-06,
+      "loss": 0.3726,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08247885157651884,
+      "grad_norm": 2.491496676280438,
+      "learning_rate": 1e-06,
+      "loss": 0.37,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08254293770827993,
+      "grad_norm": 2.6730510178537608,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08260702384004101,
+      "grad_norm": 2.607473514155551,
+      "learning_rate": 1e-06,
+      "loss": 0.4547,
+      "step": 1289
+    },
+    {
+      "epoch": 0.0826711099718021,
+      "grad_norm": 2.664953352790864,
+      "learning_rate": 1e-06,
+      "loss": 0.4457,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08273519610356318,
+      "grad_norm": 2.9065089288452675,
+      "learning_rate": 1e-06,
+      "loss": 0.4133,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08279928223532428,
+      "grad_norm": 2.761268347894189,
+      "learning_rate": 1e-06,
+      "loss": 0.4013,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08286336836708537,
+      "grad_norm": 2.53874487571015,
+      "learning_rate": 1e-06,
+      "loss": 0.4745,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08292745449884645,
+      "grad_norm": 2.5396881198425603,
+      "learning_rate": 1e-06,
+      "loss": 0.4437,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08299154063060754,
+      "grad_norm": 2.547768008887249,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08305562676236862,
+      "grad_norm": 2.499185905675141,
+      "learning_rate": 1e-06,
+      "loss": 0.4153,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08311971289412971,
+      "grad_norm": 2.4836012038697457,
+      "learning_rate": 1e-06,
+      "loss": 0.4259,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08318379902589079,
+      "grad_norm": 2.621155242220239,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08324788515765189,
+      "grad_norm": 2.592565931013178,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08331197128941296,
+      "grad_norm": 2.5498345680795924,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08337605742117406,
+      "grad_norm": 2.4929209233901295,
+      "learning_rate": 1e-06,
+      "loss": 0.4552,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08344014355293515,
+      "grad_norm": 2.6882977211963905,
+      "learning_rate": 1e-06,
+      "loss": 0.4062,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08350422968469623,
+      "grad_norm": 2.473705037392068,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08356831581645732,
+      "grad_norm": 2.5317476990261465,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 1304
+    },
+    {
+      "epoch": 0.0836324019482184,
+      "grad_norm": 2.5664114423982074,
+      "learning_rate": 1e-06,
+      "loss": 0.4016,
+      "step": 1305
+    },
+    {
+      "epoch": 0.0836964880799795,
+      "grad_norm": 2.5912791482520943,
+      "learning_rate": 1e-06,
+      "loss": 0.392,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08376057421174057,
+      "grad_norm": 2.9901280232724976,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08382466034350167,
+      "grad_norm": 2.5967756023713995,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08388874647526276,
+      "grad_norm": 2.73521407336545,
+      "learning_rate": 1e-06,
+      "loss": 0.4241,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08395283260702384,
+      "grad_norm": 2.919428684494416,
+      "learning_rate": 1e-06,
+      "loss": 0.4205,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08401691873878493,
+      "grad_norm": 2.7758438993320484,
+      "learning_rate": 1e-06,
+      "loss": 0.4519,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08408100487054601,
+      "grad_norm": 2.715976314244008,
+      "learning_rate": 1e-06,
+      "loss": 0.3993,
+      "step": 1312
+    },
+    {
+      "epoch": 0.0841450910023071,
+      "grad_norm": 2.5924326736887213,
+      "learning_rate": 1e-06,
+      "loss": 0.4025,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08420917713406818,
+      "grad_norm": 2.7752283664217017,
+      "learning_rate": 1e-06,
+      "loss": 0.3956,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08427326326582928,
+      "grad_norm": 2.4160129671872337,
+      "learning_rate": 1e-06,
+      "loss": 0.4106,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08433734939759036,
+      "grad_norm": 2.5407463488739053,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08440143552935145,
+      "grad_norm": 2.801380710310921,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08446552166111254,
+      "grad_norm": 2.6269565817549925,
+      "learning_rate": 1e-06,
+      "loss": 0.4203,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08452960779287362,
+      "grad_norm": 2.5243065786237944,
+      "learning_rate": 1e-06,
+      "loss": 0.373,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08459369392463471,
+      "grad_norm": 2.8116644364078986,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08465778005639579,
+      "grad_norm": 2.575799701620765,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08472186618815689,
+      "grad_norm": 2.75865288811887,
+      "learning_rate": 1e-06,
+      "loss": 0.4687,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08478595231991796,
+      "grad_norm": 2.7042258614953467,
+      "learning_rate": 1e-06,
+      "loss": 0.4051,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08485003845167906,
+      "grad_norm": 2.6044252860407213,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08491412458344014,
+      "grad_norm": 2.655140452465128,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08497821071520123,
+      "grad_norm": 2.6694681152353072,
+      "learning_rate": 1e-06,
+      "loss": 0.4693,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08504229684696232,
+      "grad_norm": 2.6321438736595364,
+      "learning_rate": 1e-06,
+      "loss": 0.447,
+      "step": 1327
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 2.4939365988778572,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 1328
+    },
+    {
+      "epoch": 0.0851704691104845,
+      "grad_norm": 2.7686831102984186,
+      "learning_rate": 1e-06,
+      "loss": 0.4096,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08523455524224557,
+      "grad_norm": 2.4734804109635533,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08529864137400667,
+      "grad_norm": 2.3310389802867224,
+      "learning_rate": 1e-06,
+      "loss": 0.4499,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08536272750576775,
+      "grad_norm": 2.5427992914352733,
+      "learning_rate": 1e-06,
+      "loss": 0.5163,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08542681363752884,
+      "grad_norm": 2.550742214022169,
+      "learning_rate": 1e-06,
+      "loss": 0.3351,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08549089976928993,
+      "grad_norm": 2.3491321424112823,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08555498590105101,
+      "grad_norm": 7.183934088342243,
+      "learning_rate": 1e-06,
+      "loss": 0.4585,
+      "step": 1335
+    },
+    {
+      "epoch": 0.0856190720328121,
+      "grad_norm": 2.727510164110277,
+      "learning_rate": 1e-06,
+      "loss": 0.403,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08568315816457318,
+      "grad_norm": 2.463122951016014,
+      "learning_rate": 1e-06,
+      "loss": 0.3738,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08574724429633428,
+      "grad_norm": 2.4246794004524856,
+      "learning_rate": 1e-06,
+      "loss": 0.4131,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08581133042809536,
+      "grad_norm": 2.505910282661219,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08587541655985645,
+      "grad_norm": 2.7449797274573906,
+      "learning_rate": 1e-06,
+      "loss": 0.4332,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08593950269161753,
+      "grad_norm": 2.602286871420609,
+      "learning_rate": 1e-06,
+      "loss": 0.4568,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08600358882337862,
+      "grad_norm": 2.4878615903294303,
+      "learning_rate": 1e-06,
+      "loss": 0.3549,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08606767495513971,
+      "grad_norm": 2.377510719482693,
+      "learning_rate": 1e-06,
+      "loss": 0.4109,
+      "step": 1343
+    },
+    {
+      "epoch": 0.0861317610869008,
+      "grad_norm": 2.5420518872600377,
+      "learning_rate": 1e-06,
+      "loss": 0.402,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08619584721866189,
+      "grad_norm": 2.619739587147752,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08625993335042297,
+      "grad_norm": 2.5009317514537237,
+      "learning_rate": 1e-06,
+      "loss": 0.433,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08632401948218406,
+      "grad_norm": 3.129107498488318,
+      "learning_rate": 1e-06,
+      "loss": 0.3725,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08638810561394514,
+      "grad_norm": 2.475758470143281,
+      "learning_rate": 1e-06,
+      "loss": 0.45,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08645219174570623,
+      "grad_norm": 2.181341026400154,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08651627787746731,
+      "grad_norm": 2.4954104610238024,
+      "learning_rate": 1e-06,
+      "loss": 0.4054,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0865803640092284,
+      "grad_norm": 2.6628673858751255,
+      "learning_rate": 1e-06,
+      "loss": 0.4717,
+      "step": 1351
+    },
+    {
+      "epoch": 0.0866444501409895,
+      "grad_norm": 2.6090665119674967,
+      "learning_rate": 1e-06,
+      "loss": 0.4544,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08670853627275057,
+      "grad_norm": 2.6018719333061986,
+      "learning_rate": 1e-06,
+      "loss": 0.4399,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08677262240451167,
+      "grad_norm": 2.4434199739785827,
+      "learning_rate": 1e-06,
+      "loss": 0.415,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08683670853627275,
+      "grad_norm": 2.769828363674665,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08690079466803384,
+      "grad_norm": 2.7397769472949247,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08696488079979492,
+      "grad_norm": 2.6172378352859726,
+      "learning_rate": 1e-06,
+      "loss": 0.4165,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08702896693155601,
+      "grad_norm": 2.539813707503197,
+      "learning_rate": 1e-06,
+      "loss": 0.4665,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08709305306331709,
+      "grad_norm": 2.6327431750979695,
+      "learning_rate": 1e-06,
+      "loss": 0.4685,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08715713919507818,
+      "grad_norm": 2.7709842385532455,
+      "learning_rate": 1e-06,
+      "loss": 0.495,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08722122532683928,
+      "grad_norm": 2.5015074677105753,
+      "learning_rate": 1e-06,
+      "loss": 0.4131,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08728531145860036,
+      "grad_norm": 2.5041650372229305,
+      "learning_rate": 1e-06,
+      "loss": 0.4544,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08734939759036145,
+      "grad_norm": 2.71401356392254,
+      "learning_rate": 1e-06,
+      "loss": 0.4458,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08741348372212253,
+      "grad_norm": 2.531157702107666,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08747756985388362,
+      "grad_norm": 2.626679235090447,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 1365
+    },
+    {
+      "epoch": 0.0875416559856447,
+      "grad_norm": 2.5951080452068704,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 1366
+    },
+    {
+      "epoch": 0.0876057421174058,
+      "grad_norm": 2.646236280137177,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08766982824916689,
+      "grad_norm": 2.565685965978931,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08773391438092797,
+      "grad_norm": 2.566570629622735,
+      "learning_rate": 1e-06,
+      "loss": 0.406,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08779800051268906,
+      "grad_norm": 2.558331006355462,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08786208664445014,
+      "grad_norm": 2.812026841029743,
+      "learning_rate": 1e-06,
+      "loss": 0.4102,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08792617277621123,
+      "grad_norm": 2.593914945874754,
+      "learning_rate": 1e-06,
+      "loss": 0.4115,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08799025890797231,
+      "grad_norm": 2.673191899027595,
+      "learning_rate": 1e-06,
+      "loss": 0.4581,
+      "step": 1373
+    },
+    {
+      "epoch": 0.0880543450397334,
+      "grad_norm": 2.6943890741457452,
+      "learning_rate": 1e-06,
+      "loss": 0.3679,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08811843117149448,
+      "grad_norm": 2.7105087759870248,
+      "learning_rate": 1e-06,
+      "loss": 0.3861,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08818251730325558,
+      "grad_norm": 2.6588773058390567,
+      "learning_rate": 1e-06,
+      "loss": 0.4262,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08824660343501667,
+      "grad_norm": 2.624668121030694,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08831068956677775,
+      "grad_norm": 2.567696520761241,
+      "learning_rate": 1e-06,
+      "loss": 0.3944,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08837477569853884,
+      "grad_norm": 2.525791516217746,
+      "learning_rate": 1e-06,
+      "loss": 0.3737,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08843886183029992,
+      "grad_norm": 2.5589736672483916,
+      "learning_rate": 1e-06,
+      "loss": 0.376,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08850294796206101,
+      "grad_norm": 2.8070743006208843,
+      "learning_rate": 1e-06,
+      "loss": 0.456,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08856703409382209,
+      "grad_norm": 2.5993874207613006,
+      "learning_rate": 1e-06,
+      "loss": 0.4197,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08863112022558318,
+      "grad_norm": 2.5051066249655825,
+      "learning_rate": 1e-06,
+      "loss": 0.385,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08869520635734426,
+      "grad_norm": 2.405748485125088,
+      "learning_rate": 1e-06,
+      "loss": 0.4049,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08875929248910536,
+      "grad_norm": 2.6145829424746965,
+      "learning_rate": 1e-06,
+      "loss": 0.4247,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08882337862086645,
+      "grad_norm": 2.6173900351223964,
+      "learning_rate": 1e-06,
+      "loss": 0.4221,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08888746475262753,
+      "grad_norm": 2.6806113776153246,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08895155088438862,
+      "grad_norm": 2.796773269226567,
+      "learning_rate": 1e-06,
+      "loss": 0.4554,
+      "step": 1388
+    },
+    {
+      "epoch": 0.0890156370161497,
+      "grad_norm": 2.570155349023824,
+      "learning_rate": 1e-06,
+      "loss": 0.4609,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0890797231479108,
+      "grad_norm": 2.651717663982742,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08914380927967187,
+      "grad_norm": 2.9890601911555574,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08920789541143297,
+      "grad_norm": 2.816018482470761,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08927198154319406,
+      "grad_norm": 2.806504598338548,
+      "learning_rate": 1e-06,
+      "loss": 0.4205,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08933606767495514,
+      "grad_norm": 2.773138983571822,
+      "learning_rate": 1e-06,
+      "loss": 0.439,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08940015380671623,
+      "grad_norm": 2.685056882343485,
+      "learning_rate": 1e-06,
+      "loss": 0.4444,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08946423993847731,
+      "grad_norm": 2.9257692535431485,
+      "learning_rate": 1e-06,
+      "loss": 0.4641,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0895283260702384,
+      "grad_norm": 2.4710812395258905,
+      "learning_rate": 1e-06,
+      "loss": 0.4463,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08959241220199948,
+      "grad_norm": 2.70288826605684,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08965649833376058,
+      "grad_norm": 2.4331622087280045,
+      "learning_rate": 1e-06,
+      "loss": 0.452,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08972058446552166,
+      "grad_norm": 2.324146163478173,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08978467059728275,
+      "grad_norm": 2.6508507688510754,
+      "learning_rate": 1e-06,
+      "loss": 0.4437,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08984875672904384,
+      "grad_norm": 2.6863035227529344,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08991284286080492,
+      "grad_norm": 2.5644169132895493,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08997692899256601,
+      "grad_norm": 2.4924959261581043,
+      "learning_rate": 1e-06,
+      "loss": 0.4128,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09004101512432709,
+      "grad_norm": 2.680924691244528,
+      "learning_rate": 1e-06,
+      "loss": 0.4216,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09010510125608819,
+      "grad_norm": 2.5648825557723898,
+      "learning_rate": 1e-06,
+      "loss": 0.3999,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09016918738784926,
+      "grad_norm": 2.6750967093694578,
+      "learning_rate": 1e-06,
+      "loss": 0.3778,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09023327351961036,
+      "grad_norm": 2.6244024961560157,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09029735965137144,
+      "grad_norm": 2.632057969960539,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09036144578313253,
+      "grad_norm": 2.6751906186648946,
+      "learning_rate": 1e-06,
+      "loss": 0.413,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09042553191489362,
+      "grad_norm": 2.561674070651734,
+      "learning_rate": 1e-06,
+      "loss": 0.4063,
+      "step": 1411
+    },
+    {
+      "epoch": 0.0904896180466547,
+      "grad_norm": 2.6877044774483827,
+      "learning_rate": 1e-06,
+      "loss": 0.4689,
+      "step": 1412
+    },
+    {
+      "epoch": 0.0905537041784158,
+      "grad_norm": 2.5896953889358607,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09061779031017687,
+      "grad_norm": 2.421302707798909,
+      "learning_rate": 1e-06,
+      "loss": 0.3714,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09068187644193797,
+      "grad_norm": 2.7539329450540277,
+      "learning_rate": 1e-06,
+      "loss": 0.3832,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09074596257369905,
+      "grad_norm": 2.6356933313255277,
+      "learning_rate": 1e-06,
+      "loss": 0.3381,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09081004870546014,
+      "grad_norm": 2.36289210209112,
+      "learning_rate": 1e-06,
+      "loss": 0.4371,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09087413483722123,
+      "grad_norm": 2.6024801679680905,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09093822096898231,
+      "grad_norm": 2.5452100945357117,
+      "learning_rate": 1e-06,
+      "loss": 0.4905,
+      "step": 1419
+    },
+    {
+      "epoch": 0.0910023071007434,
+      "grad_norm": 2.471533347199074,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09106639323250448,
+      "grad_norm": 2.684599051779771,
+      "learning_rate": 1e-06,
+      "loss": 0.4455,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09113047936426558,
+      "grad_norm": 2.6706289104425345,
+      "learning_rate": 1e-06,
+      "loss": 0.4298,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09119456549602666,
+      "grad_norm": 2.7372861347057706,
+      "learning_rate": 1e-06,
+      "loss": 0.4044,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09125865162778775,
+      "grad_norm": 2.6311831051474974,
+      "learning_rate": 1e-06,
+      "loss": 0.4472,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09132273775954883,
+      "grad_norm": 2.505239288524563,
+      "learning_rate": 1e-06,
+      "loss": 0.4608,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09138682389130992,
+      "grad_norm": 2.455646364875542,
+      "learning_rate": 1e-06,
+      "loss": 0.4488,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09145091002307101,
+      "grad_norm": 2.732934220116128,
+      "learning_rate": 1e-06,
+      "loss": 0.4206,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09151499615483209,
+      "grad_norm": 2.74323981935251,
+      "learning_rate": 1e-06,
+      "loss": 0.4234,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09157908228659319,
+      "grad_norm": 2.7158136731219646,
+      "learning_rate": 1e-06,
+      "loss": 0.4209,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09164316841835426,
+      "grad_norm": 2.6247631977279595,
+      "learning_rate": 1e-06,
+      "loss": 0.4421,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09170725455011536,
+      "grad_norm": 2.637304136296822,
+      "learning_rate": 1e-06,
+      "loss": 0.484,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09177134068187644,
+      "grad_norm": 2.852998156294522,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09183542681363753,
+      "grad_norm": 2.781417431625871,
+      "learning_rate": 1e-06,
+      "loss": 0.4231,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09189951294539861,
+      "grad_norm": 2.628138673747872,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 1434
+    },
+    {
+      "epoch": 0.0919635990771597,
+      "grad_norm": 2.667218051474725,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 1435
+    },
+    {
+      "epoch": 0.0920276852089208,
+      "grad_norm": 2.786734610663218,
+      "learning_rate": 1e-06,
+      "loss": 0.4462,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09209177134068187,
+      "grad_norm": 2.5652996360198013,
+      "learning_rate": 1e-06,
+      "loss": 0.4285,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09215585747244297,
+      "grad_norm": 2.614618685141593,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09221994360420405,
+      "grad_norm": 2.60209536356399,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09228402973596514,
+      "grad_norm": 2.669408041556916,
+      "learning_rate": 1e-06,
+      "loss": 0.4278,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09234811586772622,
+      "grad_norm": 2.5842411481731635,
+      "learning_rate": 1e-06,
+      "loss": 0.3888,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09241220199948731,
+      "grad_norm": 2.4305355503889077,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 1442
+    },
+    {
+      "epoch": 0.0924762881312484,
+      "grad_norm": 2.6165578444721995,
+      "learning_rate": 1e-06,
+      "loss": 0.4132,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09254037426300948,
+      "grad_norm": 2.559076650522681,
+      "learning_rate": 1e-06,
+      "loss": 0.468,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09260446039477058,
+      "grad_norm": 2.789845249548872,
+      "learning_rate": 1e-06,
+      "loss": 0.4552,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09266854652653166,
+      "grad_norm": 2.602140727223901,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09273263265829275,
+      "grad_norm": 2.7240550984204632,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09279671879005383,
+      "grad_norm": 2.5188910583037294,
+      "learning_rate": 1e-06,
+      "loss": 0.4161,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09286080492181492,
+      "grad_norm": 2.5994000439724054,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 1449
+    },
+    {
+      "epoch": 0.092924891053576,
+      "grad_norm": 2.6047446760108954,
+      "learning_rate": 1e-06,
+      "loss": 0.4495,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0929889771853371,
+      "grad_norm": 2.537159921941756,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09305306331709819,
+      "grad_norm": 2.7029522825334475,
+      "learning_rate": 1e-06,
+      "loss": 0.4348,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09311714944885927,
+      "grad_norm": 2.5101563201833286,
+      "learning_rate": 1e-06,
+      "loss": 0.3314,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09318123558062036,
+      "grad_norm": 2.5375880368645523,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09324532171238144,
+      "grad_norm": 2.630162183349999,
+      "learning_rate": 1e-06,
+      "loss": 0.4482,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09330940784414253,
+      "grad_norm": 2.770034028918006,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09337349397590361,
+      "grad_norm": 2.5993245566678507,
+      "learning_rate": 1e-06,
+      "loss": 0.3816,
+      "step": 1457
+    },
+    {
+      "epoch": 0.0934375801076647,
+      "grad_norm": 2.64778391952047,
+      "learning_rate": 1e-06,
+      "loss": 0.4181,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09350166623942578,
+      "grad_norm": 2.7001931669814008,
+      "learning_rate": 1e-06,
+      "loss": 0.3884,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09356575237118687,
+      "grad_norm": 2.7310231499002313,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09362983850294797,
+      "grad_norm": 2.5333368747726444,
+      "learning_rate": 1e-06,
+      "loss": 0.4217,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09369392463470905,
+      "grad_norm": 2.573504068195423,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09375801076647014,
+      "grad_norm": 2.586567658193003,
+      "learning_rate": 1e-06,
+      "loss": 0.4025,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09382209689823122,
+      "grad_norm": 2.561777711659588,
+      "learning_rate": 1e-06,
+      "loss": 0.4438,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09388618302999231,
+      "grad_norm": 2.8983923631902617,
+      "learning_rate": 1e-06,
+      "loss": 0.4852,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09395026916175339,
+      "grad_norm": 2.6375367410550017,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09401435529351448,
+      "grad_norm": 2.8361863431121375,
+      "learning_rate": 1e-06,
+      "loss": 0.3983,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09407844142527556,
+      "grad_norm": 2.6342268308154058,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09414252755703666,
+      "grad_norm": 2.608017548825659,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09420661368879775,
+      "grad_norm": 2.449627254791894,
+      "learning_rate": 1e-06,
+      "loss": 0.4113,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09427069982055883,
+      "grad_norm": 2.4811525375719556,
+      "learning_rate": 1e-06,
+      "loss": 0.377,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09433478595231992,
+      "grad_norm": 2.6371030366074617,
+      "learning_rate": 1e-06,
+      "loss": 0.389,
+      "step": 1472
+    },
+    {
+      "epoch": 0.094398872084081,
+      "grad_norm": 2.499025005200448,
+      "learning_rate": 1e-06,
+      "loss": 0.4178,
+      "step": 1473
+    },
+    {
+      "epoch": 0.0944629582158421,
+      "grad_norm": 2.449095256001988,
+      "learning_rate": 1e-06,
+      "loss": 0.4191,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09452704434760317,
+      "grad_norm": 2.5906968749902073,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09459113047936427,
+      "grad_norm": 2.7404480397879336,
+      "learning_rate": 1e-06,
+      "loss": 0.456,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09465521661112536,
+      "grad_norm": 2.72078552266344,
+      "learning_rate": 1e-06,
+      "loss": 0.4425,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09471930274288644,
+      "grad_norm": 2.6857231786650315,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09478338887464753,
+      "grad_norm": 2.788718665825192,
+      "learning_rate": 1e-06,
+      "loss": 0.4309,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09484747500640861,
+      "grad_norm": 2.5435811015181735,
+      "learning_rate": 1e-06,
+      "loss": 0.4508,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0949115611381697,
+      "grad_norm": 2.6178528404278567,
+      "learning_rate": 1e-06,
+      "loss": 0.375,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09497564726993078,
+      "grad_norm": 2.690631217688418,
+      "learning_rate": 1e-06,
+      "loss": 0.4573,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09503973340169188,
+      "grad_norm": 2.8404450272707704,
+      "learning_rate": 1e-06,
+      "loss": 0.4252,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09510381953345295,
+      "grad_norm": 2.546535336347818,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09516790566521405,
+      "grad_norm": 2.825421925183025,
+      "learning_rate": 1e-06,
+      "loss": 0.4806,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09523199179697514,
+      "grad_norm": 2.6961291272224654,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09529607792873622,
+      "grad_norm": 2.7748192803256577,
+      "learning_rate": 1e-06,
+      "loss": 0.4424,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09536016406049731,
+      "grad_norm": 2.598558063958087,
+      "learning_rate": 1e-06,
+      "loss": 0.483,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09542425019225839,
+      "grad_norm": 2.5230579044355554,
+      "learning_rate": 1e-06,
+      "loss": 0.4463,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09548833632401948,
+      "grad_norm": 2.7678832307087347,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09555242245578056,
+      "grad_norm": 2.6987595821256365,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09561650858754166,
+      "grad_norm": 2.5431991351880656,
+      "learning_rate": 1e-06,
+      "loss": 0.4091,
+      "step": 1492
+    },
+    {
+      "epoch": 0.09568059471930274,
+      "grad_norm": 2.684245256638835,
+      "learning_rate": 1e-06,
+      "loss": 0.4585,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09574468085106383,
+      "grad_norm": 2.552815133009136,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09580876698282492,
+      "grad_norm": 2.547549931917095,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 1495
+    },
+    {
+      "epoch": 0.095872853114586,
+      "grad_norm": 2.583837338026024,
+      "learning_rate": 1e-06,
+      "loss": 0.4405,
+      "step": 1496
+    },
+    {
+      "epoch": 0.0959369392463471,
+      "grad_norm": 2.6677862575379168,
+      "learning_rate": 1e-06,
+      "loss": 0.473,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09600102537810817,
+      "grad_norm": 2.9204657147492794,
+      "learning_rate": 1e-06,
+      "loss": 0.5309,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09606511150986927,
+      "grad_norm": 2.677637058756192,
+      "learning_rate": 1e-06,
+      "loss": 0.3905,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09612919764163035,
+      "grad_norm": 2.7688834587766435,
+      "learning_rate": 1e-06,
+      "loss": 0.4399,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09619328377339144,
+      "grad_norm": 2.740388435781337,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09625736990515253,
+      "grad_norm": 2.3972218665471514,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09632145603691361,
+      "grad_norm": 2.6633730067165526,
+      "learning_rate": 1e-06,
+      "loss": 0.4054,
+      "step": 1503
+    },
+    {
+      "epoch": 0.0963855421686747,
+      "grad_norm": 2.456909671813905,
+      "learning_rate": 1e-06,
+      "loss": 0.4143,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09644962830043578,
+      "grad_norm": 2.6087008055470235,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09651371443219688,
+      "grad_norm": 2.934760294530396,
+      "learning_rate": 1e-06,
+      "loss": 0.4752,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09657780056395796,
+      "grad_norm": 2.7777807926127185,
+      "learning_rate": 1e-06,
+      "loss": 0.3972,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09664188669571905,
+      "grad_norm": 2.5245228852446124,
+      "learning_rate": 1e-06,
+      "loss": 0.4838,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09670597282748013,
+      "grad_norm": 2.556699635104518,
+      "learning_rate": 1e-06,
+      "loss": 0.3796,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09677005895924122,
+      "grad_norm": 2.6734250594176294,
+      "learning_rate": 1e-06,
+      "loss": 0.3988,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09683414509100231,
+      "grad_norm": 2.779651783350114,
+      "learning_rate": 1e-06,
+      "loss": 0.4375,
+      "step": 1511
+    },
+    {
+      "epoch": 0.09689823122276339,
+      "grad_norm": 2.7072349412684886,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 1512
+    },
+    {
+      "epoch": 0.09696231735452449,
+      "grad_norm": 2.9895637028649746,
+      "learning_rate": 1e-06,
+      "loss": 0.4021,
+      "step": 1513
+    },
+    {
+      "epoch": 0.09702640348628556,
+      "grad_norm": 2.6235342457993416,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 1514
+    },
+    {
+      "epoch": 0.09709048961804666,
+      "grad_norm": 2.6685946420553273,
+      "learning_rate": 1e-06,
+      "loss": 0.491,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09715457574980774,
+      "grad_norm": 2.6912226523276086,
+      "learning_rate": 1e-06,
+      "loss": 0.3626,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09721866188156883,
+      "grad_norm": 2.5055739358842346,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09728274801332991,
+      "grad_norm": 3.0575599131857327,
+      "learning_rate": 1e-06,
+      "loss": 0.3804,
+      "step": 1518
+    },
+    {
+      "epoch": 0.097346834145091,
+      "grad_norm": 2.4144353166108634,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 1519
+    },
+    {
+      "epoch": 0.0974109202768521,
+      "grad_norm": 2.485527122413705,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09747500640861317,
+      "grad_norm": 2.727634679983602,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09753909254037427,
+      "grad_norm": 2.688199231816162,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 1522
+    },
+    {
+      "epoch": 0.09760317867213535,
+      "grad_norm": 2.597138289171914,
+      "learning_rate": 1e-06,
+      "loss": 0.436,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09766726480389644,
+      "grad_norm": 2.4989362972544966,
+      "learning_rate": 1e-06,
+      "loss": 0.3662,
+      "step": 1524
+    },
+    {
+      "epoch": 0.09773135093565752,
+      "grad_norm": 2.6467220685385926,
+      "learning_rate": 1e-06,
+      "loss": 0.4828,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09779543706741861,
+      "grad_norm": 2.570899529787848,
+      "learning_rate": 1e-06,
+      "loss": 0.4307,
+      "step": 1526
+    },
+    {
+      "epoch": 0.0978595231991797,
+      "grad_norm": 2.5504651035630803,
+      "learning_rate": 1e-06,
+      "loss": 0.4346,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09792360933094078,
+      "grad_norm": 2.658388218479122,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09798769546270188,
+      "grad_norm": 2.5854205257184955,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09805178159446296,
+      "grad_norm": 2.549513145186835,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09811586772622405,
+      "grad_norm": 2.7201381698108715,
+      "learning_rate": 1e-06,
+      "loss": 0.4101,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09817995385798513,
+      "grad_norm": 2.5040689871867934,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09824403998974622,
+      "grad_norm": 2.714762788781154,
+      "learning_rate": 1e-06,
+      "loss": 0.4831,
+      "step": 1533
+    },
+    {
+      "epoch": 0.0983081261215073,
+      "grad_norm": 2.552541268474821,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 1534
+    },
+    {
+      "epoch": 0.09837221225326839,
+      "grad_norm": 2.57284676055446,
+      "learning_rate": 1e-06,
+      "loss": 0.3585,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09843629838502949,
+      "grad_norm": 2.721429265477203,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09850038451679057,
+      "grad_norm": 2.774050751464011,
+      "learning_rate": 1e-06,
+      "loss": 0.4887,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09856447064855166,
+      "grad_norm": 2.7744727611174964,
+      "learning_rate": 1e-06,
+      "loss": 0.3775,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09862855678031274,
+      "grad_norm": 2.510003290063987,
+      "learning_rate": 1e-06,
+      "loss": 0.3722,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09869264291207383,
+      "grad_norm": 2.535552620114136,
+      "learning_rate": 1e-06,
+      "loss": 0.4426,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09875672904383491,
+      "grad_norm": 2.564367168182366,
+      "learning_rate": 1e-06,
+      "loss": 0.4083,
+      "step": 1541
+    },
+    {
+      "epoch": 0.098820815175596,
+      "grad_norm": 2.7394137242469947,
+      "learning_rate": 1e-06,
+      "loss": 0.4465,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09888490130735708,
+      "grad_norm": 2.545975890654039,
+      "learning_rate": 1e-06,
+      "loss": 0.3485,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09894898743911817,
+      "grad_norm": 2.774458084017365,
+      "learning_rate": 1e-06,
+      "loss": 0.4045,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09901307357087927,
+      "grad_norm": 2.5574985197779805,
+      "learning_rate": 1e-06,
+      "loss": 0.4363,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09907715970264035,
+      "grad_norm": 2.6854709411400104,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09914124583440144,
+      "grad_norm": 2.596155628657547,
+      "learning_rate": 1e-06,
+      "loss": 0.3991,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09920533196616252,
+      "grad_norm": 2.725319331359241,
+      "learning_rate": 1e-06,
+      "loss": 0.4757,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09926941809792361,
+      "grad_norm": 2.4312408844513835,
+      "learning_rate": 1e-06,
+      "loss": 0.3759,
+      "step": 1549
+    },
+    {
+      "epoch": 0.09933350422968469,
+      "grad_norm": 3.0975479055506665,
+      "learning_rate": 1e-06,
+      "loss": 0.405,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09939759036144578,
+      "grad_norm": 2.6897850500620626,
+      "learning_rate": 1e-06,
+      "loss": 0.478,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09946167649320686,
+      "grad_norm": 2.4014922518328414,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09952576262496796,
+      "grad_norm": 2.5781390912257507,
+      "learning_rate": 1e-06,
+      "loss": 0.3929,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09958984875672905,
+      "grad_norm": 2.6814168690887734,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09965393488849013,
+      "grad_norm": 2.8229320999460366,
+      "learning_rate": 1e-06,
+      "loss": 0.4424,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09971802102025122,
+      "grad_norm": 2.671523799350974,
+      "learning_rate": 1e-06,
+      "loss": 0.417,
+      "step": 1556
+    },
+    {
+      "epoch": 0.0997821071520123,
+      "grad_norm": 2.816317550333446,
+      "learning_rate": 1e-06,
+      "loss": 0.4308,
+      "step": 1557
+    },
+    {
+      "epoch": 0.0998461932837734,
