diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4465 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 0,
+  "global_step": 633,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001579778830963665,
+      "grad_norm": 3.3021833896636963,
+      "learning_rate": 1e-05,
+      "loss": 0.8142,
+      "step": 1
+    },
+    {
+      "epoch": 0.00315955766192733,
+      "grad_norm": 0.5667713284492493,
+      "learning_rate": 9.984202211690363e-06,
+      "loss": 0.4081,
+      "step": 2
+    },
+    {
+      "epoch": 0.004739336492890996,
+      "grad_norm": 7.904314994812012,
+      "learning_rate": 9.968404423380728e-06,
+      "loss": 1.1876,
+      "step": 3
+    },
+    {
+      "epoch": 0.00631911532385466,
+      "grad_norm": 10.157713890075684,
+      "learning_rate": 9.95260663507109e-06,
+      "loss": 1.4092,
+      "step": 4
+    },
+    {
+      "epoch": 0.007898894154818325,
+      "grad_norm": 4.723056316375732,
+      "learning_rate": 9.936808846761454e-06,
+      "loss": 0.7578,
+      "step": 5
+    },
+    {
+      "epoch": 0.009478672985781991,
+      "grad_norm": 7.033465385437012,
+      "learning_rate": 9.921011058451816e-06,
+      "loss": 0.5175,
+      "step": 6
+    },
+    {
+      "epoch": 0.011058451816745656,
+      "grad_norm": 0.800440788269043,
+      "learning_rate": 9.905213270142182e-06,
+      "loss": 0.4077,
+      "step": 7
+    },
+    {
+      "epoch": 0.01263823064770932,
+      "grad_norm": 0.6944026350975037,
+      "learning_rate": 9.889415481832544e-06,
+      "loss": 0.4686,
+      "step": 8
+    },
+    {
+      "epoch": 0.014218009478672985,
+      "grad_norm": 0.5700448751449585,
+      "learning_rate": 9.873617693522908e-06,
+      "loss": 0.3623,
+      "step": 9
+    },
+    {
+      "epoch": 0.01579778830963665,
+      "grad_norm": 0.7115408778190613,
+      "learning_rate": 9.85781990521327e-06,
+      "loss": 0.4727,
+      "step": 10
+    },
+    {
+      "epoch": 0.017377567140600316,
+      "grad_norm": 0.5764197707176208,
+      "learning_rate": 9.842022116903635e-06,
+      "loss": 0.4054,
+      "step": 11
+    },
+    {
+      "epoch": 0.018957345971563982,
+      "grad_norm": 0.615205705165863,
+      "learning_rate": 9.826224328593997e-06,
+      "loss": 0.3798,
+      "step": 12
+    },
+    {
+      "epoch": 0.020537124802527645,
+      "grad_norm": 0.6402739882469177,
+      "learning_rate": 9.810426540284361e-06,
+      "loss": 0.3966,
+      "step": 13
+    },
+    {
+      "epoch": 0.022116903633491312,
+      "grad_norm": 0.6007937788963318,
+      "learning_rate": 9.794628751974725e-06,
+      "loss": 0.4158,
+      "step": 14
+    },
+    {
+      "epoch": 0.023696682464454975,
+      "grad_norm": 0.5462563037872314,
+      "learning_rate": 9.778830963665089e-06,
+      "loss": 0.4795,
+      "step": 15
+    },
+    {
+      "epoch": 0.02527646129541864,
+      "grad_norm": 0.6038461923599243,
+      "learning_rate": 9.76303317535545e-06,
+      "loss": 0.4142,
+      "step": 16
+    },
+    {
+      "epoch": 0.026856240126382307,
+      "grad_norm": 0.514258861541748,
+      "learning_rate": 9.747235387045815e-06,
+      "loss": 0.4139,
+      "step": 17
+    },
+    {
+      "epoch": 0.02843601895734597,
+      "grad_norm": 0.728235125541687,
+      "learning_rate": 9.731437598736178e-06,
+      "loss": 0.3129,
+      "step": 18
+    },
+    {
+      "epoch": 0.030015797788309637,
+      "grad_norm": 0.7013534307479858,
+      "learning_rate": 9.715639810426542e-06,
+      "loss": 0.4275,
+      "step": 19
+    },
+    {
+      "epoch": 0.0315955766192733,
+      "grad_norm": 0.6062476634979248,
+      "learning_rate": 9.699842022116904e-06,
+      "loss": 0.3961,
+      "step": 20
+    },
+    {
+      "epoch": 0.03317535545023697,
+      "grad_norm": 0.6089779138565063,
+      "learning_rate": 9.684044233807268e-06,
+      "loss": 0.4972,
+      "step": 21
+    },
+    {
+      "epoch": 0.03475513428120063,
+      "grad_norm": 0.6651365756988525,
+      "learning_rate": 9.668246445497632e-06,
+      "loss": 0.4714,
+      "step": 22
+    },
+    {
+      "epoch": 0.036334913112164295,
+      "grad_norm": 0.6064260601997375,
+      "learning_rate": 9.652448657187995e-06,
+      "loss": 0.4358,
+      "step": 23
+    },
+    {
+      "epoch": 0.037914691943127965,
+      "grad_norm": 0.5868542790412903,
+      "learning_rate": 9.636650868878358e-06,
+      "loss": 0.5178,
+      "step": 24
+    },
+    {
+      "epoch": 0.03949447077409163,
+      "grad_norm": 0.6516690850257874,
+      "learning_rate": 9.620853080568721e-06,
+      "loss": 0.4281,
+      "step": 25
+    },
+    {
+      "epoch": 0.04107424960505529,
+      "grad_norm": 0.7721027731895447,
+      "learning_rate": 9.605055292259085e-06,
+      "loss": 0.4979,
+      "step": 26
+    },
+    {
+      "epoch": 0.04265402843601896,
+      "grad_norm": 0.6200973987579346,
+      "learning_rate": 9.589257503949447e-06,
+      "loss": 0.347,
+      "step": 27
+    },
+    {
+      "epoch": 0.044233807266982623,
+      "grad_norm": 0.6557235717773438,
+      "learning_rate": 9.573459715639811e-06,
+      "loss": 0.3422,
+      "step": 28
+    },
+    {
+      "epoch": 0.045813586097946286,
+      "grad_norm": 1.0422502756118774,
+      "learning_rate": 9.557661927330175e-06,
+      "loss": 0.4955,
+      "step": 29
+    },
+    {
+      "epoch": 0.04739336492890995,
+      "grad_norm": 0.8272190093994141,
+      "learning_rate": 9.541864139020539e-06,
+      "loss": 0.434,
+      "step": 30
+    },
+    {
+      "epoch": 0.04897314375987362,
+      "grad_norm": 0.5929948091506958,
+      "learning_rate": 9.5260663507109e-06,
+      "loss": 0.5042,
+      "step": 31
+    },
+    {
+      "epoch": 0.05055292259083728,
+      "grad_norm": 0.7872880101203918,
+      "learning_rate": 9.510268562401264e-06,
+      "loss": 0.5175,
+      "step": 32
+    },
+    {
+      "epoch": 0.052132701421800945,
+      "grad_norm": 0.6884463429450989,
+      "learning_rate": 9.494470774091628e-06,
+      "loss": 0.5104,
+      "step": 33
+    },
+    {
+      "epoch": 0.053712480252764615,
+      "grad_norm": 1.215976357460022,
+      "learning_rate": 9.478672985781992e-06,
+      "loss": 0.4742,
+      "step": 34
+    },
+    {
+      "epoch": 0.05529225908372828,
+      "grad_norm": 0.7471550107002258,
+      "learning_rate": 9.462875197472354e-06,
+      "loss": 0.4374,
+      "step": 35
+    },
+    {
+      "epoch": 0.05687203791469194,
+      "grad_norm": 0.6779741048812866,
+      "learning_rate": 9.447077409162718e-06,
+      "loss": 0.4337,
+      "step": 36
+    },
+    {
+      "epoch": 0.05845181674565561,
+      "grad_norm": 0.5205997824668884,
+      "learning_rate": 9.431279620853082e-06,
+      "loss": 0.4296,
+      "step": 37
+    },
+    {
+      "epoch": 0.06003159557661927,
+      "grad_norm": 0.381757527589798,
+      "learning_rate": 9.415481832543445e-06,
+      "loss": 0.2223,
+      "step": 38
+    },
+    {
+      "epoch": 0.061611374407582936,
+      "grad_norm": 0.650593101978302,
+      "learning_rate": 9.399684044233807e-06,
+      "loss": 0.5066,
+      "step": 39
+    },
+    {
+      "epoch": 0.0631911532385466,
+      "grad_norm": 0.5445153117179871,
+      "learning_rate": 9.383886255924171e-06,
+      "loss": 0.4998,
+      "step": 40
+    },
+    {
+      "epoch": 0.06477093206951026,
+      "grad_norm": 0.5024020671844482,
+      "learning_rate": 9.368088467614535e-06,
+      "loss": 0.4121,
+      "step": 41
+    },
+    {
+      "epoch": 0.06635071090047394,
+      "grad_norm": 0.6259915232658386,
+      "learning_rate": 9.352290679304899e-06,
+      "loss": 0.4969,
+      "step": 42
+    },
+    {
+      "epoch": 0.0679304897314376,
+      "grad_norm": 0.49405789375305176,
+      "learning_rate": 9.336492890995261e-06,
+      "loss": 0.4121,
+      "step": 43
+    },
+    {
+      "epoch": 0.06951026856240126,
+      "grad_norm": 0.7586628198623657,
+      "learning_rate": 9.320695102685625e-06,
+      "loss": 0.4782,
+      "step": 44
+    },
+    {
+      "epoch": 0.07109004739336493,
+      "grad_norm": 0.6203773021697998,
+      "learning_rate": 9.304897314375988e-06,
+      "loss": 0.3579,
+      "step": 45
+    },
+    {
+      "epoch": 0.07266982622432859,
+      "grad_norm": 0.6982845067977905,
+      "learning_rate": 9.289099526066352e-06,
+      "loss": 0.3876,
+      "step": 46
+    },
+    {
+      "epoch": 0.07424960505529225,
+      "grad_norm": 0.5712842345237732,
+      "learning_rate": 9.273301737756714e-06,
+      "loss": 0.4288,
+      "step": 47
+    },
+    {
+      "epoch": 0.07582938388625593,
+      "grad_norm": 0.6829891204833984,
+      "learning_rate": 9.257503949447078e-06,
+      "loss": 0.4939,
+      "step": 48
+    },
+    {
+      "epoch": 0.07740916271721959,
+      "grad_norm": 0.5508958101272583,
+      "learning_rate": 9.241706161137442e-06,
+      "loss": 0.372,
+      "step": 49
+    },
+    {
+      "epoch": 0.07898894154818326,
+      "grad_norm": 0.9345032572746277,
+      "learning_rate": 9.225908372827806e-06,
+      "loss": 0.4896,
+      "step": 50
+    },
+    {
+      "epoch": 0.08056872037914692,
+      "grad_norm": 0.6280492544174194,
+      "learning_rate": 9.210110584518168e-06,
+      "loss": 0.4375,
+      "step": 51
+    },
+    {
+      "epoch": 0.08214849921011058,
+      "grad_norm": 0.6853601336479187,
+      "learning_rate": 9.194312796208532e-06,
+      "loss": 0.4294,
+      "step": 52
+    },
+    {
+      "epoch": 0.08372827804107424,
+      "grad_norm": 0.6665984392166138,
+      "learning_rate": 9.178515007898895e-06,
+      "loss": 0.5894,
+      "step": 53
+    },
+    {
+      "epoch": 0.08530805687203792,
+      "grad_norm": 0.5088407397270203,
+      "learning_rate": 9.162717219589257e-06,
+      "loss": 0.3853,
+      "step": 54
+    },
+    {
+      "epoch": 0.08688783570300158,
+      "grad_norm": 0.5319867730140686,
+      "learning_rate": 9.146919431279621e-06,
+      "loss": 0.4791,
+      "step": 55
+    },
+    {
+      "epoch": 0.08846761453396525,
+      "grad_norm": 0.6452597975730896,
+      "learning_rate": 9.131121642969985e-06,
+      "loss": 0.4056,
+      "step": 56
+    },
+    {
+      "epoch": 0.09004739336492891,
+      "grad_norm": 0.6769601106643677,
+      "learning_rate": 9.115323854660349e-06,
+      "loss": 0.4253,
+      "step": 57
+    },
+    {
+      "epoch": 0.09162717219589257,
+      "grad_norm": 0.5170547962188721,
+      "learning_rate": 9.09952606635071e-06,
+      "loss": 0.4211,
+      "step": 58
+    },
+    {
+      "epoch": 0.09320695102685624,
+      "grad_norm": 0.5035193562507629,
+      "learning_rate": 9.083728278041075e-06,
+      "loss": 0.3144,
+      "step": 59
+    },
+    {
+      "epoch": 0.0947867298578199,
+      "grad_norm": 0.5919070243835449,
+      "learning_rate": 9.067930489731438e-06,
+      "loss": 0.4533,
+      "step": 60
+    },
+    {
+      "epoch": 0.09636650868878358,
+      "grad_norm": 0.6510637998580933,
+      "learning_rate": 9.052132701421802e-06,
+      "loss": 0.4701,
+      "step": 61
+    },
+    {
+      "epoch": 0.09794628751974724,
+      "grad_norm": 0.5784177780151367,
+      "learning_rate": 9.036334913112164e-06,
+      "loss": 0.3896,
+      "step": 62
+    },
+    {
+      "epoch": 0.0995260663507109,
+      "grad_norm": 0.7009139060974121,
+      "learning_rate": 9.020537124802528e-06,
+      "loss": 0.5018,
+      "step": 63
+    },
+    {
+      "epoch": 0.10110584518167456,
+      "grad_norm": 0.5086057186126709,
+      "learning_rate": 9.004739336492892e-06,
+      "loss": 0.4305,
+      "step": 64
+    },
+    {
+      "epoch": 0.10268562401263823,
+      "grad_norm": 0.5124595761299133,
+      "learning_rate": 8.988941548183256e-06,
+      "loss": 0.4473,
+      "step": 65
+    },
+    {
+      "epoch": 0.10426540284360189,
+      "grad_norm": 0.6409702897071838,
+      "learning_rate": 8.973143759873618e-06,
+      "loss": 0.429,
+      "step": 66
+    },
+    {
+      "epoch": 0.10584518167456557,
+      "grad_norm": 0.5651409029960632,
+      "learning_rate": 8.957345971563981e-06,
+      "loss": 0.4036,
+      "step": 67
+    },
+    {
+      "epoch": 0.10742496050552923,
+      "grad_norm": 0.6658238172531128,
+      "learning_rate": 8.941548183254345e-06,
+      "loss": 0.4726,
+      "step": 68
+    },
+    {
+      "epoch": 0.10900473933649289,
+      "grad_norm": 0.444815993309021,
+      "learning_rate": 8.925750394944709e-06,
+      "loss": 0.4016,
+      "step": 69
+    },
+    {
+      "epoch": 0.11058451816745656,
+      "grad_norm": 0.5855506658554077,
+      "learning_rate": 8.909952606635071e-06,
+      "loss": 0.4531,
+      "step": 70
+    },
+    {
+      "epoch": 0.11216429699842022,
+      "grad_norm": 0.693794310092926,
+      "learning_rate": 8.894154818325435e-06,
+      "loss": 0.4382,
+      "step": 71
+    },
+    {
+      "epoch": 0.11374407582938388,
+      "grad_norm": 0.6658089756965637,
+      "learning_rate": 8.878357030015799e-06,
+      "loss": 0.4571,
+      "step": 72
+    },
+    {
+      "epoch": 0.11532385466034756,
+      "grad_norm": 1.0504828691482544,
+      "learning_rate": 8.862559241706162e-06,
+      "loss": 0.4311,
+      "step": 73
+    },
+    {
+      "epoch": 0.11690363349131122,
+      "grad_norm": 0.5297814607620239,
+      "learning_rate": 8.846761453396524e-06,
+      "loss": 0.4391,
+      "step": 74
+    },
+    {
+      "epoch": 0.11848341232227488,
+      "grad_norm": 0.6601409316062927,
+      "learning_rate": 8.830963665086888e-06,
+      "loss": 0.5125,
+      "step": 75
+    },
+    {
+      "epoch": 0.12006319115323855,
+      "grad_norm": 0.6345618963241577,
+      "learning_rate": 8.815165876777252e-06,
+      "loss": 0.4471,
+      "step": 76
+    },
+    {
+      "epoch": 0.12164296998420221,
+      "grad_norm": 0.5008222460746765,
+      "learning_rate": 8.799368088467614e-06,
+      "loss": 0.3845,
+      "step": 77
+    },
+    {
+      "epoch": 0.12322274881516587,
+      "grad_norm": 0.5394203066825867,
+      "learning_rate": 8.783570300157978e-06,
+      "loss": 0.4117,
+      "step": 78
+    },
+    {
+      "epoch": 0.12480252764612954,
+      "grad_norm": 0.6255345940589905,
+      "learning_rate": 8.767772511848342e-06,
+      "loss": 0.512,
+      "step": 79
+    },
+    {
+      "epoch": 0.1263823064770932,
+      "grad_norm": 0.6215748190879822,
+      "learning_rate": 8.751974723538705e-06,
+      "loss": 0.509,
+      "step": 80
+    },
+    {
+      "epoch": 0.12796208530805686,
+      "grad_norm": 0.611587405204773,
+      "learning_rate": 8.736176935229068e-06,
+      "loss": 0.4036,
+      "step": 81
+    },
+    {
+      "epoch": 0.12954186413902052,
+      "grad_norm": 0.5373330116271973,
+      "learning_rate": 8.720379146919431e-06,
+      "loss": 0.393,