+      "grad_norm": 2.8263116326882476,
+      "learning_rate": 1e-06,
+      "loss": 0.461,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09991027941553447,
+      "grad_norm": 2.773664845095712,
+      "learning_rate": 1e-06,
+      "loss": 0.3831,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09997436554729557,
+      "grad_norm": 2.465271500495268,
+      "learning_rate": 1e-06,
+      "loss": 0.4111,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10003845167905666,
+      "grad_norm": 2.5630490824452696,
+      "learning_rate": 1e-06,
+      "loss": 0.3999,
+      "step": 1561
+    },
+    {
+      "epoch": 0.10010253781081774,
+      "grad_norm": 2.443896329725649,
+      "learning_rate": 1e-06,
+      "loss": 0.3384,
+      "step": 1562
+    },
+    {
+      "epoch": 0.10016662394257883,
+      "grad_norm": 2.679277621913989,
+      "learning_rate": 1e-06,
+      "loss": 0.4006,
+      "step": 1563
+    },
+    {
+      "epoch": 0.10023071007433991,
+      "grad_norm": 2.687935697149264,
+      "learning_rate": 1e-06,
+      "loss": 0.3993,
+      "step": 1564
+    },
+    {
+      "epoch": 0.100294796206101,
+      "grad_norm": 2.672827135397584,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10035888233786208,
+      "grad_norm": 2.5859584068225305,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 1566
+    },
+    {
+      "epoch": 0.10042296846962318,
+      "grad_norm": 2.6623095578704845,
+      "learning_rate": 1e-06,
+      "loss": 0.4869,
+      "step": 1567
+    },
+    {
+      "epoch": 0.10048705460138425,
+      "grad_norm": 2.751166514445602,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 1568
+    },
+    {
+      "epoch": 0.10055114073314535,
+      "grad_norm": 2.5789596779897996,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 1569
+    },
+    {
+      "epoch": 0.10061522686490644,
+      "grad_norm": 2.733584791811764,
+      "learning_rate": 1e-06,
+      "loss": 0.4257,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10067931299666752,
+      "grad_norm": 2.771740197510259,
+      "learning_rate": 1e-06,
+      "loss": 0.4188,
+      "step": 1571
+    },
+    {
+      "epoch": 0.10074339912842861,
+      "grad_norm": 2.8752303657524387,
+      "learning_rate": 1e-06,
+      "loss": 0.4678,
+      "step": 1572
+    },
+    {
+      "epoch": 0.10080748526018969,
+      "grad_norm": 2.558260771757404,
+      "learning_rate": 1e-06,
+      "loss": 0.3632,
+      "step": 1573
+    },
+    {
+      "epoch": 0.10087157139195078,
+      "grad_norm": 2.5885122104751193,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 1574
+    },
+    {
+      "epoch": 0.10093565752371186,
+      "grad_norm": 2.535083774597389,
+      "learning_rate": 1e-06,
+      "loss": 0.413,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10099974365547296,
+      "grad_norm": 2.4010853501994966,
+      "learning_rate": 1e-06,
+      "loss": 0.451,
+      "step": 1576
+    },
+    {
+      "epoch": 0.10106382978723404,
+      "grad_norm": 2.649504940578061,
+      "learning_rate": 1e-06,
+      "loss": 0.4474,
+      "step": 1577
+    },
+    {
+      "epoch": 0.10112791591899513,
+      "grad_norm": 2.410664702100842,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 1578
+    },
+    {
+      "epoch": 0.10119200205075622,
+      "grad_norm": 2.614304760449613,
+      "learning_rate": 1e-06,
+      "loss": 0.4721,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1012560881825173,
+      "grad_norm": 2.5811032100103417,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1013201743142784,
+      "grad_norm": 2.479567803200828,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 1581
+    },
+    {
+      "epoch": 0.10138426044603947,
+      "grad_norm": 2.822694534916483,
+      "learning_rate": 1e-06,
+      "loss": 0.4662,
+      "step": 1582
+    },
+    {
+      "epoch": 0.10144834657780057,
+      "grad_norm": 2.7523444736975406,
+      "learning_rate": 1e-06,
+      "loss": 0.4016,
+      "step": 1583
+    },
+    {
+      "epoch": 0.10151243270956165,
+      "grad_norm": 2.7855545296692408,
+      "learning_rate": 1e-06,
+      "loss": 0.4851,
+      "step": 1584
+    },
+    {
+      "epoch": 0.10157651884132274,
+      "grad_norm": 2.69264141081618,
+      "learning_rate": 1e-06,
+      "loss": 0.4044,
+      "step": 1585
+    },
+    {
+      "epoch": 0.10164060497308383,
+      "grad_norm": 2.5649058822118764,
+      "learning_rate": 1e-06,
+      "loss": 0.3722,
+      "step": 1586
+    },
+    {
+      "epoch": 0.10170469110484491,
+      "grad_norm": 2.5866879553520246,
+      "learning_rate": 1e-06,
+      "loss": 0.4265,
+      "step": 1587
+    },
+    {
+      "epoch": 0.101768777236606,
+      "grad_norm": 2.4330295355820577,
+      "learning_rate": 1e-06,
+      "loss": 0.4152,
+      "step": 1588
+    },
+    {
+      "epoch": 0.10183286336836708,
+      "grad_norm": 2.570725714010572,
+      "learning_rate": 1e-06,
+      "loss": 0.3802,
+      "step": 1589
+    },
+    {
+      "epoch": 0.10189694950012818,
+      "grad_norm": 2.6275641529435845,
+      "learning_rate": 1e-06,
+      "loss": 0.3901,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10196103563188925,
+      "grad_norm": 2.7067442146798943,
+      "learning_rate": 1e-06,
+      "loss": 0.4823,
+      "step": 1591
+    },
+    {
+      "epoch": 0.10202512176365035,
+      "grad_norm": 2.4594883517967885,
+      "learning_rate": 1e-06,
+      "loss": 0.3733,
+      "step": 1592
+    },
+    {
+      "epoch": 0.10208920789541143,
+      "grad_norm": 2.588977297029969,
+      "learning_rate": 1e-06,
+      "loss": 0.4596,
+      "step": 1593
+    },
+    {
+      "epoch": 0.10215329402717252,
+      "grad_norm": 2.6727644456390935,
+      "learning_rate": 1e-06,
+      "loss": 0.4501,
+      "step": 1594
+    },
+    {
+      "epoch": 0.10221738015893361,
+      "grad_norm": 2.5011109727208813,
+      "learning_rate": 1e-06,
+      "loss": 0.3772,
+      "step": 1595
+    },
+    {
+      "epoch": 0.10228146629069469,
+      "grad_norm": 2.729101743448196,
+      "learning_rate": 1e-06,
+      "loss": 0.3365,
+      "step": 1596
+    },
+    {
+      "epoch": 0.10234555242245578,
+      "grad_norm": 2.7459192458327633,
+      "learning_rate": 1e-06,
+      "loss": 0.4438,
+      "step": 1597
+    },
+    {
+      "epoch": 0.10240963855421686,
+      "grad_norm": 2.9193827751910133,
+      "learning_rate": 1e-06,
+      "loss": 0.4325,
+      "step": 1598
+    },
+    {
+      "epoch": 0.10247372468597796,
+      "grad_norm": 2.5884629470491247,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 1599
+    },
+    {
+      "epoch": 0.10253781081773904,
+      "grad_norm": 2.6180046140478512,
+      "learning_rate": 1e-06,
+      "loss": 0.4598,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10260189694950013,
+      "grad_norm": 2.6215411725840774,
+      "learning_rate": 1e-06,
+      "loss": 0.4109,
+      "step": 1601
+    },
+    {
+      "epoch": 0.10266598308126121,
+      "grad_norm": 2.7502317600306654,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1027300692130223,
+      "grad_norm": 2.473642725106791,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1027941553447834,
+      "grad_norm": 3.027137495649638,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 1604
+    },
+    {
+      "epoch": 0.10285824147654447,
+      "grad_norm": 2.589351540135811,
+      "learning_rate": 1e-06,
+      "loss": 0.3928,
+      "step": 1605
+    },
+    {
+      "epoch": 0.10292232760830557,
+      "grad_norm": 2.605232656127217,
+      "learning_rate": 1e-06,
+      "loss": 0.3833,
+      "step": 1606
+    },
+    {
+      "epoch": 0.10298641374006665,
+      "grad_norm": 2.5160966475689373,
+      "learning_rate": 1e-06,
+      "loss": 0.4443,
+      "step": 1607
+    },
+    {
+      "epoch": 0.10305049987182774,
+      "grad_norm": 2.9291249601398945,
+      "learning_rate": 1e-06,
+      "loss": 0.3807,
+      "step": 1608
+    },
+    {
+      "epoch": 0.10311458600358882,
+      "grad_norm": 2.5567951767508417,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 1609
+    },
+    {
+      "epoch": 0.10317867213534991,
+      "grad_norm": 2.73170975100072,
+      "learning_rate": 1e-06,
+      "loss": 0.4131,
+      "step": 1610
+    },
+    {
+      "epoch": 0.103242758267111,
+      "grad_norm": 2.635090816171602,
+      "learning_rate": 1e-06,
+      "loss": 0.3814,
+      "step": 1611
+    },
+    {
+      "epoch": 0.10330684439887208,
+      "grad_norm": 2.461341775873606,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 1612
+    },
+    {
+      "epoch": 0.10337093053063318,
+      "grad_norm": 2.62340434297987,
+      "learning_rate": 1e-06,
+      "loss": 0.4303,
+      "step": 1613
+    },
+    {
+      "epoch": 0.10343501666239426,
+      "grad_norm": 2.418377756799505,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 1614
+    },
+    {
+      "epoch": 0.10349910279415535,
+      "grad_norm": 2.480451698537808,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 1615
+    },
+    {
+      "epoch": 0.10356318892591643,
+      "grad_norm": 2.5930294009627506,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 1616
+    },
+    {
+      "epoch": 0.10362727505767752,
+      "grad_norm": 2.5585904418749568,
+      "learning_rate": 1e-06,
+      "loss": 0.4271,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1036913611894386,
+      "grad_norm": 2.8192930639241442,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 1618
+    },
+    {
+      "epoch": 0.10375544732119969,
+      "grad_norm": 2.6233847466921216,
+      "learning_rate": 1e-06,
+      "loss": 0.4337,
+      "step": 1619
+    },
+    {
+      "epoch": 0.10381953345296079,
+      "grad_norm": 2.476019776497116,
+      "learning_rate": 1e-06,
+      "loss": 0.3443,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10388361958472186,
+      "grad_norm": 2.65079762175131,
+      "learning_rate": 1e-06,
+      "loss": 0.4388,
+      "step": 1621
+    },
+    {
+      "epoch": 0.10394770571648296,
+      "grad_norm": 2.617383658663692,
+      "learning_rate": 1e-06,
+      "loss": 0.4207,
+      "step": 1622
+    },
+    {
+      "epoch": 0.10401179184824404,
+      "grad_norm": 2.6226899184245305,
+      "learning_rate": 1e-06,
+      "loss": 0.4373,
+      "step": 1623
+    },
+    {
+      "epoch": 0.10407587798000513,
+      "grad_norm": 2.6357591101547806,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 1624
+    },
+    {
+      "epoch": 0.10413996411176621,
+      "grad_norm": 2.6520449138547604,
+      "learning_rate": 1e-06,
+      "loss": 0.4487,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1042040502435273,
+      "grad_norm": 2.6677348499168803,
+      "learning_rate": 1e-06,
+      "loss": 0.3955,
+      "step": 1626
+    },
+    {
+      "epoch": 0.10426813637528838,
+      "grad_norm": 2.485785549204256,
+      "learning_rate": 1e-06,
+      "loss": 0.4121,
+      "step": 1627
+    },
+    {
+      "epoch": 0.10433222250704947,
+      "grad_norm": 2.6469670405473655,
+      "learning_rate": 1e-06,
+      "loss": 0.4526,
+      "step": 1628
+    },
+    {
+      "epoch": 0.10439630863881057,
+      "grad_norm": 2.6217926498103394,
+      "learning_rate": 1e-06,
+      "loss": 0.4237,
+      "step": 1629
+    },
+    {
+      "epoch": 0.10446039477057165,
+      "grad_norm": 2.5329404663917767,
+      "learning_rate": 1e-06,
+      "loss": 0.4267,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10452448090233274,
+      "grad_norm": 2.799258811339821,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 1631
+    },
+    {
+      "epoch": 0.10458856703409382,
+      "grad_norm": 2.6785670772725885,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 1632
+    },
+    {
+      "epoch": 0.10465265316585491,
+      "grad_norm": 2.8053785359525754,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 1633
+    },
+    {
+      "epoch": 0.10471673929761599,
+      "grad_norm": 2.617661148969151,
+      "learning_rate": 1e-06,
+      "loss": 0.4612,
+      "step": 1634
+    },
+    {
+      "epoch": 0.10478082542937708,
+      "grad_norm": 2.7631606730879077,
+      "learning_rate": 1e-06,
+      "loss": 0.4398,
+      "step": 1635
+    },
+    {
+      "epoch": 0.10484491156113818,
+      "grad_norm": 2.6745121058036068,
+      "learning_rate": 1e-06,
+      "loss": 0.3933,
+      "step": 1636
+    },
+    {
+      "epoch": 0.10490899769289926,
+      "grad_norm": 2.6261615730588677,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 1637
+    },
+    {
+      "epoch": 0.10497308382466035,
+      "grad_norm": 2.4662550009165747,
+      "learning_rate": 1e-06,
+      "loss": 0.3995,
+      "step": 1638
+    },
+    {
+      "epoch": 0.10503716995642143,
+      "grad_norm": 2.586298212854455,
+      "learning_rate": 1e-06,
+      "loss": 0.3858,
+      "step": 1639
+    },
+    {
+      "epoch": 0.10510125608818252,
+      "grad_norm": 2.6534127259916374,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1051653422199436,
+      "grad_norm": 2.795573062823909,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 1641
+    },
+    {
+      "epoch": 0.10522942835170469,
+      "grad_norm": 2.7539546932567767,
+      "learning_rate": 1e-06,
+      "loss": 0.4529,
+      "step": 1642
+    },
+    {
+      "epoch": 0.10529351448346577,
+      "grad_norm": 2.6383536806239123,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 1643
+    },
+    {
+      "epoch": 0.10535760061522687,
+      "grad_norm": 2.5524878339737436,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 1644
+    },
+    {
+      "epoch": 0.10542168674698796,
+      "grad_norm": 2.365115184404631,
+      "learning_rate": 1e-06,
+      "loss": 0.4012,
+      "step": 1645
+    },
+    {
+      "epoch": 0.10548577287874904,
+      "grad_norm": 2.562456499015141,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 1646
+    },
+    {
+      "epoch": 0.10554985901051013,
+      "grad_norm": 2.540218757528349,
+      "learning_rate": 1e-06,
+      "loss": 0.3861,
+      "step": 1647
+    },
+    {
+      "epoch": 0.10561394514227121,
+      "grad_norm": 2.7507708144988645,
+      "learning_rate": 1e-06,
+      "loss": 0.4476,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1056780312740323,
+      "grad_norm": 2.569575329697874,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 1649
+    },
+    {
+      "epoch": 0.10574211740579338,
+      "grad_norm": 2.820273901570955,
+      "learning_rate": 1e-06,
+      "loss": 0.3955,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10580620353755447,
+      "grad_norm": 2.5636949127298423,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 1651
+    },
+    {
+      "epoch": 0.10587028966931555,
+      "grad_norm": 2.4593346132223326,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 1652
+    },
+    {
+      "epoch": 0.10593437580107665,
+      "grad_norm": 2.4987644523028387,
+      "learning_rate": 1e-06,
+      "loss": 0.3765,
+      "step": 1653
+    },
+    {
+      "epoch": 0.10599846193283774,
+      "grad_norm": 2.658818194022743,
+      "learning_rate": 1e-06,
+      "loss": 0.4174,
+      "step": 1654
+    },
+    {
+      "epoch": 0.10606254806459882,
+      "grad_norm": 2.5958535933853213,
+      "learning_rate": 1e-06,
+      "loss": 0.3846,
+      "step": 1655
+    },
+    {
+      "epoch": 0.10612663419635991,
+      "grad_norm": 2.664277085039511,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 1656
+    },
+    {
+      "epoch": 0.10619072032812099,
+      "grad_norm": 2.6513772408963665,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 1657
+    },
+    {
+      "epoch": 0.10625480645988208,
+      "grad_norm": 2.7469582827871566,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 1658
+    },
+    {
+      "epoch": 0.10631889259164316,
+      "grad_norm": 2.7045725526549576,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 1659
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 2.623229446283425,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10644706485516534,
+      "grad_norm": 2.586774207872217,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 1661
+    },
+    {
+      "epoch": 0.10651115098692643,
+      "grad_norm": 2.953381892885236,
+      "learning_rate": 1e-06,
+      "loss": 0.4256,
+      "step": 1662
+    },
+    {
+      "epoch": 0.10657523711868752,
+      "grad_norm": 2.769340797223471,
+      "learning_rate": 1e-06,
+      "loss": 0.4559,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1066393232504486,
+      "grad_norm": 2.638792643534101,
+      "learning_rate": 1e-06,
+      "loss": 0.4563,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1067034093822097,
+      "grad_norm": 2.780562883204113,
+      "learning_rate": 1e-06,
+      "loss": 0.4674,
+      "step": 1665
+    },
+    {
+      "epoch": 0.10676749551397077,
+      "grad_norm": 2.6954405632484923,
+      "learning_rate": 1e-06,
+      "loss": 0.4098,
+      "step": 1666
+    },
+    {
+      "epoch": 0.10683158164573187,
+      "grad_norm": 2.7618401225967273,
+      "learning_rate": 1e-06,
+      "loss": 0.4703,
+      "step": 1667
+    },
+    {
+      "epoch": 0.10689566777749294,
+      "grad_norm": 2.5574016056264837,
+      "learning_rate": 1e-06,
+      "loss": 0.4101,
+      "step": 1668
+    },
+    {
+      "epoch": 0.10695975390925404,
+      "grad_norm": 2.8001123266252077,
+      "learning_rate": 1e-06,
+      "loss": 0.3912,
+      "step": 1669
+    },
+    {
+      "epoch": 0.10702384004101513,
+      "grad_norm": 2.378615768916347,
+      "learning_rate": 1e-06,
+      "loss": 0.3575,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10708792617277621,
+      "grad_norm": 2.8999182844615885,
+      "learning_rate": 1e-06,
+      "loss": 0.4575,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1071520123045373,
+      "grad_norm": 2.698293857397677,
+      "learning_rate": 1e-06,
+      "loss": 0.4647,
+      "step": 1672
+    },
+    {
+      "epoch": 0.10721609843629838,
+      "grad_norm": 2.7334643214109278,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 1673
+    },
+    {
+      "epoch": 0.10728018456805948,
+      "grad_norm": 2.711296438189429,
+      "learning_rate": 1e-06,
+      "loss": 0.3844,
+      "step": 1674
+    },
+    {
+      "epoch": 0.10734427069982055,
+      "grad_norm": 2.3686553393233916,
+      "learning_rate": 1e-06,
+      "loss": 0.404,
+      "step": 1675
+    },
+    {
+      "epoch": 0.10740835683158165,
+      "grad_norm": 2.4640210766920694,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 1676
+    },
+    {
+      "epoch": 0.10747244296334273,
+      "grad_norm": 2.694372214956423,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 1677
+    },
+    {
+      "epoch": 0.10753652909510382,
+      "grad_norm": 2.629781628214181,
+      "learning_rate": 1e-06,
+      "loss": 0.464,
+      "step": 1678
+    },
+    {
+      "epoch": 0.10760061522686491,
+      "grad_norm": 2.491325092259068,
+      "learning_rate": 1e-06,
+      "loss": 0.4328,
+      "step": 1679
+    },
+    {
+      "epoch": 0.10766470135862599,
+      "grad_norm": 2.788493606737418,
+      "learning_rate": 1e-06,
+      "loss": 0.3886,
+      "step": 1680
+    },
+    {
+      "epoch": 0.10772878749038708,
+      "grad_norm": 2.916286988190873,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 1681
+    },
+    {
+      "epoch": 0.10779287362214816,
+      "grad_norm": 2.583616470716141,
+      "learning_rate": 1e-06,
+      "loss": 0.4642,
+      "step": 1682
+    },
+    {
+      "epoch": 0.10785695975390926,
+      "grad_norm": 2.4721597951784906,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 1683
+    },
+    {
+      "epoch": 0.10792104588567034,
+      "grad_norm": 2.67527569790069,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 1684
+    },
+    {
+      "epoch": 0.10798513201743143,
+      "grad_norm": 2.5895651711542,
+      "learning_rate": 1e-06,
+      "loss": 0.3923,
+      "step": 1685
+    },
+    {
+      "epoch": 0.10804921814919251,
+      "grad_norm": 2.5846016622522403,
+      "learning_rate": 1e-06,
+      "loss": 0.4352,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1081133042809536,
+      "grad_norm": 2.777517387535592,
+      "learning_rate": 1e-06,
+      "loss": 0.3758,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1081773904127147,
+      "grad_norm": 2.6384771823414557,
+      "learning_rate": 1e-06,
+      "loss": 0.395,
+      "step": 1688
+    },
+    {
+      "epoch": 0.10824147654447577,
+      "grad_norm": 2.3574131829286817,
+      "learning_rate": 1e-06,
+      "loss": 0.455,
+      "step": 1689
+    },
+    {
+      "epoch": 0.10830556267623687,
+      "grad_norm": 2.6567719560903567,
+      "learning_rate": 1e-06,
+      "loss": 0.3988,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10836964880799795,
+      "grad_norm": 2.5692755223481507,
+      "learning_rate": 1e-06,
+      "loss": 0.4512,
+      "step": 1691
+    },
+    {
+      "epoch": 0.10843373493975904,
+      "grad_norm": 2.499408417636477,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 1692
+    },
+    {
+      "epoch": 0.10849782107152012,
+      "grad_norm": 2.7581677889012046,
+      "learning_rate": 1e-06,
+      "loss": 0.4781,
+      "step": 1693
+    },
+    {
+      "epoch": 0.10856190720328121,
+      "grad_norm": 2.7449113590760703,
+      "learning_rate": 1e-06,
+      "loss": 0.3819,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1086259933350423,
+      "grad_norm": 2.4758292398159507,
+      "learning_rate": 1e-06,
+      "loss": 0.4012,
+      "step": 1695
+    },
+    {
+      "epoch": 0.10869007946680338,
+      "grad_norm": 2.776264244021826,
+      "learning_rate": 1e-06,
+      "loss": 0.4143,
+      "step": 1696
+    },
+    {
+      "epoch": 0.10875416559856448,
+      "grad_norm": 2.638478281868802,
+      "learning_rate": 1e-06,
+      "loss": 0.4193,
+      "step": 1697
+    },
+    {
+      "epoch": 0.10881825173032555,
+      "grad_norm": 2.481263674878327,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 1698
+    },
+    {
+      "epoch": 0.10888233786208665,
+      "grad_norm": 2.552295863623659,
+      "learning_rate": 1e-06,
+      "loss": 0.433,
+      "step": 1699
+    },
+    {
+      "epoch": 0.10894642399384773,
+      "grad_norm": 2.52105190848856,
+      "learning_rate": 1e-06,
+      "loss": 0.3908,
+      "step": 1700
+    },
+    {
+      "epoch": 0.10901051012560882,
+      "grad_norm": 2.518675874240785,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1090745962573699,
+      "grad_norm": 2.614317095507154,
+      "learning_rate": 1e-06,
+      "loss": 0.387,
+      "step": 1702
+    },
+    {
+      "epoch": 0.10913868238913099,
+      "grad_norm": 2.670338340620114,
+      "learning_rate": 1e-06,
+      "loss": 0.4327,
+      "step": 1703
+    },
+    {
+      "epoch": 0.10920276852089209,
+      "grad_norm": 2.679878055563821,
+      "learning_rate": 1e-06,
+      "loss": 0.4729,
+      "step": 1704
+    },
+    {
+      "epoch": 0.10926685465265316,
+      "grad_norm": 2.493372948578064,
+      "learning_rate": 1e-06,
+      "loss": 0.4325,
+      "step": 1705
+    },
+    {
+      "epoch": 0.10933094078441426,
+      "grad_norm": 2.5642512118942187,
+      "learning_rate": 1e-06,
+      "loss": 0.4222,
+      "step": 1706
+    },
+    {
+      "epoch": 0.10939502691617534,
+      "grad_norm": 2.5169199510593927,
+      "learning_rate": 1e-06,
+      "loss": 0.426,
+      "step": 1707
+    },
+    {
+      "epoch": 0.10945911304793643,
+      "grad_norm": 2.4887182192261665,
+      "learning_rate": 1e-06,
+      "loss": 0.409,
+      "step": 1708
+    },
+    {
+      "epoch": 0.10952319917969751,
+      "grad_norm": 2.7289202425480417,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1095872853114586,
+      "grad_norm": 2.61721341106042,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 1710
+    },
+    {
+      "epoch": 0.10965137144321968,
+      "grad_norm": 2.72774723157587,
+      "learning_rate": 1e-06,
+      "loss": 0.4001,
+      "step": 1711
+    },
+    {
+      "epoch": 0.10971545757498077,
+      "grad_norm": 2.413345581192089,
+      "learning_rate": 1e-06,
+      "loss": 0.3807,
+      "step": 1712
+    },
+    {
+      "epoch": 0.10977954370674187,
+      "grad_norm": 2.7211350379570955,
+      "learning_rate": 1e-06,
+      "loss": 0.3684,
+      "step": 1713
+    },
+    {
+      "epoch": 0.10984362983850295,
+      "grad_norm": 2.737851682076553,
+      "learning_rate": 1e-06,
+      "loss": 0.4509,
+      "step": 1714
+    },
+    {
+      "epoch": 0.10990771597026404,
+      "grad_norm": 2.5769530061758275,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 1715
+    },
+    {
+      "epoch": 0.10997180210202512,
+      "grad_norm": 2.570743845028058,
+      "learning_rate": 1e-06,
+      "loss": 0.4236,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11003588823378621,
+      "grad_norm": 2.756598918829375,
+      "learning_rate": 1e-06,
+      "loss": 0.4437,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11009997436554729,
+      "grad_norm": 2.57919783563559,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11016406049730838,
+      "grad_norm": 2.5580986528574887,
+      "learning_rate": 1e-06,
+      "loss": 0.4406,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11022814662906948,
+      "grad_norm": 2.668776644544978,
+      "learning_rate": 1e-06,
+      "loss": 0.451,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11029223276083056,
+      "grad_norm": 2.523514496339625,
+      "learning_rate": 1e-06,
+      "loss": 0.41,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11035631889259165,
+      "grad_norm": 2.603462247414295,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11042040502435273,
+      "grad_norm": 2.855953751352866,
+      "learning_rate": 1e-06,
+      "loss": 0.4488,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11048449115611382,
+      "grad_norm": 2.459019889378647,
+      "learning_rate": 1e-06,
+      "loss": 0.3524,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1105485772878749,
+      "grad_norm": 2.6543139169949668,
+      "learning_rate": 1e-06,
+      "loss": 0.3758,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11061266341963599,
+      "grad_norm": 2.732939934915263,
+      "learning_rate": 1e-06,
+      "loss": 0.4017,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11067674955139707,
+      "grad_norm": 2.5573541578391663,
+      "learning_rate": 1e-06,
+      "loss": 0.4032,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11074083568315816,
+      "grad_norm": 2.48054846948909,
+      "learning_rate": 1e-06,
+      "loss": 0.4306,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11080492181491926,
+      "grad_norm": 2.521020872733791,
+      "learning_rate": 1e-06,
+      "loss": 0.435,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11086900794668034,
+      "grad_norm": 2.401450393189567,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11093309407844143,
+      "grad_norm": 2.617773165324972,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11099718021020251,
+      "grad_norm": 2.6692631322553035,
+      "learning_rate": 1e-06,
+      "loss": 0.442,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1110612663419636,
+      "grad_norm": 2.6413046769270627,
+      "learning_rate": 1e-06,
+      "loss": 0.4591,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11112535247372468,
+      "grad_norm": 2.5232036530580353,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11118943860548577,
+      "grad_norm": 2.598170322205431,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11125352473724685,
+      "grad_norm": 2.5204863843861713,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11131761086900795,
+      "grad_norm": 2.531925371710578,
+      "learning_rate": 1e-06,
+      "loss": 0.4006,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11138169700076904,
+      "grad_norm": 2.5551159173093456,
+      "learning_rate": 1e-06,
+      "loss": 0.382,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11144578313253012,
+      "grad_norm": 2.76432440791053,
+      "learning_rate": 1e-06,
+      "loss": 0.4738,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11150986926429121,
+      "grad_norm": 2.7599097775430863,
+      "learning_rate": 1e-06,
+      "loss": 0.4049,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11157395539605229,
+      "grad_norm": 2.727649566997873,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11163804152781338,
+      "grad_norm": 2.548484134262224,
+      "learning_rate": 1e-06,
+      "loss": 0.3875,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11170212765957446,
+      "grad_norm": 2.4432877629217225,
+      "learning_rate": 1e-06,
+      "loss": 0.3908,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11176621379133556,
+      "grad_norm": 3.1019322307948913,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11183029992309665,
+      "grad_norm": 3.0096044794329613,
+      "learning_rate": 1e-06,
+      "loss": 0.4709,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11189438605485773,
+      "grad_norm": 2.469796239735023,
+      "learning_rate": 1e-06,
+      "loss": 0.4759,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11195847218661882,
+      "grad_norm": 2.5208667157610196,
+      "learning_rate": 1e-06,
+      "loss": 0.4266,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1120225583183799,
+      "grad_norm": 2.5523207697871175,
+      "learning_rate": 1e-06,
+      "loss": 0.4218,
+      "step": 1748
+    },
+    {
+      "epoch": 0.112086644450141,
+      "grad_norm": 2.517061409866147,
+      "learning_rate": 1e-06,
+      "loss": 0.3933,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11215073058190207,
+      "grad_norm": 2.5726688003570772,
+      "learning_rate": 1e-06,
+      "loss": 0.4698,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11221481671366317,
+      "grad_norm": 2.660408633103373,
+      "learning_rate": 1e-06,
+      "loss": 0.4223,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11227890284542424,
+      "grad_norm": 2.4592087263046505,
+      "learning_rate": 1e-06,
+      "loss": 0.3798,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11234298897718534,
+      "grad_norm": 2.583126620448884,
+      "learning_rate": 1e-06,
+      "loss": 0.444,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11240707510894643,
+      "grad_norm": 2.385137220467907,
+      "learning_rate": 1e-06,
+      "loss": 0.3733,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11247116124070751,
+      "grad_norm": 2.583408504897648,
+      "learning_rate": 1e-06,
+      "loss": 0.3652,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1125352473724686,
+      "grad_norm": 2.528928465202496,
+      "learning_rate": 1e-06,
+      "loss": 0.4839,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11259933350422968,
+      "grad_norm": 2.4844004028727578,
+      "learning_rate": 1e-06,
+      "loss": 0.3589,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11266341963599077,
+      "grad_norm": 2.6836038257816552,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11272750576775185,
+      "grad_norm": 2.5384485116106017,
+      "learning_rate": 1e-06,
+      "loss": 0.4044,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11279159189951295,
+      "grad_norm": 2.753583038091332,
+      "learning_rate": 1e-06,
+      "loss": 0.4455,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11285567803127403,
+      "grad_norm": 2.9345543598136827,
+      "learning_rate": 1e-06,
+      "loss": 0.4116,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11291976416303512,
+      "grad_norm": 2.7326619237198067,
+      "learning_rate": 1e-06,
+      "loss": 0.3968,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11298385029479621,
+      "grad_norm": 2.64669411763611,
+      "learning_rate": 1e-06,
+      "loss": 0.3625,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11304793642655729,
+      "grad_norm": 2.552350284148369,
+      "learning_rate": 1e-06,
+      "loss": 0.3777,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11311202255831838,
+      "grad_norm": 2.8295097832904132,
+      "learning_rate": 1e-06,
+      "loss": 0.4645,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11317610869007946,
+      "grad_norm": 2.77247592560161,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11324019482184056,
+      "grad_norm": 2.5665589634227968,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11330428095360164,
+      "grad_norm": 2.5795506039181593,
+      "learning_rate": 1e-06,
+      "loss": 0.4332,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11336836708536273,
+      "grad_norm": 2.509627569113794,
+      "learning_rate": 1e-06,
+      "loss": 0.3402,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11343245321712381,
+      "grad_norm": 2.8115985620573087,
+      "learning_rate": 1e-06,
+      "loss": 0.4901,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1134965393488849,
+      "grad_norm": 2.839001751369005,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 1771
+    },
+    {
+      "epoch": 0.113560625480646,
+      "grad_norm": 2.565272919502216,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11362471161240707,
+      "grad_norm": 2.5032995137142797,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11368879774416817,
+      "grad_norm": 2.5599129867744015,
+      "learning_rate": 1e-06,
+      "loss": 0.4237,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11375288387592924,
+      "grad_norm": 2.8282808159804227,
+      "learning_rate": 1e-06,
+      "loss": 0.397,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11381697000769034,
+      "grad_norm": 2.435898564425463,
+      "learning_rate": 1e-06,
+      "loss": 0.4122,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11388105613945142,
+      "grad_norm": 2.590842457237392,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11394514227121251,
+      "grad_norm": 2.78652717011761,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1140092284029736,
+      "grad_norm": 2.8160506518694404,
+      "learning_rate": 1e-06,
+      "loss": 0.4206,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11407331453473468,
+      "grad_norm": 2.7610789996086043,
+      "learning_rate": 1e-06,
+      "loss": 0.4261,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11413740066649578,
+      "grad_norm": 2.6617844651355846,
+      "learning_rate": 1e-06,
+      "loss": 0.3811,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11420148679825685,
+      "grad_norm": 2.496233276933666,
+      "learning_rate": 1e-06,
+      "loss": 0.4673,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11426557293001795,
+      "grad_norm": 2.455381531806013,
+      "learning_rate": 1e-06,
+      "loss": 0.3537,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11432965906177903,
+      "grad_norm": 2.7512371176528867,
+      "learning_rate": 1e-06,
+      "loss": 0.3862,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11439374519354012,
+      "grad_norm": 2.455432753073087,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1144578313253012,
+      "grad_norm": 2.5204235042326544,
+      "learning_rate": 1e-06,
+      "loss": 0.3941,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11452191745706229,
+      "grad_norm": 2.8503009414302665,
+      "learning_rate": 1e-06,
+      "loss": 0.3999,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11458600358882338,
+      "grad_norm": 2.6189598564105228,
+      "learning_rate": 1e-06,
+      "loss": 0.4742,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11465008972058446,
+      "grad_norm": 2.605962177702438,
+      "learning_rate": 1e-06,
+      "loss": 0.4198,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11471417585234556,
+      "grad_norm": 2.6260862237551996,
+      "learning_rate": 1e-06,
+      "loss": 0.4471,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11477826198410664,
+      "grad_norm": 2.9144544195926447,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11484234811586773,
+      "grad_norm": 2.7848484757662604,
+      "learning_rate": 1e-06,
+      "loss": 0.4851,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11490643424762881,
+      "grad_norm": 2.6432933600668926,
+      "learning_rate": 1e-06,
+      "loss": 0.4358,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1149705203793899,
+      "grad_norm": 2.5969710533282697,
+      "learning_rate": 1e-06,
+      "loss": 0.3901,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11503460651115098,
+      "grad_norm": 2.6038648664893076,
+      "learning_rate": 1e-06,
+      "loss": 0.3698,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11509869264291207,
+      "grad_norm": 2.664492510496455,
+      "learning_rate": 1e-06,
+      "loss": 0.4344,
+      "step": 1796
+    },
+    {
+      "epoch": 0.11516277877467317,
+      "grad_norm": 2.4865877780180248,
+      "learning_rate": 1e-06,
+      "loss": 0.4152,
+      "step": 1797
+    },
+    {
+      "epoch": 0.11522686490643425,
+      "grad_norm": 2.6980174755024864,
+      "learning_rate": 1e-06,
+      "loss": 0.4049,
+      "step": 1798
+    },
+    {
+      "epoch": 0.11529095103819534,
+      "grad_norm": 2.7132888219218447,
+      "learning_rate": 1e-06,
+      "loss": 0.4101,
+      "step": 1799
+    },
+    {
+      "epoch": 0.11535503716995642,
+      "grad_norm": 2.6163019786785755,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11541912330171751,
+      "grad_norm": 2.438070830380359,
+      "learning_rate": 1e-06,
+      "loss": 0.3896,
+      "step": 1801
+    },
+    {
+      "epoch": 0.11548320943347859,
+      "grad_norm": 2.777361514104172,
+      "learning_rate": 1e-06,
+      "loss": 0.4361,
+      "step": 1802
+    },
+    {
+      "epoch": 0.11554729556523968,
+      "grad_norm": 2.4615952090613153,
+      "learning_rate": 1e-06,
+      "loss": 0.4725,
+      "step": 1803
+    },
+    {
+      "epoch": 0.11561138169700078,
+      "grad_norm": 2.786019550812383,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 1804
+    },
+    {
+      "epoch": 0.11567546782876185,
+      "grad_norm": 2.582580806012161,
+      "learning_rate": 1e-06,
+      "loss": 0.4174,
+      "step": 1805
+    },
+    {
+      "epoch": 0.11573955396052295,
+      "grad_norm": 2.6551798788814307,
+      "learning_rate": 1e-06,
+      "loss": 0.4031,
+      "step": 1806
+    },
+    {
+      "epoch": 0.11580364009228403,
+      "grad_norm": 2.628264393086419,
+      "learning_rate": 1e-06,
+      "loss": 0.4183,
+      "step": 1807
+    },
+    {
+      "epoch": 0.11586772622404512,
+      "grad_norm": 2.554298651583823,
+      "learning_rate": 1e-06,
+      "loss": 0.4543,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1159318123558062,
+      "grad_norm": 2.5870692588572477,
+      "learning_rate": 1e-06,
+      "loss": 0.4007,
+      "step": 1809
+    },
+    {
+      "epoch": 0.11599589848756729,
+      "grad_norm": 2.536498615958139,
+      "learning_rate": 1e-06,
+      "loss": 0.4527,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11605998461932837,
+      "grad_norm": 2.7350417193936605,
+      "learning_rate": 1e-06,
+      "loss": 0.4948,
+      "step": 1811
+    },
+    {
+      "epoch": 0.11612407075108946,
+      "grad_norm": 2.8962172722341446,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 1812
+    },
+    {
+      "epoch": 0.11618815688285056,
+      "grad_norm": 2.447551642523489,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 1813
+    },
+    {
+      "epoch": 0.11625224301461164,
+      "grad_norm": 2.7581062937740324,
+      "learning_rate": 1e-06,
+      "loss": 0.4046,
+      "step": 1814
+    },
+    {
+      "epoch": 0.11631632914637273,
+      "grad_norm": 2.6613780988530387,
+      "learning_rate": 1e-06,
+      "loss": 0.4378,
+      "step": 1815
+    },
+    {
+      "epoch": 0.11638041527813381,
+      "grad_norm": 2.7024783765058067,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1164445014098949,
+      "grad_norm": 2.610733885000389,
+      "learning_rate": 1e-06,
+      "loss": 0.4581,
+      "step": 1817
+    },
+    {
+      "epoch": 0.11650858754165598,
+      "grad_norm": 2.7818604058811367,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 1818
+    },
+    {
+      "epoch": 0.11657267367341707,
+      "grad_norm": 2.852704189136639,
+      "learning_rate": 1e-06,
+      "loss": 0.4626,
+      "step": 1819
+    },
+    {
+      "epoch": 0.11663675980517815,
+      "grad_norm": 2.6925423286197496,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11670084593693925,
+      "grad_norm": 2.8192484239080056,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 1821
+    },
+    {
+      "epoch": 0.11676493206870034,
+      "grad_norm": 2.8485436665584536,
+      "learning_rate": 1e-06,
+      "loss": 0.4731,
+      "step": 1822
+    },
+    {
+      "epoch": 0.11682901820046142,
+      "grad_norm": 2.6475121457375645,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 1823
+    },
+    {
+      "epoch": 0.11689310433222251,
+      "grad_norm": 2.8733685444969903,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 1824
+    },
+    {
+      "epoch": 0.11695719046398359,
+      "grad_norm": 2.6062579224564595,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 1825
+    },
+    {
+      "epoch": 0.11702127659574468,
+      "grad_norm": 2.6613936407149485,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 1826
+    },
+    {
+      "epoch": 0.11708536272750576,
+      "grad_norm": 2.607041472301362,
+      "learning_rate": 1e-06,
+      "loss": 0.4213,
+      "step": 1827
+    },
+    {
+      "epoch": 0.11714944885926686,
+      "grad_norm": 2.595697100885704,
+      "learning_rate": 1e-06,
+      "loss": 0.3763,
+      "step": 1828
+    },
+    {
+      "epoch": 0.11721353499102795,
+      "grad_norm": 2.627079136159903,
+      "learning_rate": 1e-06,
+      "loss": 0.4152,
+      "step": 1829
+    },
+    {
+      "epoch": 0.11727762112278903,
+      "grad_norm": 2.8799355704445695,
+      "learning_rate": 1e-06,
+      "loss": 0.4191,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11734170725455012,
+      "grad_norm": 2.66793046478117,
+      "learning_rate": 1e-06,
+      "loss": 0.4221,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1174057933863112,
+      "grad_norm": 2.5435157561676003,
+      "learning_rate": 1e-06,
+      "loss": 0.4753,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11746987951807229,
+      "grad_norm": 2.6227753499816684,
+      "learning_rate": 1e-06,
+      "loss": 0.4436,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11753396564983337,
+      "grad_norm": 2.6787886244585413,
+      "learning_rate": 1e-06,
+      "loss": 0.3946,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11759805178159446,
+      "grad_norm": 2.6989424714818435,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11766213791335554,
+      "grad_norm": 2.693869459395887,
+      "learning_rate": 1e-06,
+      "loss": 0.3903,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11772622404511664,
+      "grad_norm": 2.5534714081744165,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11779031017687773,
+      "grad_norm": 2.5396904316255298,
+      "learning_rate": 1e-06,
+      "loss": 0.4625,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11785439630863881,
+      "grad_norm": 2.6864914728899834,
+      "learning_rate": 1e-06,
+      "loss": 0.4628,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1179184824403999,
+      "grad_norm": 2.4031481570736,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11798256857216098,
+      "grad_norm": 2.8510808627208566,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11804665470392207,
+      "grad_norm": 2.625121495397035,
+      "learning_rate": 1e-06,
+      "loss": 0.4293,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11811074083568315,
+      "grad_norm": 2.379041716484206,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 1843
+    },
+    {
+      "epoch": 0.11817482696744425,
+      "grad_norm": 2.941854719930437,
+      "learning_rate": 1e-06,
+      "loss": 0.4159,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11823891309920533,
+      "grad_norm": 2.659123817605973,
+      "learning_rate": 1e-06,
+      "loss": 0.3979,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11830299923096642,
+      "grad_norm": 2.664461683615228,
+      "learning_rate": 1e-06,
+      "loss": 0.372,
+      "step": 1846
+    },
+    {
+      "epoch": 0.11836708536272751,
+      "grad_norm": 2.4454388383128145,
+      "learning_rate": 1e-06,
+      "loss": 0.3694,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11843117149448859,
+      "grad_norm": 2.9853732892969007,
+      "learning_rate": 1e-06,
+      "loss": 0.4965,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11849525762624968,
+      "grad_norm": 2.5862583091122637,
+      "learning_rate": 1e-06,
+      "loss": 0.429,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11855934375801076,
+      "grad_norm": 2.650195995580348,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11862342988977186,
+      "grad_norm": 2.4840409762343776,
+      "learning_rate": 1e-06,
+      "loss": 0.4511,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11868751602153294,
+      "grad_norm": 2.8313952437027763,
+      "learning_rate": 1e-06,
+      "loss": 0.3909,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11875160215329403,
+      "grad_norm": 2.598131250280352,
+      "learning_rate": 1e-06,
+      "loss": 0.4383,
+      "step": 1853
+    },
+    {
+      "epoch": 0.11881568828505511,
+      "grad_norm": 2.5956430213471644,
+      "learning_rate": 1e-06,
+      "loss": 0.3624,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1188797744168162,
+      "grad_norm": 2.819100294041442,
+      "learning_rate": 1e-06,
+      "loss": 0.4846,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1189438605485773,
+      "grad_norm": 2.659323824569532,
+      "learning_rate": 1e-06,
+      "loss": 0.4342,
+      "step": 1856
+    },
+    {
+      "epoch": 0.11900794668033837,
+      "grad_norm": 2.5236617424640784,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11907203281209947,
+      "grad_norm": 2.4090615191739526,
+      "learning_rate": 1e-06,
+      "loss": 0.4327,
+      "step": 1858
+    },
+    {
+      "epoch": 0.11913611894386054,
+      "grad_norm": 2.769880388165618,
+      "learning_rate": 1e-06,
+      "loss": 0.381,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11920020507562164,
+      "grad_norm": 2.819628662213885,
+      "learning_rate": 1e-06,
+      "loss": 0.4547,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11926429120738272,
+      "grad_norm": 2.5754514969457873,
+      "learning_rate": 1e-06,
+      "loss": 0.4594,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11932837733914381,
+      "grad_norm": 2.5022880249203325,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1193924634709049,
+      "grad_norm": 2.4699745018799404,
+      "learning_rate": 1e-06,
+      "loss": 0.3806,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11945654960266598,
+      "grad_norm": 2.6456565415557374,
+      "learning_rate": 1e-06,
+      "loss": 0.4065,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11952063573442707,
+      "grad_norm": 2.4005550753618015,
+      "learning_rate": 1e-06,
+      "loss": 0.4158,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11958472186618815,
+      "grad_norm": 2.5060801713618983,
+      "learning_rate": 1e-06,
+      "loss": 0.4542,
+      "step": 1866
+    },
+    {
+      "epoch": 0.11964880799794925,
+      "grad_norm": 2.4513719291485585,
+      "learning_rate": 1e-06,
+      "loss": 0.4339,
+      "step": 1867
+    },
+    {
+      "epoch": 0.11971289412971033,
+      "grad_norm": 2.6769921439010753,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11977698026147142,
+      "grad_norm": 2.6840492380158834,
+      "learning_rate": 1e-06,
+      "loss": 0.4007,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1198410663932325,
+      "grad_norm": 2.803156210251872,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11990515252499359,
+      "grad_norm": 2.6987854999821597,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11996923865675468,
+      "grad_norm": 2.665721870443163,
+      "learning_rate": 1e-06,
+      "loss": 0.4221,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12003332478851576,
+      "grad_norm": 2.6560564490065093,
+      "learning_rate": 1e-06,
+      "loss": 0.3591,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12009741092027686,
+      "grad_norm": 2.7021179621334004,
+      "learning_rate": 1e-06,
+      "loss": 0.4218,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12016149705203794,
+      "grad_norm": 2.6055942460970827,
+      "learning_rate": 1e-06,
+      "loss": 0.4245,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12022558318379903,
+      "grad_norm": 2.416483571969232,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12028966931556011,
+      "grad_norm": 2.8828531541836506,
+      "learning_rate": 1e-06,
+      "loss": 0.4091,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1203537554473212,
+      "grad_norm": 2.6443922479248787,
+      "learning_rate": 1e-06,
+      "loss": 0.4548,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12041784157908228,
+      "grad_norm": 2.688894934206754,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12048192771084337,
+      "grad_norm": 2.633924101866232,
+      "learning_rate": 1e-06,
+      "loss": 0.3938,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12054601384260447,
+      "grad_norm": 2.807137697859837,
+      "learning_rate": 1e-06,
+      "loss": 0.413,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12061009997436554,
+      "grad_norm": 2.552820173004839,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12067418610612664,
+      "grad_norm": 2.6265690322793978,
+      "learning_rate": 1e-06,
+      "loss": 0.3926,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12073827223788772,
+      "grad_norm": 2.678479232293037,
+      "learning_rate": 1e-06,
+      "loss": 0.4639,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12080235836964881,
+      "grad_norm": 2.6050744958084264,
+      "learning_rate": 1e-06,
+      "loss": 0.4349,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12086644450140989,
+      "grad_norm": 2.7082918851876427,
+      "learning_rate": 1e-06,
+      "loss": 0.4765,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12093053063317098,
+      "grad_norm": 2.526845919082137,
+      "learning_rate": 1e-06,
+      "loss": 0.3662,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12099461676493208,
+      "grad_norm": 2.6302833083000703,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12105870289669315,
+      "grad_norm": 2.553849692507201,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12112278902845425,
+      "grad_norm": 2.683491080670172,
+      "learning_rate": 1e-06,
+      "loss": 0.4514,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12118687516021533,
+      "grad_norm": 2.6364753658326125,
+      "learning_rate": 1e-06,
+      "loss": 0.4189,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12125096129197642,
+      "grad_norm": 2.7165916312599423,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1213150474237375,
+      "grad_norm": 2.549170313734264,
+      "learning_rate": 1e-06,
+      "loss": 0.4106,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12137913355549859,
+      "grad_norm": 2.6838182220230906,
+      "learning_rate": 1e-06,
+      "loss": 0.3677,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12144321968725967,
+      "grad_norm": 2.6414043011658865,
+      "learning_rate": 1e-06,
+      "loss": 0.3912,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12150730581902076,
+      "grad_norm": 2.5632629923340775,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12157139195078186,
+      "grad_norm": 2.544984397970424,
+      "learning_rate": 1e-06,
+      "loss": 0.4227,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12163547808254294,
+      "grad_norm": 2.5847715613939455,
+      "learning_rate": 1e-06,
+      "loss": 0.4267,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12169956421430403,
+      "grad_norm": 2.65803391386428,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12176365034606511,
+      "grad_norm": 2.6897903682954847,
+      "learning_rate": 1e-06,
+      "loss": 0.4652,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1218277364778262,
+      "grad_norm": 2.56277118816797,
+      "learning_rate": 1e-06,
+      "loss": 0.4529,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12189182260958728,
+      "grad_norm": 2.5517501695285016,
+      "learning_rate": 1e-06,
+      "loss": 0.465,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12195590874134837,
+      "grad_norm": 2.648252581171726,
+      "learning_rate": 1e-06,
+      "loss": 0.455,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12201999487310945,
+      "grad_norm": 2.6367007538455938,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12208408100487055,
+      "grad_norm": 2.6880014376963044,
+      "learning_rate": 1e-06,
+      "loss": 0.4533,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12214816713663164,
+      "grad_norm": 2.6551564910990058,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12221225326839272,
+      "grad_norm": 2.814098153913346,
+      "learning_rate": 1e-06,
+      "loss": 0.37,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12227633940015381,
+      "grad_norm": 2.5032472998077466,
+      "learning_rate": 1e-06,
+      "loss": 0.3843,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12234042553191489,
+      "grad_norm": 2.8306387124319192,
+      "learning_rate": 1e-06,
+      "loss": 0.4515,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12240451166367598,
+      "grad_norm": 2.42122372122656,
+      "learning_rate": 1e-06,
+      "loss": 0.4164,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12246859779543706,
+      "grad_norm": 2.537249205277831,
+      "learning_rate": 1e-06,
+      "loss": 0.4379,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12253268392719815,
+      "grad_norm": 2.6442746159357924,
+      "learning_rate": 1e-06,
+      "loss": 0.4344,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12259677005895925,
+      "grad_norm": 2.656606308028338,
+      "learning_rate": 1e-06,
+      "loss": 0.379,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12266085619072033,
+      "grad_norm": 2.8088794752870014,
+      "learning_rate": 1e-06,
+      "loss": 0.4047,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12272494232248142,
+      "grad_norm": 2.582575451213974,
+      "learning_rate": 1e-06,
+      "loss": 0.4024,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1227890284542425,
+      "grad_norm": 2.710660924430504,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12285311458600359,
+      "grad_norm": 2.42215293270404,
+      "learning_rate": 1e-06,
+      "loss": 0.4298,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12291720071776467,
+      "grad_norm": 2.5921801959218316,
+      "learning_rate": 1e-06,
+      "loss": 0.3942,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12298128684952576,
+      "grad_norm": 2.5278890165586057,
+      "learning_rate": 1e-06,
+      "loss": 0.3819,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12304537298128684,
+      "grad_norm": 2.524488936361231,
+      "learning_rate": 1e-06,
+      "loss": 0.4046,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12310945911304794,
+      "grad_norm": 2.4328966559195675,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12317354524480903,
+      "grad_norm": 2.774360894158552,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12323763137657011,
+      "grad_norm": 2.518717450808227,
+      "learning_rate": 1e-06,
+      "loss": 0.4391,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1233017175083312,
+      "grad_norm": 2.832159155238884,
+      "learning_rate": 1e-06,
+      "loss": 0.4229,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12336580364009228,
+      "grad_norm": 2.511577878519737,
+      "learning_rate": 1e-06,
+      "loss": 0.4165,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12342988977185337,
+      "grad_norm": 3.5208364019882885,
+      "learning_rate": 1e-06,
+      "loss": 0.3873,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12349397590361445,
+      "grad_norm": 2.427688682125614,
+      "learning_rate": 1e-06,
+      "loss": 0.377,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12355806203537555,
+      "grad_norm": 2.6404501242098837,
+      "learning_rate": 1e-06,
+      "loss": 0.454,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12362214816713663,
+      "grad_norm": 2.441849984145885,
+      "learning_rate": 1e-06,
+      "loss": 0.3963,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12368623429889772,
+      "grad_norm": 2.443122503103709,
+      "learning_rate": 1e-06,
+      "loss": 0.4631,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12375032043065881,
+      "grad_norm": 2.8171252209762474,
+      "learning_rate": 1e-06,
+      "loss": 0.4887,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12381440656241989,
+      "grad_norm": 2.4480320743078887,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12387849269418098,
+      "grad_norm": 2.6932808980425107,
+      "learning_rate": 1e-06,
+      "loss": 0.4569,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12394257882594206,
+      "grad_norm": 2.820935235598579,
+      "learning_rate": 1e-06,
+      "loss": 0.4038,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12400666495770316,
+      "grad_norm": 2.6341412685655983,
+      "learning_rate": 1e-06,
+      "loss": 0.4254,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12407075108946423,
+      "grad_norm": 2.6019885732357326,
+      "learning_rate": 1e-06,
+      "loss": 0.4036,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12413483722122533,
+      "grad_norm": 2.6751182734908965,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12419892335298642,
+      "grad_norm": 2.5870567275864325,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1242630094847475,
+      "grad_norm": 2.5353320352240525,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12432709561650859,
+      "grad_norm": 2.598518940629256,
+      "learning_rate": 1e-06,
+      "loss": 0.3844,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12439118174826967,
+      "grad_norm": 2.562879451548365,
+      "learning_rate": 1e-06,
+      "loss": 0.3962,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12445526788003076,
+      "grad_norm": 2.645786720650379,
+      "learning_rate": 1e-06,
+      "loss": 0.4838,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12451935401179184,
+      "grad_norm": 2.6403502102882808,
+      "learning_rate": 1e-06,
+      "loss": 0.4711,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12458344014355294,
+      "grad_norm": 2.876774406457867,
+      "learning_rate": 1e-06,
+      "loss": 0.4092,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12464752627531402,
+      "grad_norm": 2.751100020052121,
+      "learning_rate": 1e-06,
+      "loss": 0.4066,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12471161240707511,
+      "grad_norm": 2.721494548343405,
+      "learning_rate": 1e-06,
+      "loss": 0.4333,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1247756985388362,
+      "grad_norm": 3.0707931228709318,
+      "learning_rate": 1e-06,
+      "loss": 0.4444,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12483978467059728,
+      "grad_norm": 2.5536278780711874,
+      "learning_rate": 1e-06,
+      "loss": 0.4432,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12490387080235837,
+      "grad_norm": 2.5952295096194327,
+      "learning_rate": 1e-06,
+      "loss": 0.4064,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12496795693411945,
+      "grad_norm": 2.804966831875095,
+      "learning_rate": 1e-06,
+      "loss": 0.4377,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12503204306588053,
+      "grad_norm": 2.6254645986466016,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12509612919764163,
+      "grad_norm": 2.8575547875752734,
+      "learning_rate": 1e-06,
+      "loss": 0.4404,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12516021532940272,
+      "grad_norm": 2.890946909114489,
+      "learning_rate": 1e-06,
+      "loss": 0.3916,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1252243014611638,
+      "grad_norm": 2.4936005818287588,
+      "learning_rate": 1e-06,
+      "loss": 0.4075,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1252883875929249,
+      "grad_norm": 2.6488281056040917,
+      "learning_rate": 1e-06,
+      "loss": 0.4452,
+      "step": 1955
+    },
+    {
+      "epoch": 0.12535247372468597,
+      "grad_norm": 2.5861471919909875,
+      "learning_rate": 1e-06,
+      "loss": 0.3942,
+      "step": 1956
+    },
+    {
+      "epoch": 0.12541655985644706,
+      "grad_norm": 3.5318637594090085,
+      "learning_rate": 1e-06,
+      "loss": 0.3976,
+      "step": 1957
+    },
+    {
+      "epoch": 0.12548064598820816,
+      "grad_norm": 2.399050718582829,
+      "learning_rate": 1e-06,
+      "loss": 0.3801,
+      "step": 1958
+    },
+    {
+      "epoch": 0.12554473211996925,
+      "grad_norm": 2.597714681002882,
+      "learning_rate": 1e-06,
+      "loss": 0.3961,
+      "step": 1959
+    },
+    {
+      "epoch": 0.12560881825173031,
+      "grad_norm": 2.4663048518140362,
+      "learning_rate": 1e-06,
+      "loss": 0.4151,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1256729043834914,
+      "grad_norm": 2.5624663402691126,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1257369905152525,
+      "grad_norm": 3.140524750218864,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1258010766470136,
+      "grad_norm": 2.810127821809344,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1258651627787747,
+      "grad_norm": 2.5209202909604,
+      "learning_rate": 1e-06,
+      "loss": 0.4337,
+      "step": 1964
+    },
+    {
+      "epoch": 0.12592924891053575,
+      "grad_norm": 2.651740149209284,
+      "learning_rate": 1e-06,
+      "loss": 0.4091,
+      "step": 1965
+    },
+    {
+      "epoch": 0.12599333504229684,
+      "grad_norm": 2.6638550091846,
+      "learning_rate": 1e-06,
+      "loss": 0.3951,
+      "step": 1966
+    },
+    {
+      "epoch": 0.12605742117405794,
+      "grad_norm": 2.7390919328674106,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 1967
+    },
+    {
+      "epoch": 0.12612150730581903,
+      "grad_norm": 2.568954836564736,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1261855934375801,
+      "grad_norm": 2.623192224449949,
+      "learning_rate": 1e-06,
+      "loss": 0.4082,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1262496795693412,
+      "grad_norm": 2.772705656163098,
+      "learning_rate": 1e-06,
+      "loss": 0.4291,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12631376570110228,
+      "grad_norm": 3.0653029335133293,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 1971
+    },
+    {
+      "epoch": 0.12637785183286337,
+      "grad_norm": 2.71401827712792,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 1972
+    },
+    {
+      "epoch": 0.12644193796462447,
+      "grad_norm": 2.3069147395316767,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 1973
+    },
+    {
+      "epoch": 0.12650602409638553,
+      "grad_norm": 2.944709378967964,
+      "learning_rate": 1e-06,
+      "loss": 0.4009,
+      "step": 1974
+    },
+    {
+      "epoch": 0.12657011022814663,
+      "grad_norm": 2.6377150914839675,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 1975
+    },
+    {
+      "epoch": 0.12663419635990772,
+      "grad_norm": 2.8380873293773226,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1266982824916688,
+      "grad_norm": 2.677246943205837,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 1977
+    },
+    {
+      "epoch": 0.12676236862342988,
+      "grad_norm": 2.5127787900075402,
+      "learning_rate": 1e-06,
+      "loss": 0.3851,
+      "step": 1978
+    },
+    {
+      "epoch": 0.12682645475519097,
+      "grad_norm": 2.4144112563671793,
+      "learning_rate": 1e-06,
+      "loss": 0.4408,
+      "step": 1979
+    },
+    {
+      "epoch": 0.12689054088695206,
+      "grad_norm": 2.592467710784995,
+      "learning_rate": 1e-06,
+      "loss": 0.4076,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12695462701871316,
+      "grad_norm": 2.702943687034177,
+      "learning_rate": 1e-06,
+      "loss": 0.4539,
+      "step": 1981
+    },
+    {
+      "epoch": 0.12701871315047425,
+      "grad_norm": 2.555904697910308,
+      "learning_rate": 1e-06,
+      "loss": 0.3962,
+      "step": 1982
+    },
+    {
+      "epoch": 0.12708279928223531,
+      "grad_norm": 2.6617695924556775,
+      "learning_rate": 1e-06,
+      "loss": 0.4315,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1271468854139964,
+      "grad_norm": 2.5865121829306155,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1272109715457575,
+      "grad_norm": 2.4733075250800067,
+      "learning_rate": 1e-06,
+      "loss": 0.4381,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1272750576775186,
+      "grad_norm": 2.7177984913035558,
+      "learning_rate": 1e-06,
+      "loss": 0.4293,
+      "step": 1986
+    },
+    {
+      "epoch": 0.12733914380927966,
+      "grad_norm": 2.68776734487849,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 1987
+    },
+    {
+      "epoch": 0.12740322994104075,
+      "grad_norm": 2.619007004179668,
+      "learning_rate": 1e-06,
+      "loss": 0.4024,
+      "step": 1988
+    },
+    {
+      "epoch": 0.12746731607280185,
+      "grad_norm": 2.65050880017839,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 1989
+    },
+    {
+      "epoch": 0.12753140220456294,
+      "grad_norm": 2.6588622202293077,
+      "learning_rate": 1e-06,
+      "loss": 0.4207,
+      "step": 1990
+    },
+    {
+      "epoch": 0.12759548833632403,
+      "grad_norm": 2.708910525725563,
+      "learning_rate": 1e-06,
+      "loss": 0.4158,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 2.5687777946786285,
+      "learning_rate": 1e-06,
+      "loss": 0.3429,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1277236605998462,
+      "grad_norm": 2.8445974840088857,
+      "learning_rate": 1e-06,
+      "loss": 0.3966,
+      "step": 1993
+    },
+    {
+      "epoch": 0.12778774673160728,
+      "grad_norm": 2.4812860715534737,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 1994
+    },
+    {
+      "epoch": 0.12785183286336838,
+      "grad_norm": 2.5067894831823274,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 1995
+    },
+    {
+      "epoch": 0.12791591899512944,
+      "grad_norm": 3.120477088832285,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 1996
+    },
+    {
+      "epoch": 0.12798000512689053,
+      "grad_norm": 2.68728983577007,
+      "learning_rate": 1e-06,
+      "loss": 0.3998,
+      "step": 1997
+    },
+    {
+      "epoch": 0.12804409125865163,
+      "grad_norm": 2.5708000190381006,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 1998
+    },
+    {
+      "epoch": 0.12810817739041272,
+      "grad_norm": 2.816934859878245,
+      "learning_rate": 1e-06,
+      "loss": 0.4359,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1281722635221738,
+      "grad_norm": 2.7079638576190903,
+      "learning_rate": 1e-06,
+      "loss": 0.4471,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12823634965393488,
+      "grad_norm": 2.5800674482671635,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 2001
+    },
+    {
+      "epoch": 0.12830043578569597,
+      "grad_norm": 2.4591428885818507,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 2002
+    },
+    {
+      "epoch": 0.12836452191745706,
+      "grad_norm": 2.6827533850410337,
+      "learning_rate": 1e-06,
+      "loss": 0.3498,
+      "step": 2003
+    },
+    {
+      "epoch": 0.12842860804921816,
+      "grad_norm": 2.709188331291668,
+      "learning_rate": 1e-06,
+      "loss": 0.4272,
+      "step": 2004
+    },
+    {
+      "epoch": 0.12849269418097922,
+      "grad_norm": 2.499208486274247,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 2005
+    },
+    {
+      "epoch": 0.12855678031274032,
+      "grad_norm": 2.6909496268940214,
+      "learning_rate": 1e-06,
+      "loss": 0.4524,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1286208664445014,
+      "grad_norm": 2.802587101271318,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1286849525762625,
+      "grad_norm": 2.4601661186060673,
+      "learning_rate": 1e-06,
+      "loss": 0.3981,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1287490387080236,
+      "grad_norm": 2.654138037986819,
+      "learning_rate": 1e-06,
+      "loss": 0.3885,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12881312483978466,
+      "grad_norm": 2.8370521642356876,
+      "learning_rate": 1e-06,
+      "loss": 0.434,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12887721097154575,
+      "grad_norm": 2.5252795408324964,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 2011
+    },
+    {
+      "epoch": 0.12894129710330685,
+      "grad_norm": 2.4799437101808866,
+      "learning_rate": 1e-06,
+      "loss": 0.3828,
+      "step": 2012
+    },
+    {
+      "epoch": 0.12900538323506794,
+      "grad_norm": 2.7494549099164614,
+      "learning_rate": 1e-06,
+      "loss": 0.4594,
+      "step": 2013
+    },
+    {
+      "epoch": 0.12906946936682903,
+      "grad_norm": 2.4795137457784926,
+      "learning_rate": 1e-06,
+      "loss": 0.4005,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1291335554985901,
+      "grad_norm": 2.7520680465218175,
+      "learning_rate": 1e-06,
+      "loss": 0.4466,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1291976416303512,
+      "grad_norm": 2.722284636249325,
+      "learning_rate": 1e-06,
+      "loss": 0.3849,
+      "step": 2016
+    },
+    {
+      "epoch": 0.12926172776211228,
+      "grad_norm": 2.6524413526819237,
+      "learning_rate": 1e-06,
+      "loss": 0.4324,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12932581389387338,
+      "grad_norm": 2.66292985459627,
+      "learning_rate": 1e-06,
+      "loss": 0.4615,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12938990002563444,
+      "grad_norm": 2.6159139687609576,
+      "learning_rate": 1e-06,
+      "loss": 0.4115,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12945398615739553,
+      "grad_norm": 2.7173243701716823,
+      "learning_rate": 1e-06,
+      "loss": 0.4108,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12951807228915663,
+      "grad_norm": 2.7996012785431357,
+      "learning_rate": 1e-06,
+      "loss": 0.418,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12958215842091772,
+      "grad_norm": 2.6421635527383063,
+      "learning_rate": 1e-06,
+      "loss": 0.5005,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1296462445526788,
+      "grad_norm": 2.701234561198063,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12971033068443988,
+      "grad_norm": 2.5982870193962793,
+      "learning_rate": 1e-06,
+      "loss": 0.4336,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12977441681620097,
+      "grad_norm": 3.0523581053313165,
+      "learning_rate": 1e-06,
+      "loss": 0.4677,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12983850294796206,
+      "grad_norm": 2.498398555961267,
+      "learning_rate": 1e-06,
+      "loss": 0.4339,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12990258907972316,
+      "grad_norm": 2.6602278770071623,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12996667521148422,
+      "grad_norm": 2.4848270044860645,
+      "learning_rate": 1e-06,
+      "loss": 0.4618,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13003076134324532,
+      "grad_norm": 2.6426560908155805,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1300948474750064,
+      "grad_norm": 2.685119281293971,
+      "learning_rate": 1e-06,
+      "loss": 0.4239,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1301589336067675,
+      "grad_norm": 2.653893626434556,
+      "learning_rate": 1e-06,
+      "loss": 0.4626,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1302230197385286,
+      "grad_norm": 2.693815977647124,
+      "learning_rate": 1e-06,
+      "loss": 0.4073,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13028710587028966,
+      "grad_norm": 2.7742930167368316,
+      "learning_rate": 1e-06,
+      "loss": 0.4358,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13035119200205075,
+      "grad_norm": 2.4521871241228532,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13041527813381185,
+      "grad_norm": 2.5778926021918656,
+      "learning_rate": 1e-06,
+      "loss": 0.4419,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13047936426557294,
+      "grad_norm": 2.5233030916242307,
+      "learning_rate": 1e-06,
+      "loss": 0.4635,
+      "step": 2036
+    },
+    {
+      "epoch": 0.130543450397334,
+      "grad_norm": 2.7769656735415227,
+      "learning_rate": 1e-06,
+      "loss": 0.4503,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1306075365290951,
+      "grad_norm": 2.6061311223083528,
+      "learning_rate": 1e-06,
+      "loss": 0.4037,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1306716226608562,
+      "grad_norm": 2.6441536424858416,
+      "learning_rate": 1e-06,
+      "loss": 0.4084,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13073570879261728,
+      "grad_norm": 2.646813268165125,
+      "learning_rate": 1e-06,
+      "loss": 0.404,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13079979492437838,
+      "grad_norm": 2.718568711377987,
+      "learning_rate": 1e-06,
+      "loss": 0.3933,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13086388105613944,
+      "grad_norm": 2.4331819490330076,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13092796718790053,
+      "grad_norm": 2.5525577204591787,
+      "learning_rate": 1e-06,
+      "loss": 0.4228,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13099205331966163,
+      "grad_norm": 2.591680461417647,
+      "learning_rate": 1e-06,
+      "loss": 0.3663,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13105613945142272,
+      "grad_norm": 2.7473732680920464,
+      "learning_rate": 1e-06,
+      "loss": 0.4217,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13112022558318379,
+      "grad_norm": 2.512421236575192,
+      "learning_rate": 1e-06,
+      "loss": 0.4137,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13118431171494488,
+      "grad_norm": 2.745025052977934,
+      "learning_rate": 1e-06,
+      "loss": 0.4289,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13124839784670597,
+      "grad_norm": 2.6990060791654327,
+      "learning_rate": 1e-06,
+      "loss": 0.4014,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13131248397846706,
+      "grad_norm": 2.603256256896644,
+      "learning_rate": 1e-06,
+      "loss": 0.4655,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13137657011022816,
+      "grad_norm": 2.6494921431521687,
+      "learning_rate": 1e-06,
+      "loss": 0.3774,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13144065624198922,
+      "grad_norm": 2.8543194864950965,
+      "learning_rate": 1e-06,
+      "loss": 0.4587,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13150474237375032,
+      "grad_norm": 2.7739591966852166,
+      "learning_rate": 1e-06,
+      "loss": 0.3972,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1315688285055114,
+      "grad_norm": 2.4982817995739444,
+      "learning_rate": 1e-06,
+      "loss": 0.3863,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1316329146372725,
+      "grad_norm": 2.6312308133622286,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13169700076903357,
+      "grad_norm": 3.0636204926026425,
+      "learning_rate": 1e-06,
+      "loss": 0.4028,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13176108690079466,
+      "grad_norm": 2.7556518659223626,
+      "learning_rate": 1e-06,
+      "loss": 0.4212,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13182517303255575,
+      "grad_norm": 2.7385712365433936,
+      "learning_rate": 1e-06,
+      "loss": 0.4222,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13188925916431685,
+      "grad_norm": 2.533880989808742,
+      "learning_rate": 1e-06,
+      "loss": 0.4315,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13195334529607794,
+      "grad_norm": 2.426764055890877,
+      "learning_rate": 1e-06,
+      "loss": 0.4003,
+      "step": 2059
+    },
+    {
+      "epoch": 0.132017431427839,
+      "grad_norm": 2.5769365252329246,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1320815175596001,
+      "grad_norm": 2.5672902578483447,
+      "learning_rate": 1e-06,
+      "loss": 0.3979,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1321456036913612,
+      "grad_norm": 2.727113752369044,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13220968982312228,
+      "grad_norm": 2.598448088421263,
+      "learning_rate": 1e-06,
+      "loss": 0.3973,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13227377595488338,
+      "grad_norm": 2.808346442547483,
+      "learning_rate": 1e-06,
+      "loss": 0.3907,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13233786208664444,
+      "grad_norm": 2.5537562254410817,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13240194821840554,
+      "grad_norm": 2.507844085434063,
+      "learning_rate": 1e-06,
+      "loss": 0.3912,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13246603435016663,
+      "grad_norm": 2.6236845142825556,
+      "learning_rate": 1e-06,
+      "loss": 0.3805,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13253012048192772,
+      "grad_norm": 2.5809839489336666,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1325942066136888,
+      "grad_norm": 2.5456272047937096,
+      "learning_rate": 1e-06,
+      "loss": 0.3996,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13265829274544988,
+      "grad_norm": 2.7029859623452235,
+      "learning_rate": 1e-06,
+      "loss": 0.3932,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13272237887721097,
+      "grad_norm": 3.113893501887302,
+      "learning_rate": 1e-06,
+      "loss": 0.4189,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13278646500897207,
+      "grad_norm": 2.895251806197545,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13285055114073316,
+      "grad_norm": 2.692808938525751,
+      "learning_rate": 1e-06,
+      "loss": 0.3822,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13291463727249422,
+      "grad_norm": 2.666108543162362,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13297872340425532,
+      "grad_norm": 2.628539864622391,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1330428095360164,
+      "grad_norm": 2.6457888313754854,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1331068956677775,
+      "grad_norm": 2.73962603052359,
+      "learning_rate": 1e-06,
+      "loss": 0.4013,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13317098179953857,
+      "grad_norm": 2.73176395407253,
+      "learning_rate": 1e-06,
+      "loss": 0.4025,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13323506793129966,
+      "grad_norm": 2.698908843828074,
+      "learning_rate": 1e-06,
+      "loss": 0.4199,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13329915406306075,
+      "grad_norm": 2.717511171326012,
+      "learning_rate": 1e-06,
+      "loss": 0.4055,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13336324019482185,
+      "grad_norm": 2.663032680662114,
+      "learning_rate": 1e-06,
+      "loss": 0.4314,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13342732632658294,