+      "step": 82
+    },
+    {
+      "epoch": 0.13112164296998421,
+      "grad_norm": 0.5936598181724548,
+      "learning_rate": 8.704581358609795e-06,
+      "loss": 0.4092,
+      "step": 83
+    },
+    {
+      "epoch": 0.13270142180094788,
+      "grad_norm": 0.576614260673523,
+      "learning_rate": 8.688783570300159e-06,
+      "loss": 0.5513,
+      "step": 84
+    },
+    {
+      "epoch": 0.13428120063191154,
+      "grad_norm": 0.5715078711509705,
+      "learning_rate": 8.672985781990521e-06,
+      "loss": 0.4403,
+      "step": 85
+    },
+    {
+      "epoch": 0.1358609794628752,
+      "grad_norm": 0.6212042570114136,
+      "learning_rate": 8.657187993680885e-06,
+      "loss": 0.391,
+      "step": 86
+    },
+    {
+      "epoch": 0.13744075829383887,
+      "grad_norm": 0.5439122319221497,
+      "learning_rate": 8.641390205371249e-06,
+      "loss": 0.4764,
+      "step": 87
+    },
+    {
+      "epoch": 0.13902053712480253,
+      "grad_norm": 0.6808428168296814,
+      "learning_rate": 8.625592417061612e-06,
+      "loss": 0.512,
+      "step": 88
+    },
+    {
+      "epoch": 0.1406003159557662,
+      "grad_norm": 0.7429847717285156,
+      "learning_rate": 8.609794628751974e-06,
+      "loss": 0.3834,
+      "step": 89
+    },
+    {
+      "epoch": 0.14218009478672985,
+      "grad_norm": 0.6030511260032654,
+      "learning_rate": 8.59399684044234e-06,
+      "loss": 0.4631,
+      "step": 90
+    },
+    {
+      "epoch": 0.14375987361769352,
+      "grad_norm": 0.6499682068824768,
+      "learning_rate": 8.578199052132702e-06,
+      "loss": 0.4484,
+      "step": 91
+    },
+    {
+      "epoch": 0.14533965244865718,
+      "grad_norm": 0.6490275859832764,
+      "learning_rate": 8.562401263823066e-06,
+      "loss": 0.414,
+      "step": 92
+    },
+    {
+      "epoch": 0.14691943127962084,
+      "grad_norm": 0.6859791874885559,
+      "learning_rate": 8.546603475513428e-06,
+      "loss": 0.386,
+      "step": 93
+    },
+    {
+      "epoch": 0.1484992101105845,
+      "grad_norm": 0.5281291007995605,
+      "learning_rate": 8.530805687203793e-06,
+      "loss": 0.4036,
+      "step": 94
+    },
+    {
+      "epoch": 0.1500789889415482,
+      "grad_norm": 0.5261964797973633,
+      "learning_rate": 8.515007898894155e-06,
+      "loss": 0.33,
+      "step": 95
+    },
+    {
+      "epoch": 0.15165876777251186,
+      "grad_norm": 0.4350665211677551,
+      "learning_rate": 8.499210110584519e-06,
+      "loss": 0.3347,
+      "step": 96
+    },
+    {
+      "epoch": 0.15323854660347552,
+      "grad_norm": 0.8448456525802612,
+      "learning_rate": 8.483412322274883e-06,
+      "loss": 0.4253,
+      "step": 97
+    },
+    {
+      "epoch": 0.15481832543443919,
+      "grad_norm": 0.6256837248802185,
+      "learning_rate": 8.467614533965247e-06,
+      "loss": 0.4464,
+      "step": 98
+    },
+    {
+      "epoch": 0.15639810426540285,
+      "grad_norm": 0.7007749676704407,
+      "learning_rate": 8.451816745655609e-06,
+      "loss": 0.4641,
+      "step": 99
+    },
+    {
+      "epoch": 0.1579778830963665,
+      "grad_norm": 0.6551494002342224,
+      "learning_rate": 8.436018957345973e-06,
+      "loss": 0.5097,
+      "step": 100
+    },
+    {
+      "epoch": 0.15955766192733017,
+      "grad_norm": 0.5944113731384277,
+      "learning_rate": 8.420221169036336e-06,
+      "loss": 0.4554,
+      "step": 101
+    },
+    {
+      "epoch": 0.16113744075829384,
+      "grad_norm": 0.5755615234375,
+      "learning_rate": 8.4044233807267e-06,
+      "loss": 0.443,
+      "step": 102
+    },
+    {
+      "epoch": 0.1627172195892575,
+      "grad_norm": 0.5263962745666504,
+      "learning_rate": 8.388625592417062e-06,
+      "loss": 0.4355,
+      "step": 103
+    },
+    {
+      "epoch": 0.16429699842022116,
+      "grad_norm": 0.6115814447402954,
+      "learning_rate": 8.372827804107424e-06,
+      "loss": 0.4863,
+      "step": 104
+    },
+    {
+      "epoch": 0.16587677725118483,
+      "grad_norm": 0.5544970631599426,
+      "learning_rate": 8.35703001579779e-06,
+      "loss": 0.3979,
+      "step": 105
+    },
+    {
+      "epoch": 0.1674565560821485,
+      "grad_norm": 0.5588533878326416,
+      "learning_rate": 8.341232227488152e-06,
+      "loss": 0.4073,
+      "step": 106
+    },
+    {
+      "epoch": 0.16903633491311215,
+      "grad_norm": 0.578982949256897,
+      "learning_rate": 8.325434439178516e-06,
+      "loss": 0.3745,
+      "step": 107
+    },
+    {
+      "epoch": 0.17061611374407584,
+      "grad_norm": 0.4955246150493622,
+      "learning_rate": 8.30963665086888e-06,
+      "loss": 0.438,
+      "step": 108
+    },
+    {
+      "epoch": 0.1721958925750395,
+      "grad_norm": 0.593362033367157,
+      "learning_rate": 8.293838862559243e-06,
+      "loss": 0.4161,
+      "step": 109
+    },
+    {
+      "epoch": 0.17377567140600317,
+      "grad_norm": 0.5000883340835571,
+      "learning_rate": 8.278041074249605e-06,
+      "loss": 0.432,
+      "step": 110
+    },
+    {
+      "epoch": 0.17535545023696683,
+      "grad_norm": 0.5794082880020142,
+      "learning_rate": 8.262243285939969e-06,
+      "loss": 0.4431,
+      "step": 111
+    },
+    {
+      "epoch": 0.1769352290679305,
+      "grad_norm": 0.6179563999176025,
+      "learning_rate": 8.246445497630333e-06,
+      "loss": 0.3871,
+      "step": 112
+    },
+    {
+      "epoch": 0.17851500789889416,
+      "grad_norm": 0.6540956497192383,
+      "learning_rate": 8.230647709320697e-06,
+      "loss": 0.3706,
+      "step": 113
+    },
+    {
+      "epoch": 0.18009478672985782,
+      "grad_norm": 0.7029737234115601,
+      "learning_rate": 8.214849921011059e-06,
+      "loss": 0.5077,
+      "step": 114
+    },
+    {
+      "epoch": 0.18167456556082148,
+      "grad_norm": 0.5466600656509399,
+      "learning_rate": 8.199052132701422e-06,
+      "loss": 0.4634,
+      "step": 115
+    },
+    {
+      "epoch": 0.18325434439178515,
+      "grad_norm": 0.5513831973075867,
+      "learning_rate": 8.183254344391786e-06,
+      "loss": 0.4457,
+      "step": 116
+    },
+    {
+      "epoch": 0.1848341232227488,
+      "grad_norm": 0.7652455568313599,
+      "learning_rate": 8.16745655608215e-06,
+      "loss": 0.4376,
+      "step": 117
+    },
+    {
+      "epoch": 0.18641390205371247,
+      "grad_norm": 0.6213077902793884,
+      "learning_rate": 8.151658767772512e-06,
+      "loss": 0.3988,
+      "step": 118
+    },
+    {
+      "epoch": 0.18799368088467613,
+      "grad_norm": 0.50051349401474,
+      "learning_rate": 8.135860979462876e-06,
+      "loss": 0.4142,
+      "step": 119
+    },
+    {
+      "epoch": 0.1895734597156398,
+      "grad_norm": 0.8015328049659729,
+      "learning_rate": 8.12006319115324e-06,
+      "loss": 0.4474,
+      "step": 120
+    },
+    {
+      "epoch": 0.1911532385466035,
+      "grad_norm": 0.6595532298088074,
+      "learning_rate": 8.104265402843603e-06,
+      "loss": 0.5173,
+      "step": 121
+    },
+    {
+      "epoch": 0.19273301737756715,
+      "grad_norm": 0.7859697937965393,
+      "learning_rate": 8.088467614533966e-06,
+      "loss": 0.4465,
+      "step": 122
+    },
+    {
+      "epoch": 0.1943127962085308,
+      "grad_norm": 0.6508023738861084,
+      "learning_rate": 8.07266982622433e-06,
+      "loss": 0.4448,
+      "step": 123
+    },
+    {
+      "epoch": 0.19589257503949448,