+      "grad_norm": 2.67227144538236,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 2082
+    },
+    {
+      "epoch": 0.133491412458344,
+      "grad_norm": 2.7192184119233347,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1335554985901051,
+      "grad_norm": 2.80607421933027,
+      "learning_rate": 1e-06,
+      "loss": 0.442,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1336195847218662,
+      "grad_norm": 2.4282833747159365,
+      "learning_rate": 1e-06,
+      "loss": 0.3403,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13368367085362728,
+      "grad_norm": 2.779501854115342,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13374775698538835,
+      "grad_norm": 2.605891767733823,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13381184311714944,
+      "grad_norm": 2.4367002776811315,
+      "learning_rate": 1e-06,
+      "loss": 0.4342,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13387592924891054,
+      "grad_norm": 2.462742753214077,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13394001538067163,
+      "grad_norm": 2.717385353303187,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13400410151243272,
+      "grad_norm": 2.5506542547042583,
+      "learning_rate": 1e-06,
+      "loss": 0.3751,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1340681876441938,
+      "grad_norm": 2.48170579062675,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13413227377595488,
+      "grad_norm": 2.5466247727125646,
+      "learning_rate": 1e-06,
+      "loss": 0.3816,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13419635990771597,
+      "grad_norm": 2.792497217280625,
+      "learning_rate": 1e-06,
+      "loss": 0.4217,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13426044603947707,
+      "grad_norm": 2.8605300019220246,
+      "learning_rate": 1e-06,
+      "loss": 0.4964,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13432453217123813,
+      "grad_norm": 2.5038991233158363,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13438861830299922,
+      "grad_norm": 2.868839229735299,
+      "learning_rate": 1e-06,
+      "loss": 0.4488,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13445270443476032,
+      "grad_norm": 2.545144883407565,
+      "learning_rate": 1e-06,
+      "loss": 0.4652,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1345167905665214,
+      "grad_norm": 2.5425577896010565,
+      "learning_rate": 1e-06,
+      "loss": 0.3585,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1345808766982825,
+      "grad_norm": 2.6384130176202962,
+      "learning_rate": 1e-06,
+      "loss": 0.4988,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13464496283004357,
+      "grad_norm": 2.6027499158121215,
+      "learning_rate": 1e-06,
+      "loss": 0.4604,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13470904896180466,
+      "grad_norm": 2.6048158509628614,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13477313509356575,
+      "grad_norm": 2.8293034564567248,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13483722122532685,
+      "grad_norm": 2.592521713422449,
+      "learning_rate": 1e-06,
+      "loss": 0.4517,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1349013073570879,
+      "grad_norm": 2.6693886764853407,
+      "learning_rate": 1e-06,
+      "loss": 0.3943,
+      "step": 2105
+    },
+    {
+      "epoch": 0.134965393488849,
+      "grad_norm": 2.504808703738117,
+      "learning_rate": 1e-06,
+      "loss": 0.3935,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1350294796206101,
+      "grad_norm": 2.907032906917195,
+      "learning_rate": 1e-06,
+      "loss": 0.4508,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1350935657523712,
+      "grad_norm": 2.679610851400339,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13515765188413228,
+      "grad_norm": 2.748289961872713,
+      "learning_rate": 1e-06,
+      "loss": 0.4391,
+      "step": 2109
+    },
+    {
+      "epoch": 0.13522173801589335,
+      "grad_norm": 2.7234870239625493,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13528582414765444,
+      "grad_norm": 2.7140754867383023,
+      "learning_rate": 1e-06,
+      "loss": 0.4414,
+      "step": 2111
+    },
+    {
+      "epoch": 0.13534991027941554,
+      "grad_norm": 2.6496771127358527,
+      "learning_rate": 1e-06,
+      "loss": 0.4467,
+      "step": 2112
+    },
+    {
+      "epoch": 0.13541399641117663,
+      "grad_norm": 2.730252973494939,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1354780825429377,
+      "grad_norm": 2.655324313984564,
+      "learning_rate": 1e-06,
+      "loss": 0.4708,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1355421686746988,
+      "grad_norm": 2.8728419806844454,
+      "learning_rate": 1e-06,
+      "loss": 0.473,
+      "step": 2115
+    },
+    {
+      "epoch": 0.13560625480645988,
+      "grad_norm": 2.5073844860924184,
+      "learning_rate": 1e-06,
+      "loss": 0.3807,
+      "step": 2116
+    },
+    {
+      "epoch": 0.13567034093822097,
+      "grad_norm": 2.60753256323996,
+      "learning_rate": 1e-06,
+      "loss": 0.4838,
+      "step": 2117
+    },
+    {
+      "epoch": 0.13573442706998207,
+      "grad_norm": 2.7082299427616907,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 2118
+    },
+    {
+      "epoch": 0.13579851320174313,
+      "grad_norm": 2.4118401233619253,
+      "learning_rate": 1e-06,
+      "loss": 0.4567,
+      "step": 2119
+    },
+    {
+      "epoch": 0.13586259933350422,
+      "grad_norm": 2.4643325105841503,
+      "learning_rate": 1e-06,
+      "loss": 0.4006,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13592668546526532,
+      "grad_norm": 2.5275100662919554,
+      "learning_rate": 1e-06,
+      "loss": 0.4766,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1359907715970264,
+      "grad_norm": 2.655859438369582,
+      "learning_rate": 1e-06,
+      "loss": 0.4157,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1360548577287875,
+      "grad_norm": 2.6427058505139187,
+      "learning_rate": 1e-06,
+      "loss": 0.3924,
+      "step": 2123
+    },
+    {
+      "epoch": 0.13611894386054857,
+      "grad_norm": 2.6632577787547,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 2124
+    },
+    {
+      "epoch": 0.13618302999230966,
+      "grad_norm": 2.6287929102425003,
+      "learning_rate": 1e-06,
+      "loss": 0.3918,
+      "step": 2125
+    },
+    {
+      "epoch": 0.13624711612407076,
+      "grad_norm": 2.8016371088295475,
+      "learning_rate": 1e-06,
+      "loss": 0.4356,
+      "step": 2126
+    },
+    {
+      "epoch": 0.13631120225583185,
+      "grad_norm": 3.3386006555368057,
+      "learning_rate": 1e-06,
+      "loss": 0.3933,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1363752883875929,
+      "grad_norm": 2.7097965814412226,
+      "learning_rate": 1e-06,
+      "loss": 0.4021,
+      "step": 2128
+    },
+    {
+      "epoch": 0.136439374519354,
+      "grad_norm": 2.7644699066410983,
+      "learning_rate": 1e-06,
+      "loss": 0.4464,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1365034606511151,
+      "grad_norm": 2.4771892454096793,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1365675467828762,
+      "grad_norm": 2.760564997369234,
+      "learning_rate": 1e-06,
+      "loss": 0.3566,
+      "step": 2131
+    },
+    {
+      "epoch": 0.13663163291463729,
+      "grad_norm": 2.661760302416408,
+      "learning_rate": 1e-06,
+      "loss": 0.5215,
+      "step": 2132
+    },
+    {
+      "epoch": 0.13669571904639835,
+      "grad_norm": 2.623671240870864,
+      "learning_rate": 1e-06,
+      "loss": 0.4042,
+      "step": 2133
+    },
+    {
+      "epoch": 0.13675980517815944,
+      "grad_norm": 2.9925947621163673,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 2134
+    },
+    {
+      "epoch": 0.13682389130992054,
+      "grad_norm": 2.7417367216554207,
+      "learning_rate": 1e-06,
+      "loss": 0.4468,
+      "step": 2135
+    },
+    {
+      "epoch": 0.13688797744168163,
+      "grad_norm": 2.669990654570271,
+      "learning_rate": 1e-06,
+      "loss": 0.4087,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1369520635734427,
+      "grad_norm": 2.8533638654559805,
+      "learning_rate": 1e-06,
+      "loss": 0.4808,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1370161497052038,
+      "grad_norm": 2.7643977807163793,
+      "learning_rate": 1e-06,
+      "loss": 0.3754,
+      "step": 2138
+    },
+    {
+      "epoch": 0.13708023583696488,
+      "grad_norm": 2.536365532104418,
+      "learning_rate": 1e-06,
+      "loss": 0.3998,
+      "step": 2139
+    },
+    {
+      "epoch": 0.13714432196872597,
+      "grad_norm": 2.637414739617339,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13720840810048707,
+      "grad_norm": 2.5290361793526963,
+      "learning_rate": 1e-06,
+      "loss": 0.4058,
+      "step": 2141
+    },
+    {
+      "epoch": 0.13727249423224813,
+      "grad_norm": 2.600362258487816,
+      "learning_rate": 1e-06,
+      "loss": 0.3776,
+      "step": 2142
+    },
+    {
+      "epoch": 0.13733658036400923,
+      "grad_norm": 2.5573978268464623,
+      "learning_rate": 1e-06,
+      "loss": 0.3914,
+      "step": 2143
+    },
+    {
+      "epoch": 0.13740066649577032,
+      "grad_norm": 2.680943994912968,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1374647526275314,
+      "grad_norm": 2.6925962306306146,
+      "learning_rate": 1e-06,
+      "loss": 0.4573,
+      "step": 2145
+    },
+    {
+      "epoch": 0.13752883875929248,
+      "grad_norm": 2.689427306845986,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 2146
+    },
+    {
+      "epoch": 0.13759292489105357,
+      "grad_norm": 2.7293189543709766,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 2147
+    },
+    {
+      "epoch": 0.13765701102281466,
+      "grad_norm": 2.8777615193581676,
+      "learning_rate": 1e-06,
+      "loss": 0.4442,
+      "step": 2148
+    },
+    {
+      "epoch": 0.13772109715457576,
+      "grad_norm": 2.6444638826899634,
+      "learning_rate": 1e-06,
+      "loss": 0.4287,
+      "step": 2149
+    },
+    {
+      "epoch": 0.13778518328633685,
+      "grad_norm": 2.7759250640439217,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 2150
+    },
+    {
+      "epoch": 0.13784926941809791,
+      "grad_norm": 2.616339376376333,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 2151
+    },
+    {
+      "epoch": 0.137913355549859,
+      "grad_norm": 2.538884599658828,
+      "learning_rate": 1e-06,
+      "loss": 0.4364,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1379774416816201,
+      "grad_norm": 2.75943965508161,
+      "learning_rate": 1e-06,
+      "loss": 0.404,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1380415278133812,
+      "grad_norm": 2.5818273891275867,
+      "learning_rate": 1e-06,
+      "loss": 0.455,
+      "step": 2154
+    },
+    {
+      "epoch": 0.13810561394514226,
+      "grad_norm": 2.9759449959751176,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 2155
+    },
+    {
+      "epoch": 0.13816970007690335,
+      "grad_norm": 2.595854159746892,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2156
+    },
+    {
+      "epoch": 0.13823378620866444,
+      "grad_norm": 2.970107649172304,
+      "learning_rate": 1e-06,
+      "loss": 0.3991,
+      "step": 2157
+    },
+    {
+      "epoch": 0.13829787234042554,
+      "grad_norm": 2.5620174383001,
+      "learning_rate": 1e-06,
+      "loss": 0.3853,
+      "step": 2158
+    },
+    {
+      "epoch": 0.13836195847218663,
+      "grad_norm": 2.682996794741283,
+      "learning_rate": 1e-06,
+      "loss": 0.3926,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1384260446039477,
+      "grad_norm": 2.640623080908026,
+      "learning_rate": 1e-06,
+      "loss": 0.4437,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1384901307357088,
+      "grad_norm": 2.562450172958601,
+      "learning_rate": 1e-06,
+      "loss": 0.4001,
+      "step": 2161
+    },
+    {
+      "epoch": 0.13855421686746988,
+      "grad_norm": 2.625741101049455,
+      "learning_rate": 1e-06,
+      "loss": 0.4063,
+      "step": 2162
+    },
+    {
+      "epoch": 0.13861830299923097,
+      "grad_norm": 2.926628566586209,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 2163
+    },
+    {
+      "epoch": 0.13868238913099204,
+      "grad_norm": 2.532631132337974,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 2164
+    },
+    {
+      "epoch": 0.13874647526275313,
+      "grad_norm": 2.41138765437668,
+      "learning_rate": 1e-06,
+      "loss": 0.3966,
+      "step": 2165
+    },
+    {
+      "epoch": 0.13881056139451423,
+      "grad_norm": 2.7163462717552247,
+      "learning_rate": 1e-06,
+      "loss": 0.4605,
+      "step": 2166
+    },
+    {
+      "epoch": 0.13887464752627532,
+      "grad_norm": 2.7349179779173682,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1389387336580364,
+      "grad_norm": 2.705844709704702,
+      "learning_rate": 1e-06,
+      "loss": 0.3875,
+      "step": 2168
+    },
+    {
+      "epoch": 0.13900281978979748,
+      "grad_norm": 3.077466061475745,
+      "learning_rate": 1e-06,
+      "loss": 0.461,
+      "step": 2169
+    },
+    {
+      "epoch": 0.13906690592155857,
+      "grad_norm": 2.7230463846895163,
+      "learning_rate": 1e-06,
+      "loss": 0.4436,
+      "step": 2170
+    },
+    {
+      "epoch": 0.13913099205331966,
+      "grad_norm": 2.8623601827737355,
+      "learning_rate": 1e-06,
+      "loss": 0.4651,
+      "step": 2171
+    },
+    {
+      "epoch": 0.13919507818508076,
+      "grad_norm": 2.681961203424956,
+      "learning_rate": 1e-06,
+      "loss": 0.3852,
+      "step": 2172
+    },
+    {
+      "epoch": 0.13925916431684185,
+      "grad_norm": 2.646391464509139,
+      "learning_rate": 1e-06,
+      "loss": 0.3898,
+      "step": 2173
+    },
+    {
+      "epoch": 0.13932325044860291,
+      "grad_norm": 2.7593356672279805,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 2174
+    },
+    {
+      "epoch": 0.139387336580364,
+      "grad_norm": 2.742273416019991,
+      "learning_rate": 1e-06,
+      "loss": 0.4334,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1394514227121251,
+      "grad_norm": 2.609949808118975,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1395155088438862,
+      "grad_norm": 2.688282306305331,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 2177
+    },
+    {
+      "epoch": 0.13957959497564726,
+      "grad_norm": 2.664368149793294,
+      "learning_rate": 1e-06,
+      "loss": 0.3988,
+      "step": 2178
+    },
+    {
+      "epoch": 0.13964368110740835,
+      "grad_norm": 2.5754163962200782,
+      "learning_rate": 1e-06,
+      "loss": 0.3572,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13970776723916944,
+      "grad_norm": 2.7535770236048647,
+      "learning_rate": 1e-06,
+      "loss": 0.4486,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13977185337093054,
+      "grad_norm": 2.5586750005179812,
+      "learning_rate": 1e-06,
+      "loss": 0.4398,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13983593950269163,
+      "grad_norm": 2.5801312411637634,
+      "learning_rate": 1e-06,
+      "loss": 0.4103,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1399000256344527,
+      "grad_norm": 2.6418380770398233,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1399641117662138,
+      "grad_norm": 2.8406851039525205,
+      "learning_rate": 1e-06,
+      "loss": 0.4637,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14002819789797488,
+      "grad_norm": 2.6629292430069524,
+      "learning_rate": 1e-06,
+      "loss": 0.4501,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14009228402973598,
+      "grad_norm": 2.654798177824672,
+      "learning_rate": 1e-06,
+      "loss": 0.4054,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14015637016149704,
+      "grad_norm": 2.9168291610332195,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14022045629325813,
+      "grad_norm": 2.600267663410967,
+      "learning_rate": 1e-06,
+      "loss": 0.4414,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14028454242501923,
+      "grad_norm": 2.5417581238364906,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14034862855678032,
+      "grad_norm": 2.454596888462331,
+      "learning_rate": 1e-06,
+      "loss": 0.3628,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1404127146885414,
+      "grad_norm": 2.8569961486821867,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14047680082030248,
+      "grad_norm": 2.8200076888696706,
+      "learning_rate": 1e-06,
+      "loss": 0.3914,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14054088695206357,
+      "grad_norm": 2.7538541707394315,
+      "learning_rate": 1e-06,
+      "loss": 0.3932,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14060497308382466,
+      "grad_norm": 2.756896816994633,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14066905921558576,
+      "grad_norm": 2.387845561333487,
+      "learning_rate": 1e-06,
+      "loss": 0.418,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14073314534734682,
+      "grad_norm": 2.7006071057783876,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14079723147910791,
+      "grad_norm": 2.6467935141543086,
+      "learning_rate": 1e-06,
+      "loss": 0.4038,
+      "step": 2197
+    },
+    {
+      "epoch": 0.140861317610869,
+      "grad_norm": 2.6198363896498282,
+      "learning_rate": 1e-06,
+      "loss": 0.425,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1409254037426301,
+      "grad_norm": 2.5636411622366264,
+      "learning_rate": 1e-06,
+      "loss": 0.4789,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1409894898743912,
+      "grad_norm": 2.4631479300062114,
+      "learning_rate": 1e-06,
+      "loss": 0.4166,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14105357600615226,
+      "grad_norm": 2.691147262921431,
+      "learning_rate": 1e-06,
+      "loss": 0.4125,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14111766213791335,
+      "grad_norm": 2.653751571043374,
+      "learning_rate": 1e-06,
+      "loss": 0.4041,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14118174826967445,
+      "grad_norm": 2.457050148938937,
+      "learning_rate": 1e-06,
+      "loss": 0.3652,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14124583440143554,
+      "grad_norm": 2.3973561946298103,
+      "learning_rate": 1e-06,
+      "loss": 0.3735,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1413099205331966,
+      "grad_norm": 2.7407883027389155,
+      "learning_rate": 1e-06,
+      "loss": 0.4125,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1413740066649577,
+      "grad_norm": 2.6688132463232317,
+      "learning_rate": 1e-06,
+      "loss": 0.4421,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1414380927967188,
+      "grad_norm": 2.643644919200164,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14150217892847988,
+      "grad_norm": 2.65080615124718,
+      "learning_rate": 1e-06,
+      "loss": 0.4559,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14156626506024098,
+      "grad_norm": 2.776708320652998,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14163035119200204,
+      "grad_norm": 2.7020700688413513,
+      "learning_rate": 1e-06,
+      "loss": 0.4608,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14169443732376313,
+      "grad_norm": 2.4552041930115904,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14175852345552423,
+      "grad_norm": 2.5767945349775854,
+      "learning_rate": 1e-06,
+      "loss": 0.401,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14182260958728532,
+      "grad_norm": 2.6559174118536046,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14188669571904639,
+      "grad_norm": 2.7023876372387985,
+      "learning_rate": 1e-06,
+      "loss": 0.4348,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14195078185080748,
+      "grad_norm": 2.610694095494199,
+      "learning_rate": 1e-06,
+      "loss": 0.4084,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14201486798256857,
+      "grad_norm": 2.6285159451990623,
+      "learning_rate": 1e-06,
+      "loss": 0.4375,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14207895411432966,
+      "grad_norm": 2.640627442850314,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14214304024609076,
+      "grad_norm": 2.6134124413492756,
+      "learning_rate": 1e-06,
+      "loss": 0.4073,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14220712637785182,
+      "grad_norm": 2.6094590815898457,
+      "learning_rate": 1e-06,
+      "loss": 0.4346,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14227121250961292,
+      "grad_norm": 2.6490584464866247,
+      "learning_rate": 1e-06,
+      "loss": 0.4433,
+      "step": 2220
+    },
+    {
+      "epoch": 0.142335298641374,
+      "grad_norm": 2.5338550549194956,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1423993847731351,
+      "grad_norm": 2.6873280903762553,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14246347090489617,
+      "grad_norm": 2.882287571490444,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14252755703665726,
+      "grad_norm": 2.6621684855736145,
+      "learning_rate": 1e-06,
+      "loss": 0.4098,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14259164316841835,
+      "grad_norm": 2.5399516145610406,
+      "learning_rate": 1e-06,
+      "loss": 0.4198,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14265572930017945,
+      "grad_norm": 5.304960869663389,
+      "learning_rate": 1e-06,
+      "loss": 0.4114,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14271981543194054,
+      "grad_norm": 2.6465131106356563,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1427839015637016,
+      "grad_norm": 2.823785893975285,
+      "learning_rate": 1e-06,
+      "loss": 0.35,
+      "step": 2228
+    },
+    {
+      "epoch": 0.1428479876954627,
+      "grad_norm": 2.6252988135275728,
+      "learning_rate": 1e-06,
+      "loss": 0.4251,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1429120738272238,
+      "grad_norm": 2.5715196655427395,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14297615995898488,
+      "grad_norm": 2.5374478270268215,
+      "learning_rate": 1e-06,
+      "loss": 0.392,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14304024609074598,
+      "grad_norm": 2.693764343659205,
+      "learning_rate": 1e-06,
+      "loss": 0.4185,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14310433222250704,
+      "grad_norm": 2.5538527593794145,
+      "learning_rate": 1e-06,
+      "loss": 0.4159,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14316841835426813,
+      "grad_norm": 2.643995614450662,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14323250448602923,
+      "grad_norm": 2.525937144073715,
+      "learning_rate": 1e-06,
+      "loss": 0.4341,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14329659061779032,
+      "grad_norm": 2.451146215447884,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14336067674955139,
+      "grad_norm": 2.9775043321883623,
+      "learning_rate": 1e-06,
+      "loss": 0.4192,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14342476288131248,
+      "grad_norm": 2.6372943700404314,
+      "learning_rate": 1e-06,
+      "loss": 0.3781,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14348884901307357,
+      "grad_norm": 2.681809857926576,
+      "learning_rate": 1e-06,
+      "loss": 0.3501,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14355293514483466,
+      "grad_norm": 2.6330684169629888,
+      "learning_rate": 1e-06,
+      "loss": 0.4125,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14361702127659576,
+      "grad_norm": 2.5635021280630146,
+      "learning_rate": 1e-06,
+      "loss": 0.4494,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14368110740835682,
+      "grad_norm": 2.5420088820932087,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14374519354011792,
+      "grad_norm": 2.5737910098068064,
+      "learning_rate": 1e-06,
+      "loss": 0.3646,
+      "step": 2243
+    },
+    {
+      "epoch": 0.143809279671879,
+      "grad_norm": 2.824372043653846,
+      "learning_rate": 1e-06,
+      "loss": 0.4076,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1438733658036401,
+      "grad_norm": 2.4381851437931386,
+      "learning_rate": 1e-06,
+      "loss": 0.377,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14393745193540117,
+      "grad_norm": 2.508371697735693,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14400153806716226,
+      "grad_norm": 2.720177214825913,
+      "learning_rate": 1e-06,
+      "loss": 0.4186,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14406562419892335,
+      "grad_norm": 2.6059820662859936,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14412971033068445,
+      "grad_norm": 2.6734684323623865,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14419379646244554,
+      "grad_norm": 2.71227587875518,
+      "learning_rate": 1e-06,
+      "loss": 0.3907,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1442578825942066,
+      "grad_norm": 2.63605657042964,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 2251
+    },
+    {
+      "epoch": 0.1443219687259677,
+      "grad_norm": 2.596026537683679,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1443860548577288,
+      "grad_norm": 2.6134149718188047,
+      "learning_rate": 1e-06,
+      "loss": 0.4782,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14445014098948988,
+      "grad_norm": 2.692257498204221,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14451422712125095,
+      "grad_norm": 2.4847848114936006,
+      "learning_rate": 1e-06,
+      "loss": 0.3496,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14457831325301204,
+      "grad_norm": 2.642340400044151,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14464239938477313,
+      "grad_norm": 2.8905906794526293,
+      "learning_rate": 1e-06,
+      "loss": 0.4426,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14470648551653423,
+      "grad_norm": 2.6608506377012247,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14477057164829532,
+      "grad_norm": 2.6869403811241064,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1448346577800564,
+      "grad_norm": 2.512352252639373,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14489874391181748,
+      "grad_norm": 2.622494628319042,
+      "learning_rate": 1e-06,
+      "loss": 0.4209,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14496283004357857,
+      "grad_norm": 2.7770276907918974,
+      "learning_rate": 1e-06,
+      "loss": 0.4578,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14502691617533967,
+      "grad_norm": 2.666828543397899,
+      "learning_rate": 1e-06,
+      "loss": 0.3942,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14509100230710073,
+      "grad_norm": 2.765447332478381,
+      "learning_rate": 1e-06,
+      "loss": 0.4545,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14515508843886182,
+      "grad_norm": 2.4732133303198967,
+      "learning_rate": 1e-06,
+      "loss": 0.4159,
+      "step": 2265
+    },
+    {
+      "epoch": 0.14521917457062292,
+      "grad_norm": 2.715075164848874,
+      "learning_rate": 1e-06,
+      "loss": 0.4377,
+      "step": 2266
+    },
+    {
+      "epoch": 0.145283260702384,
+      "grad_norm": 2.647220900812757,
+      "learning_rate": 1e-06,
+      "loss": 0.414,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1453473468341451,
+      "grad_norm": 2.799960861199678,
+      "learning_rate": 1e-06,
+      "loss": 0.3925,
+      "step": 2268
+    },
+    {
+      "epoch": 0.14541143296590617,
+      "grad_norm": 2.4458853797244573,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 2269
+    },
+    {
+      "epoch": 0.14547551909766726,
+      "grad_norm": 2.6759906456774503,
+      "learning_rate": 1e-06,
+      "loss": 0.4692,
+      "step": 2270
+    },
+    {
+      "epoch": 0.14553960522942835,
+      "grad_norm": 2.570308914424076,
+      "learning_rate": 1e-06,
+      "loss": 0.3719,
+      "step": 2271
+    },
+    {
+      "epoch": 0.14560369136118945,
+      "grad_norm": 2.4318165071673215,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1456677774929505,
+      "grad_norm": 2.747814037626672,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1457318636247116,
+      "grad_norm": 2.6257289323903743,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1457959497564727,
+      "grad_norm": 2.4752797780084985,
+      "learning_rate": 1e-06,
+      "loss": 0.4473,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1458600358882338,
+      "grad_norm": 2.3987218490349473,
+      "learning_rate": 1e-06,
+      "loss": 0.3736,
+      "step": 2276
+    },
+    {
+      "epoch": 0.14592412201999488,
+      "grad_norm": 2.8867281314680997,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 2277
+    },
+    {
+      "epoch": 0.14598820815175595,
+      "grad_norm": 2.7716528014364146,
+      "learning_rate": 1e-06,
+      "loss": 0.3906,
+      "step": 2278
+    },
+    {
+      "epoch": 0.14605229428351704,
+      "grad_norm": 2.8974582112907363,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 2279
+    },
+    {
+      "epoch": 0.14611638041527814,
+      "grad_norm": 2.880131925690081,
+      "learning_rate": 1e-06,
+      "loss": 0.4175,
+      "step": 2280
+    },
+    {
+      "epoch": 0.14618046654703923,
+      "grad_norm": 2.8585496372328723,
+      "learning_rate": 1e-06,
+      "loss": 0.3892,
+      "step": 2281
+    },
+    {
+      "epoch": 0.14624455267880032,
+      "grad_norm": 2.4339756990712744,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1463086388105614,
+      "grad_norm": 2.5992029115648316,
+      "learning_rate": 1e-06,
+      "loss": 0.3717,
+      "step": 2283
+    },
+    {
+      "epoch": 0.14637272494232248,
+      "grad_norm": 2.797688498581479,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 2284
+    },
+    {
+      "epoch": 0.14643681107408357,
+      "grad_norm": 2.4833588908902557,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 2285
+    },
+    {
+      "epoch": 0.14650089720584467,
+      "grad_norm": 2.58446876123646,
+      "learning_rate": 1e-06,
+      "loss": 0.3839,
+      "step": 2286
+    },
+    {
+      "epoch": 0.14656498333760573,
+      "grad_norm": 2.606881969638494,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 2287
+    },
+    {
+      "epoch": 0.14662906946936682,
+      "grad_norm": 2.6604866091293475,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 2288
+    },
+    {
+      "epoch": 0.14669315560112792,
+      "grad_norm": 2.6989920957953246,
+      "learning_rate": 1e-06,
+      "loss": 0.4315,
+      "step": 2289
+    },
+    {
+      "epoch": 0.146757241732889,
+      "grad_norm": 2.725493235912062,
+      "learning_rate": 1e-06,
+      "loss": 0.3907,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1468213278646501,
+      "grad_norm": 2.69132660874459,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 2291
+    },
+    {
+      "epoch": 0.14688541399641117,
+      "grad_norm": 2.610103572519288,
+      "learning_rate": 1e-06,
+      "loss": 0.4651,
+      "step": 2292
+    },
+    {
+      "epoch": 0.14694950012817226,
+      "grad_norm": 2.4991515605638455,
+      "learning_rate": 1e-06,
+      "loss": 0.4226,
+      "step": 2293
+    },
+    {
+      "epoch": 0.14701358625993335,
+      "grad_norm": 2.538974285721927,
+      "learning_rate": 1e-06,
+      "loss": 0.4055,
+      "step": 2294
+    },
+    {
+      "epoch": 0.14707767239169445,
+      "grad_norm": 2.8066736531056464,
+      "learning_rate": 1e-06,
+      "loss": 0.387,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1471417585234555,
+      "grad_norm": 2.855173144029738,
+      "learning_rate": 1e-06,
+      "loss": 0.4252,
+      "step": 2296
+    },
+    {
+      "epoch": 0.1472058446552166,
+      "grad_norm": 2.8441153351557293,
+      "learning_rate": 1e-06,
+      "loss": 0.3698,
+      "step": 2297
+    },
+    {
+      "epoch": 0.1472699307869777,
+      "grad_norm": 2.6372999046635357,
+      "learning_rate": 1e-06,
+      "loss": 0.4374,
+      "step": 2298
+    },
+    {
+      "epoch": 0.1473340169187388,
+      "grad_norm": 2.561119349595809,
+      "learning_rate": 1e-06,
+      "loss": 0.4233,
+      "step": 2299
+    },
+    {
+      "epoch": 0.14739810305049988,
+      "grad_norm": 2.6508050362160485,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.14746218918226095,
+      "grad_norm": 2.627610268613044,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 2301
+    },
+    {
+      "epoch": 0.14752627531402204,
+      "grad_norm": 2.811581016043246,
+      "learning_rate": 1e-06,
+      "loss": 0.4672,
+      "step": 2302
+    },
+    {
+      "epoch": 0.14759036144578314,
+      "grad_norm": 2.8176789877873993,
+      "learning_rate": 1e-06,
+      "loss": 0.44,
+      "step": 2303
+    },
+    {
+      "epoch": 0.14765444757754423,
+      "grad_norm": 2.6136882447196848,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1477185337093053,
+      "grad_norm": 2.631618330649024,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 2305
+    },
+    {
+      "epoch": 0.1477826198410664,
+      "grad_norm": 2.8307020910722445,
+      "learning_rate": 1e-06,
+      "loss": 0.445,
+      "step": 2306
+    },
+    {
+      "epoch": 0.14784670597282748,
+      "grad_norm": 2.859794204166223,
+      "learning_rate": 1e-06,
+      "loss": 0.3921,
+      "step": 2307
+    },
+    {
+      "epoch": 0.14791079210458857,
+      "grad_norm": 3.029930250277341,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 2308
+    },
+    {
+      "epoch": 0.14797487823634967,
+      "grad_norm": 2.769615320491857,
+      "learning_rate": 1e-06,
+      "loss": 0.467,
+      "step": 2309
+    },
+    {
+      "epoch": 0.14803896436811073,
+      "grad_norm": 2.8163409163128414,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 2310
+    },
+    {
+      "epoch": 0.14810305049987182,
+      "grad_norm": 2.9181548672688438,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 2311
+    },
+    {
+      "epoch": 0.14816713663163292,
+      "grad_norm": 2.515786063454079,
+      "learning_rate": 1e-06,
+      "loss": 0.396,
+      "step": 2312
+    },
+    {
+      "epoch": 0.148231222763394,
+      "grad_norm": 2.6501859655841136,
+      "learning_rate": 1e-06,
+      "loss": 0.3726,
+      "step": 2313
+    },
+    {
+      "epoch": 0.14829530889515508,
+      "grad_norm": 2.727336603534406,
+      "learning_rate": 1e-06,
+      "loss": 0.3959,
+      "step": 2314
+    },
+    {
+      "epoch": 0.14835939502691617,
+      "grad_norm": 2.7322829607818786,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 2315
+    },
+    {
+      "epoch": 0.14842348115867726,
+      "grad_norm": 2.4220749790430496,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 2316
+    },
+    {
+      "epoch": 0.14848756729043835,
+      "grad_norm": 2.6338981053065598,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 2317
+    },
+    {
+      "epoch": 0.14855165342219945,
+      "grad_norm": 2.773079409430958,
+      "learning_rate": 1e-06,
+      "loss": 0.4168,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1486157395539605,
+      "grad_norm": 2.5455067815807704,
+      "learning_rate": 1e-06,
+      "loss": 0.4192,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1486798256857216,
+      "grad_norm": 2.5151484665388555,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1487439118174827,
+      "grad_norm": 2.6818970061144043,
+      "learning_rate": 1e-06,
+      "loss": 0.4192,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1488079979492438,
+      "grad_norm": 2.831179594461404,
+      "learning_rate": 1e-06,
+      "loss": 0.4075,
+      "step": 2322
+    },
+    {
+      "epoch": 0.14887208408100486,
+      "grad_norm": 2.8459896863335987,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 2323
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 2.393045917617131,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 2324
+    },
+    {
+      "epoch": 0.14900025634452704,
+      "grad_norm": 2.592468826362506,
+      "learning_rate": 1e-06,
+      "loss": 0.448,
+      "step": 2325
+    },
+    {
+      "epoch": 0.14906434247628814,
+      "grad_norm": 2.5695432484816036,
+      "learning_rate": 1e-06,
+      "loss": 0.4202,
+      "step": 2326
+    },
+    {
+      "epoch": 0.14912842860804923,
+      "grad_norm": 2.7467870217923966,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2327
+    },
+    {
+      "epoch": 0.1491925147398103,
+      "grad_norm": 2.7219181762815765,
+      "learning_rate": 1e-06,
+      "loss": 0.3861,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1492566008715714,
+      "grad_norm": 2.6606309439094575,
+      "learning_rate": 1e-06,
+      "loss": 0.3892,
+      "step": 2329
+    },
+    {
+      "epoch": 0.14932068700333248,
+      "grad_norm": 2.4477701980906903,
+      "learning_rate": 1e-06,
+      "loss": 0.3845,
+      "step": 2330
+    },
+    {
+      "epoch": 0.14938477313509357,
+      "grad_norm": 2.5710900386400555,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 2331
+    },
+    {
+      "epoch": 0.14944885926685464,
+      "grad_norm": 2.9202947763246847,
+      "learning_rate": 1e-06,
+      "loss": 0.4593,
+      "step": 2332
+    },
+    {
+      "epoch": 0.14951294539861573,
+      "grad_norm": 2.8278221339598812,
+      "learning_rate": 1e-06,
+      "loss": 0.4257,
+      "step": 2333
+    },
+    {
+      "epoch": 0.14957703153037682,
+      "grad_norm": 2.8006536711305854,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 2334
+    },
+    {
+      "epoch": 0.14964111766213792,
+      "grad_norm": 2.5025824047107026,
+      "learning_rate": 1e-06,
+      "loss": 0.4401,
+      "step": 2335
+    },
+    {
+      "epoch": 0.149705203793899,
+      "grad_norm": 2.647992832106811,
+      "learning_rate": 1e-06,
+      "loss": 0.3943,
+      "step": 2336
+    },
+    {
+      "epoch": 0.14976928992566008,
+      "grad_norm": 2.534047410212519,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 2337
+    },
+    {
+      "epoch": 0.14983337605742117,
+      "grad_norm": 2.544201360530356,
+      "learning_rate": 1e-06,
+      "loss": 0.4222,
+      "step": 2338
+    },
+    {
+      "epoch": 0.14989746218918226,
+      "grad_norm": 2.598334654859618,
+      "learning_rate": 1e-06,
+      "loss": 0.4778,
+      "step": 2339
+    },
+    {
+      "epoch": 0.14996154832094336,
+      "grad_norm": 2.6037963285409806,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15002563445270445,
+      "grad_norm": 2.5511618288184152,
+      "learning_rate": 1e-06,
+      "loss": 0.4308,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1500897205844655,
+      "grad_norm": 2.596673832687979,
+      "learning_rate": 1e-06,
+      "loss": 0.392,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1501538067162266,
+      "grad_norm": 2.5935959167733493,
+      "learning_rate": 1e-06,
+      "loss": 0.3799,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1502178928479877,
+      "grad_norm": 2.8157616888072776,
+      "learning_rate": 1e-06,
+      "loss": 0.3875,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1502819789797488,
+      "grad_norm": 2.7239095516119964,
+      "learning_rate": 1e-06,
+      "loss": 0.3579,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15034606511150986,
+      "grad_norm": 2.8732628594396914,
+      "learning_rate": 1e-06,
+      "loss": 0.4426,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15041015124327095,
+      "grad_norm": 2.6053261102088596,
+      "learning_rate": 1e-06,
+      "loss": 0.4361,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15047423737503204,
+      "grad_norm": 2.7501538950837006,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15053832350679314,
+      "grad_norm": 2.6748933692579016,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15060240963855423,
+      "grad_norm": 2.5140175391129467,
+      "learning_rate": 1e-06,
+      "loss": 0.396,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1506664957703153,
+      "grad_norm": 2.51220473120615,
+      "learning_rate": 1e-06,
+      "loss": 0.4299,
+      "step": 2351
+    },
+    {
+      "epoch": 0.1507305819020764,
+      "grad_norm": 2.5533477749431985,
+      "learning_rate": 1e-06,
+      "loss": 0.4048,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15079466803383748,
+      "grad_norm": 2.7622224664376276,
+      "learning_rate": 1e-06,
+      "loss": 0.3982,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15085875416559857,
+      "grad_norm": 2.745535809237263,
+      "learning_rate": 1e-06,
+      "loss": 0.4543,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15092284029735964,
+      "grad_norm": 2.7300386583191885,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15098692642912073,
+      "grad_norm": 2.7017399198890613,
+      "learning_rate": 1e-06,
+      "loss": 0.3976,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15105101256088183,
+      "grad_norm": 2.641428987366016,
+      "learning_rate": 1e-06,
+      "loss": 0.4643,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15111509869264292,
+      "grad_norm": 2.6569814357613186,
+      "learning_rate": 1e-06,
+      "loss": 0.4498,
+      "step": 2358
+    },
+    {
+      "epoch": 0.151179184824404,
+      "grad_norm": 2.5401208849561416,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15124327095616508,
+      "grad_norm": 2.5054149135493136,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15130735708792617,
+      "grad_norm": 2.6694292857092905,
+      "learning_rate": 1e-06,
+      "loss": 0.4528,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15137144321968726,
+      "grad_norm": 2.5114041686713646,
+      "learning_rate": 1e-06,
+      "loss": 0.4483,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15143552935144836,
+      "grad_norm": 2.6480189800249803,
+      "learning_rate": 1e-06,
+      "loss": 0.4174,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15149961548320942,
+      "grad_norm": 2.4534126730482573,
+      "learning_rate": 1e-06,
+      "loss": 0.3561,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15156370161497051,
+      "grad_norm": 2.6671292785945866,
+      "learning_rate": 1e-06,
+      "loss": 0.4201,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1516277877467316,
+      "grad_norm": 2.726065605093534,
+      "learning_rate": 1e-06,
+      "loss": 0.4679,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1516918738784927,
+      "grad_norm": 2.545323534174155,
+      "learning_rate": 1e-06,
+      "loss": 0.401,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1517559600102538,
+      "grad_norm": 2.791402105635689,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15182004614201486,
+      "grad_norm": 2.6468669705500565,
+      "learning_rate": 1e-06,
+      "loss": 0.4129,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15188413227377595,
+      "grad_norm": 2.5476833598403528,
+      "learning_rate": 1e-06,
+      "loss": 0.4278,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15194821840553704,
+      "grad_norm": 2.506687865478058,
+      "learning_rate": 1e-06,
+      "loss": 0.3836,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15201230453729814,
+      "grad_norm": 2.750669796319302,
+      "learning_rate": 1e-06,
+      "loss": 0.4487,
+      "step": 2372
+    },
+    {
+      "epoch": 0.1520763906690592,
+      "grad_norm": 2.6545688375212264,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1521404768008203,
+      "grad_norm": 2.5413046028139235,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 2374
+    },
+    {
+      "epoch": 0.1522045629325814,
+      "grad_norm": 2.554444200234787,
+      "learning_rate": 1e-06,
+      "loss": 0.395,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15226864906434248,
+      "grad_norm": 2.579414741406172,
+      "learning_rate": 1e-06,
+      "loss": 0.3627,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15233273519610357,
+      "grad_norm": 2.6671368075780446,
+      "learning_rate": 1e-06,
+      "loss": 0.4369,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15239682132786464,
+      "grad_norm": 2.6810079593267346,
+      "learning_rate": 1e-06,
+      "loss": 0.4162,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15246090745962573,
+      "grad_norm": 2.683487202560093,
+      "learning_rate": 1e-06,
+      "loss": 0.4557,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15252499359138683,
+      "grad_norm": 2.589917499641482,
+      "learning_rate": 1e-06,
+      "loss": 0.3939,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15258907972314792,
+      "grad_norm": 2.8205977099230477,
+      "learning_rate": 1e-06,
+      "loss": 0.4398,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15265316585490898,
+      "grad_norm": 2.592663049668745,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15271725198667008,
+      "grad_norm": 2.5460240252707265,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15278133811843117,
+      "grad_norm": 2.5645214681159696,
+      "learning_rate": 1e-06,
+      "loss": 0.3777,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15284542425019226,
+      "grad_norm": 2.7015014505643604,
+      "learning_rate": 1e-06,
+      "loss": 0.4084,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15290951038195336,
+      "grad_norm": 2.6132337605579186,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15297359651371442,
+      "grad_norm": 2.565897987852246,
+      "learning_rate": 1e-06,
+      "loss": 0.4156,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15303768264547551,
+      "grad_norm": 2.5809983715596645,
+      "learning_rate": 1e-06,
+      "loss": 0.3894,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1531017687772366,
+      "grad_norm": 2.5560680323288376,
+      "learning_rate": 1e-06,
+      "loss": 0.3725,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1531658549089977,
+      "grad_norm": 2.6934001340487073,
+      "learning_rate": 1e-06,
+      "loss": 0.4946,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1532299410407588,
+      "grad_norm": 2.4350422689075937,
+      "learning_rate": 1e-06,
+      "loss": 0.3635,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15329402717251986,
+      "grad_norm": 2.5978625164002436,
+      "learning_rate": 1e-06,
+      "loss": 0.3755,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15335811330428095,
+      "grad_norm": 2.7060153807077065,
+      "learning_rate": 1e-06,
+      "loss": 0.3844,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15342219943604204,
+      "grad_norm": 2.5977471758004533,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15348628556780314,
+      "grad_norm": 2.8192146974315007,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 2395
+    },
+    {
+      "epoch": 0.1535503716995642,
+      "grad_norm": 2.8507772617205984,
+      "learning_rate": 1e-06,
+      "loss": 0.4593,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1536144578313253,
+      "grad_norm": 2.757912230882947,
+      "learning_rate": 1e-06,
+      "loss": 0.376,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1536785439630864,
+      "grad_norm": 2.914634273117128,
+      "learning_rate": 1e-06,
+      "loss": 0.3884,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15374263009484748,
+      "grad_norm": 2.7028190039958164,
+      "learning_rate": 1e-06,
+      "loss": 0.429,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15380671622660858,
+      "grad_norm": 2.805975220284593,
+      "learning_rate": 1e-06,
+      "loss": 0.4336,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15387080235836964,
+      "grad_norm": 2.612350568348285,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15393488849013073,
+      "grad_norm": 2.6235487697848012,
+      "learning_rate": 1e-06,
+      "loss": 0.3941,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15399897462189183,
+      "grad_norm": 2.5075163297958416,
+      "learning_rate": 1e-06,
+      "loss": 0.443,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15406306075365292,
+      "grad_norm": 2.5941254002232106,
+      "learning_rate": 1e-06,
+      "loss": 0.4505,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15412714688541398,
+      "grad_norm": 2.646111568843153,
+      "learning_rate": 1e-06,
+      "loss": 0.4259,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15419123301717508,
+      "grad_norm": 2.627844708101551,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15425531914893617,
+      "grad_norm": 2.6155267950712315,
+      "learning_rate": 1e-06,
+      "loss": 0.3833,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15431940528069726,
+      "grad_norm": 2.80650626799486,
+      "learning_rate": 1e-06,
+      "loss": 0.4245,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15438349141245836,
+      "grad_norm": 2.61101795676931,
+      "learning_rate": 1e-06,
+      "loss": 0.359,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15444757754421942,
+      "grad_norm": 2.707694683825272,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15451166367598052,
+      "grad_norm": 2.5831471926173624,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 2411
+    },
+    {
+      "epoch": 0.1545757498077416,
+      "grad_norm": 2.6242498641680334,
+      "learning_rate": 1e-06,
+      "loss": 0.448,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1546398359395027,
+      "grad_norm": 2.72655226953322,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15470392207126377,
+      "grad_norm": 2.7383815110827583,
+      "learning_rate": 1e-06,
+      "loss": 0.4047,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15476800820302486,
+      "grad_norm": 2.5026739932341155,
+      "learning_rate": 1e-06,
+      "loss": 0.3867,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15483209433478595,
+      "grad_norm": 2.674607203930584,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15489618046654705,
+      "grad_norm": 2.754064762987339,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15496026659830814,
+      "grad_norm": 2.570533524925514,
+      "learning_rate": 1e-06,
+      "loss": 0.4598,
+      "step": 2418
+    },
+    {
+      "epoch": 0.1550243527300692,
+      "grad_norm": 2.597622703572672,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1550884388618303,
+      "grad_norm": 2.8288481023355776,
+      "learning_rate": 1e-06,
+      "loss": 0.4526,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1551525249935914,
+      "grad_norm": 2.4192969540719527,
+      "learning_rate": 1e-06,
+      "loss": 0.4239,
+      "step": 2421
+    },
+    {
+      "epoch": 0.15521661112535248,
+      "grad_norm": 2.570303124194049,
+      "learning_rate": 1e-06,
+      "loss": 0.4859,
+      "step": 2422
+    },
+    {
+      "epoch": 0.15528069725711355,
+      "grad_norm": 2.5962623992860463,
+      "learning_rate": 1e-06,
+      "loss": 0.414,
+      "step": 2423
+    },
+    {
+      "epoch": 0.15534478338887464,
+      "grad_norm": 2.6407161187844586,
+      "learning_rate": 1e-06,
+      "loss": 0.4539,
+      "step": 2424
+    },
+    {
+      "epoch": 0.15540886952063573,
+      "grad_norm": 2.7032853850977983,
+      "learning_rate": 1e-06,
+      "loss": 0.4852,
+      "step": 2425
+    },
+    {
+      "epoch": 0.15547295565239683,
+      "grad_norm": 2.6186822936277574,
+      "learning_rate": 1e-06,
+      "loss": 0.4252,
+      "step": 2426
+    },
+    {
+      "epoch": 0.15553704178415792,
+      "grad_norm": 2.7692792260713768,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2427
+    },
+    {
+      "epoch": 0.15560112791591899,
+      "grad_norm": 2.819212145403548,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2428
+    },
+    {
+      "epoch": 0.15566521404768008,
+      "grad_norm": 2.7820635144861376,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 2429
+    },
+    {
+      "epoch": 0.15572930017944117,
+      "grad_norm": 2.540855770044114,
+      "learning_rate": 1e-06,
+      "loss": 0.3847,
+      "step": 2430
+    },
+    {
+      "epoch": 0.15579338631120226,
+      "grad_norm": 2.5881439472756202,
+      "learning_rate": 1e-06,
+      "loss": 0.3905,
+      "step": 2431
+    },
+    {
+      "epoch": 0.15585747244296333,
+      "grad_norm": 2.628112799287706,
+      "learning_rate": 1e-06,
+      "loss": 0.4897,
+      "step": 2432
+    },
+    {
+      "epoch": 0.15592155857472442,
+      "grad_norm": 2.8470469351320666,
+      "learning_rate": 1e-06,
+      "loss": 0.4372,
+      "step": 2433
+    },
+    {
+      "epoch": 0.15598564470648552,
+      "grad_norm": 2.959718617534213,
+      "learning_rate": 1e-06,
+      "loss": 0.3556,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1560497308382466,
+      "grad_norm": 2.63349545605712,
+      "learning_rate": 1e-06,
+      "loss": 0.3966,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1561138169700077,
+      "grad_norm": 2.6920193790596687,
+      "learning_rate": 1e-06,
+      "loss": 0.4002,
+      "step": 2436
+    },
+    {
+      "epoch": 0.15617790310176877,
+      "grad_norm": 2.5515125234377916,
+      "learning_rate": 1e-06,
+      "loss": 0.4704,
+      "step": 2437
+    },
+    {
+      "epoch": 0.15624198923352986,
+      "grad_norm": 2.7845164985771738,
+      "learning_rate": 1e-06,
+      "loss": 0.3741,
+      "step": 2438
+    },
+    {
+      "epoch": 0.15630607536529095,
+      "grad_norm": 2.7051788981886715,
+      "learning_rate": 1e-06,
+      "loss": 0.4248,
+      "step": 2439
+    },
+    {
+      "epoch": 0.15637016149705205,
+      "grad_norm": 2.5027319502036103,
+      "learning_rate": 1e-06,
+      "loss": 0.4275,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1564342476288131,
+      "grad_norm": 2.67872865736227,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 2441
+    },
+    {
+      "epoch": 0.1564983337605742,
+      "grad_norm": 2.6234605610607797,
+      "learning_rate": 1e-06,
+      "loss": 0.4122,
+      "step": 2442
+    },
+    {
+      "epoch": 0.1565624198923353,
+      "grad_norm": 2.660891907681123,
+      "learning_rate": 1e-06,
+      "loss": 0.4079,
+      "step": 2443
+    },
+    {
+      "epoch": 0.1566265060240964,
+      "grad_norm": 2.469269141511082,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 2444
+    },
+    {
+      "epoch": 0.15669059215585748,
+      "grad_norm": 2.9128965615675084,
+      "learning_rate": 1e-06,
+      "loss": 0.4011,
+      "step": 2445
+    },
+    {
+      "epoch": 0.15675467828761855,
+      "grad_norm": 2.3634946944799338,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 2446
+    },
+    {
+      "epoch": 0.15681876441937964,
+      "grad_norm": 2.487803849131197,
+      "learning_rate": 1e-06,
+      "loss": 0.381,
+      "step": 2447
+    },
+    {
+      "epoch": 0.15688285055114073,
+      "grad_norm": 2.73280436377226,
+      "learning_rate": 1e-06,
+      "loss": 0.4247,
+      "step": 2448
+    },
+    {
+      "epoch": 0.15694693668290183,
+      "grad_norm": 2.645994090911685,
+      "learning_rate": 1e-06,
+      "loss": 0.4245,
+      "step": 2449
+    },
+    {
+      "epoch": 0.15701102281466292,
+      "grad_norm": 2.7045428079749887,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 2450
+    },
+    {
+      "epoch": 0.15707510894642399,
+      "grad_norm": 2.6721213176800283,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 2451
+    },
+    {
+      "epoch": 0.15713919507818508,
+      "grad_norm": 2.5446333446323575,
+      "learning_rate": 1e-06,
+      "loss": 0.4501,
+      "step": 2452
+    },
+    {
+      "epoch": 0.15720328120994617,
+      "grad_norm": 2.7025878221632214,
+      "learning_rate": 1e-06,
+      "loss": 0.4365,
+      "step": 2453
+    },
+    {
+      "epoch": 0.15726736734170726,
+      "grad_norm": 2.4377600454198673,
+      "learning_rate": 1e-06,
+      "loss": 0.393,
+      "step": 2454
+    },
+    {
+      "epoch": 0.15733145347346833,
+      "grad_norm": 2.3395968765113704,
+      "learning_rate": 1e-06,
+      "loss": 0.4026,
+      "step": 2455
+    },
+    {
+      "epoch": 0.15739553960522942,
+      "grad_norm": 2.5508080145151895,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 2456
+    },
+    {
+      "epoch": 0.15745962573699052,
+      "grad_norm": 2.9131013972018893,
+      "learning_rate": 1e-06,
+      "loss": 0.406,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1575237118687516,
+      "grad_norm": 2.6926586987036005,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1575877980005127,
+      "grad_norm": 2.661373242133559,
+      "learning_rate": 1e-06,
+      "loss": 0.4544,
+      "step": 2459
+    },
+    {
+      "epoch": 0.15765188413227377,
+      "grad_norm": 2.865019871633759,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 2460
+    },
+    {
+      "epoch": 0.15771597026403486,
+      "grad_norm": 2.7673917992905857,
+      "learning_rate": 1e-06,
+      "loss": 0.3693,
+      "step": 2461
+    },
+    {
+      "epoch": 0.15778005639579595,
+      "grad_norm": 2.6496428816256183,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 2462
+    },
+    {
+      "epoch": 0.15784414252755705,
+      "grad_norm": 2.5833209913336788,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 2463
+    },
+    {
+      "epoch": 0.1579082286593181,
+      "grad_norm": 2.9927054051853594,
+      "learning_rate": 1e-06,
+      "loss": 0.4455,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1579723147910792,
+      "grad_norm": 2.688891946243902,
+      "learning_rate": 1e-06,
+      "loss": 0.4572,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1580364009228403,
+      "grad_norm": 2.8014701175146275,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 2466
+    },
+    {
+      "epoch": 0.1581004870546014,
+      "grad_norm": 2.6209516151437016,
+      "learning_rate": 1e-06,
+      "loss": 0.4293,
+      "step": 2467
+    },
+    {
+      "epoch": 0.15816457318636248,
+      "grad_norm": 2.6968176829286077,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 2468
+    },
+    {
+      "epoch": 0.15822865931812355,
+      "grad_norm": 2.716996681830155,
+      "learning_rate": 1e-06,
+      "loss": 0.4425,
+      "step": 2469
+    },
+    {
+      "epoch": 0.15829274544988464,
+      "grad_norm": 2.7238337604748373,
+      "learning_rate": 1e-06,
+      "loss": 0.4307,
+      "step": 2470
+    },
+    {
+      "epoch": 0.15835683158164574,
+      "grad_norm": 2.739047909106007,
+      "learning_rate": 1e-06,
+      "loss": 0.4299,
+      "step": 2471
+    },
+    {
+      "epoch": 0.15842091771340683,
+      "grad_norm": 2.526469438645234,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 2472
+    },
+    {
+      "epoch": 0.1584850038451679,
+      "grad_norm": 2.6278675097925763,
+      "learning_rate": 1e-06,
+      "loss": 0.4373,
+      "step": 2473
+    },
+    {
+      "epoch": 0.158549089976929,
+      "grad_norm": 2.5826759075688055,
+      "learning_rate": 1e-06,
+      "loss": 0.4258,
+      "step": 2474
+    },
+    {
+      "epoch": 0.15861317610869008,
+      "grad_norm": 2.77590367710701,
+      "learning_rate": 1e-06,
+      "loss": 0.4196,
+      "step": 2475
+    },
+    {
+      "epoch": 0.15867726224045117,
+      "grad_norm": 2.7835205766483893,
+      "learning_rate": 1e-06,
+      "loss": 0.3705,
+      "step": 2476
+    },
+    {
+      "epoch": 0.15874134837221227,
+      "grad_norm": 2.863480380954405,
+      "learning_rate": 1e-06,
+      "loss": 0.4371,
+      "step": 2477
+    },
+    {
+      "epoch": 0.15880543450397333,
+      "grad_norm": 2.8855879891557374,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 2478
+    },
+    {
+      "epoch": 0.15886952063573442,
+      "grad_norm": 2.721920994882,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 2479
+    },
+    {
+      "epoch": 0.15893360676749552,
+      "grad_norm": 2.6621503936105686,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1589976928992566,
+      "grad_norm": 2.625862393236496,
+      "learning_rate": 1e-06,
+      "loss": 0.4072,
+      "step": 2481
+    },
+    {
+      "epoch": 0.15906177903101767,
+      "grad_norm": 2.698437048521841,
+      "learning_rate": 1e-06,
+      "loss": 0.4212,
+      "step": 2482
+    },
+    {
+      "epoch": 0.15912586516277877,
+      "grad_norm": 2.747135808007895,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 2483
+    },
+    {
+      "epoch": 0.15918995129453986,
+      "grad_norm": 2.6924238591324436,
+      "learning_rate": 1e-06,
+      "loss": 0.4419,
+      "step": 2484
+    },
+    {
+      "epoch": 0.15925403742630095,
+      "grad_norm": 2.864874679432206,
+      "learning_rate": 1e-06,
+      "loss": 0.4116,
+      "step": 2485
+    },
+    {
+      "epoch": 0.15931812355806205,
+      "grad_norm": 2.584090760115717,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1593822096898231,
+      "grad_norm": 2.3501320571795565,
+      "learning_rate": 1e-06,
+      "loss": 0.4432,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1594462958215842,
+      "grad_norm": 2.62447790662663,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1595103819533453,
+      "grad_norm": 2.809122378803803,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1595744680851064,
+      "grad_norm": 2.560302012329596,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 2490
+    },
+    {
+      "epoch": 0.15963855421686746,
+      "grad_norm": 2.7315313434872364,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 2491
+    },
+    {
+      "epoch": 0.15970264034862855,
+      "grad_norm": 2.7487724870084707,
+      "learning_rate": 1e-06,
+      "loss": 0.4438,
+      "step": 2492
+    },
+    {
+      "epoch": 0.15976672648038964,
+      "grad_norm": 2.5211532370360317,
+      "learning_rate": 1e-06,
+      "loss": 0.4091,
+      "step": 2493
+    },
+    {
+      "epoch": 0.15983081261215074,
+      "grad_norm": 2.69198583244194,
+      "learning_rate": 1e-06,
+      "loss": 0.3919,
+      "step": 2494
+    },
+    {
+      "epoch": 0.15989489874391183,
+      "grad_norm": 2.6710258015363175,
+      "learning_rate": 1e-06,
+      "loss": 0.3612,
+      "step": 2495
+    },
+    {
+      "epoch": 0.1599589848756729,
+      "grad_norm": 2.6456233748049787,
+      "learning_rate": 1e-06,
+      "loss": 0.4136,
+      "step": 2496
+    },
+    {
+      "epoch": 0.160023071007434,
+      "grad_norm": 2.5859243386442037,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16008715713919508,
+      "grad_norm": 2.7520692263098416,
+      "learning_rate": 1e-06,
+      "loss": 0.4324,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16015124327095617,
+      "grad_norm": 2.7816574091909723,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16021532940271727,
+      "grad_norm": 2.6344550709051044,
+      "learning_rate": 1e-06,
+      "loss": 0.4278,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16027941553447833,
+      "grad_norm": 3.197606757914543,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16034350166623942,
+      "grad_norm": 2.6471620383147676,
+      "learning_rate": 1e-06,
+      "loss": 0.4028,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16040758779800052,
+      "grad_norm": 2.882593526216847,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1604716739297616,
+      "grad_norm": 2.6876055696015246,
+      "learning_rate": 1e-06,
+      "loss": 0.4112,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16053576006152268,
+      "grad_norm": 2.6908732798019157,
+      "learning_rate": 1e-06,
+      "loss": 0.4084,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16059984619328377,
+      "grad_norm": 2.696421366995068,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16066393232504486,
+      "grad_norm": 2.5708012883800184,
+      "learning_rate": 1e-06,
+      "loss": 0.4442,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16072801845680595,
+      "grad_norm": 2.5577193706830594,
+      "learning_rate": 1e-06,
+      "loss": 0.3862,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16079210458856705,
+      "grad_norm": 2.7187413971424266,
+      "learning_rate": 1e-06,
+      "loss": 0.4327,
+      "step": 2509
+    },
+    {
+      "epoch": 0.1608561907203281,
+      "grad_norm": 2.5817179456414037,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1609202768520892,
+      "grad_norm": 2.7269113290519593,
+      "learning_rate": 1e-06,
+      "loss": 0.4325,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1609843629838503,
+      "grad_norm": 2.572578117655859,
+      "learning_rate": 1e-06,
+      "loss": 0.3851,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1610484491156114,
+      "grad_norm": 2.6450180571019937,
+      "learning_rate": 1e-06,
+      "loss": 0.4411,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16111253524737246,
+      "grad_norm": 2.5789710591909545,
+      "learning_rate": 1e-06,
+      "loss": 0.4102,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16117662137913355,
+      "grad_norm": 2.4838537947972177,
+      "learning_rate": 1e-06,
+      "loss": 0.405,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16124070751089464,
+      "grad_norm": 2.534907881809524,
+      "learning_rate": 1e-06,
+      "loss": 0.3469,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16130479364265574,
+      "grad_norm": 2.500800182833469,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16136887977441683,
+      "grad_norm": 2.6929560857938974,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2518
+    },
+    {
+      "epoch": 0.1614329659061779,
+      "grad_norm": 2.6113591154458287,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 2519
+    },
+    {
+      "epoch": 0.161497052037939,
+      "grad_norm": 2.537840773620322,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16156113816970008,
+      "grad_norm": 2.7707118455203767,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16162522430146117,
+      "grad_norm": 2.7225252383237275,
+      "learning_rate": 1e-06,
+      "loss": 0.3935,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16168931043322224,
+      "grad_norm": 2.828852689766831,
+      "learning_rate": 1e-06,
+      "loss": 0.4267,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16175339656498333,
+      "grad_norm": 2.603720737428551,
+      "learning_rate": 1e-06,
+      "loss": 0.4425,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16181748269674442,
+      "grad_norm": 2.5478451431697033,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16188156882850552,
+      "grad_norm": 2.661807124509187,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 2526
+    },
+    {
+      "epoch": 0.1619456549602666,
+      "grad_norm": 2.6041595365534604,
+      "learning_rate": 1e-06,
+      "loss": 0.3774,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16200974109202768,
+      "grad_norm": 2.570551916458256,
+      "learning_rate": 1e-06,
+      "loss": 0.4089,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16207382722378877,
+      "grad_norm": 2.691936352986267,
+      "learning_rate": 1e-06,
+      "loss": 0.3846,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16213791335554986,
+      "grad_norm": 2.630771561034015,
+      "learning_rate": 1e-06,
+      "loss": 0.4067,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16220199948731095,
+      "grad_norm": 2.812901706927222,
+      "learning_rate": 1e-06,
+      "loss": 0.4493,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16226608561907202,
+      "grad_norm": 2.895255542589514,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1623301717508331,
+      "grad_norm": 2.689542921264475,
+      "learning_rate": 1e-06,
+      "loss": 0.4611,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1623942578825942,
+      "grad_norm": 2.7578900917283318,
+      "learning_rate": 1e-06,
+      "loss": 0.4068,
+      "step": 2534
+    },
+    {
+      "epoch": 0.1624583440143553,
+      "grad_norm": 2.7909225863700104,
+      "learning_rate": 1e-06,
+      "loss": 0.4066,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1625224301461164,
+      "grad_norm": 2.7296440103447535,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16258651627787746,
+      "grad_norm": 2.7413253180770747,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16265060240963855,
+      "grad_norm": 2.968023303872132,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16271468854139964,
+      "grad_norm": 2.5240106724224285,
+      "learning_rate": 1e-06,
+      "loss": 0.4667,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16277877467316074,
+      "grad_norm": 2.8291621239977434,
+      "learning_rate": 1e-06,
+      "loss": 0.4637,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1628428608049218,
+      "grad_norm": 2.685586301138915,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1629069469366829,
+      "grad_norm": 2.7011354343541827,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 2542
+    },
+    {
+      "epoch": 0.162971033068444,
+      "grad_norm": 2.652397024902033,
+      "learning_rate": 1e-06,
+      "loss": 0.454,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16303511920020508,
+      "grad_norm": 2.977816057807566,
+      "learning_rate": 1e-06,
+      "loss": 0.4084,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16309920533196617,
+      "grad_norm": 2.7882639087445233,
+      "learning_rate": 1e-06,
+      "loss": 0.4856,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16316329146372724,
+      "grad_norm": 2.5045074481642526,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16322737759548833,
+      "grad_norm": 2.617452267392621,
+      "learning_rate": 1e-06,
+      "loss": 0.4572,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16329146372724943,
+      "grad_norm": 2.658894789581529,
+      "learning_rate": 1e-06,
+      "loss": 0.4675,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16335554985901052,
+      "grad_norm": 2.852357932664824,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16341963599077158,
+      "grad_norm": 2.6360746418969017,
+      "learning_rate": 1e-06,
+      "loss": 0.3874,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16348372212253268,
+      "grad_norm": 2.8766537155424214,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16354780825429377,
+      "grad_norm": 2.625135910693668,
+      "learning_rate": 1e-06,
+      "loss": 0.4223,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16361189438605486,
+      "grad_norm": 2.564132710733575,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16367598051781596,
+      "grad_norm": 2.533594805547048,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16374006664957702,
+      "grad_norm": 2.6563807475777854,
+      "learning_rate": 1e-06,
+      "loss": 0.474,
+      "step": 2555
+    },
+    {
+      "epoch": 0.1638041527813381,
+      "grad_norm": 2.6555639273705953,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1638682389130992,
+      "grad_norm": 3.423681981284491,
+      "learning_rate": 1e-06,
+      "loss": 0.3963,
+      "step": 2557
+    },
+    {
+      "epoch": 0.1639323250448603,
+      "grad_norm": 2.6343181562972364,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 2558
+    },
+    {
+      "epoch": 0.1639964111766214,
+      "grad_norm": 2.558480507768559,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16406049730838246,
+      "grad_norm": 2.5653732110638154,
+      "learning_rate": 1e-06,
+      "loss": 0.4264,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16412458344014355,
+      "grad_norm": 2.7535149483462025,
+      "learning_rate": 1e-06,
+      "loss": 0.408,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16418866957190464,
+      "grad_norm": 2.7378653936579553,
+      "learning_rate": 1e-06,
+      "loss": 0.4314,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16425275570366574,
+      "grad_norm": 2.6625159731761268,
+      "learning_rate": 1e-06,
+      "loss": 0.3955,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1643168418354268,
+      "grad_norm": 2.6482824209454137,
+      "learning_rate": 1e-06,
+      "loss": 0.3903,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1643809279671879,
+      "grad_norm": 2.8188480793044666,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 2565
+    },
+    {
+      "epoch": 0.164445014098949,
+      "grad_norm": 2.563521159898662,
+      "learning_rate": 1e-06,
+      "loss": 0.401,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16450910023071008,
+      "grad_norm": 2.7131990194512623,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16457318636247117,
+      "grad_norm": 2.4797567023822937,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16463727249423224,
+      "grad_norm": 2.467398013163161,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16470135862599333,
+      "grad_norm": 2.544381041213776,
+      "learning_rate": 1e-06,
+      "loss": 0.4389,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16476544475775443,
+      "grad_norm": 2.5132164527889427,
+      "learning_rate": 1e-06,
+      "loss": 0.3867,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16482953088951552,
+      "grad_norm": 2.5655755742496074,
+      "learning_rate": 1e-06,
+      "loss": 0.3894,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16489361702127658,
+      "grad_norm": 2.5545087984725794,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16495770315303768,
+      "grad_norm": 2.724930949542696,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16502178928479877,
+      "grad_norm": 2.9083920762744464,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16508587541655986,
+      "grad_norm": 2.647945420853656,
+      "learning_rate": 1e-06,
+      "loss": 0.4454,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16514996154832096,
+      "grad_norm": 2.656887468380951,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 2577
+    },
+    {
+      "epoch": 0.16521404768008202,
+      "grad_norm": 2.4060271256432095,
+      "learning_rate": 1e-06,
+      "loss": 0.3968,
+      "step": 2578
+    },
+    {
+      "epoch": 0.16527813381184311,
+      "grad_norm": 2.673050335073366,
+      "learning_rate": 1e-06,
+      "loss": 0.4298,
+      "step": 2579
+    },
+    {
+      "epoch": 0.1653422199436042,
+      "grad_norm": 2.6629568710429488,
+      "learning_rate": 1e-06,
+      "loss": 0.4281,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1654063060753653,
+      "grad_norm": 2.6124306527589414,
+      "learning_rate": 1e-06,
+      "loss": 0.4531,
+      "step": 2581
+    },
+    {
+      "epoch": 0.16547039220712637,
+      "grad_norm": 2.7811035047254764,
+      "learning_rate": 1e-06,
+      "loss": 0.4631,
+      "step": 2582
+    },
+    {
+      "epoch": 0.16553447833888746,
+      "grad_norm": 2.5917185980322937,
+      "learning_rate": 1e-06,
+      "loss": 0.4477,
+      "step": 2583
+    },
+    {
+      "epoch": 0.16559856447064855,
+      "grad_norm": 2.591034780645144,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 2584
+    },
+    {
+      "epoch": 0.16566265060240964,
+      "grad_norm": 2.794625908520945,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 2585
+    },
+    {
+      "epoch": 0.16572673673417074,
+      "grad_norm": 2.6106494060829046,
+      "learning_rate": 1e-06,
+      "loss": 0.4439,
+      "step": 2586
+    },
+    {
+      "epoch": 0.1657908228659318,
+      "grad_norm": 2.6363465931070245,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 2587
+    },
+    {
+      "epoch": 0.1658549089976929,
+      "grad_norm": 2.8385488313877363,
+      "learning_rate": 1e-06,
+      "loss": 0.4257,
+      "step": 2588
+    },
+    {
+      "epoch": 0.165918995129454,
+      "grad_norm": 2.5718888642010556,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 2589
+    },
+    {
+      "epoch": 0.16598308126121508,
+      "grad_norm": 2.824032193073902,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2590
+    },
+    {
+      "epoch": 0.16604716739297615,
+      "grad_norm": 2.6419319673150614,
+      "learning_rate": 1e-06,
+      "loss": 0.3877,
+      "step": 2591
+    },
+    {
+      "epoch": 0.16611125352473724,
+      "grad_norm": 2.6349648425656222,
+      "learning_rate": 1e-06,
+      "loss": 0.4523,
+      "step": 2592
+    },
+    {
+      "epoch": 0.16617533965649833,
+      "grad_norm": 2.5687018277447673,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 2593
+    },
+    {
+      "epoch": 0.16623942578825943,
+      "grad_norm": 2.7502135847373,
+      "learning_rate": 1e-06,
+      "loss": 0.3938,
+      "step": 2594
+    },
+    {
+      "epoch": 0.16630351192002052,
+      "grad_norm": 2.7870082246775536,
+      "learning_rate": 1e-06,
+      "loss": 0.3712,
+      "step": 2595
+    },
+    {
+      "epoch": 0.16636759805178158,
+      "grad_norm": 2.664471483180359,
+      "learning_rate": 1e-06,
+      "loss": 0.3785,
+      "step": 2596
+    },
+    {
+      "epoch": 0.16643168418354268,
+      "grad_norm": 2.8006579800788405,
+      "learning_rate": 1e-06,
+      "loss": 0.4322,
+      "step": 2597
+    },
+    {
+      "epoch": 0.16649577031530377,
+      "grad_norm": 2.651191581602123,
+      "learning_rate": 1e-06,
+      "loss": 0.428,
+      "step": 2598
+    },
+    {
+      "epoch": 0.16655985644706486,
+      "grad_norm": 2.7140854600700695,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 2599
+    },
+    {
+      "epoch": 0.16662394257882593,
+      "grad_norm": 2.7856278395057856,
+      "learning_rate": 1e-06,
+      "loss": 0.4592,
+      "step": 2600
+    },
+    {
+      "epoch": 0.16668802871058702,
+      "grad_norm": 2.607410882542063,
+      "learning_rate": 1e-06,
+      "loss": 0.4324,
+      "step": 2601
+    },
+    {
+      "epoch": 0.16675211484234811,
+      "grad_norm": 2.732714333661932,
+      "learning_rate": 1e-06,
+      "loss": 0.4252,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1668162009741092,
+      "grad_norm": 2.921716514174632,
+      "learning_rate": 1e-06,
+      "loss": 0.4345,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1668802871058703,
+      "grad_norm": 2.4597493378828497,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 2604
+    },
+    {
+      "epoch": 0.16694437323763137,
+      "grad_norm": 2.6806802003367127,
+      "learning_rate": 1e-06,
+      "loss": 0.3903,
+      "step": 2605
+    },
+    {
+      "epoch": 0.16700845936939246,
+      "grad_norm": 2.5394049202373004,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 2606
+    },
+    {
+      "epoch": 0.16707254550115355,
+      "grad_norm": 2.537658335645286,
+      "learning_rate": 1e-06,
+      "loss": 0.4692,
+      "step": 2607
+    },
+    {
+      "epoch": 0.16713663163291465,
+      "grad_norm": 2.486190723866445,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1672007177646757,
+      "grad_norm": 2.6499647012087504,
+      "learning_rate": 1e-06,
+      "loss": 0.3738,
+      "step": 2609
+    },
+    {
+      "epoch": 0.1672648038964368,
+      "grad_norm": 2.4785537911649365,
+      "learning_rate": 1e-06,
+      "loss": 0.415,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1673288900281979,
+      "grad_norm": 2.6183916526906894,
+      "learning_rate": 1e-06,
+      "loss": 0.5018,
+      "step": 2611
+    },
+    {
+      "epoch": 0.167392976159959,
+      "grad_norm": 2.6916594666182996,
+      "learning_rate": 1e-06,
+      "loss": 0.4613,
+      "step": 2612
+    },
+    {
+      "epoch": 0.16745706229172008,
+      "grad_norm": 2.731776274248501,
+      "learning_rate": 1e-06,
+      "loss": 0.338,
+      "step": 2613
+    },
+    {
+      "epoch": 0.16752114842348115,
+      "grad_norm": 2.5633563923010367,
+      "learning_rate": 1e-06,
+      "loss": 0.4063,
+      "step": 2614
+    },
+    {
+      "epoch": 0.16758523455524224,
+      "grad_norm": 2.6792211962059986,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 2615