+      "grad_norm": 0.49232304096221924,
+      "learning_rate": 8.056872037914693e-06,
+      "loss": 0.4005,
+      "step": 124
+    },
+    {
+      "epoch": 0.19747235387045814,
+      "grad_norm": 0.6464349031448364,
+      "learning_rate": 8.041074249605057e-06,
+      "loss": 0.47,
+      "step": 125
+    },
+    {
+      "epoch": 0.1990521327014218,
+      "grad_norm": 0.5296919345855713,
+      "learning_rate": 8.025276461295419e-06,
+      "loss": 0.4247,
+      "step": 126
+    },
+    {
+      "epoch": 0.20063191153238547,
+      "grad_norm": 0.6270297765731812,
+      "learning_rate": 8.009478672985783e-06,
+      "loss": 0.5397,
+      "step": 127
+    },
+    {
+      "epoch": 0.20221169036334913,
+      "grad_norm": 0.6148909330368042,
+      "learning_rate": 7.993680884676147e-06,
+      "loss": 0.4133,
+      "step": 128
+    },
+    {
+      "epoch": 0.2037914691943128,
+      "grad_norm": 0.7778130173683167,
+      "learning_rate": 7.977883096366509e-06,
+      "loss": 0.5119,
+      "step": 129
+    },
+    {
+      "epoch": 0.20537124802527645,
+      "grad_norm": 0.47952044010162354,
+      "learning_rate": 7.962085308056872e-06,
+      "loss": 0.386,
+      "step": 130
+    },
+    {
+      "epoch": 0.20695102685624012,
+      "grad_norm": 0.5951160788536072,
+      "learning_rate": 7.946287519747236e-06,
+      "loss": 0.5101,
+      "step": 131
+    },
+    {
+      "epoch": 0.20853080568720378,
+      "grad_norm": 0.6209789514541626,
+      "learning_rate": 7.9304897314376e-06,
+      "loss": 0.4988,
+      "step": 132
+    },
+    {
+      "epoch": 0.21011058451816747,
+      "grad_norm": 0.5093654990196228,
+      "learning_rate": 7.914691943127962e-06,
+      "loss": 0.374,
+      "step": 133
+    },
+    {
+      "epoch": 0.21169036334913113,
+      "grad_norm": 0.5125884413719177,
+      "learning_rate": 7.898894154818326e-06,
+      "loss": 0.4097,
+      "step": 134
+    },
+    {
+      "epoch": 0.2132701421800948,
+      "grad_norm": 0.5116066932678223,
+      "learning_rate": 7.88309636650869e-06,
+      "loss": 0.4643,
+      "step": 135
+    },
+    {
+      "epoch": 0.21484992101105846,
+      "grad_norm": 0.5778034329414368,
+      "learning_rate": 7.867298578199053e-06,
+      "loss": 0.4645,
+      "step": 136
+    },
+    {
+      "epoch": 0.21642969984202212,
+      "grad_norm": 0.6490422487258911,
+      "learning_rate": 7.851500789889415e-06,
+      "loss": 0.4825,
+      "step": 137
+    },
+    {
+      "epoch": 0.21800947867298578,
+      "grad_norm": 0.644008219242096,
+      "learning_rate": 7.83570300157978e-06,
+      "loss": 0.3954,
+      "step": 138
+    },
+    {
+      "epoch": 0.21958925750394945,
+      "grad_norm": 0.8628047704696655,
+      "learning_rate": 7.819905213270143e-06,
+      "loss": 0.5322,
+      "step": 139
+    },
+    {
+      "epoch": 0.2211690363349131,
+      "grad_norm": 0.6286507844924927,
+      "learning_rate": 7.804107424960507e-06,
+      "loss": 0.3741,
+      "step": 140
+    },
+    {
+      "epoch": 0.22274881516587677,
+      "grad_norm": 0.6210809350013733,
+      "learning_rate": 7.788309636650869e-06,
+      "loss": 0.4572,
+      "step": 141
+    },
+    {
+      "epoch": 0.22432859399684044,
+      "grad_norm": 0.5337722897529602,
+      "learning_rate": 7.772511848341233e-06,
+      "loss": 0.3788,
+      "step": 142
+    },
+    {
+      "epoch": 0.2259083728278041,
+      "grad_norm": 0.5743194818496704,
+      "learning_rate": 7.756714060031596e-06,
+      "loss": 0.3963,
+      "step": 143
+    },
+    {
+      "epoch": 0.22748815165876776,
+      "grad_norm": 0.4972652792930603,
+      "learning_rate": 7.74091627172196e-06,
+      "loss": 0.2906,
+      "step": 144
+    },
+    {
+      "epoch": 0.22906793048973143,
+      "grad_norm": 0.5239664316177368,
+      "learning_rate": 7.725118483412322e-06,
+      "loss": 0.4009,
+      "step": 145
+    },
+    {
+      "epoch": 0.23064770932069512,
+      "grad_norm": 0.5151936411857605,
+      "learning_rate": 7.709320695102686e-06,
+      "loss": 0.4208,
+      "step": 146
+    },
+    {
+      "epoch": 0.23222748815165878,
+      "grad_norm": 0.6128547191619873,
+      "learning_rate": 7.69352290679305e-06,
+      "loss": 0.4779,
+      "step": 147
+    },
+    {
+      "epoch": 0.23380726698262244,
+      "grad_norm": 0.5268502235412598,
+      "learning_rate": 7.677725118483414e-06,
+      "loss": 0.4219,
+      "step": 148
+    },
+    {
+      "epoch": 0.2353870458135861,
+      "grad_norm": 0.5439866185188293,
+      "learning_rate": 7.661927330173776e-06,
+      "loss": 0.4436,
+      "step": 149
+    },
+    {
+      "epoch": 0.23696682464454977,
+      "grad_norm": 0.5291867852210999,
+      "learning_rate": 7.64612954186414e-06,
+      "loss": 0.407,
+      "step": 150
+    },
+    {
+      "epoch": 0.23854660347551343,
+      "grad_norm": 0.6638155579566956,
+      "learning_rate": 7.630331753554503e-06,
+      "loss": 0.403,
+      "step": 151
+    },
+    {
+      "epoch": 0.2401263823064771,
+      "grad_norm": 0.5501230955123901,
+      "learning_rate": 7.614533965244867e-06,
+      "loss": 0.5004,
+      "step": 152
+    },
+    {
+      "epoch": 0.24170616113744076,
+      "grad_norm": 0.5949499011039734,
+      "learning_rate": 7.59873617693523e-06,
+      "loss": 0.4708,
+      "step": 153
+    },
+    {
+      "epoch": 0.24328593996840442,
+      "grad_norm": 0.5841517448425293,
+      "learning_rate": 7.582938388625593e-06,
+      "loss": 0.4836,
+      "step": 154
+    },
+    {
+      "epoch": 0.24486571879936808,
+      "grad_norm": 0.6298154592514038,
+      "learning_rate": 7.567140600315957e-06,
+      "loss": 0.4728,
+      "step": 155
+    },
+    {
+      "epoch": 0.24644549763033174,
+      "grad_norm": 0.6107637882232666,
+      "learning_rate": 7.55134281200632e-06,
+      "loss": 0.4243,
+      "step": 156
+    },
+    {
+      "epoch": 0.2480252764612954,
+      "grad_norm": 0.5174968838691711,
+      "learning_rate": 7.535545023696683e-06,
+      "loss": 0.4657,
+      "step": 157
+    },
+    {
+      "epoch": 0.24960505529225907,
+      "grad_norm": 0.5588591694831848,
+      "learning_rate": 7.519747235387046e-06,
+      "loss": 0.4567,
+      "step": 158
+    },
+    {
+      "epoch": 0.25118483412322273,
+      "grad_norm": 0.8415222764015198,
+      "learning_rate": 7.50394944707741e-06,
+      "loss": 0.4625,
+      "step": 159
+    },
+    {
+      "epoch": 0.2527646129541864,
+      "grad_norm": 0.6054974794387817,
+      "learning_rate": 7.488151658767773e-06,
+      "loss": 0.3843,
+      "step": 160
+    },
+    {
+      "epoch": 0.25434439178515006,
+      "grad_norm": 0.5117557644844055,
+      "learning_rate": 7.472353870458137e-06,
+      "loss": 0.3887,
+      "step": 161
+    },
+    {
+      "epoch": 0.2559241706161137,
+      "grad_norm": 0.5849332213401794,
+      "learning_rate": 7.4565560821485e-06,
+      "loss": 0.4528,
+      "step": 162
+    },
+    {
+      "epoch": 0.2575039494470774,
+      "grad_norm": 0.5625325441360474,
+      "learning_rate": 7.4407582938388635e-06,
+      "loss": 0.4542,
+      "step": 163
+    },
+    {
+      "epoch": 0.25908372827804105,
+      "grad_norm": 0.5406492352485657,
+      "learning_rate": 7.4249605055292264e-06,
+      "loss": 0.4592,
+      "step": 164
+    },
+    {
+      "epoch": 0.26066350710900477,
+      "grad_norm": 0.6318654417991638,
+      "learning_rate": 7.40916271721959e-06,
+      "loss": 0.4361,
+      "step": 165
+    },
+    {
+      "epoch": 0.26224328593996843,
+      "grad_norm": 0.5719902515411377,
+      "learning_rate": 7.393364928909953e-06,
+      "loss": 0.4799,
+      "step": 166
+    },
+    {
+      "epoch": 0.2638230647709321,
+      "grad_norm": 0.5211177468299866,
+      "learning_rate": 7.377567140600317e-06,
+      "loss": 0.33,
+      "step": 167
+    },
+    {
+      "epoch": 0.26540284360189575,
+      "grad_norm": 0.6400920152664185,
+      "learning_rate": 7.36176935229068e-06,
+      "loss": 0.4235,
+      "step": 168
+    },
+    {
+      "epoch": 0.2669826224328594,
+      "grad_norm": 0.5302186608314514,
+      "learning_rate": 7.345971563981044e-06,
+      "loss": 0.4342,
+      "step": 169
+    },
+    {
+      "epoch": 0.2685624012638231,
+      "grad_norm": 0.5393325686454773,
+      "learning_rate": 7.3301737756714066e-06,
+      "loss": 0.3632,
+      "step": 170
+    },
+    {
+      "epoch": 0.27014218009478674,
+      "grad_norm": 0.5409063696861267,
+      "learning_rate": 7.31437598736177e-06,
+      "loss": 0.4076,
+      "step": 171
+    },
+    {
+      "epoch": 0.2717219589257504,
+      "grad_norm": 0.5056774616241455,
+      "learning_rate": 7.298578199052133e-06,
+      "loss": 0.4821,
+      "step": 172
+    },
+    {
+      "epoch": 0.27330173775671407,
+      "grad_norm": 0.6061700582504272,
+      "learning_rate": 7.282780410742497e-06,
+      "loss": 0.5137,
+      "step": 173
+    },
+    {
+      "epoch": 0.27488151658767773,
+      "grad_norm": 0.5524815917015076,
+      "learning_rate": 7.26698262243286e-06,
+      "loss": 0.4116,
+      "step": 174
+    },
+    {
+      "epoch": 0.2764612954186414,
+      "grad_norm": 0.5045567750930786,
+      "learning_rate": 7.251184834123224e-06,
+      "loss": 0.3969,
+      "step": 175
+    },
+    {
+      "epoch": 0.27804107424960506,
+      "grad_norm": 0.604505717754364,
+      "learning_rate": 7.235387045813587e-06,
+      "loss": 0.5176,
+      "step": 176
+    },
+    {
+      "epoch": 0.2796208530805687,
+      "grad_norm": 0.6067575812339783,
+      "learning_rate": 7.2195892575039505e-06,
+      "loss": 0.4438,
+      "step": 177
+    },
+    {
+      "epoch": 0.2812006319115324,
+      "grad_norm": 0.6412494778633118,
+      "learning_rate": 7.203791469194313e-06,
+      "loss": 0.4758,
+      "step": 178
+    },
+    {
+      "epoch": 0.28278041074249605,
+      "grad_norm": 0.5432886481285095,
+      "learning_rate": 7.187993680884676e-06,
+      "loss": 0.4387,
+      "step": 179
+    },
+    {
+      "epoch": 0.2843601895734597,
+      "grad_norm": 0.4622472822666168,
+      "learning_rate": 7.17219589257504e-06,
+      "loss": 0.4775,
+      "step": 180
+    },
+    {
+      "epoch": 0.2859399684044234,
+      "grad_norm": 0.643259584903717,
+      "learning_rate": 7.156398104265403e-06,
+      "loss": 0.4479,
+      "step": 181
+    },
+    {
+      "epoch": 0.28751974723538704,
+      "grad_norm": 0.48998138308525085,
+      "learning_rate": 7.140600315955767e-06,
+      "loss": 0.399,
+      "step": 182
+    },
+    {
+      "epoch": 0.2890995260663507,
+      "grad_norm": 0.5146614909172058,
+      "learning_rate": 7.12480252764613e-06,
+      "loss": 0.4475,
+      "step": 183
+    },
+    {
+      "epoch": 0.29067930489731436,
+      "grad_norm": 0.5386670231819153,
+      "learning_rate": 7.1090047393364935e-06,
+      "loss": 0.3892,
+      "step": 184
+    },
+    {
+      "epoch": 0.292259083728278,
+      "grad_norm": 0.5147759318351746,
+      "learning_rate": 7.0932069510268565e-06,
+      "loss": 0.3755,
+      "step": 185
+    },
+    {
+      "epoch": 0.2938388625592417,
+      "grad_norm": 0.5141321420669556,
+      "learning_rate": 7.07740916271722e-06,
+      "loss": 0.355,
+      "step": 186
+    },
+    {
+      "epoch": 0.29541864139020535,
+      "grad_norm": 0.9518134593963623,
+      "learning_rate": 7.061611374407583e-06,
+      "loss": 0.4021,
+      "step": 187
+    },
+    {
+      "epoch": 0.296998420221169,
+      "grad_norm": 0.5844981670379639,
+      "learning_rate": 7.045813586097947e-06,
+      "loss": 0.4233,
+      "step": 188
+    },
+    {
+      "epoch": 0.2985781990521327,
+      "grad_norm": 0.6381546854972839,
+      "learning_rate": 7.03001579778831e-06,
+      "loss": 0.4862,
+      "step": 189
+    },
+    {
+      "epoch": 0.3001579778830964,
+      "grad_norm": 0.7311195135116577,
+      "learning_rate": 7.014218009478674e-06,
+      "loss": 0.4822,
+      "step": 190
+    },
+    {
+      "epoch": 0.30173775671406006,
+      "grad_norm": 0.5827596783638,
+      "learning_rate": 6.998420221169037e-06,
+      "loss": 0.4027,
+      "step": 191
+    },
+    {
+      "epoch": 0.3033175355450237,
+      "grad_norm": 0.6907688975334167,
+      "learning_rate": 6.9826224328594e-06,
+      "loss": 0.4374,
+      "step": 192
+    },
+    {
+      "epoch": 0.3048973143759874,
+      "grad_norm": 0.5060120820999146,
+      "learning_rate": 6.966824644549763e-06,
+      "loss": 0.4226,
+      "step": 193
+    },
+    {
+      "epoch": 0.30647709320695105,
+      "grad_norm": 0.41480544209480286,
+      "learning_rate": 6.951026856240127e-06,
+      "loss": 0.3766,
+      "step": 194
+    },
+    {
+      "epoch": 0.3080568720379147,
+      "grad_norm": 0.5637404322624207,
+      "learning_rate": 6.93522906793049e-06,
+      "loss": 0.4365,
+      "step": 195
+    },
+    {
+      "epoch": 0.30963665086887837,
+      "grad_norm": 0.6389409899711609,
+      "learning_rate": 6.919431279620854e-06,
+      "loss": 0.4186,
+      "step": 196
+    },
+    {
+      "epoch": 0.31121642969984203,
+      "grad_norm": 0.48588162660598755,
+      "learning_rate": 6.903633491311217e-06,
+      "loss": 0.4023,
+      "step": 197
+    },
+    {
+      "epoch": 0.3127962085308057,
+      "grad_norm": 0.6066514253616333,
+      "learning_rate": 6.8878357030015805e-06,
+      "loss": 0.4652,
+      "step": 198
+    },
+    {
+      "epoch": 0.31437598736176936,
+      "grad_norm": 0.6308689117431641,
+      "learning_rate": 6.8720379146919435e-06,
+      "loss": 0.3885,
+      "step": 199
+    },
+    {
+      "epoch": 0.315955766192733,
+      "grad_norm": 0.4883437752723694,
+      "learning_rate": 6.856240126382307e-06,
+      "loss": 0.4128,
+      "step": 200
+    },
+    {
+      "epoch": 0.3175355450236967,
+      "grad_norm": 0.720086932182312,
+      "learning_rate": 6.84044233807267e-06,
+      "loss": 0.4333,
+      "step": 201
+    },
+    {
+      "epoch": 0.31911532385466035,
+      "grad_norm": 0.6698761582374573,
+      "learning_rate": 6.824644549763034e-06,
+      "loss": 0.3967,
+      "step": 202
+    },
+    {
+      "epoch": 0.320695102685624,
+      "grad_norm": 0.5240082740783691,
+      "learning_rate": 6.808846761453397e-06,
+      "loss": 0.4055,
+      "step": 203
+    },
+    {
+      "epoch": 0.3222748815165877,
+      "grad_norm": 0.6142946481704712,
+      "learning_rate": 6.79304897314376e-06,
+      "loss": 0.3645,
+      "step": 204
+    },
+    {
+      "epoch": 0.32385466034755134,
+      "grad_norm": 0.6439379453659058,
+      "learning_rate": 6.777251184834124e-06,
+      "loss": 0.3207,
+      "step": 205
+    },
+    {
+      "epoch": 0.325434439178515,
+      "grad_norm": 0.6862720847129822,
+      "learning_rate": 6.7614533965244865e-06,