+    },
+    {
+      "epoch": 0.16764932068700333,
+      "grad_norm": 2.704490572915443,
+      "learning_rate": 1e-06,
+      "loss": 0.3596,
+      "step": 2616
+    },
+    {
+      "epoch": 0.16771340681876443,
+      "grad_norm": 2.7040420901466127,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 2617
+    },
+    {
+      "epoch": 0.16777749295052552,
+      "grad_norm": 2.508603329270507,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 2618
+    },
+    {
+      "epoch": 0.16784157908228658,
+      "grad_norm": 2.667668834729716,
+      "learning_rate": 1e-06,
+      "loss": 0.3944,
+      "step": 2619
+    },
+    {
+      "epoch": 0.16790566521404768,
+      "grad_norm": 2.788009681704992,
+      "learning_rate": 1e-06,
+      "loss": 0.393,
+      "step": 2620
+    },
+    {
+      "epoch": 0.16796975134580877,
+      "grad_norm": 2.9943683712234295,
+      "learning_rate": 1e-06,
+      "loss": 0.4375,
+      "step": 2621
+    },
+    {
+      "epoch": 0.16803383747756986,
+      "grad_norm": 2.4580934140045576,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 2622
+    },
+    {
+      "epoch": 0.16809792360933093,
+      "grad_norm": 2.682255157750206,
+      "learning_rate": 1e-06,
+      "loss": 0.415,
+      "step": 2623
+    },
+    {
+      "epoch": 0.16816200974109202,
+      "grad_norm": 2.6326858723207915,
+      "learning_rate": 1e-06,
+      "loss": 0.4016,
+      "step": 2624
+    },
+    {
+      "epoch": 0.16822609587285312,
+      "grad_norm": 2.7531460345683945,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1682901820046142,
+      "grad_norm": 2.4582924292223063,
+      "learning_rate": 1e-06,
+      "loss": 0.3714,
+      "step": 2626
+    },
+    {
+      "epoch": 0.1683542681363753,
+      "grad_norm": 2.4189133463074413,
+      "learning_rate": 1e-06,
+      "loss": 0.3395,
+      "step": 2627
+    },
+    {
+      "epoch": 0.16841835426813637,
+      "grad_norm": 2.4978464701468854,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 2628
+    },
+    {
+      "epoch": 0.16848244039989746,
+      "grad_norm": 2.6293887606762003,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 2629
+    },
+    {
+      "epoch": 0.16854652653165855,
+      "grad_norm": 2.688163877974156,
+      "learning_rate": 1e-06,
+      "loss": 0.4566,
+      "step": 2630
+    },
+    {
+      "epoch": 0.16861061266341965,
+      "grad_norm": 2.6615797889998385,
+      "learning_rate": 1e-06,
+      "loss": 0.4286,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1686746987951807,
+      "grad_norm": 2.6150907943144452,
+      "learning_rate": 1e-06,
+      "loss": 0.3935,
+      "step": 2632
+    },
+    {
+      "epoch": 0.1687387849269418,
+      "grad_norm": 2.920403244685697,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1688028710587029,
+      "grad_norm": 2.748707499569203,
+      "learning_rate": 1e-06,
+      "loss": 0.4312,
+      "step": 2634
+    },
+    {
+      "epoch": 0.168866957190464,
+      "grad_norm": 2.7044982203747323,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 2635
+    },
+    {
+      "epoch": 0.16893104332222508,
+      "grad_norm": 2.695912823716649,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 2636
+    },
+    {
+      "epoch": 0.16899512945398615,
+      "grad_norm": 2.5569026627371625,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 2637
+    },
+    {
+      "epoch": 0.16905921558574724,
+      "grad_norm": 2.617933717691224,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 2638
+    },
+    {
+      "epoch": 0.16912330171750833,
+      "grad_norm": 2.828806100247617,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 2639
+    },
+    {
+      "epoch": 0.16918738784926943,
+      "grad_norm": 2.6495628081292613,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1692514739810305,
+      "grad_norm": 2.762389598809753,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2641
+    },
+    {
+      "epoch": 0.16931556011279159,
+      "grad_norm": 2.5771748842832647,
+      "learning_rate": 1e-06,
+      "loss": 0.3541,
+      "step": 2642
+    },
+    {
+      "epoch": 0.16937964624455268,
+      "grad_norm": 2.5680137462087944,
+      "learning_rate": 1e-06,
+      "loss": 0.3764,
+      "step": 2643
+    },
+    {
+      "epoch": 0.16944373237631377,
+      "grad_norm": 2.533760471382165,
+      "learning_rate": 1e-06,
+      "loss": 0.3492,
+      "step": 2644
+    },
+    {
+      "epoch": 0.16950781850807486,
+      "grad_norm": 2.537156219708976,
+      "learning_rate": 1e-06,
+      "loss": 0.4071,
+      "step": 2645
+    },
+    {
+      "epoch": 0.16957190463983593,
+      "grad_norm": 2.662693914933848,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 2646
+    },
+    {
+      "epoch": 0.16963599077159702,
+      "grad_norm": 2.6963640114525202,
+      "learning_rate": 1e-06,
+      "loss": 0.3973,
+      "step": 2647
+    },
+    {
+      "epoch": 0.16970007690335812,
+      "grad_norm": 2.381564264533936,
+      "learning_rate": 1e-06,
+      "loss": 0.393,
+      "step": 2648
+    },
+    {
+      "epoch": 0.1697641630351192,
+      "grad_norm": 2.666993056664914,
+      "learning_rate": 1e-06,
+      "loss": 0.4031,
+      "step": 2649
+    },
+    {
+      "epoch": 0.16982824916688027,
+      "grad_norm": 2.7697832526098214,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 2650
+    },
+    {
+      "epoch": 0.16989233529864137,
+      "grad_norm": 2.77774648574453,
+      "learning_rate": 1e-06,
+      "loss": 0.3955,
+      "step": 2651
+    },
+    {
+      "epoch": 0.16995642143040246,
+      "grad_norm": 2.773426081646378,
+      "learning_rate": 1e-06,
+      "loss": 0.4396,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17002050756216355,
+      "grad_norm": 2.349409710796212,
+      "learning_rate": 1e-06,
+      "loss": 0.3712,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17008459369392465,
+      "grad_norm": 2.6672881479515587,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2654
+    },
+    {
+      "epoch": 0.1701486798256857,
+      "grad_norm": 2.5934156121316865,
+      "learning_rate": 1e-06,
+      "loss": 0.4574,
+      "step": 2655
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 3.0540709948188893,
+      "learning_rate": 1e-06,
+      "loss": 0.4623,
+      "step": 2656
+    },
+    {
+      "epoch": 0.1702768520892079,
+      "grad_norm": 2.5464630959332926,
+      "learning_rate": 1e-06,
+      "loss": 0.3893,
+      "step": 2657
+    },
+    {
+      "epoch": 0.170340938220969,
+      "grad_norm": 2.6894763480997854,
+      "learning_rate": 1e-06,
+      "loss": 0.4528,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17040502435273006,
+      "grad_norm": 2.640883933595137,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17046911048449115,
+      "grad_norm": 2.84348627663285,
+      "learning_rate": 1e-06,
+      "loss": 0.3694,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17053319661625224,
+      "grad_norm": 2.7193478058528147,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17059728274801333,
+      "grad_norm": 2.3724490315311173,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17066136887977443,
+      "grad_norm": 2.7397106390320998,
+      "learning_rate": 1e-06,
+      "loss": 0.4298,
+      "step": 2663
+    },
+    {
+      "epoch": 0.1707254550115355,
+      "grad_norm": 2.928854559865521,
+      "learning_rate": 1e-06,
+      "loss": 0.3542,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17078954114329659,
+      "grad_norm": 3.174576889900393,
+      "learning_rate": 1e-06,
+      "loss": 0.4487,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17085362727505768,
+      "grad_norm": 2.768162935078266,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17091771340681877,
+      "grad_norm": 2.6517696398341006,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17098179953857987,
+      "grad_norm": 2.4838150857756656,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17104588567034093,
+      "grad_norm": 2.6833635728441023,
+      "learning_rate": 1e-06,
+      "loss": 0.4171,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17110997180210202,
+      "grad_norm": 2.603487459743295,
+      "learning_rate": 1e-06,
+      "loss": 0.3906,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17117405793386312,
+      "grad_norm": 2.5774146996288136,
+      "learning_rate": 1e-06,
+      "loss": 0.3956,
+      "step": 2671
+    },
+    {
+      "epoch": 0.1712381440656242,
+      "grad_norm": 2.938289624792572,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17130223019738527,
+      "grad_norm": 2.8039532904551585,
+      "learning_rate": 1e-06,
+      "loss": 0.4422,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17136631632914637,
+      "grad_norm": 2.750175646020322,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17143040246090746,
+      "grad_norm": 2.4505574568016857,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17149448859266855,
+      "grad_norm": 2.5900624275998285,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17155857472442965,
+      "grad_norm": 2.573632710786655,
+      "learning_rate": 1e-06,
+      "loss": 0.4297,
+      "step": 2677
+    },
+    {
+      "epoch": 0.1716226608561907,
+      "grad_norm": 2.868666362706998,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 2678
+    },
+    {
+      "epoch": 0.1716867469879518,
+      "grad_norm": 2.653904345921513,
+      "learning_rate": 1e-06,
+      "loss": 0.433,
+      "step": 2679
+    },
+    {
+      "epoch": 0.1717508331197129,
+      "grad_norm": 2.55586101100343,
+      "learning_rate": 1e-06,
+      "loss": 0.3604,
+      "step": 2680
+    },
+    {
+      "epoch": 0.171814919251474,
+      "grad_norm": 2.8104819328821535,
+      "learning_rate": 1e-06,
+      "loss": 0.4086,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17187900538323506,
+      "grad_norm": 2.706876954065412,
+      "learning_rate": 1e-06,
+      "loss": 0.506,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17194309151499615,
+      "grad_norm": 2.510943495458629,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17200717764675724,
+      "grad_norm": 2.9015099476835777,
+      "learning_rate": 1e-06,
+      "loss": 0.3541,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17207126377851834,
+      "grad_norm": 2.6421129558259504,
+      "learning_rate": 1e-06,
+      "loss": 0.4685,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17213534991027943,
+      "grad_norm": 2.6910006860551583,
+      "learning_rate": 1e-06,
+      "loss": 0.4363,
+      "step": 2686
+    },
+    {
+      "epoch": 0.1721994360420405,
+      "grad_norm": 3.002895992145651,
+      "learning_rate": 1e-06,
+      "loss": 0.4241,
+      "step": 2687
+    },
+    {
+      "epoch": 0.1722635221738016,
+      "grad_norm": 2.630505341919852,
+      "learning_rate": 1e-06,
+      "loss": 0.3906,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17232760830556268,
+      "grad_norm": 2.5058688409268504,
+      "learning_rate": 1e-06,
+      "loss": 0.4239,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17239169443732377,
+      "grad_norm": 2.9556913711850963,
+      "learning_rate": 1e-06,
+      "loss": 0.4375,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17245578056908484,
+      "grad_norm": 2.7664247622675684,
+      "learning_rate": 1e-06,
+      "loss": 0.4922,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17251986670084593,
+      "grad_norm": 2.619465355014572,
+      "learning_rate": 1e-06,
+      "loss": 0.4407,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17258395283260702,
+      "grad_norm": 2.686744915784644,
+      "learning_rate": 1e-06,
+      "loss": 0.4147,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17264803896436812,
+      "grad_norm": 2.6546237480998762,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1727121250961292,
+      "grad_norm": 2.898017643183714,
+      "learning_rate": 1e-06,
+      "loss": 0.4154,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17277621122789028,
+      "grad_norm": 2.6544734467667817,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17284029735965137,
+      "grad_norm": 2.6309737535988593,
+      "learning_rate": 1e-06,
+      "loss": 0.4016,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17290438349141246,
+      "grad_norm": 2.643895300769825,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17296846962317355,
+      "grad_norm": 2.6365834195624682,
+      "learning_rate": 1e-06,
+      "loss": 0.466,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17303255575493462,
+      "grad_norm": 2.5310055725501313,
+      "learning_rate": 1e-06,
+      "loss": 0.4427,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1730966418866957,
+      "grad_norm": 2.6053804421251727,
+      "learning_rate": 1e-06,
+      "loss": 0.3722,
+      "step": 2701
+    },
+    {
+      "epoch": 0.1731607280184568,
+      "grad_norm": 2.5949485014712925,
+      "learning_rate": 1e-06,
+      "loss": 0.3716,
+      "step": 2702
+    },
+    {
+      "epoch": 0.1732248141502179,
+      "grad_norm": 2.461700301283681,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 2703
+    },
+    {
+      "epoch": 0.173288900281979,
+      "grad_norm": 2.667172217132449,
+      "learning_rate": 1e-06,
+      "loss": 0.4409,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17335298641374006,
+      "grad_norm": 2.676509271450921,
+      "learning_rate": 1e-06,
+      "loss": 0.3738,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17341707254550115,
+      "grad_norm": 2.8242422751607963,
+      "learning_rate": 1e-06,
+      "loss": 0.4045,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17348115867726224,
+      "grad_norm": 2.476650612552705,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17354524480902334,
+      "grad_norm": 2.993760001961813,
+      "learning_rate": 1e-06,
+      "loss": 0.4191,
+      "step": 2708
+    },
+    {
+      "epoch": 0.1736093309407844,
+      "grad_norm": 3.0046453722930457,
+      "learning_rate": 1e-06,
+      "loss": 0.4066,
+      "step": 2709
+    },
+    {
+      "epoch": 0.1736734170725455,
+      "grad_norm": 2.9357275097901137,
+      "learning_rate": 1e-06,
+      "loss": 0.4118,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1737375032043066,
+      "grad_norm": 2.610844895489968,
+      "learning_rate": 1e-06,
+      "loss": 0.4334,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17380158933606768,
+      "grad_norm": 2.825996241401419,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17386567546782877,
+      "grad_norm": 2.8467469193619923,
+      "learning_rate": 1e-06,
+      "loss": 0.4601,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17392976159958984,
+      "grad_norm": 2.908167454995991,
+      "learning_rate": 1e-06,
+      "loss": 0.4204,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17399384773135093,
+      "grad_norm": 2.67761276554012,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17405793386311202,
+      "grad_norm": 2.6317302005851384,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17412201999487312,
+      "grad_norm": 2.747385895018015,
+      "learning_rate": 1e-06,
+      "loss": 0.3776,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17418610612663418,
+      "grad_norm": 2.3660503979944356,
+      "learning_rate": 1e-06,
+      "loss": 0.3765,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17425019225839528,
+      "grad_norm": 2.5496448902297444,
+      "learning_rate": 1e-06,
+      "loss": 0.4378,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17431427839015637,
+      "grad_norm": 2.6347752728926244,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17437836452191746,
+      "grad_norm": 2.5835556645401136,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17444245065367855,
+      "grad_norm": 2.712552510654322,
+      "learning_rate": 1e-06,
+      "loss": 0.4285,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17450653678543962,
+      "grad_norm": 2.641969170333069,
+      "learning_rate": 1e-06,
+      "loss": 0.4099,
+      "step": 2723
+    },
+    {
+      "epoch": 0.1745706229172007,
+      "grad_norm": 3.0637998879801187,
+      "learning_rate": 1e-06,
+      "loss": 0.4272,
+      "step": 2724
+    },
+    {
+      "epoch": 0.1746347090489618,
+      "grad_norm": 2.727006901877962,
+      "learning_rate": 1e-06,
+      "loss": 0.374,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1746987951807229,
+      "grad_norm": 3.023247813234385,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 2726
+    },
+    {
+      "epoch": 0.174762881312484,
+      "grad_norm": 2.814083489198847,
+      "learning_rate": 1e-06,
+      "loss": 0.393,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17482696744424506,
+      "grad_norm": 2.449041367738676,
+      "learning_rate": 1e-06,
+      "loss": 0.4182,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17489105357600615,
+      "grad_norm": 2.8231543234962806,
+      "learning_rate": 1e-06,
+      "loss": 0.3886,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17495513970776724,
+      "grad_norm": 2.7766829388477325,
+      "learning_rate": 1e-06,
+      "loss": 0.4284,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17501922583952834,
+      "grad_norm": 2.847867939571184,
+      "learning_rate": 1e-06,
+      "loss": 0.4885,
+      "step": 2731
+    },
+    {
+      "epoch": 0.1750833119712894,
+      "grad_norm": 2.4603559459072897,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1751473981030505,
+      "grad_norm": 2.9335988133488913,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 2733
+    },
+    {
+      "epoch": 0.1752114842348116,
+      "grad_norm": 2.801463474487573,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 2734
+    },
+    {
+      "epoch": 0.17527557036657268,
+      "grad_norm": 2.6157190091020146,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 2735
+    },
+    {
+      "epoch": 0.17533965649833377,
+      "grad_norm": 2.860532946073638,
+      "learning_rate": 1e-06,
+      "loss": 0.4914,
+      "step": 2736
+    },
+    {
+      "epoch": 0.17540374263009484,
+      "grad_norm": 2.6380029352671186,
+      "learning_rate": 1e-06,
+      "loss": 0.4389,
+      "step": 2737
+    },
+    {
+      "epoch": 0.17546782876185593,
+      "grad_norm": 2.7757439534033725,
+      "learning_rate": 1e-06,
+      "loss": 0.3723,
+      "step": 2738
+    },
+    {
+      "epoch": 0.17553191489361702,
+      "grad_norm": 2.5525342561888293,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 2739
+    },
+    {
+      "epoch": 0.17559600102537812,
+      "grad_norm": 2.8172344485968264,
+      "learning_rate": 1e-06,
+      "loss": 0.447,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17566008715713918,
+      "grad_norm": 2.609296582257566,
+      "learning_rate": 1e-06,
+      "loss": 0.3896,
+      "step": 2741
+    },
+    {
+      "epoch": 0.17572417328890028,
+      "grad_norm": 2.4000086904756732,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 2742
+    },
+    {
+      "epoch": 0.17578825942066137,
+      "grad_norm": 2.8350571223146908,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 2743
+    },
+    {
+      "epoch": 0.17585234555242246,
+      "grad_norm": 2.7509441663604552,
+      "learning_rate": 1e-06,
+      "loss": 0.4094,
+      "step": 2744
+    },
+    {
+      "epoch": 0.17591643168418356,
+      "grad_norm": 2.5301745440674135,
+      "learning_rate": 1e-06,
+      "loss": 0.3968,
+      "step": 2745
+    },
+    {
+      "epoch": 0.17598051781594462,
+      "grad_norm": 2.798905110335789,
+      "learning_rate": 1e-06,
+      "loss": 0.3853,
+      "step": 2746
+    },
+    {
+      "epoch": 0.1760446039477057,
+      "grad_norm": 2.705049111037033,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1761086900794668,
+      "grad_norm": 2.8010432178773823,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1761727762112279,
+      "grad_norm": 2.4673188572577125,
+      "learning_rate": 1e-06,
+      "loss": 0.3891,
+      "step": 2749
+    },
+    {
+      "epoch": 0.17623686234298896,
+      "grad_norm": 2.345419411995933,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 2750
+    },
+    {
+      "epoch": 0.17630094847475006,
+      "grad_norm": 2.652929007808521,
+      "learning_rate": 1e-06,
+      "loss": 0.402,
+      "step": 2751
+    },
+    {
+      "epoch": 0.17636503460651115,
+      "grad_norm": 2.4664570988941397,
+      "learning_rate": 1e-06,
+      "loss": 0.3661,
+      "step": 2752
+    },
+    {
+      "epoch": 0.17642912073827224,
+      "grad_norm": 2.5442145233577733,
+      "learning_rate": 1e-06,
+      "loss": 0.387,
+      "step": 2753
+    },
+    {
+      "epoch": 0.17649320687003334,
+      "grad_norm": 2.663665531636393,
+      "learning_rate": 1e-06,
+      "loss": 0.5124,
+      "step": 2754
+    },
+    {
+      "epoch": 0.1765572930017944,
+      "grad_norm": 2.6084492727596986,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1766213791335555,
+      "grad_norm": 2.6078674777982007,
+      "learning_rate": 1e-06,
+      "loss": 0.477,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1766854652653166,
+      "grad_norm": 2.548574378570834,
+      "learning_rate": 1e-06,
+      "loss": 0.3738,
+      "step": 2757
+    },
+    {
+      "epoch": 0.17674955139707768,
+      "grad_norm": 2.5789832769642036,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 2758
+    },
+    {
+      "epoch": 0.17681363752883875,
+      "grad_norm": 2.717502081315403,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 2759
+    },
+    {
+      "epoch": 0.17687772366059984,
+      "grad_norm": 2.568127521668392,
+      "learning_rate": 1e-06,
+      "loss": 0.3814,
+      "step": 2760
+    },
+    {
+      "epoch": 0.17694180979236093,
+      "grad_norm": 2.789005011677915,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 2761
+    },
+    {
+      "epoch": 0.17700589592412203,
+      "grad_norm": 2.477575732446859,
+      "learning_rate": 1e-06,
+      "loss": 0.4279,
+      "step": 2762
+    },
+    {
+      "epoch": 0.17706998205588312,
+      "grad_norm": 2.7186598813432994,
+      "learning_rate": 1e-06,
+      "loss": 0.4073,
+      "step": 2763
+    },
+    {
+      "epoch": 0.17713406818764418,
+      "grad_norm": 2.7710934300979773,
+      "learning_rate": 1e-06,
+      "loss": 0.4294,
+      "step": 2764
+    },
+    {
+      "epoch": 0.17719815431940528,
+      "grad_norm": 2.7693788880209076,
+      "learning_rate": 1e-06,
+      "loss": 0.3503,
+      "step": 2765
+    },
+    {
+      "epoch": 0.17726224045116637,
+      "grad_norm": 2.6829776861796404,
+      "learning_rate": 1e-06,
+      "loss": 0.4504,
+      "step": 2766
+    },
+    {
+      "epoch": 0.17732632658292746,
+      "grad_norm": 2.4899668268787583,
+      "learning_rate": 1e-06,
+      "loss": 0.4042,
+      "step": 2767
+    },
+    {
+      "epoch": 0.17739041271468853,
+      "grad_norm": 2.5376629009155374,
+      "learning_rate": 1e-06,
+      "loss": 0.4113,
+      "step": 2768
+    },
+    {
+      "epoch": 0.17745449884644962,
+      "grad_norm": 2.6190589519635847,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 2769
+    },
+    {
+      "epoch": 0.17751858497821071,
+      "grad_norm": 2.733587140196812,
+      "learning_rate": 1e-06,
+      "loss": 0.3711,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1775826711099718,
+      "grad_norm": 2.7646668438152378,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 2771
+    },
+    {
+      "epoch": 0.1776467572417329,
+      "grad_norm": 2.595584633191652,
+      "learning_rate": 1e-06,
+      "loss": 0.3571,
+      "step": 2772
+    },
+    {
+      "epoch": 0.17771084337349397,
+      "grad_norm": 2.7361207178039266,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 2773
+    },
+    {
+      "epoch": 0.17777492950525506,
+      "grad_norm": 2.876956812662481,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 2774
+    },
+    {
+      "epoch": 0.17783901563701615,
+      "grad_norm": 2.5906483777638485,
+      "learning_rate": 1e-06,
+      "loss": 0.4096,
+      "step": 2775
+    },
+    {
+      "epoch": 0.17790310176877724,
+      "grad_norm": 2.604898683351699,
+      "learning_rate": 1e-06,
+      "loss": 0.3946,
+      "step": 2776
+    },
+    {
+      "epoch": 0.17796718790053834,
+      "grad_norm": 2.6280129282501243,
+      "learning_rate": 1e-06,
+      "loss": 0.4328,
+      "step": 2777
+    },
+    {
+      "epoch": 0.1780312740322994,
+      "grad_norm": 2.773509257973345,
+      "learning_rate": 1e-06,
+      "loss": 0.4847,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1780953601640605,
+      "grad_norm": 2.669028302656756,
+      "learning_rate": 1e-06,
+      "loss": 0.4049,
+      "step": 2779
+    },
+    {
+      "epoch": 0.1781594462958216,
+      "grad_norm": 2.5119037751744466,
+      "learning_rate": 1e-06,
+      "loss": 0.3751,
+      "step": 2780
+    },
+    {
+      "epoch": 0.17822353242758268,
+      "grad_norm": 2.719489579389454,
+      "learning_rate": 1e-06,
+      "loss": 0.4567,
+      "step": 2781
+    },
+    {
+      "epoch": 0.17828761855934375,
+      "grad_norm": 2.802127986585095,
+      "learning_rate": 1e-06,
+      "loss": 0.4662,
+      "step": 2782
+    },
+    {
+      "epoch": 0.17835170469110484,
+      "grad_norm": 3.488029860507148,
+      "learning_rate": 1e-06,
+      "loss": 0.419,
+      "step": 2783
+    },
+    {
+      "epoch": 0.17841579082286593,
+      "grad_norm": 2.7624002719807117,
+      "learning_rate": 1e-06,
+      "loss": 0.4514,
+      "step": 2784
+    },
+    {
+      "epoch": 0.17847987695462703,
+      "grad_norm": 2.7263045947433677,
+      "learning_rate": 1e-06,
+      "loss": 0.4049,
+      "step": 2785
+    },
+    {
+      "epoch": 0.17854396308638812,
+      "grad_norm": 2.664083750414536,
+      "learning_rate": 1e-06,
+      "loss": 0.4095,
+      "step": 2786
+    },
+    {
+      "epoch": 0.17860804921814918,
+      "grad_norm": 2.5985944580963025,
+      "learning_rate": 1e-06,
+      "loss": 0.4391,
+      "step": 2787
+    },
+    {
+      "epoch": 0.17867213534991028,
+      "grad_norm": 2.6064454548319884,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 2788
+    },
+    {
+      "epoch": 0.17873622148167137,
+      "grad_norm": 2.5802910154946668,
+      "learning_rate": 1e-06,
+      "loss": 0.3561,
+      "step": 2789
+    },
+    {
+      "epoch": 0.17880030761343246,
+      "grad_norm": 2.571872682594762,
+      "learning_rate": 1e-06,
+      "loss": 0.3856,
+      "step": 2790
+    },
+    {
+      "epoch": 0.17886439374519353,
+      "grad_norm": 2.6983518939151896,
+      "learning_rate": 1e-06,
+      "loss": 0.4202,
+      "step": 2791
+    },
+    {
+      "epoch": 0.17892847987695462,
+      "grad_norm": 2.599966184965254,
+      "learning_rate": 1e-06,
+      "loss": 0.4119,
+      "step": 2792
+    },
+    {
+      "epoch": 0.17899256600871571,
+      "grad_norm": 3.207940867000454,
+      "learning_rate": 1e-06,
+      "loss": 0.4075,
+      "step": 2793
+    },
+    {
+      "epoch": 0.1790566521404768,
+      "grad_norm": 2.714492976153829,
+      "learning_rate": 1e-06,
+      "loss": 0.406,
+      "step": 2794
+    },
+    {
+      "epoch": 0.1791207382722379,
+      "grad_norm": 2.6020945100146124,
+      "learning_rate": 1e-06,
+      "loss": 0.4085,
+      "step": 2795
+    },
+    {
+      "epoch": 0.17918482440399897,
+      "grad_norm": 2.751146216088707,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 2796
+    },
+    {
+      "epoch": 0.17924891053576006,
+      "grad_norm": 2.7125630629926847,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 2797
+    },
+    {
+      "epoch": 0.17931299666752115,
+      "grad_norm": 2.5691944151852666,
+      "learning_rate": 1e-06,
+      "loss": 0.3925,
+      "step": 2798
+    },
+    {
+      "epoch": 0.17937708279928224,
+      "grad_norm": 2.4952783268939,
+      "learning_rate": 1e-06,
+      "loss": 0.3993,
+      "step": 2799
+    },
+    {
+      "epoch": 0.1794411689310433,
+      "grad_norm": 2.6788205945919157,
+      "learning_rate": 1e-06,
+      "loss": 0.3886,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1795052550628044,
+      "grad_norm": 2.6060361462237736,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1795693411945655,
+      "grad_norm": 2.825696798513869,
+      "learning_rate": 1e-06,
+      "loss": 0.4254,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1796334273263266,
+      "grad_norm": 2.6233720527199287,
+      "learning_rate": 1e-06,
+      "loss": 0.3849,
+      "step": 2803
+    },
+    {
+      "epoch": 0.17969751345808768,
+      "grad_norm": 2.72767923666059,
+      "learning_rate": 1e-06,
+      "loss": 0.3825,
+      "step": 2804
+    },
+    {
+      "epoch": 0.17976159958984875,
+      "grad_norm": 2.6288577192437903,
+      "learning_rate": 1e-06,
+      "loss": 0.4019,
+      "step": 2805
+    },
+    {
+      "epoch": 0.17982568572160984,
+      "grad_norm": 2.897506293696981,
+      "learning_rate": 1e-06,
+      "loss": 0.4932,
+      "step": 2806
+    },
+    {
+      "epoch": 0.17988977185337093,
+      "grad_norm": 2.6155573925589124,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 2807
+    },
+    {
+      "epoch": 0.17995385798513203,
+      "grad_norm": 2.7440180522487627,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 2808
+    },
+    {
+      "epoch": 0.1800179441168931,
+      "grad_norm": 2.8646368360187378,
+      "learning_rate": 1e-06,
+      "loss": 0.4188,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18008203024865418,
+      "grad_norm": 2.569865671345972,
+      "learning_rate": 1e-06,
+      "loss": 0.382,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18014611638041528,
+      "grad_norm": 2.5413764281681055,
+      "learning_rate": 1e-06,
+      "loss": 0.3716,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18021020251217637,
+      "grad_norm": 2.438466070103246,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18027428864393746,
+      "grad_norm": 2.5496759695348814,
+      "learning_rate": 1e-06,
+      "loss": 0.4054,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18033837477569853,
+      "grad_norm": 2.6402962681310065,
+      "learning_rate": 1e-06,
+      "loss": 0.4161,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18040246090745962,
+      "grad_norm": 2.670866895541432,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18046654703922071,
+      "grad_norm": 2.791334110746202,
+      "learning_rate": 1e-06,
+      "loss": 0.3661,
+      "step": 2816
+    },
+    {
+      "epoch": 0.1805306331709818,
+      "grad_norm": 2.509276363670558,
+      "learning_rate": 1e-06,
+      "loss": 0.4076,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18059471930274287,
+      "grad_norm": 2.5760759381567775,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18065880543450397,
+      "grad_norm": 2.823011500253391,
+      "learning_rate": 1e-06,
+      "loss": 0.3794,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18072289156626506,
+      "grad_norm": 2.5784308273181895,
+      "learning_rate": 1e-06,
+      "loss": 0.4494,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18078697769802615,
+      "grad_norm": 2.8354990552703767,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18085106382978725,
+      "grad_norm": 2.545519235918881,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1809151499615483,
+      "grad_norm": 2.650679339899046,
+      "learning_rate": 1e-06,
+      "loss": 0.4316,
+      "step": 2823
+    },
+    {
+      "epoch": 0.1809792360933094,
+      "grad_norm": 2.5115419988434167,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1810433222250705,
+      "grad_norm": 2.5266254302391333,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 2825
+    },
+    {
+      "epoch": 0.1811074083568316,
+      "grad_norm": 2.7285294453224895,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18117149448859265,
+      "grad_norm": 2.6957667643101453,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18123558062035375,
+      "grad_norm": 2.5773499213332327,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18129966675211484,
+      "grad_norm": 2.605720550620889,
+      "learning_rate": 1e-06,
+      "loss": 0.4316,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18136375288387593,
+      "grad_norm": 2.852712735022763,
+      "learning_rate": 1e-06,
+      "loss": 0.4289,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18142783901563703,
+      "grad_norm": 2.75059295048179,
+      "learning_rate": 1e-06,
+      "loss": 0.3943,
+      "step": 2831
+    },
+    {
+      "epoch": 0.1814919251473981,
+      "grad_norm": 3.210091013374015,
+      "learning_rate": 1e-06,
+      "loss": 0.4356,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18155601127915919,
+      "grad_norm": 2.7176044087136577,
+      "learning_rate": 1e-06,
+      "loss": 0.4638,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18162009741092028,
+      "grad_norm": 2.6859048781918156,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18168418354268137,
+      "grad_norm": 2.690898630883437,
+      "learning_rate": 1e-06,
+      "loss": 0.4282,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18174826967444246,
+      "grad_norm": 2.670934781807855,
+      "learning_rate": 1e-06,
+      "loss": 0.3632,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18181235580620353,
+      "grad_norm": 2.5683697269090775,
+      "learning_rate": 1e-06,
+      "loss": 0.3937,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18187644193796462,
+      "grad_norm": 2.721953999228212,
+      "learning_rate": 1e-06,
+      "loss": 0.4822,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18194052806972572,
+      "grad_norm": 2.761994453753278,
+      "learning_rate": 1e-06,
+      "loss": 0.4528,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1820046142014868,
+      "grad_norm": 2.4924789215698047,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18206870033324787,
+      "grad_norm": 2.6911603552831953,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18213278646500897,
+      "grad_norm": 2.4914369771221514,
+      "learning_rate": 1e-06,
+      "loss": 0.4142,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18219687259677006,
+      "grad_norm": 2.670264209025931,
+      "learning_rate": 1e-06,
+      "loss": 0.4526,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18226095872853115,
+      "grad_norm": 2.655186995685657,
+      "learning_rate": 1e-06,
+      "loss": 0.4319,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18232504486029225,
+      "grad_norm": 2.7994660696485654,
+      "learning_rate": 1e-06,
+      "loss": 0.3766,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1823891309920533,
+      "grad_norm": 2.6672566061037983,
+      "learning_rate": 1e-06,
+      "loss": 0.3663,
+      "step": 2846
+    },
+    {
+      "epoch": 0.1824532171238144,
+      "grad_norm": 2.4995281084688403,
+      "learning_rate": 1e-06,
+      "loss": 0.3993,
+      "step": 2847
+    },
+    {
+      "epoch": 0.1825173032555755,
+      "grad_norm": 2.9307889954258664,
+      "learning_rate": 1e-06,
+      "loss": 0.4265,
+      "step": 2848
+    },
+    {
+      "epoch": 0.1825813893873366,
+      "grad_norm": 2.418817204994986,
+      "learning_rate": 1e-06,
+      "loss": 0.399,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18264547551909766,
+      "grad_norm": 2.683877473241475,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18270956165085875,
+      "grad_norm": 2.4528568962910615,
+      "learning_rate": 1e-06,
+      "loss": 0.4044,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18277364778261984,
+      "grad_norm": 2.507911995843844,
+      "learning_rate": 1e-06,
+      "loss": 0.4158,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18283773391438093,
+      "grad_norm": 2.6538955269222164,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18290182004614203,
+      "grad_norm": 2.3812884712864055,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 2854
+    },
+    {
+      "epoch": 0.1829659061779031,
+      "grad_norm": 2.7519692295648457,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18302999230966419,
+      "grad_norm": 2.5769076247826717,
+      "learning_rate": 1e-06,
+      "loss": 0.4546,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18309407844142528,
+      "grad_norm": 2.7671252476956343,
+      "learning_rate": 1e-06,
+      "loss": 0.4223,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18315816457318637,
+      "grad_norm": 2.7236996174804853,
+      "learning_rate": 1e-06,
+      "loss": 0.4156,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18322225070494744,
+      "grad_norm": 2.777018765978659,
+      "learning_rate": 1e-06,
+      "loss": 0.4837,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18328633683670853,
+      "grad_norm": 2.7820808558817247,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18335042296846962,
+      "grad_norm": 2.728486937802698,
+      "learning_rate": 1e-06,
+      "loss": 0.4269,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18341450910023072,
+      "grad_norm": 2.4941825007689644,
+      "learning_rate": 1e-06,
+      "loss": 0.4187,
+      "step": 2862
+    },
+    {
+      "epoch": 0.1834785952319918,
+      "grad_norm": 2.5636554116823262,
+      "learning_rate": 1e-06,
+      "loss": 0.4574,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18354268136375287,
+      "grad_norm": 2.745168426594291,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18360676749551397,
+      "grad_norm": 2.55893475455084,
+      "learning_rate": 1e-06,
+      "loss": 0.3796,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18367085362727506,
+      "grad_norm": 2.48777488032714,
+      "learning_rate": 1e-06,
+      "loss": 0.432,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18373493975903615,
+      "grad_norm": 3.1189156752141023,
+      "learning_rate": 1e-06,
+      "loss": 0.4482,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18379902589079722,
+      "grad_norm": 2.5609793040069584,
+      "learning_rate": 1e-06,
+      "loss": 0.4464,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1838631120225583,
+      "grad_norm": 2.538218566718866,
+      "learning_rate": 1e-06,
+      "loss": 0.4741,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1839271981543194,
+      "grad_norm": 2.473628489684195,
+      "learning_rate": 1e-06,
+      "loss": 0.4571,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1839912842860805,
+      "grad_norm": 2.6655771340155425,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 2871
+    },
+    {
+      "epoch": 0.1840553704178416,
+      "grad_norm": 2.5370217160344426,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18411945654960266,
+      "grad_norm": 2.4867648521029797,
+      "learning_rate": 1e-06,
+      "loss": 0.3996,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18418354268136375,
+      "grad_norm": 2.835924752123543,
+      "learning_rate": 1e-06,
+      "loss": 0.4362,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18424762881312484,
+      "grad_norm": 2.405751638794753,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18431171494488593,
+      "grad_norm": 2.8384661298225087,
+      "learning_rate": 1e-06,
+      "loss": 0.3972,
+      "step": 2876
+    },
+    {
+      "epoch": 0.184375801076647,
+      "grad_norm": 2.4594890443427184,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 2877
+    },
+    {
+      "epoch": 0.1844398872084081,
+      "grad_norm": 2.526104120739134,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 2878
+    },
+    {
+      "epoch": 0.1845039733401692,
+      "grad_norm": 2.6250238766805585,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18456805947193028,
+      "grad_norm": 2.633967543674042,
+      "learning_rate": 1e-06,
+      "loss": 0.435,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18463214560369137,
+      "grad_norm": 2.7900500444927507,
+      "learning_rate": 1e-06,
+      "loss": 0.462,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18469623173545244,
+      "grad_norm": 2.9289848277770814,
+      "learning_rate": 1e-06,
+      "loss": 0.453,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18476031786721353,
+      "grad_norm": 2.6918411760353536,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18482440399897462,
+      "grad_norm": 2.9527899042701615,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18488849013073572,
+      "grad_norm": 2.6012306739317785,
+      "learning_rate": 1e-06,
+      "loss": 0.4306,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1849525762624968,
+      "grad_norm": 2.9987158571479977,
+      "learning_rate": 1e-06,
+      "loss": 0.365,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18501666239425787,
+      "grad_norm": 2.800871847179712,
+      "learning_rate": 1e-06,
+      "loss": 0.3935,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18508074852601897,
+      "grad_norm": 2.828246130679259,
+      "learning_rate": 1e-06,
+      "loss": 0.3956,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18514483465778006,
+      "grad_norm": 2.619825070399056,
+      "learning_rate": 1e-06,
+      "loss": 0.4172,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18520892078954115,
+      "grad_norm": 2.624254162054244,
+      "learning_rate": 1e-06,
+      "loss": 0.4387,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18527300692130222,
+      "grad_norm": 2.669571750177714,
+      "learning_rate": 1e-06,
+      "loss": 0.4586,
+      "step": 2891
+    },
+    {
+      "epoch": 0.1853370930530633,
+      "grad_norm": 2.58413530972886,
+      "learning_rate": 1e-06,
+      "loss": 0.4605,
+      "step": 2892
+    },
+    {
+      "epoch": 0.1854011791848244,
+      "grad_norm": 2.750414580107718,
+      "learning_rate": 1e-06,
+      "loss": 0.4133,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1854652653165855,
+      "grad_norm": 2.5738479042844635,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 2894
+    },
+    {
+      "epoch": 0.1855293514483466,
+      "grad_norm": 2.5740432194741394,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 2895
+    },
+    {
+      "epoch": 0.18559343758010766,
+      "grad_norm": 2.7328320766355745,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 2896
+    },
+    {
+      "epoch": 0.18565752371186875,
+      "grad_norm": 2.6266822727179617,
+      "learning_rate": 1e-06,
+      "loss": 0.4322,
+      "step": 2897
+    },
+    {
+      "epoch": 0.18572160984362984,
+      "grad_norm": 2.7423123147741624,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 2898
+    },
+    {
+      "epoch": 0.18578569597539094,
+      "grad_norm": 2.6644779759023405,
+      "learning_rate": 1e-06,
+      "loss": 0.3878,
+      "step": 2899
+    },
+    {
+      "epoch": 0.185849782107152,
+      "grad_norm": 2.8416891170218164,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1859138682389131,
+      "grad_norm": 2.651614683525343,
+      "learning_rate": 1e-06,
+      "loss": 0.4604,
+      "step": 2901
+    },
+    {
+      "epoch": 0.1859779543706742,
+      "grad_norm": 2.6711252166907253,
+      "learning_rate": 1e-06,
+      "loss": 0.4333,
+      "step": 2902
+    },
+    {
+      "epoch": 0.18604204050243528,
+      "grad_norm": 2.6775615588910275,
+      "learning_rate": 1e-06,
+      "loss": 0.3517,
+      "step": 2903
+    },
+    {
+      "epoch": 0.18610612663419637,
+      "grad_norm": 2.702059561627033,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 2904
+    },
+    {
+      "epoch": 0.18617021276595744,
+      "grad_norm": 2.536263599797668,
+      "learning_rate": 1e-06,
+      "loss": 0.3772,
+      "step": 2905
+    },
+    {
+      "epoch": 0.18623429889771853,
+      "grad_norm": 2.845862941191916,
+      "learning_rate": 1e-06,
+      "loss": 0.4259,
+      "step": 2906
+    },
+    {
+      "epoch": 0.18629838502947962,
+      "grad_norm": 2.700477381709106,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 2907
+    },
+    {
+      "epoch": 0.18636247116124072,
+      "grad_norm": 2.682607738754587,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 2908
+    },
+    {
+      "epoch": 0.18642655729300178,
+      "grad_norm": 2.554052321832966,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 2909
+    },
+    {
+      "epoch": 0.18649064342476288,
+      "grad_norm": 2.6403351465968274,
+      "learning_rate": 1e-06,
+      "loss": 0.4517,
+      "step": 2910
+    },
+    {
+      "epoch": 0.18655472955652397,
+      "grad_norm": 2.5477057026439476,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 2911
+    },
+    {
+      "epoch": 0.18661881568828506,
+      "grad_norm": 2.5704484585325926,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 2912
+    },
+    {
+      "epoch": 0.18668290182004615,
+      "grad_norm": 2.6724804046778594,
+      "learning_rate": 1e-06,
+      "loss": 0.3791,
+      "step": 2913
+    },
+    {
+      "epoch": 0.18674698795180722,
+      "grad_norm": 2.705919945837271,
+      "learning_rate": 1e-06,
+      "loss": 0.4193,
+      "step": 2914
+    },
+    {
+      "epoch": 0.1868110740835683,
+      "grad_norm": 2.6275085521905037,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1868751602153294,
+      "grad_norm": 2.5644893129504993,
+      "learning_rate": 1e-06,
+      "loss": 0.385,
+      "step": 2916
+    },
+    {
+      "epoch": 0.1869392463470905,
+      "grad_norm": 2.6374626451766843,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 2917
+    },
+    {
+      "epoch": 0.18700333247885156,
+      "grad_norm": 2.916365013642046,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 2918
+    },
+    {
+      "epoch": 0.18706741861061266,
+      "grad_norm": 2.5881011852417557,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 2919
+    },
+    {
+      "epoch": 0.18713150474237375,
+      "grad_norm": 2.455367104725631,
+      "learning_rate": 1e-06,
+      "loss": 0.4351,
+      "step": 2920
+    },
+    {
+      "epoch": 0.18719559087413484,
+      "grad_norm": 2.675434496637276,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 2921
+    },
+    {
+      "epoch": 0.18725967700589594,
+      "grad_norm": 2.7514060307588823,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 2922
+    },
+    {
+      "epoch": 0.187323763137657,
+      "grad_norm": 2.8818081578888135,
+      "learning_rate": 1e-06,
+      "loss": 0.4365,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1873878492694181,
+      "grad_norm": 2.722378179353567,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 2924
+    },
+    {
+      "epoch": 0.1874519354011792,
+      "grad_norm": 2.6386298247760442,
+      "learning_rate": 1e-06,
+      "loss": 0.4116,
+      "step": 2925
+    },
+    {
+      "epoch": 0.18751602153294028,
+      "grad_norm": 2.6530040293298693,
+      "learning_rate": 1e-06,
+      "loss": 0.3865,
+      "step": 2926
+    },
+    {
+      "epoch": 0.18758010766470135,
+      "grad_norm": 2.8852720002345364,
+      "learning_rate": 1e-06,
+      "loss": 0.4578,
+      "step": 2927
+    },
+    {
+      "epoch": 0.18764419379646244,
+      "grad_norm": 2.624833979077509,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 2928
+    },
+    {
+      "epoch": 0.18770827992822353,
+      "grad_norm": 2.5224928226685557,
+      "learning_rate": 1e-06,
+      "loss": 0.3461,
+      "step": 2929
+    },
+    {
+      "epoch": 0.18777236605998462,
+      "grad_norm": 2.783588647866934,
+      "learning_rate": 1e-06,
+      "loss": 0.349,
+      "step": 2930
+    },
+    {
+      "epoch": 0.18783645219174572,
+      "grad_norm": 2.5465407705725,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 2931
+    },
+    {
+      "epoch": 0.18790053832350678,
+      "grad_norm": 2.5538413117725964,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 2932
+    },
+    {
+      "epoch": 0.18796462445526788,
+      "grad_norm": 2.551250691707148,
+      "learning_rate": 1e-06,
+      "loss": 0.4594,
+      "step": 2933
+    },
+    {
+      "epoch": 0.18802871058702897,
+      "grad_norm": 3.1452067199513345,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 2934
+    },
+    {
+      "epoch": 0.18809279671879006,
+      "grad_norm": 2.743176730793466,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 2935
+    },
+    {
+      "epoch": 0.18815688285055113,
+      "grad_norm": 2.8114938453740113,
+      "learning_rate": 1e-06,
+      "loss": 0.4454,
+      "step": 2936
+    },
+    {
+      "epoch": 0.18822096898231222,
+      "grad_norm": 2.8125690365615457,
+      "learning_rate": 1e-06,
+      "loss": 0.3852,
+      "step": 2937
+    },
+    {
+      "epoch": 0.1882850551140733,
+      "grad_norm": 2.7093968409248954,
+      "learning_rate": 1e-06,
+      "loss": 0.4598,
+      "step": 2938
+    },
+    {
+      "epoch": 0.1883491412458344,
+      "grad_norm": 2.7084723079218977,
+      "learning_rate": 1e-06,
+      "loss": 0.3802,
+      "step": 2939
+    },
+    {
+      "epoch": 0.1884132273775955,
+      "grad_norm": 2.924608356651168,
+      "learning_rate": 1e-06,
+      "loss": 0.4494,
+      "step": 2940
+    },
+    {
+      "epoch": 0.18847731350935656,
+      "grad_norm": 2.7623725533237296,
+      "learning_rate": 1e-06,
+      "loss": 0.4018,
+      "step": 2941
+    },
+    {
+      "epoch": 0.18854139964111766,
+      "grad_norm": 2.570143623445525,
+      "learning_rate": 1e-06,
+      "loss": 0.3848,
+      "step": 2942
+    },
+    {
+      "epoch": 0.18860548577287875,
+      "grad_norm": 2.590379184495267,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 2943
+    },
+    {
+      "epoch": 0.18866957190463984,
+      "grad_norm": 2.6091991566587796,
+      "learning_rate": 1e-06,
+      "loss": 0.3754,
+      "step": 2944
+    },
+    {
+      "epoch": 0.18873365803640094,
+      "grad_norm": 2.5426130853423685,
+      "learning_rate": 1e-06,
+      "loss": 0.4436,
+      "step": 2945
+    },
+    {
+      "epoch": 0.188797744168162,
+      "grad_norm": 2.4880023974397534,
+      "learning_rate": 1e-06,
+      "loss": 0.3936,
+      "step": 2946
+    },
+    {
+      "epoch": 0.1888618302999231,
+      "grad_norm": 2.7254911181430734,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 2947
+    },
+    {
+      "epoch": 0.1889259164316842,
+      "grad_norm": 2.855985760623235,
+      "learning_rate": 1e-06,
+      "loss": 0.4405,
+      "step": 2948
+    },
+    {
+      "epoch": 0.18899000256344528,
+      "grad_norm": 2.738984275461709,
+      "learning_rate": 1e-06,
+      "loss": 0.4785,
+      "step": 2949
+    },
+    {
+      "epoch": 0.18905408869520635,
+      "grad_norm": 2.693467333245631,
+      "learning_rate": 1e-06,
+      "loss": 0.4415,
+      "step": 2950
+    },
+    {
+      "epoch": 0.18911817482696744,
+      "grad_norm": 2.725268201313213,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 2951
+    },
+    {
+      "epoch": 0.18918226095872853,
+      "grad_norm": 2.649478170221162,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 2952
+    },
+    {
+      "epoch": 0.18924634709048962,
+      "grad_norm": 2.7059280275569053,
+      "learning_rate": 1e-06,
+      "loss": 0.4168,
+      "step": 2953
+    },
+    {
+      "epoch": 0.18931043322225072,
+      "grad_norm": 2.9448804667390966,
+      "learning_rate": 1e-06,
+      "loss": 0.4091,
+      "step": 2954
+    },
+    {
+      "epoch": 0.18937451935401178,
+      "grad_norm": 2.6088760199937036,
+      "learning_rate": 1e-06,
+      "loss": 0.4026,
+      "step": 2955
+    },
+    {
+      "epoch": 0.18943860548577288,
+      "grad_norm": 2.792982465429231,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 2956
+    },
+    {
+      "epoch": 0.18950269161753397,
+      "grad_norm": 2.856731613527467,
+      "learning_rate": 1e-06,
+      "loss": 0.4374,
+      "step": 2957
+    },
+    {
+      "epoch": 0.18956677774929506,
+      "grad_norm": 2.6591552161854044,
+      "learning_rate": 1e-06,
+      "loss": 0.4328,
+      "step": 2958
+    },
+    {
+      "epoch": 0.18963086388105613,
+      "grad_norm": 2.820863474787736,
+      "learning_rate": 1e-06,
+      "loss": 0.4068,
+      "step": 2959
+    },
+    {
+      "epoch": 0.18969495001281722,
+      "grad_norm": 3.1564803034432067,
+      "learning_rate": 1e-06,
+      "loss": 0.4894,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1897590361445783,
+      "grad_norm": 2.7775467867516723,
+      "learning_rate": 1e-06,
+      "loss": 0.4219,
+      "step": 2961
+    },
+    {
+      "epoch": 0.1898231222763394,
+      "grad_norm": 2.570709104763904,
+      "learning_rate": 1e-06,
+      "loss": 0.3836,
+      "step": 2962
+    },
+    {
+      "epoch": 0.1898872084081005,
+      "grad_norm": 3.02681267671875,
+      "learning_rate": 1e-06,
+      "loss": 0.4426,
+      "step": 2963
+    },
+    {
+      "epoch": 0.18995129453986156,
+      "grad_norm": 2.510522158823479,
+      "learning_rate": 1e-06,
+      "loss": 0.4054,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19001538067162266,
+      "grad_norm": 2.693774060885273,
+      "learning_rate": 1e-06,
+      "loss": 0.4705,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19007946680338375,
+      "grad_norm": 2.748528804109493,
+      "learning_rate": 1e-06,
+      "loss": 0.4332,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19014355293514484,
+      "grad_norm": 2.776594086545492,
+      "learning_rate": 1e-06,
+      "loss": 0.4031,
+      "step": 2967
+    },
+    {
+      "epoch": 0.1902076390669059,
+      "grad_norm": 2.797526799261737,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 2968
+    },
+    {
+      "epoch": 0.190271725198667,
+      "grad_norm": 2.716540472496669,
+      "learning_rate": 1e-06,
+      "loss": 0.3527,
+      "step": 2969
+    },
+    {
+      "epoch": 0.1903358113304281,
+      "grad_norm": 2.5129230641901255,
+      "learning_rate": 1e-06,
+      "loss": 0.4134,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1903998974621892,
+      "grad_norm": 2.600501120594301,
+      "learning_rate": 1e-06,
+      "loss": 0.4419,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19046398359395028,
+      "grad_norm": 2.976313279269352,
+      "learning_rate": 1e-06,
+      "loss": 0.4721,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19052806972571135,
+      "grad_norm": 2.771766068236897,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19059215585747244,
+      "grad_norm": 2.6489145777810394,
+      "learning_rate": 1e-06,
+      "loss": 0.4546,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19065624198923353,
+      "grad_norm": 2.580312074382169,
+      "learning_rate": 1e-06,
+      "loss": 0.3992,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19072032812099463,
+      "grad_norm": 2.7443330078441237,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 2976
+    },
+    {
+      "epoch": 0.1907844142527557,
+      "grad_norm": 2.633306061091404,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19084850038451678,
+      "grad_norm": 2.604761994040819,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19091258651627788,
+      "grad_norm": 2.683095016513032,
+      "learning_rate": 1e-06,
+      "loss": 0.4331,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19097667264803897,
+      "grad_norm": 2.6835183303069003,
+      "learning_rate": 1e-06,
+      "loss": 0.4427,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19104075877980006,
+      "grad_norm": 2.502521760012333,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19110484491156113,
+      "grad_norm": 2.4357388695095055,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19116893104332222,
+      "grad_norm": 2.7305965671623036,
+      "learning_rate": 1e-06,
+      "loss": 0.4131,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19123301717508331,
+      "grad_norm": 2.362646675213141,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1912971033068444,
+      "grad_norm": 2.7384238709786413,
+      "learning_rate": 1e-06,
+      "loss": 0.413,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19136118943860547,
+      "grad_norm": 2.5007223098863647,
+      "learning_rate": 1e-06,
+      "loss": 0.3885,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19142527557036657,
+      "grad_norm": 2.700176270926466,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 2.7085371035105346,
+      "learning_rate": 1e-06,
+      "loss": 0.4206,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19155344783388875,
+      "grad_norm": 2.79283849751298,
+      "learning_rate": 1e-06,
+      "loss": 0.4543,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19161753396564984,
+      "grad_norm": 2.7877343008577946,
+      "learning_rate": 1e-06,
+      "loss": 0.4294,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1916816200974109,
+      "grad_norm": 2.7778817346697493,
+      "learning_rate": 1e-06,
+      "loss": 0.3539,
+      "step": 2991
+    },
+    {
+      "epoch": 0.191745706229172,
+      "grad_norm": 2.6074940866398157,
+      "learning_rate": 1e-06,
+      "loss": 0.4145,
+      "step": 2992
+    },
+    {
+      "epoch": 0.1918097923609331,
+      "grad_norm": 2.7947299787206887,
+      "learning_rate": 1e-06,
+      "loss": 0.4358,
+      "step": 2993
+    },
+    {
+      "epoch": 0.1918738784926942,
+      "grad_norm": 2.550657685199485,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19193796462445528,
+      "grad_norm": 2.6333770331684154,
+      "learning_rate": 1e-06,
+      "loss": 0.3853,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19200205075621635,
+      "grad_norm": 2.6548130755758126,
+      "learning_rate": 1e-06,
+      "loss": 0.3742,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19206613688797744,
+      "grad_norm": 2.5175421052166915,
+      "learning_rate": 1e-06,
+      "loss": 0.4083,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19213022301973853,
+      "grad_norm": 2.6092498555990433,
+      "learning_rate": 1e-06,
+      "loss": 0.4065,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19219430915149963,
+      "grad_norm": 2.6746199446994057,
+      "learning_rate": 1e-06,
+      "loss": 0.3676,
+      "step": 2999
+    },
+    {
+      "epoch": 0.1922583952832607,
+      "grad_norm": 2.689102338246197,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19232248141502178,
+      "grad_norm": 2.8545617517948174,
+      "learning_rate": 1e-06,
+      "loss": 0.3947,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19238656754678288,
+      "grad_norm": 2.72399379976401,
+      "learning_rate": 1e-06,
+      "loss": 0.4357,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19245065367854397,
+      "grad_norm": 2.6992917271041943,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19251473981030506,
+      "grad_norm": 2.7469258281355975,
+      "learning_rate": 1e-06,
+      "loss": 0.4817,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19257882594206613,
+      "grad_norm": 2.8151942792068714,
+      "learning_rate": 1e-06,
+      "loss": 0.4743,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19264291207382722,
+      "grad_norm": 2.717535241339008,
+      "learning_rate": 1e-06,
+      "loss": 0.4756,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19270699820558831,
+      "grad_norm": 2.6518470118186936,
+      "learning_rate": 1e-06,
+      "loss": 0.4416,
+      "step": 3007
+    },
+    {
+      "epoch": 0.1927710843373494,
+      "grad_norm": 2.745889107117633,
+      "learning_rate": 1e-06,
+      "loss": 0.4452,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19283517046911047,
+      "grad_norm": 2.6299402934811225,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19289925660087157,
+      "grad_norm": 3.3013282687495265,
+      "learning_rate": 1e-06,
+      "loss": 0.4228,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19296334273263266,
+      "grad_norm": 2.7392676652603045,
+      "learning_rate": 1e-06,
+      "loss": 0.439,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19302742886439375,
+      "grad_norm": 2.5699388413274815,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19309151499615484,
+      "grad_norm": 2.527230563956275,
+      "learning_rate": 1e-06,
+      "loss": 0.4066,
+      "step": 3013
+    },
+    {
+      "epoch": 0.1931556011279159,
+      "grad_norm": 2.4490387549595685,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 3014
+    },
+    {
+      "epoch": 0.193219687259677,
+      "grad_norm": 2.7023787880147108,
+      "learning_rate": 1e-06,
+      "loss": 0.4121,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1932837733914381,
+      "grad_norm": 2.6746716675529654,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1933478595231992,
+      "grad_norm": 2.6345823971450373,
+      "learning_rate": 1e-06,
+      "loss": 0.4575,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19341194565496025,
+      "grad_norm": 2.7711286043104755,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19347603178672135,
+      "grad_norm": 2.4680613743712616,
+      "learning_rate": 1e-06,
+      "loss": 0.4018,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19354011791848244,
+      "grad_norm": 2.6003860886706556,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19360420405024353,
+      "grad_norm": 2.6108666448050384,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19366829018200463,
+      "grad_norm": 2.758026508393086,
+      "learning_rate": 1e-06,
+      "loss": 0.3995,
+      "step": 3022
+    },
+    {
+      "epoch": 0.1937323763137657,
+      "grad_norm": 2.4812630624240546,
+      "learning_rate": 1e-06,
+      "loss": 0.373,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19379646244552678,
+      "grad_norm": 2.5059074556288685,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19386054857728788,
+      "grad_norm": 2.700687588961143,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19392463470904897,
+      "grad_norm": 2.5778023098429625,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19398872084081004,
+      "grad_norm": 2.8092589531817236,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19405280697257113,
+      "grad_norm": 2.6755152209276702,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19411689310433222,
+      "grad_norm": 2.717840638318553,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19418097923609332,
+      "grad_norm": 2.8100797610129864,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1942450653678544,
+      "grad_norm": 2.8468932816579846,
+      "learning_rate": 1e-06,
+      "loss": 0.4597,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19430915149961547,
+      "grad_norm": 2.7200125501208388,
+      "learning_rate": 1e-06,
+      "loss": 0.4227,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19437323763137657,
+      "grad_norm": 2.745674467926342,
+      "learning_rate": 1e-06,
+      "loss": 0.3926,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19443732376313766,
+      "grad_norm": 2.66764518189243,
+      "learning_rate": 1e-06,
+      "loss": 0.4487,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19450140989489875,
+      "grad_norm": 2.72148919183853,
+      "learning_rate": 1e-06,
+      "loss": 0.431,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19456549602665982,
+      "grad_norm": 2.766399378265474,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1946295821584209,
+      "grad_norm": 2.4568489231818256,
+      "learning_rate": 1e-06,
+      "loss": 0.3821,
+      "step": 3037
+    },
+    {
+      "epoch": 0.194693668290182,
+      "grad_norm": 2.8245628648840047,
+      "learning_rate": 1e-06,
+      "loss": 0.3791,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1947577544219431,
+      "grad_norm": 2.8867994809759514,
+      "learning_rate": 1e-06,
+      "loss": 0.4492,
+      "step": 3039
+    },
+    {
+      "epoch": 0.1948218405537042,
+      "grad_norm": 2.7372903443602343,
+      "learning_rate": 1e-06,
+      "loss": 0.3816,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19488592668546525,
+      "grad_norm": 2.654619328065438,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19495001281722635,
+      "grad_norm": 2.6189824595138953,
+      "learning_rate": 1e-06,
+      "loss": 0.4104,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19501409894898744,
+      "grad_norm": 2.549845796501405,
+      "learning_rate": 1e-06,
+      "loss": 0.3848,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19507818508074853,
+      "grad_norm": 2.7252239626651438,
+      "learning_rate": 1e-06,
+      "loss": 0.4229,
+      "step": 3044
+    },
+    {
+      "epoch": 0.1951422712125096,
+      "grad_norm": 2.8890697357641875,
+      "learning_rate": 1e-06,
+      "loss": 0.4447,
+      "step": 3045
+    },
+    {
+      "epoch": 0.1952063573442707,
+      "grad_norm": 2.72838732384927,
+      "learning_rate": 1e-06,
+      "loss": 0.4355,
+      "step": 3046
+    },
+    {
+      "epoch": 0.19527044347603179,
+      "grad_norm": 2.656321398304245,
+      "learning_rate": 1e-06,
+      "loss": 0.422,
+      "step": 3047
+    },
+    {
+      "epoch": 0.19533452960779288,
+      "grad_norm": 2.6185422333681867,
+      "learning_rate": 1e-06,
+      "loss": 0.4524,
+      "step": 3048
+    },
+    {
+      "epoch": 0.19539861573955397,
+      "grad_norm": 2.5406421196673437,
+      "learning_rate": 1e-06,
+      "loss": 0.3771,
+      "step": 3049
+    },
+    {
+      "epoch": 0.19546270187131504,
+      "grad_norm": 2.7469839128125497,
+      "learning_rate": 1e-06,
+      "loss": 0.3864,
+      "step": 3050
+    },
+    {
+      "epoch": 0.19552678800307613,
+      "grad_norm": 2.4965398898043585,
+      "learning_rate": 1e-06,
+      "loss": 0.4065,
+      "step": 3051
+    },
+    {
+      "epoch": 0.19559087413483722,
+      "grad_norm": 2.4349462204869314,
+      "learning_rate": 1e-06,
+      "loss": 0.3925,
+      "step": 3052
+    },
+    {
+      "epoch": 0.19565496026659832,
+      "grad_norm": 2.466053052427371,
+      "learning_rate": 1e-06,
+      "loss": 0.4485,
+      "step": 3053
+    },
+    {
+      "epoch": 0.1957190463983594,
+      "grad_norm": 2.644641993283081,
+      "learning_rate": 1e-06,
+      "loss": 0.3925,
+      "step": 3054
+    },
+    {
+      "epoch": 0.19578313253012047,
+      "grad_norm": 2.7286313655607666,
+      "learning_rate": 1e-06,
+      "loss": 0.3796,
+      "step": 3055
+    },
+    {
+      "epoch": 0.19584721866188157,
+      "grad_norm": 3.288638532373253,
+      "learning_rate": 1e-06,
+      "loss": 0.5013,
+      "step": 3056
+    },
+    {
+      "epoch": 0.19591130479364266,
+      "grad_norm": 2.6024355715319603,
+      "learning_rate": 1e-06,
+      "loss": 0.4452,
+      "step": 3057
+    },
+    {
+      "epoch": 0.19597539092540375,
+      "grad_norm": 2.781418640960001,
+      "learning_rate": 1e-06,
+      "loss": 0.4034,
+      "step": 3058
+    },
+    {
+      "epoch": 0.19603947705716482,
+      "grad_norm": 2.5251976082178884,
+      "learning_rate": 1e-06,
+      "loss": 0.3554,
+      "step": 3059
+    },
+    {
+      "epoch": 0.1961035631889259,
+      "grad_norm": 2.7218225486270264,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 3060
+    },
+    {
+      "epoch": 0.196167649320687,
+      "grad_norm": 2.465634369754827,
+      "learning_rate": 1e-06,
+      "loss": 0.3811,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1962317354524481,
+      "grad_norm": 2.637439083827366,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 3062
+    },
+    {
+      "epoch": 0.1962958215842092,
+      "grad_norm": 2.615911737590689,
+      "learning_rate": 1e-06,
+      "loss": 0.4588,
+      "step": 3063
+    },
+    {
+      "epoch": 0.19635990771597026,
+      "grad_norm": 2.5254899516331526,
+      "learning_rate": 1e-06,
+      "loss": 0.4321,
+      "step": 3064
+    },
+    {
+      "epoch": 0.19642399384773135,
+      "grad_norm": 2.6078553888979394,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 3065
+    },
+    {
+      "epoch": 0.19648807997949244,
+      "grad_norm": 2.7298428056608466,
+      "learning_rate": 1e-06,
+      "loss": 0.4116,
+      "step": 3066
+    },
+    {
+      "epoch": 0.19655216611125353,
+      "grad_norm": 2.636498777465297,
+      "learning_rate": 1e-06,
+      "loss": 0.4114,
+      "step": 3067
+    },
+    {
+      "epoch": 0.1966162522430146,
+      "grad_norm": 2.6319615141742934,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1966803383747757,
+      "grad_norm": 2.5563489648377313,
+      "learning_rate": 1e-06,
+      "loss": 0.468,
+      "step": 3069
+    },
+    {
+      "epoch": 0.19674442450653679,
+      "grad_norm": 2.4495929287604232,
+      "learning_rate": 1e-06,
+      "loss": 0.3965,
+      "step": 3070
+    },
+    {
+      "epoch": 0.19680851063829788,
+      "grad_norm": 2.4717484721741254,
+      "learning_rate": 1e-06,
+      "loss": 0.4132,
+      "step": 3071
+    },
+    {
+      "epoch": 0.19687259677005897,
+      "grad_norm": 2.590478233813598,
+      "learning_rate": 1e-06,
+      "loss": 0.4312,
+      "step": 3072
+    },
+    {
+      "epoch": 0.19693668290182004,
+      "grad_norm": 2.7616833115828707,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 3073
+    },
+    {
+      "epoch": 0.19700076903358113,
+      "grad_norm": 2.5261292650531884,
+      "learning_rate": 1e-06,
+      "loss": 0.3813,
+      "step": 3074
+    },
+    {
+      "epoch": 0.19706485516534222,
+      "grad_norm": 2.7011738929384284,
+      "learning_rate": 1e-06,
+      "loss": 0.3816,
+      "step": 3075
+    },
+    {
+      "epoch": 0.19712894129710332,
+      "grad_norm": 2.879203428383538,
+      "learning_rate": 1e-06,
+      "loss": 0.4445,
+      "step": 3076
+    },
+    {
+      "epoch": 0.19719302742886438,
+      "grad_norm": 2.595528866437978,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 3077
+    },
+    {
+      "epoch": 0.19725711356062547,
+      "grad_norm": 2.94515771596298,
+      "learning_rate": 1e-06,
+      "loss": 0.4529,
+      "step": 3078
+    },
+    {
+      "epoch": 0.19732119969238657,
+      "grad_norm": 2.6736790736746445,
+      "learning_rate": 1e-06,
+      "loss": 0.4309,
+      "step": 3079
+    },
+    {
+      "epoch": 0.19738528582414766,
+      "grad_norm": 2.5140252504503726,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 3080
+    },
+    {
+      "epoch": 0.19744937195590875,
+      "grad_norm": 2.81911651094041,
+      "learning_rate": 1e-06,
+      "loss": 0.4125,
+      "step": 3081
+    },
+    {
+      "epoch": 0.19751345808766982,
+      "grad_norm": 2.5216675333964185,
+      "learning_rate": 1e-06,
+      "loss": 0.4035,
+      "step": 3082
+    },
+    {
+      "epoch": 0.1975775442194309,
+      "grad_norm": 2.597791352413284,
+      "learning_rate": 1e-06,
+      "loss": 0.4371,
+      "step": 3083
+    },
+    {
+      "epoch": 0.197641630351192,
+      "grad_norm": 2.6051238849878824,
+      "learning_rate": 1e-06,
+      "loss": 0.4179,
+      "step": 3084
+    },
+    {
+      "epoch": 0.1977057164829531,
+      "grad_norm": 2.4626622081291916,
+      "learning_rate": 1e-06,
+      "loss": 0.3564,
+      "step": 3085
+    },
+    {
+      "epoch": 0.19776980261471416,
+      "grad_norm": 2.6135316427042787,
+      "learning_rate": 1e-06,
+      "loss": 0.4043,
+      "step": 3086
+    },
+    {
+      "epoch": 0.19783388874647526,
+      "grad_norm": 2.497469648863096,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 3087
+    },
+    {
+      "epoch": 0.19789797487823635,
+      "grad_norm": 2.5303900994698476,
+      "learning_rate": 1e-06,
+      "loss": 0.3694,
+      "step": 3088
+    },
+    {
+      "epoch": 0.19796206100999744,
+      "grad_norm": 2.6047418213304003,
+      "learning_rate": 1e-06,
+      "loss": 0.4524,
+      "step": 3089
+    },
+    {
+      "epoch": 0.19802614714175854,
+      "grad_norm": 2.6321807982006797,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1980902332735196,
+      "grad_norm": 2.5343685708188834,
+      "learning_rate": 1e-06,
+      "loss": 0.4256,
+      "step": 3091
+    },
+    {
+      "epoch": 0.1981543194052807,
+      "grad_norm": 2.8626866297721123,
+      "learning_rate": 1e-06,
+      "loss": 0.3833,
+      "step": 3092
+    },
+    {
+      "epoch": 0.1982184055370418,
+      "grad_norm": 2.648616331118623,
+      "learning_rate": 1e-06,
+      "loss": 0.3815,
+      "step": 3093
+    },
+    {
+      "epoch": 0.19828249166880288,
+      "grad_norm": 2.7536516489830976,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 3094
+    },
+    {
+      "epoch": 0.19834657780056394,
+      "grad_norm": 2.7451936420156873,
+      "learning_rate": 1e-06,
+      "loss": 0.4143,
+      "step": 3095
+    },
+    {
+      "epoch": 0.19841066393232504,
+      "grad_norm": 2.559391512260891,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 3096
+    },
+    {
+      "epoch": 0.19847475006408613,
+      "grad_norm": 2.5924661584948376,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 3097
+    },
+    {
+      "epoch": 0.19853883619584722,
+      "grad_norm": 2.7921584349548145,
+      "learning_rate": 1e-06,
+      "loss": 0.45,
+      "step": 3098
+    },
+    {
+      "epoch": 0.19860292232760832,
+      "grad_norm": 2.659738608230625,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 3099
+    },
+    {
+      "epoch": 0.19866700845936938,
+      "grad_norm": 2.5305396379037197,
+      "learning_rate": 1e-06,
+      "loss": 0.3791,
+      "step": 3100
+    },
+    {
+      "epoch": 0.19873109459113047,
+      "grad_norm": 2.756190416272402,
+      "learning_rate": 1e-06,
+      "loss": 0.3839,
+      "step": 3101
+    },
+    {
+      "epoch": 0.19879518072289157,
+      "grad_norm": 2.539580583309545,
+      "learning_rate": 1e-06,
+      "loss": 0.4275,
+      "step": 3102
+    },
+    {
+      "epoch": 0.19885926685465266,
+      "grad_norm": 2.67813254086445,
+      "learning_rate": 1e-06,
+      "loss": 0.3512,
+      "step": 3103
+    },
+    {
+      "epoch": 0.19892335298641373,
+      "grad_norm": 2.640991640157843,
+      "learning_rate": 1e-06,
+      "loss": 0.4512,
+      "step": 3104
+    },
+    {
+      "epoch": 0.19898743911817482,
+      "grad_norm": 2.5533029619456853,
+      "learning_rate": 1e-06,
+      "loss": 0.3762,
+      "step": 3105
+    },
+    {
+      "epoch": 0.1990515252499359,
+      "grad_norm": 2.6127542430810635,
+      "learning_rate": 1e-06,
+      "loss": 0.3938,
+      "step": 3106
+    },
+    {
+      "epoch": 0.199115611381697,
+      "grad_norm": 2.6179136331616215,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 3107
+    },
+    {
+      "epoch": 0.1991796975134581,
+      "grad_norm": 2.6944181569516914,
+      "learning_rate": 1e-06,
+      "loss": 0.4073,
+      "step": 3108
+    },
+    {
+      "epoch": 0.19924378364521916,
+      "grad_norm": 2.8686676163236613,
+      "learning_rate": 1e-06,
+      "loss": 0.4284,
+      "step": 3109
+    },
+    {
+      "epoch": 0.19930786977698026,
+      "grad_norm": 2.679658713052307,
+      "learning_rate": 1e-06,
+      "loss": 0.3844,
+      "step": 3110
+    },
+    {
+      "epoch": 0.19937195590874135,
+      "grad_norm": 2.905490286430153,
+      "learning_rate": 1e-06,
+      "loss": 0.4576,
+      "step": 3111
+    },
+    {
+      "epoch": 0.19943604204050244,
+      "grad_norm": 2.77202972977206,
+      "learning_rate": 1e-06,
+      "loss": 0.4071,
+      "step": 3112
+    },
+    {
+      "epoch": 0.19950012817226354,
+      "grad_norm": 2.6987682861458877,
+      "learning_rate": 1e-06,
+      "loss": 0.3778,
+      "step": 3113
+    },
+    {
+      "epoch": 0.1995642143040246,
+      "grad_norm": 2.9302419214602664,
+      "learning_rate": 1e-06,
+      "loss": 0.4353,
+      "step": 3114
+    },
+    {
+      "epoch": 0.1996283004357857,
+      "grad_norm": 2.7112189817286887,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1996923865675468,
+      "grad_norm": 2.6842811724796687,
+      "learning_rate": 1e-06,
+      "loss": 0.4416,
+      "step": 3116
+    },
+    {
+      "epoch": 0.19975647269930788,
+      "grad_norm": 2.5186808347832708,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 3117
+    },
+    {
+      "epoch": 0.19982055883106895,
+      "grad_norm": 2.891265829188758,
+      "learning_rate": 1e-06,
+      "loss": 0.4248,
+      "step": 3118
+    },
+    {
+      "epoch": 0.19988464496283004,
+      "grad_norm": 2.445277384717998,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 3119
+    },
+    {
+      "epoch": 0.19994873109459113,
+      "grad_norm": 2.4402702476916227,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20001281722635222,
+      "grad_norm": 2.6536177038610727,
+      "learning_rate": 1e-06,
+      "loss": 0.3974,
+      "step": 3121
+    },
+    {
+      "epoch": 0.20007690335811332,
+      "grad_norm": 2.68284801164674,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 3122
+    },
+    {
+      "epoch": 0.20014098948987438,
+      "grad_norm": 2.9008015151571342,
+      "learning_rate": 1e-06,
+      "loss": 0.4362,
+      "step": 3123
+    },
+    {
+      "epoch": 0.20020507562163548,
+      "grad_norm": 2.640711993107474,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 3124
+    },
+    {
+      "epoch": 0.20026916175339657,
+      "grad_norm": 2.7448204574462416,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 3125
+    },
+    {
+      "epoch": 0.20033324788515766,
+      "grad_norm": 2.6303804716884183,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 3126
+    },
+    {
+      "epoch": 0.20039733401691873,
+      "grad_norm": 2.7394724745610897,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 3127
+    },
+    {
+      "epoch": 0.20046142014867982,
+      "grad_norm": 2.407575755511553,
+      "learning_rate": 1e-06,
+      "loss": 0.3786,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2005255062804409,
+      "grad_norm": 2.7833525732649353,
+      "learning_rate": 1e-06,
+      "loss": 0.4346,
+      "step": 3129
+    },
+    {
+      "epoch": 0.200589592412202,
+      "grad_norm": 2.7948666514800657,
+      "learning_rate": 1e-06,
+      "loss": 0.4425,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2006536785439631,
+      "grad_norm": 2.619044236684681,
+      "learning_rate": 1e-06,
+      "loss": 0.3851,
+      "step": 3131
+    },
+    {
+      "epoch": 0.20071776467572416,
+      "grad_norm": 2.7631433482197596,
+      "learning_rate": 1e-06,
+      "loss": 0.3831,
+      "step": 3132
+    },
+    {
+      "epoch": 0.20078185080748526,
+      "grad_norm": 2.7735001022993155,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 3133
+    },
+    {
+      "epoch": 0.20084593693924635,
+      "grad_norm": 2.6649390403366198,
+      "learning_rate": 1e-06,
+      "loss": 0.4513,
+      "step": 3134
+    },
+    {
+      "epoch": 0.20091002307100744,
+      "grad_norm": 2.642064035206794,
+      "learning_rate": 1e-06,
+      "loss": 0.4495,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2009741092027685,
+      "grad_norm": 2.6129073896361414,
+      "learning_rate": 1e-06,
+      "loss": 0.3816,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2010381953345296,
+      "grad_norm": 2.9648669735240447,
+      "learning_rate": 1e-06,
+      "loss": 0.4165,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2011022814662907,
+      "grad_norm": 2.8669137466757326,
+      "learning_rate": 1e-06,
+      "loss": 0.3595,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2011663675980518,
+      "grad_norm": 2.5460585330292003,
+      "learning_rate": 1e-06,
+      "loss": 0.3667,
+      "step": 3139
+    },
+    {
+      "epoch": 0.20123045372981288,
+      "grad_norm": 2.3987256825811274,
+      "learning_rate": 1e-06,
+      "loss": 0.3765,
+      "step": 3140
+    },
+    {
+      "epoch": 0.20129453986157395,
+      "grad_norm": 2.5263050591029867,
+      "learning_rate": 1e-06,
+      "loss": 0.4062,
+      "step": 3141
+    },
+    {
+      "epoch": 0.20135862599333504,
+      "grad_norm": 2.985354785460851,
+      "learning_rate": 1e-06,