+      "loss": 0.4944,
+      "step": 206
+    },
+    {
+      "epoch": 0.32701421800947866,
+      "grad_norm": 0.6720433235168457,
+      "learning_rate": 6.74565560821485e-06,
+      "loss": 0.4335,
+      "step": 207
+    },
+    {
+      "epoch": 0.3285939968404423,
+      "grad_norm": 0.531577467918396,
+      "learning_rate": 6.729857819905213e-06,
+      "loss": 0.5327,
+      "step": 208
+    },
+    {
+      "epoch": 0.330173775671406,
+      "grad_norm": 0.5542590022087097,
+      "learning_rate": 6.714060031595577e-06,
+      "loss": 0.3629,
+      "step": 209
+    },
+    {
+      "epoch": 0.33175355450236965,
+      "grad_norm": 0.5614448189735413,
+      "learning_rate": 6.69826224328594e-06,
+      "loss": 0.4097,
+      "step": 210
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.7383466362953186,
+      "learning_rate": 6.682464454976304e-06,
+      "loss": 0.5031,
+      "step": 211
+    },
+    {
+      "epoch": 0.334913112164297,
+      "grad_norm": 0.6345497965812683,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5029,
+      "step": 212
+    },
+    {
+      "epoch": 0.33649289099526064,
+      "grad_norm": 0.579641580581665,
+      "learning_rate": 6.6508688783570304e-06,
+      "loss": 0.4949,
+      "step": 213
+    },
+    {
+      "epoch": 0.3380726698262243,
+      "grad_norm": 0.5040780305862427,
+      "learning_rate": 6.635071090047393e-06,
+      "loss": 0.4537,
+      "step": 214
+    },
+    {
+      "epoch": 0.33965244865718797,
+      "grad_norm": 0.5917491316795349,
+      "learning_rate": 6.619273301737757e-06,
+      "loss": 0.3883,
+      "step": 215
+    },
+    {
+      "epoch": 0.3412322274881517,
+      "grad_norm": 0.7031399011611938,
+      "learning_rate": 6.60347551342812e-06,
+      "loss": 0.4554,
+      "step": 216
+    },
+    {
+      "epoch": 0.34281200631911535,
+      "grad_norm": 0.5503798127174377,
+      "learning_rate": 6.587677725118484e-06,
+      "loss": 0.352,
+      "step": 217
+    },
+    {
+      "epoch": 0.344391785150079,
+      "grad_norm": 0.5412716269493103,
+      "learning_rate": 6.571879936808847e-06,
+      "loss": 0.4191,
+      "step": 218
+    },
+    {
+      "epoch": 0.3459715639810427,
+      "grad_norm": 0.6272369623184204,
+      "learning_rate": 6.556082148499211e-06,
+      "loss": 0.4595,
+      "step": 219
+    },
+    {
+      "epoch": 0.34755134281200634,
+      "grad_norm": 0.5309504270553589,
+      "learning_rate": 6.5402843601895735e-06,
+      "loss": 0.4095,
+      "step": 220
+    },
+    {
+      "epoch": 0.34913112164297,
+      "grad_norm": 0.5687200427055359,
+      "learning_rate": 6.524486571879938e-06,
+      "loss": 0.435,
+      "step": 221
+    },
+    {
+      "epoch": 0.35071090047393366,
+      "grad_norm": 0.5819438099861145,
+      "learning_rate": 6.5086887835703e-06,
+      "loss": 0.4695,
+      "step": 222
+    },
+    {
+      "epoch": 0.3522906793048973,
+      "grad_norm": 0.6310110092163086,
+      "learning_rate": 6.492890995260665e-06,
+      "loss": 0.4346,
+      "step": 223
+    },
+    {
+      "epoch": 0.353870458135861,
+      "grad_norm": 0.5838906168937683,
+      "learning_rate": 6.477093206951027e-06,
+      "loss": 0.47,
+      "step": 224
+    },
+    {
+      "epoch": 0.35545023696682465,
+      "grad_norm": 0.6752678155899048,
+      "learning_rate": 6.4612954186413915e-06,
+      "loss": 0.3842,
+      "step": 225
+    },
+    {
+      "epoch": 0.3570300157977883,
+      "grad_norm": 0.7029111981391907,
+      "learning_rate": 6.445497630331754e-06,
+      "loss": 0.4442,
+      "step": 226
+    },
+    {
+      "epoch": 0.358609794628752,
+      "grad_norm": 0.511812686920166,
+      "learning_rate": 6.429699842022118e-06,
+      "loss": 0.5171,
+      "step": 227
+    },
+    {
+      "epoch": 0.36018957345971564,
+      "grad_norm": 0.49457868933677673,
+      "learning_rate": 6.413902053712481e-06,
+      "loss": 0.3695,
+      "step": 228
+    },
+    {
+      "epoch": 0.3617693522906793,
+      "grad_norm": 0.4521022439002991,
+      "learning_rate": 6.398104265402843e-06,
+      "loss": 0.3909,
+      "step": 229
+    },
+    {
+      "epoch": 0.36334913112164297,
+      "grad_norm": 0.45229026675224304,
+      "learning_rate": 6.382306477093208e-06,
+      "loss": 0.3417,
+      "step": 230
+    },
+    {
+      "epoch": 0.36492890995260663,
+      "grad_norm": 0.5070056915283203,
+      "learning_rate": 6.36650868878357e-06,
+      "loss": 0.3518,
+      "step": 231
+    },
+    {
+      "epoch": 0.3665086887835703,
+      "grad_norm": 0.9325531721115112,
+      "learning_rate": 6.350710900473935e-06,
+      "loss": 0.5172,
+      "step": 232
+    },
+    {
+      "epoch": 0.36808846761453395,
+      "grad_norm": 0.6027977466583252,
+      "learning_rate": 6.334913112164297e-06,
+      "loss": 0.4052,
+      "step": 233
+    },
+    {
+      "epoch": 0.3696682464454976,
+      "grad_norm": 0.7251097559928894,
+      "learning_rate": 6.319115323854661e-06,
+      "loss": 0.4739,
+      "step": 234
+    },
+    {
+      "epoch": 0.3712480252764613,
+      "grad_norm": 0.6470052003860474,
+      "learning_rate": 6.303317535545023e-06,
+      "loss": 0.4745,
+      "step": 235
+    },
+    {
+      "epoch": 0.37282780410742494,
+      "grad_norm": 0.7177411317825317,
+      "learning_rate": 6.287519747235388e-06,
+      "loss": 0.364,
+      "step": 236
+    },
+    {
+      "epoch": 0.3744075829383886,
+      "grad_norm": 0.7681677341461182,
+      "learning_rate": 6.271721958925751e-06,
+      "loss": 0.4559,
+      "step": 237
+    },
+    {
+      "epoch": 0.37598736176935227,
+      "grad_norm": 0.6160128116607666,
+      "learning_rate": 6.255924170616115e-06,
+      "loss": 0.421,
+      "step": 238
+    },
+    {
+      "epoch": 0.37756714060031593,
+      "grad_norm": 0.658981442451477,
+      "learning_rate": 6.240126382306478e-06,
+      "loss": 0.3979,
+      "step": 239
+    },
+    {
+      "epoch": 0.3791469194312796,
+      "grad_norm": 0.9422373175621033,
+      "learning_rate": 6.2243285939968414e-06,
+      "loss": 0.3586,
+      "step": 240
+    },
+    {
+      "epoch": 0.3807266982622433,
+      "grad_norm": 0.5452501773834229,
+      "learning_rate": 6.208530805687204e-06,
+      "loss": 0.4209,
+      "step": 241
+    },
+    {
+      "epoch": 0.382306477093207,
+      "grad_norm": 0.4912925660610199,
+      "learning_rate": 6.192733017377568e-06,
+      "loss": 0.4784,
+      "step": 242
+    },
+    {
+      "epoch": 0.38388625592417064,
+      "grad_norm": 0.6575455665588379,
+      "learning_rate": 6.176935229067931e-06,
+      "loss": 0.4062,
+      "step": 243
+    },
+    {
+      "epoch": 0.3854660347551343,
+      "grad_norm": 0.8840091824531555,
+      "learning_rate": 6.161137440758295e-06,
+      "loss": 0.4177,
+      "step": 244
+    },
+    {
+      "epoch": 0.38704581358609796,
+      "grad_norm": 0.5949338674545288,
+      "learning_rate": 6.145339652448658e-06,
+      "loss": 0.4477,
+      "step": 245
+    },
+    {
+      "epoch": 0.3886255924170616,
+      "grad_norm": 0.5938326120376587,
+      "learning_rate": 6.1295418641390216e-06,
+      "loss": 0.4155,
+      "step": 246
+    },
+    {
+      "epoch": 0.3902053712480253,
+      "grad_norm": 0.5401394367218018,
+      "learning_rate": 6.1137440758293845e-06,
+      "loss": 0.3873,
+      "step": 247
+    },
+    {
+      "epoch": 0.39178515007898895,
+      "grad_norm": 0.5220497846603394,
+      "learning_rate": 6.097946287519748e-06,
+      "loss": 0.3803,
+      "step": 248
+    },
+    {
+      "epoch": 0.3933649289099526,
+      "grad_norm": 0.5426644086837769,
+      "learning_rate": 6.082148499210111e-06,
+      "loss": 0.3239,
+      "step": 249
+    },
+    {
+      "epoch": 0.3949447077409163,
+      "grad_norm": 0.5215898156166077,
+      "learning_rate": 6.066350710900475e-06,
+      "loss": 0.4373,
+      "step": 250
+    },
+    {
+      "epoch": 0.39652448657187994,
+      "grad_norm": 0.5694135427474976,
+      "learning_rate": 6.050552922590838e-06,
+      "loss": 0.4948,
+      "step": 251
+    },
+    {
+      "epoch": 0.3981042654028436,
+      "grad_norm": 0.5505183339118958,
+      "learning_rate": 6.034755134281202e-06,
+      "loss": 0.4108,
+      "step": 252
+    },
+    {
+      "epoch": 0.39968404423380727,
+      "grad_norm": 0.593190610408783,
+      "learning_rate": 6.018957345971565e-06,
+      "loss": 0.429,
+      "step": 253
+    },
+    {
+      "epoch": 0.40126382306477093,
+      "grad_norm": 0.5409046411514282,
+      "learning_rate": 6.003159557661928e-06,
+      "loss": 0.4443,
+      "step": 254
+    },
+    {
+      "epoch": 0.4028436018957346,
+      "grad_norm": 0.5520291328430176,
+      "learning_rate": 5.987361769352291e-06,
+      "loss": 0.4485,
+      "step": 255
+    },
+    {
+      "epoch": 0.40442338072669826,
+      "grad_norm": 0.5622429847717285,
+      "learning_rate": 5.971563981042654e-06,
+      "loss": 0.4181,
+      "step": 256
+    },
+    {
+      "epoch": 0.4060031595576619,
+      "grad_norm": 0.5267983078956604,
+      "learning_rate": 5.955766192733018e-06,
+      "loss": 0.4235,
+      "step": 257
+    },
+    {
+      "epoch": 0.4075829383886256,
+      "grad_norm": 0.5384082198143005,
+      "learning_rate": 5.939968404423381e-06,
+      "loss": 0.4055,
+      "step": 258
+    },
+    {
+      "epoch": 0.40916271721958924,
+      "grad_norm": 0.5427289605140686,
+      "learning_rate": 5.924170616113745e-06,
+      "loss": 0.3427,
+      "step": 259
+    },
+    {
+      "epoch": 0.4107424960505529,
+      "grad_norm": 0.4936423599720001,
+      "learning_rate": 5.908372827804108e-06,
+      "loss": 0.4133,
+      "step": 260
+    },
+    {
+      "epoch": 0.41232227488151657,
+      "grad_norm": 0.5825520753860474,
+      "learning_rate": 5.8925750394944715e-06,
+      "loss": 0.377,
+      "step": 261
+    },
+    {
+      "epoch": 0.41390205371248023,
+      "grad_norm": 0.6343340277671814,
+      "learning_rate": 5.876777251184834e-06,
+      "loss": 0.441,
+      "step": 262
+    },
+    {
+      "epoch": 0.4154818325434439,
+      "grad_norm": 0.5479387044906616,
+      "learning_rate": 5.860979462875198e-06,
+      "loss": 0.4353,
+      "step": 263
+    },
+    {
+      "epoch": 0.41706161137440756,
+      "grad_norm": 0.5873805284500122,
+      "learning_rate": 5.845181674565561e-06,
+      "loss": 0.4293,
+      "step": 264
+    },
+    {
+      "epoch": 0.4186413902053712,
+      "grad_norm": 0.6624792218208313,
+      "learning_rate": 5.829383886255925e-06,
+      "loss": 0.5162,
+      "step": 265
+    },
+    {
+      "epoch": 0.42022116903633494,
+      "grad_norm": 0.5797149538993835,
+      "learning_rate": 5.813586097946288e-06,
+      "loss": 0.3651,
+      "step": 266
+    },
+    {
+      "epoch": 0.4218009478672986,
+      "grad_norm": 0.5814763903617859,
+      "learning_rate": 5.797788309636652e-06,
+      "loss": 0.3817,
+      "step": 267
+    },
+    {
+      "epoch": 0.42338072669826227,
+      "grad_norm": 0.5556735992431641,
+      "learning_rate": 5.7819905213270145e-06,
+      "loss": 0.4186,
+      "step": 268
+    },
+    {
+      "epoch": 0.42496050552922593,
+      "grad_norm": 0.5842727422714233,
+      "learning_rate": 5.766192733017378e-06,
+      "loss": 0.4343,
+      "step": 269
+    },
+    {
+      "epoch": 0.4265402843601896,
+      "grad_norm": 0.5401722192764282,
+      "learning_rate": 5.750394944707741e-06,
+      "loss": 0.4418,
+      "step": 270
+    },
+    {
+      "epoch": 0.42812006319115326,
+      "grad_norm": 0.5917039513587952,
+      "learning_rate": 5.734597156398105e-06,
+      "loss": 0.5371,
+      "step": 271
+    },
+    {
+      "epoch": 0.4296998420221169,
+      "grad_norm": 0.5991331338882446,
+      "learning_rate": 5.718799368088468e-06,
+      "loss": 0.4969,
+      "step": 272
+    },
+    {
+      "epoch": 0.4312796208530806,
+      "grad_norm": 0.4709448218345642,
+      "learning_rate": 5.703001579778832e-06,
+      "loss": 0.4139,
+      "step": 273
+    },
+    {
+      "epoch": 0.43285939968404424,
+      "grad_norm": 0.5746496319770813,
+      "learning_rate": 5.687203791469195e-06,
+      "loss": 0.4683,
+      "step": 274
+    },
+    {
+      "epoch": 0.4344391785150079,
+      "grad_norm": 0.523835301399231,
+      "learning_rate": 5.6714060031595584e-06,
+      "loss": 0.4346,
+      "step": 275
+    },
+    {
+      "epoch": 0.43601895734597157,
+      "grad_norm": 0.5292810797691345,
+      "learning_rate": 5.655608214849921e-06,
+      "loss": 0.463,
+      "step": 276
+    },
+    {
+      "epoch": 0.43759873617693523,
+      "grad_norm": 0.6543466448783875,
+      "learning_rate": 5.639810426540285e-06,
+      "loss": 0.427,
+      "step": 277
+    },
+    {
+      "epoch": 0.4391785150078989,
+      "grad_norm": 0.5543989539146423,
+      "learning_rate": 5.624012638230648e-06,
+      "loss": 0.3902,
+      "step": 278
+    },
+    {
+      "epoch": 0.44075829383886256,
+      "grad_norm": 0.5905360579490662,
+      "learning_rate": 5.608214849921012e-06,
+      "loss": 0.4266,
+      "step": 279
+    },
+    {
+      "epoch": 0.4423380726698262,
+      "grad_norm": 0.5785796046257019,
+      "learning_rate": 5.592417061611375e-06,
+      "loss": 0.4521,
+      "step": 280
+    },
+    {
+      "epoch": 0.4439178515007899,
+      "grad_norm": 0.5580607056617737,
+      "learning_rate": 5.576619273301738e-06,
+      "loss": 0.378,
+      "step": 281
+    },
+    {
+      "epoch": 0.44549763033175355,
+      "grad_norm": 0.5100966691970825,
+      "learning_rate": 5.5608214849921015e-06,
+      "loss": 0.3876,
+      "step": 282
+    },
+    {
+      "epoch": 0.4470774091627172,
+      "grad_norm": 0.5704023241996765,
+      "learning_rate": 5.5450236966824644e-06,
+      "loss": 0.4694,
+      "step": 283
+    },
+    {
+      "epoch": 0.4486571879936809,
+      "grad_norm": 0.5954383611679077,
+      "learning_rate": 5.529225908372828e-06,
+      "loss": 0.5049,
+      "step": 284
+    },
+    {
+      "epoch": 0.45023696682464454,
+      "grad_norm": 0.5239635705947876,
+      "learning_rate": 5.513428120063191e-06,
+      "loss": 0.4182,
+      "step": 285
+    },
+    {
+      "epoch": 0.4518167456556082,
+      "grad_norm": 0.6643552780151367,
+      "learning_rate": 5.497630331753555e-06,
+      "loss": 0.4434,
+      "step": 286
+    },
+    {
+      "epoch": 0.45339652448657186,
+      "grad_norm": 0.6675540804862976,
+      "learning_rate": 5.481832543443918e-06,
+      "loss": 0.3745,
+      "step": 287
+    },
+    {
+      "epoch": 0.4549763033175355,
+      "grad_norm": 0.5871401429176331,
+      "learning_rate": 5.466034755134282e-06,
+      "loss": 0.5527,
+      "step": 288
+    },
+    {
+      "epoch": 0.4565560821484992,
+      "grad_norm": 0.5936838984489441,