+      "loss": 0.396,
+      "step": 3142
+    },
+    {
+      "epoch": 0.20142271212509613,
+      "grad_norm": 2.7659888764684264,
+      "learning_rate": 1e-06,
+      "loss": 0.4453,
+      "step": 3143
+    },
+    {
+      "epoch": 0.20148679825685722,
+      "grad_norm": 2.591450612167699,
+      "learning_rate": 1e-06,
+      "loss": 0.4083,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2015508843886183,
+      "grad_norm": 2.8521468462132544,
+      "learning_rate": 1e-06,
+      "loss": 0.4404,
+      "step": 3145
+    },
+    {
+      "epoch": 0.20161497052037938,
+      "grad_norm": 2.436797817729498,
+      "learning_rate": 1e-06,
+      "loss": 0.4238,
+      "step": 3146
+    },
+    {
+      "epoch": 0.20167905665214048,
+      "grad_norm": 2.8810258226794265,
+      "learning_rate": 1e-06,
+      "loss": 0.4529,
+      "step": 3147
+    },
+    {
+      "epoch": 0.20174314278390157,
+      "grad_norm": 2.749593307275725,
+      "learning_rate": 1e-06,
+      "loss": 0.3718,
+      "step": 3148
+    },
+    {
+      "epoch": 0.20180722891566266,
+      "grad_norm": 2.5819133802039698,
+      "learning_rate": 1e-06,
+      "loss": 0.4202,
+      "step": 3149
+    },
+    {
+      "epoch": 0.20187131504742373,
+      "grad_norm": 2.814319540956971,
+      "learning_rate": 1e-06,
+      "loss": 0.4702,
+      "step": 3150
+    },
+    {
+      "epoch": 0.20193540117918482,
+      "grad_norm": 2.755121410304437,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2019994873109459,
+      "grad_norm": 2.8607357345261186,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 3152
+    },
+    {
+      "epoch": 0.202063573442707,
+      "grad_norm": 2.7644998286160414,
+      "learning_rate": 1e-06,
+      "loss": 0.4148,
+      "step": 3153
+    },
+    {
+      "epoch": 0.20212765957446807,
+      "grad_norm": 3.1791677414919937,
+      "learning_rate": 1e-06,
+      "loss": 0.4024,
+      "step": 3154
+    },
+    {
+      "epoch": 0.20219174570622916,
+      "grad_norm": 2.761445190168834,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 3155
+    },
+    {
+      "epoch": 0.20225583183799026,
+      "grad_norm": 2.6524443855935744,
+      "learning_rate": 1e-06,
+      "loss": 0.3857,
+      "step": 3156
+    },
+    {
+      "epoch": 0.20231991796975135,
+      "grad_norm": 2.6616266519428486,
+      "learning_rate": 1e-06,
+      "loss": 0.4252,
+      "step": 3157
+    },
+    {
+      "epoch": 0.20238400410151244,
+      "grad_norm": 2.5016076188972827,
+      "learning_rate": 1e-06,
+      "loss": 0.3676,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2024480902332735,
+      "grad_norm": 2.530401535505697,
+      "learning_rate": 1e-06,
+      "loss": 0.3893,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2025121763650346,
+      "grad_norm": 2.6587919579612147,
+      "learning_rate": 1e-06,
+      "loss": 0.4009,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2025762624967957,
+      "grad_norm": 2.9566194209341,
+      "learning_rate": 1e-06,
+      "loss": 0.42,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2026403486285568,
+      "grad_norm": 2.6206510556300033,
+      "learning_rate": 1e-06,
+      "loss": 0.4262,
+      "step": 3162
+    },
+    {
+      "epoch": 0.20270443476031788,
+      "grad_norm": 2.927652472371367,
+      "learning_rate": 1e-06,
+      "loss": 0.4889,
+      "step": 3163
+    },
+    {
+      "epoch": 0.20276852089207895,
+      "grad_norm": 2.65170729751374,
+      "learning_rate": 1e-06,
+      "loss": 0.4269,
+      "step": 3164
+    },
+    {
+      "epoch": 0.20283260702384004,
+      "grad_norm": 2.5867425876599865,
+      "learning_rate": 1e-06,
+      "loss": 0.4889,
+      "step": 3165
+    },
+    {
+      "epoch": 0.20289669315560113,
+      "grad_norm": 2.7063246897863285,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 3166
+    },
+    {
+      "epoch": 0.20296077928736223,
+      "grad_norm": 2.624579109415967,
+      "learning_rate": 1e-06,
+      "loss": 0.4464,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2030248654191233,
+      "grad_norm": 2.7380146988562504,
+      "learning_rate": 1e-06,
+      "loss": 0.3857,
+      "step": 3168
+    },
+    {
+      "epoch": 0.20308895155088438,
+      "grad_norm": 2.4525698832360465,
+      "learning_rate": 1e-06,
+      "loss": 0.4224,
+      "step": 3169
+    },
+    {
+      "epoch": 0.20315303768264548,
+      "grad_norm": 2.717087375931541,
+      "learning_rate": 1e-06,
+      "loss": 0.3979,
+      "step": 3170
+    },
+    {
+      "epoch": 0.20321712381440657,
+      "grad_norm": 2.7561886591445672,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 3171
+    },
+    {
+      "epoch": 0.20328120994616766,
+      "grad_norm": 2.7637234247630853,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 3172
+    },
+    {
+      "epoch": 0.20334529607792873,
+      "grad_norm": 2.660887630337373,
+      "learning_rate": 1e-06,
+      "loss": 0.3672,
+      "step": 3173
+    },
+    {
+      "epoch": 0.20340938220968982,
+      "grad_norm": 2.7012411259252924,
+      "learning_rate": 1e-06,
+      "loss": 0.5082,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2034734683414509,
+      "grad_norm": 2.7199733769770913,
+      "learning_rate": 1e-06,
+      "loss": 0.4146,
+      "step": 3175
+    },
+    {
+      "epoch": 0.203537554473212,
+      "grad_norm": 2.8716087266272314,
+      "learning_rate": 1e-06,
+      "loss": 0.3657,
+      "step": 3176
+    },
+    {
+      "epoch": 0.20360164060497307,
+      "grad_norm": 2.757204262047918,
+      "learning_rate": 1e-06,
+      "loss": 0.418,
+      "step": 3177
+    },
+    {
+      "epoch": 0.20366572673673417,
+      "grad_norm": 2.7868293714034356,
+      "learning_rate": 1e-06,
+      "loss": 0.4133,
+      "step": 3178
+    },
+    {
+      "epoch": 0.20372981286849526,
+      "grad_norm": 2.4386658649656097,
+      "learning_rate": 1e-06,
+      "loss": 0.3242,
+      "step": 3179
+    },
+    {
+      "epoch": 0.20379389900025635,
+      "grad_norm": 2.745982025953506,
+      "learning_rate": 1e-06,
+      "loss": 0.4203,
+      "step": 3180
+    },
+    {
+      "epoch": 0.20385798513201744,
+      "grad_norm": 2.8324684296501697,
+      "learning_rate": 1e-06,
+      "loss": 0.4359,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2039220712637785,
+      "grad_norm": 2.997115716317168,
+      "learning_rate": 1e-06,
+      "loss": 0.4222,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2039861573955396,
+      "grad_norm": 2.6677174069426366,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2040502435273007,
+      "grad_norm": 2.5669472730439704,
+      "learning_rate": 1e-06,
+      "loss": 0.4617,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2041143296590618,
+      "grad_norm": 2.5467043969436656,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 3185
+    },
+    {
+      "epoch": 0.20417841579082285,
+      "grad_norm": 2.6219195102759723,
+      "learning_rate": 1e-06,
+      "loss": 0.4482,
+      "step": 3186
+    },
+    {
+      "epoch": 0.20424250192258395,
+      "grad_norm": 2.753937077422682,
+      "learning_rate": 1e-06,
+      "loss": 0.4078,
+      "step": 3187
+    },
+    {
+      "epoch": 0.20430658805434504,
+      "grad_norm": 2.5606663484409355,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 3188
+    },
+    {
+      "epoch": 0.20437067418610613,
+      "grad_norm": 2.6646453207007994,
+      "learning_rate": 1e-06,
+      "loss": 0.4301,
+      "step": 3189
+    },
+    {
+      "epoch": 0.20443476031786723,
+      "grad_norm": 2.6862551366161513,
+      "learning_rate": 1e-06,
+      "loss": 0.3938,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2044988464496283,
+      "grad_norm": 2.523615636687127,
+      "learning_rate": 1e-06,
+      "loss": 0.4181,
+      "step": 3191
+    },
+    {
+      "epoch": 0.20456293258138938,
+      "grad_norm": 2.6884484446187873,
+      "learning_rate": 1e-06,
+      "loss": 0.3618,
+      "step": 3192
+    },
+    {
+      "epoch": 0.20462701871315048,
+      "grad_norm": 2.829665325764848,
+      "learning_rate": 1e-06,
+      "loss": 0.4052,
+      "step": 3193
+    },
+    {
+      "epoch": 0.20469110484491157,
+      "grad_norm": 2.5953374746890403,
+      "learning_rate": 1e-06,
+      "loss": 0.4389,
+      "step": 3194
+    },
+    {
+      "epoch": 0.20475519097667264,
+      "grad_norm": 2.8466331258627604,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 3195
+    },
+    {
+      "epoch": 0.20481927710843373,
+      "grad_norm": 2.7486183212115547,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 3196
+    },
+    {
+      "epoch": 0.20488336324019482,
+      "grad_norm": 2.801298683547888,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 3197
+    },
+    {
+      "epoch": 0.20494744937195591,
+      "grad_norm": 2.695730931752156,
+      "learning_rate": 1e-06,
+      "loss": 0.4933,
+      "step": 3198
+    },
+    {
+      "epoch": 0.205011535503717,
+      "grad_norm": 2.8125270494981027,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 3199
+    },
+    {
+      "epoch": 0.20507562163547807,
+      "grad_norm": 3.2577167938002796,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 3200
+    },
+    {
+      "epoch": 0.20513970776723917,
+      "grad_norm": 2.758478341893175,
+      "learning_rate": 1e-06,
+      "loss": 0.4352,
+      "step": 3201
+    },
+    {
+      "epoch": 0.20520379389900026,
+      "grad_norm": 2.879050943230032,
+      "learning_rate": 1e-06,
+      "loss": 0.4305,
+      "step": 3202
+    },
+    {
+      "epoch": 0.20526788003076135,
+      "grad_norm": 2.686643342867524,
+      "learning_rate": 1e-06,
+      "loss": 0.3573,
+      "step": 3203
+    },
+    {
+      "epoch": 0.20533196616252242,
+      "grad_norm": 2.721258382735068,
+      "learning_rate": 1e-06,
+      "loss": 0.4019,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2053960522942835,
+      "grad_norm": 2.7158203178493854,
+      "learning_rate": 1e-06,
+      "loss": 0.43,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2054601384260446,
+      "grad_norm": 2.624134577750634,
+      "learning_rate": 1e-06,
+      "loss": 0.3608,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2055242245578057,
+      "grad_norm": 2.720833641453419,
+      "learning_rate": 1e-06,
+      "loss": 0.4991,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2055883106895668,
+      "grad_norm": 2.7278965888939255,
+      "learning_rate": 1e-06,
+      "loss": 0.4445,
+      "step": 3208
+    },
+    {
+      "epoch": 0.20565239682132785,
+      "grad_norm": 2.9254754982044346,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 3209
+    },
+    {
+      "epoch": 0.20571648295308895,
+      "grad_norm": 2.6542985721484382,
+      "learning_rate": 1e-06,
+      "loss": 0.3975,
+      "step": 3210
+    },
+    {
+      "epoch": 0.20578056908485004,
+      "grad_norm": 2.699604037727829,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 3211
+    },
+    {
+      "epoch": 0.20584465521661113,
+      "grad_norm": 2.7276421722192525,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2059087413483722,
+      "grad_norm": 2.6977864355232715,
+      "learning_rate": 1e-06,
+      "loss": 0.4003,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2059728274801333,
+      "grad_norm": 2.6477848771201886,
+      "learning_rate": 1e-06,
+      "loss": 0.4429,
+      "step": 3214
+    },
+    {
+      "epoch": 0.20603691361189438,
+      "grad_norm": 2.672606633365545,
+      "learning_rate": 1e-06,
+      "loss": 0.3984,
+      "step": 3215
+    },
+    {
+      "epoch": 0.20610099974365548,
+      "grad_norm": 2.4604151399516945,
+      "learning_rate": 1e-06,
+      "loss": 0.4922,
+      "step": 3216
+    },
+    {
+      "epoch": 0.20616508587541657,
+      "grad_norm": 2.857586432951352,
+      "learning_rate": 1e-06,
+      "loss": 0.3976,
+      "step": 3217
+    },
+    {
+      "epoch": 0.20622917200717764,
+      "grad_norm": 2.4822416440627264,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 3218
+    },
+    {
+      "epoch": 0.20629325813893873,
+      "grad_norm": 2.6436355434453565,
+      "learning_rate": 1e-06,
+      "loss": 0.3792,
+      "step": 3219
+    },
+    {
+      "epoch": 0.20635734427069982,
+      "grad_norm": 2.6742355296426212,
+      "learning_rate": 1e-06,
+      "loss": 0.4404,
+      "step": 3220
+    },
+    {
+      "epoch": 0.20642143040246091,
+      "grad_norm": 2.8105861214534453,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 3221
+    },
+    {
+      "epoch": 0.206485516534222,
+      "grad_norm": 2.5930834712675876,
+      "learning_rate": 1e-06,
+      "loss": 0.4299,
+      "step": 3222
+    },
+    {
+      "epoch": 0.20654960266598307,
+      "grad_norm": 2.5829548868133654,
+      "learning_rate": 1e-06,
+      "loss": 0.4105,
+      "step": 3223
+    },
+    {
+      "epoch": 0.20661368879774417,
+      "grad_norm": 2.719837435467215,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 3224
+    },
+    {
+      "epoch": 0.20667777492950526,
+      "grad_norm": 2.61535271989138,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 3225
+    },
+    {
+      "epoch": 0.20674186106126635,
+      "grad_norm": 2.5771180704151755,
+      "learning_rate": 1e-06,
+      "loss": 0.4275,
+      "step": 3226
+    },
+    {
+      "epoch": 0.20680594719302742,
+      "grad_norm": 2.672183893932333,
+      "learning_rate": 1e-06,
+      "loss": 0.4107,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2068700333247885,
+      "grad_norm": 2.773792137100908,
+      "learning_rate": 1e-06,
+      "loss": 0.4827,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2069341194565496,
+      "grad_norm": 2.681428202763583,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2069982055883107,
+      "grad_norm": 2.9492158799161796,
+      "learning_rate": 1e-06,
+      "loss": 0.4708,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2070622917200718,
+      "grad_norm": 2.8897345694841534,
+      "learning_rate": 1e-06,
+      "loss": 0.4303,
+      "step": 3231
+    },
+    {
+      "epoch": 0.20712637785183285,
+      "grad_norm": 2.759597729311294,
+      "learning_rate": 1e-06,
+      "loss": 0.4035,
+      "step": 3232
+    },
+    {
+      "epoch": 0.20719046398359395,
+      "grad_norm": 2.840421542454811,
+      "learning_rate": 1e-06,
+      "loss": 0.4273,
+      "step": 3233
+    },
+    {
+      "epoch": 0.20725455011535504,
+      "grad_norm": 2.739728384977558,
+      "learning_rate": 1e-06,
+      "loss": 0.4528,
+      "step": 3234
+    },
+    {
+      "epoch": 0.20731863624711613,
+      "grad_norm": 2.824433398735102,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2073827223788772,
+      "grad_norm": 2.6354881167840505,
+      "learning_rate": 1e-06,
+      "loss": 0.3858,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2074468085106383,
+      "grad_norm": 2.8945305522820477,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 3237
+    },
+    {
+      "epoch": 0.20751089464239938,
+      "grad_norm": 2.764839911328175,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 3238
+    },
+    {
+      "epoch": 0.20757498077416048,
+      "grad_norm": 2.773496463928611,
+      "learning_rate": 1e-06,
+      "loss": 0.5385,
+      "step": 3239
+    },
+    {
+      "epoch": 0.20763906690592157,
+      "grad_norm": 2.7564360890775084,
+      "learning_rate": 1e-06,
+      "loss": 0.4388,
+      "step": 3240
+    },
+    {
+      "epoch": 0.20770315303768264,
+      "grad_norm": 2.6660067624046637,
+      "learning_rate": 1e-06,
+      "loss": 0.4451,
+      "step": 3241
+    },
+    {
+      "epoch": 0.20776723916944373,
+      "grad_norm": 2.3442950677854797,
+      "learning_rate": 1e-06,
+      "loss": 0.3913,
+      "step": 3242
+    },
+    {
+      "epoch": 0.20783132530120482,
+      "grad_norm": 2.6071034795612014,
+      "learning_rate": 1e-06,
+      "loss": 0.35,
+      "step": 3243
+    },
+    {
+      "epoch": 0.20789541143296592,
+      "grad_norm": 2.5404513415884065,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 3244
+    },
+    {
+      "epoch": 0.20795949756472698,
+      "grad_norm": 2.732016662782567,
+      "learning_rate": 1e-06,
+      "loss": 0.4083,
+      "step": 3245
+    },
+    {
+      "epoch": 0.20802358369648807,
+      "grad_norm": 2.470865179331822,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 3246
+    },
+    {
+      "epoch": 0.20808766982824917,
+      "grad_norm": 2.8312892904633227,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 3247
+    },
+    {
+      "epoch": 0.20815175596001026,
+      "grad_norm": 2.9626699856363143,
+      "learning_rate": 1e-06,
+      "loss": 0.3919,
+      "step": 3248
+    },
+    {
+      "epoch": 0.20821584209177135,
+      "grad_norm": 2.7202318287031617,
+      "learning_rate": 1e-06,
+      "loss": 0.3718,
+      "step": 3249
+    },
+    {
+      "epoch": 0.20827992822353242,
+      "grad_norm": 2.5074669078149814,
+      "learning_rate": 1e-06,
+      "loss": 0.4725,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2083440143552935,
+      "grad_norm": 2.9385941662449473,
+      "learning_rate": 1e-06,
+      "loss": 0.4403,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2084081004870546,
+      "grad_norm": 2.748212967840643,
+      "learning_rate": 1e-06,
+      "loss": 0.3987,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2084721866188157,
+      "grad_norm": 2.6400904306111572,
+      "learning_rate": 1e-06,
+      "loss": 0.4194,
+      "step": 3253
+    },
+    {
+      "epoch": 0.20853627275057676,
+      "grad_norm": 2.7657922553991554,
+      "learning_rate": 1e-06,
+      "loss": 0.4343,
+      "step": 3254
+    },
+    {
+      "epoch": 0.20860035888233786,
+      "grad_norm": 2.6880821418013046,
+      "learning_rate": 1e-06,
+      "loss": 0.3921,
+      "step": 3255
+    },
+    {
+      "epoch": 0.20866444501409895,
+      "grad_norm": 2.5586073503420756,
+      "learning_rate": 1e-06,
+      "loss": 0.3641,
+      "step": 3256
+    },
+    {
+      "epoch": 0.20872853114586004,
+      "grad_norm": 2.7781977676069314,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 3257
+    },
+    {
+      "epoch": 0.20879261727762113,
+      "grad_norm": 2.6557419528008372,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2088567034093822,
+      "grad_norm": 2.5621653160572184,
+      "learning_rate": 1e-06,
+      "loss": 0.3763,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2089207895411433,
+      "grad_norm": 2.9311715852645515,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 3260
+    },
+    {
+      "epoch": 0.20898487567290439,
+      "grad_norm": 3.0964528455275517,
+      "learning_rate": 1e-06,
+      "loss": 0.4291,
+      "step": 3261
+    },
+    {
+      "epoch": 0.20904896180466548,
+      "grad_norm": 2.9231040164499826,
+      "learning_rate": 1e-06,
+      "loss": 0.3814,
+      "step": 3262
+    },
+    {
+      "epoch": 0.20911304793642654,
+      "grad_norm": 2.449860144323429,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 3263
+    },
+    {
+      "epoch": 0.20917713406818764,
+      "grad_norm": 2.7797757734601145,
+      "learning_rate": 1e-06,
+      "loss": 0.4768,
+      "step": 3264
+    },
+    {
+      "epoch": 0.20924122019994873,
+      "grad_norm": 2.813532878048226,
+      "learning_rate": 1e-06,
+      "loss": 0.4363,
+      "step": 3265
+    },
+    {
+      "epoch": 0.20930530633170982,
+      "grad_norm": 2.66716937995186,
+      "learning_rate": 1e-06,
+      "loss": 0.4468,
+      "step": 3266
+    },
+    {
+      "epoch": 0.20936939246347092,
+      "grad_norm": 2.5170598395883825,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 3267
+    },
+    {
+      "epoch": 0.20943347859523198,
+      "grad_norm": 2.674849349647163,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 3268
+    },
+    {
+      "epoch": 0.20949756472699307,
+      "grad_norm": 2.624379949822509,
+      "learning_rate": 1e-06,
+      "loss": 0.3982,
+      "step": 3269
+    },
+    {
+      "epoch": 0.20956165085875417,
+      "grad_norm": 2.820816841414604,
+      "learning_rate": 1e-06,
+      "loss": 0.395,
+      "step": 3270
+    },
+    {
+      "epoch": 0.20962573699051526,
+      "grad_norm": 2.527956786895495,
+      "learning_rate": 1e-06,
+      "loss": 0.4627,
+      "step": 3271
+    },
+    {
+      "epoch": 0.20968982312227635,
+      "grad_norm": 2.725541845425045,
+      "learning_rate": 1e-06,
+      "loss": 0.4151,
+      "step": 3272
+    },
+    {
+      "epoch": 0.20975390925403742,
+      "grad_norm": 2.878927720490599,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2098179953857985,
+      "grad_norm": 2.9259836698705963,
+      "learning_rate": 1e-06,
+      "loss": 0.4036,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2098820815175596,
+      "grad_norm": 2.7982781093411586,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2099461676493207,
+      "grad_norm": 2.5793222625428953,
+      "learning_rate": 1e-06,
+      "loss": 0.4232,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21001025378108176,
+      "grad_norm": 2.9796808226633757,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21007433991284286,
+      "grad_norm": 2.4441804847492343,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21013842604460395,
+      "grad_norm": 2.925764791415414,
+      "learning_rate": 1e-06,
+      "loss": 0.3994,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21020251217636504,
+      "grad_norm": 2.633148554481942,
+      "learning_rate": 1e-06,
+      "loss": 0.4208,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21026659830812613,
+      "grad_norm": 2.5106080788505984,
+      "learning_rate": 1e-06,
+      "loss": 0.4771,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2103306844398872,
+      "grad_norm": 2.6172458561738865,
+      "learning_rate": 1e-06,
+      "loss": 0.4521,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2103947705716483,
+      "grad_norm": 2.5929935788049487,
+      "learning_rate": 1e-06,
+      "loss": 0.3985,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21045885670340939,
+      "grad_norm": 2.7905184923144657,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21052294283517048,
+      "grad_norm": 2.5347958919401874,
+      "learning_rate": 1e-06,
+      "loss": 0.3942,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21058702896693154,
+      "grad_norm": 2.8626532701014487,
+      "learning_rate": 1e-06,
+      "loss": 0.4876,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21065111509869264,
+      "grad_norm": 2.757452199734976,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21071520123045373,
+      "grad_norm": 2.903120359363833,
+      "learning_rate": 1e-06,
+      "loss": 0.4326,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21077928736221482,
+      "grad_norm": 2.880237709233351,
+      "learning_rate": 1e-06,
+      "loss": 0.4016,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21084337349397592,
+      "grad_norm": 2.553661952738896,
+      "learning_rate": 1e-06,
+      "loss": 0.3995,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21090745962573698,
+      "grad_norm": 2.5558492231930363,
+      "learning_rate": 1e-06,
+      "loss": 0.3656,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21097154575749807,
+      "grad_norm": 2.7074908240403066,
+      "learning_rate": 1e-06,
+      "loss": 0.4687,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21103563188925917,
+      "grad_norm": 2.505986301029715,
+      "learning_rate": 1e-06,
+      "loss": 0.4268,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21109971802102026,
+      "grad_norm": 2.663176507171353,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21116380415278133,
+      "grad_norm": 2.680302070013859,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21122789028454242,
+      "grad_norm": 2.561409369888171,
+      "learning_rate": 1e-06,
+      "loss": 0.3794,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2112919764163035,
+      "grad_norm": 2.419773725787716,
+      "learning_rate": 1e-06,
+      "loss": 0.4255,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2113560625480646,
+      "grad_norm": 2.982157224673509,
+      "learning_rate": 1e-06,
+      "loss": 0.4163,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2114201486798257,
+      "grad_norm": 2.5932585562032577,
+      "learning_rate": 1e-06,
+      "loss": 0.3908,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21148423481158676,
+      "grad_norm": 2.5018110031718033,
+      "learning_rate": 1e-06,
+      "loss": 0.4069,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21154832094334786,
+      "grad_norm": 2.742542865770265,
+      "learning_rate": 1e-06,
+      "loss": 0.4739,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21161240707510895,
+      "grad_norm": 2.5360610155104784,
+      "learning_rate": 1e-06,
+      "loss": 0.4523,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21167649320687004,
+      "grad_norm": 2.551525084190843,
+      "learning_rate": 1e-06,
+      "loss": 0.4041,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2117405793386311,
+      "grad_norm": 2.434809756110649,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2118046654703922,
+      "grad_norm": 2.411436200322004,
+      "learning_rate": 1e-06,
+      "loss": 0.4291,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2118687516021533,
+      "grad_norm": 2.905538512840825,
+      "learning_rate": 1e-06,
+      "loss": 0.4184,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2119328377339144,
+      "grad_norm": 2.9864965713027205,
+      "learning_rate": 1e-06,
+      "loss": 0.4506,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21199692386567548,
+      "grad_norm": 2.742136252581274,
+      "learning_rate": 1e-06,
+      "loss": 0.3788,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21206100999743654,
+      "grad_norm": 2.5326228134750206,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21212509612919764,
+      "grad_norm": 2.830915338256494,
+      "learning_rate": 1e-06,
+      "loss": 0.417,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21218918226095873,
+      "grad_norm": 2.562052867540231,
+      "learning_rate": 1e-06,
+      "loss": 0.4002,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21225326839271982,
+      "grad_norm": 2.7686620430274655,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2123173545244809,
+      "grad_norm": 2.7678823836226916,
+      "learning_rate": 1e-06,
+      "loss": 0.4272,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21238144065624198,
+      "grad_norm": 2.5608644724506515,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21244552678800308,
+      "grad_norm": 2.7431587712595635,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21250961291976417,
+      "grad_norm": 2.6725004227866846,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21257369905152526,
+      "grad_norm": 2.6414894647068774,
+      "learning_rate": 1e-06,
+      "loss": 0.4033,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21263778518328633,
+      "grad_norm": 2.6201335968447186,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21270187131504742,
+      "grad_norm": 2.6344494226288595,
+      "learning_rate": 1e-06,
+      "loss": 0.443,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 2.6728837469432785,
+      "learning_rate": 1e-06,
+      "loss": 0.4037,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2128300435785696,
+      "grad_norm": 2.5284088187012155,
+      "learning_rate": 1e-06,
+      "loss": 0.3615,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21289412971033067,
+      "grad_norm": 2.7400867894135406,
+      "learning_rate": 1e-06,
+      "loss": 0.4012,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21295821584209176,
+      "grad_norm": 2.8026338036476974,
+      "learning_rate": 1e-06,
+      "loss": 0.432,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21302230197385286,
+      "grad_norm": 2.733888181563427,
+      "learning_rate": 1e-06,
+      "loss": 0.4293,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21308638810561395,
+      "grad_norm": 2.587643683316493,
+      "learning_rate": 1e-06,
+      "loss": 0.374,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21315047423737504,
+      "grad_norm": 2.6521670469030143,
+      "learning_rate": 1e-06,
+      "loss": 0.4076,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2132145603691361,
+      "grad_norm": 2.858247171493016,
+      "learning_rate": 1e-06,
+      "loss": 0.4151,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2132786465008972,
+      "grad_norm": 2.9518990696313554,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2133427326326583,
+      "grad_norm": 2.5807436920995706,
+      "learning_rate": 1e-06,
+      "loss": 0.3848,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2134068187644194,
+      "grad_norm": 2.692268668879906,
+      "learning_rate": 1e-06,
+      "loss": 0.4469,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21347090489618048,
+      "grad_norm": 2.548409043303377,
+      "learning_rate": 1e-06,
+      "loss": 0.4276,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21353499102794155,
+      "grad_norm": 2.457547361289557,
+      "learning_rate": 1e-06,
+      "loss": 0.4154,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21359907715970264,
+      "grad_norm": 3.140234343189398,
+      "learning_rate": 1e-06,
+      "loss": 0.3672,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21366316329146373,
+      "grad_norm": 2.5891886639159987,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21372724942322482,
+      "grad_norm": 2.8625826627525175,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2137913355549859,
+      "grad_norm": 2.509069922553237,
+      "learning_rate": 1e-06,
+      "loss": 0.4424,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21385542168674698,
+      "grad_norm": 2.6593254396801727,
+      "learning_rate": 1e-06,
+      "loss": 0.3739,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21391950781850808,
+      "grad_norm": 2.5553613531487565,
+      "learning_rate": 1e-06,
+      "loss": 0.4033,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21398359395026917,
+      "grad_norm": 2.7534326763941133,
+      "learning_rate": 1e-06,
+      "loss": 0.4361,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21404768008203026,
+      "grad_norm": 2.809417915523082,
+      "learning_rate": 1e-06,
+      "loss": 0.4618,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21411176621379133,
+      "grad_norm": 2.5990675730161152,
+      "learning_rate": 1e-06,
+      "loss": 0.4367,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21417585234555242,
+      "grad_norm": 2.7348091981913756,
+      "learning_rate": 1e-06,
+      "loss": 0.4644,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2142399384773135,
+      "grad_norm": 2.4841998191988854,
+      "learning_rate": 1e-06,
+      "loss": 0.4541,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2143040246090746,
+      "grad_norm": 2.513354725827738,
+      "learning_rate": 1e-06,
+      "loss": 0.3605,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21436811074083567,
+      "grad_norm": 2.6741945967742975,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21443219687259676,
+      "grad_norm": 2.6481448811114743,
+      "learning_rate": 1e-06,
+      "loss": 0.4211,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21449628300435786,
+      "grad_norm": 2.881638476738493,
+      "learning_rate": 1e-06,
+      "loss": 0.4405,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21456036913611895,
+      "grad_norm": 2.5160389638004594,
+      "learning_rate": 1e-06,
+      "loss": 0.4447,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21462445526788004,
+      "grad_norm": 2.6219438331059846,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2146885413996411,
+      "grad_norm": 2.7939679761335587,
+      "learning_rate": 1e-06,
+      "loss": 0.4316,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2147526275314022,
+      "grad_norm": 2.742546179444123,
+      "learning_rate": 1e-06,
+      "loss": 0.4001,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2148167136631633,
+      "grad_norm": 2.7588570120175753,
+      "learning_rate": 1e-06,
+      "loss": 0.4035,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2148807997949244,
+      "grad_norm": 2.9727016909828863,
+      "learning_rate": 1e-06,
+      "loss": 0.4292,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21494488592668545,
+      "grad_norm": 2.700187053010241,
+      "learning_rate": 1e-06,
+      "loss": 0.377,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21500897205844655,
+      "grad_norm": 2.8029003823348546,
+      "learning_rate": 1e-06,
+      "loss": 0.4384,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21507305819020764,
+      "grad_norm": 2.6629430980581237,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21513714432196873,
+      "grad_norm": 2.757531931586532,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21520123045372982,
+      "grad_norm": 2.918032036734727,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2152653165854909,
+      "grad_norm": 2.989072875608744,
+      "learning_rate": 1e-06,
+      "loss": 0.417,
+      "step": 3359
+    },
+    {
+      "epoch": 0.21532940271725198,
+      "grad_norm": 2.754298990911609,
+      "learning_rate": 1e-06,
+      "loss": 0.4508,
+      "step": 3360
+    },
+    {
+      "epoch": 0.21539348884901308,
+      "grad_norm": 2.6404117189428407,
+      "learning_rate": 1e-06,
+      "loss": 0.4318,
+      "step": 3361
+    },
+    {
+      "epoch": 0.21545757498077417,
+      "grad_norm": 3.3312953457728525,
+      "learning_rate": 1e-06,
+      "loss": 0.3877,
+      "step": 3362
+    },
+    {
+      "epoch": 0.21552166111253523,
+      "grad_norm": 2.775685646675011,
+      "learning_rate": 1e-06,
+      "loss": 0.4369,
+      "step": 3363
+    },
+    {
+      "epoch": 0.21558574724429633,
+      "grad_norm": 2.7227277955938005,
+      "learning_rate": 1e-06,
+      "loss": 0.439,
+      "step": 3364
+    },
+    {
+      "epoch": 0.21564983337605742,
+      "grad_norm": 2.7479986060865693,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2157139195078185,
+      "grad_norm": 2.749304608901622,
+      "learning_rate": 1e-06,
+      "loss": 0.407,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2157780056395796,
+      "grad_norm": 2.677492736617638,
+      "learning_rate": 1e-06,
+      "loss": 0.4595,
+      "step": 3367
+    },
+    {
+      "epoch": 0.21584209177134067,
+      "grad_norm": 2.721007325759014,
+      "learning_rate": 1e-06,
+      "loss": 0.3814,
+      "step": 3368
+    },
+    {
+      "epoch": 0.21590617790310176,
+      "grad_norm": 2.8715102553121805,
+      "learning_rate": 1e-06,
+      "loss": 0.4608,
+      "step": 3369
+    },
+    {
+      "epoch": 0.21597026403486286,
+      "grad_norm": 2.7047494169852784,
+      "learning_rate": 1e-06,
+      "loss": 0.4142,
+      "step": 3370
+    },
+    {
+      "epoch": 0.21603435016662395,
+      "grad_norm": 2.390858891462131,
+      "learning_rate": 1e-06,
+      "loss": 0.4006,
+      "step": 3371
+    },
+    {
+      "epoch": 0.21609843629838502,
+      "grad_norm": 2.688612240628484,
+      "learning_rate": 1e-06,
+      "loss": 0.3934,
+      "step": 3372
+    },
+    {
+      "epoch": 0.2161625224301461,
+      "grad_norm": 2.3726230078498354,
+      "learning_rate": 1e-06,
+      "loss": 0.3854,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2162266085619072,
+      "grad_norm": 2.4845931896277924,
+      "learning_rate": 1e-06,
+      "loss": 0.3693,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2162906946936683,
+      "grad_norm": 2.569118902647708,
+      "learning_rate": 1e-06,
+      "loss": 0.4183,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2163547808254294,
+      "grad_norm": 2.8433001863545884,
+      "learning_rate": 1e-06,
+      "loss": 0.4537,
+      "step": 3376
+    },
+    {
+      "epoch": 0.21641886695719045,
+      "grad_norm": 2.6467296962041003,
+      "learning_rate": 1e-06,
+      "loss": 0.3823,
+      "step": 3377
+    },
+    {
+      "epoch": 0.21648295308895155,
+      "grad_norm": 2.5467529724807223,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 3378
+    },
+    {
+      "epoch": 0.21654703922071264,
+      "grad_norm": 2.6583699584853515,
+      "learning_rate": 1e-06,
+      "loss": 0.5313,
+      "step": 3379
+    },
+    {
+      "epoch": 0.21661112535247373,
+      "grad_norm": 2.7413522163195903,
+      "learning_rate": 1e-06,
+      "loss": 0.3918,
+      "step": 3380
+    },
+    {
+      "epoch": 0.21667521148423483,
+      "grad_norm": 2.850274265246485,
+      "learning_rate": 1e-06,
+      "loss": 0.3708,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2167392976159959,
+      "grad_norm": 2.4545951876671444,
+      "learning_rate": 1e-06,
+      "loss": 0.4202,
+      "step": 3382
+    },
+    {
+      "epoch": 0.21680338374775698,
+      "grad_norm": 2.7541006484296826,
+      "learning_rate": 1e-06,
+      "loss": 0.4132,
+      "step": 3383
+    },
+    {
+      "epoch": 0.21686746987951808,
+      "grad_norm": 2.654628145662869,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 3384
+    },
+    {
+      "epoch": 0.21693155601127917,
+      "grad_norm": 2.679905549390188,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 3385
+    },
+    {
+      "epoch": 0.21699564214304023,
+      "grad_norm": 2.7026520249743853,
+      "learning_rate": 1e-06,
+      "loss": 0.411,
+      "step": 3386
+    },
+    {
+      "epoch": 0.21705972827480133,
+      "grad_norm": 2.8935314931031972,
+      "learning_rate": 1e-06,
+      "loss": 0.4039,
+      "step": 3387
+    },
+    {
+      "epoch": 0.21712381440656242,
+      "grad_norm": 2.8256596007112678,
+      "learning_rate": 1e-06,
+      "loss": 0.4511,
+      "step": 3388
+    },
+    {
+      "epoch": 0.21718790053832351,
+      "grad_norm": 2.666041222942203,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2172519866700846,
+      "grad_norm": 2.6944711686890104,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 3390
+    },
+    {
+      "epoch": 0.21731607280184567,
+      "grad_norm": 2.550478836812208,
+      "learning_rate": 1e-06,
+      "loss": 0.399,
+      "step": 3391
+    },
+    {
+      "epoch": 0.21738015893360677,
+      "grad_norm": 2.614636062866319,
+      "learning_rate": 1e-06,
+      "loss": 0.4499,
+      "step": 3392
+    },
+    {
+      "epoch": 0.21744424506536786,
+      "grad_norm": 2.668976199272245,
+      "learning_rate": 1e-06,
+      "loss": 0.459,
+      "step": 3393
+    },
+    {
+      "epoch": 0.21750833119712895,
+      "grad_norm": 2.5668912370881976,
+      "learning_rate": 1e-06,
+      "loss": 0.3908,
+      "step": 3394
+    },
+    {
+      "epoch": 0.21757241732889002,
+      "grad_norm": 2.8245523377207307,
+      "learning_rate": 1e-06,
+      "loss": 0.4191,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2176365034606511,
+      "grad_norm": 2.6904693651871763,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2177005895924122,
+      "grad_norm": 2.754562661101229,
+      "learning_rate": 1e-06,
+      "loss": 0.3947,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2177646757241733,
+      "grad_norm": 2.63518445508415,
+      "learning_rate": 1e-06,
+      "loss": 0.4651,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2178287618559344,
+      "grad_norm": 2.377123367412555,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 3399
+    },
+    {
+      "epoch": 0.21789284798769545,
+      "grad_norm": 2.674782214632184,
+      "learning_rate": 1e-06,
+      "loss": 0.453,
+      "step": 3400
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15604,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 825505490534400.0,
+  "train_batch_size": 10,
+  "trial_name": null,
+  "trial_params": null
+}