+      "learning_rate": 5.4502369668246446e-06,
+      "loss": 0.4857,
+      "step": 289
+    },
+    {
+      "epoch": 0.45813586097946285,
+      "grad_norm": 0.5998191833496094,
+      "learning_rate": 5.434439178515008e-06,
+      "loss": 0.4395,
+      "step": 290
+    },
+    {
+      "epoch": 0.4597156398104265,
+      "grad_norm": 0.5102293491363525,
+      "learning_rate": 5.418641390205371e-06,
+      "loss": 0.4496,
+      "step": 291
+    },
+    {
+      "epoch": 0.46129541864139023,
+      "grad_norm": 0.6297216415405273,
+      "learning_rate": 5.402843601895735e-06,
+      "loss": 0.3555,
+      "step": 292
+    },
+    {
+      "epoch": 0.4628751974723539,
+      "grad_norm": 0.6780267953872681,
+      "learning_rate": 5.387045813586098e-06,
+      "loss": 0.3295,
+      "step": 293
+    },
+    {
+      "epoch": 0.46445497630331756,
+      "grad_norm": 0.5788872838020325,
+      "learning_rate": 5.371248025276462e-06,
+      "loss": 0.4293,
+      "step": 294
+    },
+    {
+      "epoch": 0.4660347551342812,
+      "grad_norm": 0.5679113268852234,
+      "learning_rate": 5.355450236966825e-06,
+      "loss": 0.4274,
+      "step": 295
+    },
+    {
+      "epoch": 0.4676145339652449,
+      "grad_norm": 0.5739018321037292,
+      "learning_rate": 5.3396524486571885e-06,
+      "loss": 0.3292,
+      "step": 296
+    },
+    {
+      "epoch": 0.46919431279620855,
+      "grad_norm": 0.5387299060821533,
+      "learning_rate": 5.323854660347551e-06,
+      "loss": 0.36,
+      "step": 297
+    },
+    {
+      "epoch": 0.4707740916271722,
+      "grad_norm": 0.4877624213695526,
+      "learning_rate": 5.308056872037915e-06,
+      "loss": 0.403,
+      "step": 298
+    },
+    {
+      "epoch": 0.47235387045813587,
+      "grad_norm": 0.5668107271194458,
+      "learning_rate": 5.292259083728278e-06,
+      "loss": 0.4087,
+      "step": 299
+    },
+    {
+      "epoch": 0.47393364928909953,
+      "grad_norm": 0.5592719316482544,
+      "learning_rate": 5.276461295418642e-06,
+      "loss": 0.405,
+      "step": 300
+    },
+    {
+      "epoch": 0.4755134281200632,
+      "grad_norm": 0.48879534006118774,
+      "learning_rate": 5.260663507109005e-06,
+      "loss": 0.3562,
+      "step": 301
+    },
+    {
+      "epoch": 0.47709320695102686,
+      "grad_norm": 0.5968641042709351,
+      "learning_rate": 5.244865718799369e-06,
+      "loss": 0.4216,
+      "step": 302
+    },
+    {
+      "epoch": 0.4786729857819905,
+      "grad_norm": 0.7803828120231628,
+      "learning_rate": 5.2290679304897315e-06,
+      "loss": 0.4014,
+      "step": 303
+    },
+    {
+      "epoch": 0.4802527646129542,
+      "grad_norm": 0.592827558517456,
+      "learning_rate": 5.213270142180096e-06,
+      "loss": 0.2895,
+      "step": 304
+    },
+    {
+      "epoch": 0.48183254344391785,
+      "grad_norm": 0.8070396184921265,
+      "learning_rate": 5.197472353870458e-06,
+      "loss": 0.3972,
+      "step": 305
+    },
+    {
+      "epoch": 0.4834123222748815,
+      "grad_norm": 0.5256397724151611,
+      "learning_rate": 5.181674565560821e-06,
+      "loss": 0.4384,
+      "step": 306
+    },
+    {
+      "epoch": 0.4849921011058452,
+      "grad_norm": 0.5307562947273254,
+      "learning_rate": 5.165876777251185e-06,
+      "loss": 0.3788,
+      "step": 307
+    },
+    {
+      "epoch": 0.48657187993680884,
+      "grad_norm": 0.4588807225227356,
+      "learning_rate": 5.150078988941548e-06,
+      "loss": 0.3491,
+      "step": 308
+    },
+    {
+      "epoch": 0.4881516587677725,
+      "grad_norm": 0.524919331073761,
+      "learning_rate": 5.134281200631912e-06,
+      "loss": 0.4375,
+      "step": 309
+    },
+    {
+      "epoch": 0.48973143759873616,
+      "grad_norm": 0.6611966490745544,
+      "learning_rate": 5.118483412322275e-06,
+      "loss": 0.4399,
+      "step": 310
+    },
+    {
+      "epoch": 0.4913112164296998,
+      "grad_norm": 0.5597748160362244,
+      "learning_rate": 5.102685624012638e-06,
+      "loss": 0.5073,
+      "step": 311
+    },
+    {
+      "epoch": 0.4928909952606635,
+      "grad_norm": 0.8958181738853455,
+      "learning_rate": 5.086887835703001e-06,
+      "loss": 0.4756,
+      "step": 312
+    },
+    {
+      "epoch": 0.49447077409162715,
+      "grad_norm": 0.4875742197036743,
+      "learning_rate": 5.071090047393366e-06,
+      "loss": 0.4424,
+      "step": 313
+    },
+    {
+      "epoch": 0.4960505529225908,
+      "grad_norm": 0.6110445261001587,
+      "learning_rate": 5.055292259083728e-06,
+      "loss": 0.4686,
+      "step": 314
+    },
+    {
+      "epoch": 0.4976303317535545,
+      "grad_norm": 0.5900540351867676,
+      "learning_rate": 5.039494470774093e-06,
+      "loss": 0.4,
+      "step": 315
+    },
+    {
+      "epoch": 0.49921011058451814,
+      "grad_norm": 0.624906599521637,
+      "learning_rate": 5.023696682464455e-06,
+      "loss": 0.3967,
+      "step": 316
+    },
+    {
+      "epoch": 0.5007898894154819,
+      "grad_norm": 0.6435191631317139,
+      "learning_rate": 5.007898894154819e-06,
+      "loss": 0.5104,
+      "step": 317
+    },
+    {
+      "epoch": 0.5023696682464455,
+      "grad_norm": 0.7464382648468018,
+      "learning_rate": 4.9921011058451815e-06,
+      "loss": 0.4621,
+      "step": 318
+    },
+    {
+      "epoch": 0.5039494470774092,
+      "grad_norm": 0.7912509441375732,
+      "learning_rate": 4.976303317535545e-06,
+      "loss": 0.4186,
+      "step": 319
+    },
+    {
+      "epoch": 0.5055292259083728,
+      "grad_norm": 0.6150445938110352,
+      "learning_rate": 4.960505529225908e-06,
+      "loss": 0.469,
+      "step": 320
+    },
+    {
+      "epoch": 0.5071090047393365,
+      "grad_norm": 0.5445781946182251,
+      "learning_rate": 4.944707740916272e-06,
+      "loss": 0.4111,
+      "step": 321
+    },
+    {
+      "epoch": 0.5086887835703001,
+      "grad_norm": 0.5628255605697632,
+      "learning_rate": 4.928909952606635e-06,
+      "loss": 0.4884,
+      "step": 322
+    },
+    {
+      "epoch": 0.5102685624012638,
+      "grad_norm": 0.5007054805755615,
+      "learning_rate": 4.913112164296999e-06,
+      "loss": 0.4315,
+      "step": 323
+    },
+    {
+      "epoch": 0.5118483412322274,
+      "grad_norm": 0.6346699595451355,
+      "learning_rate": 4.8973143759873624e-06,
+      "loss": 0.4033,
+      "step": 324
+    },
+    {
+      "epoch": 0.5134281200631912,
+      "grad_norm": 0.639045774936676,
+      "learning_rate": 4.881516587677725e-06,
+      "loss": 0.3748,
+      "step": 325
+    },
+    {
+      "epoch": 0.5150078988941548,
+      "grad_norm": 0.5578002333641052,
+      "learning_rate": 4.865718799368089e-06,
+      "loss": 0.5055,
+      "step": 326
+    },
+    {
+      "epoch": 0.5165876777251185,
+      "grad_norm": 0.5281325578689575,
+      "learning_rate": 4.849921011058452e-06,
+      "loss": 0.4307,
+      "step": 327
+    },
+    {
+      "epoch": 0.5181674565560821,
+      "grad_norm": 0.6557057499885559,
+      "learning_rate": 4.834123222748816e-06,
+      "loss": 0.4085,
+      "step": 328
+    },
+    {
+      "epoch": 0.5197472353870458,
+      "grad_norm": 0.5667731761932373,
+      "learning_rate": 4.818325434439179e-06,
+      "loss": 0.4774,
+      "step": 329
+    },
+    {
+      "epoch": 0.5213270142180095,
+      "grad_norm": 0.5362856984138489,
+      "learning_rate": 4.8025276461295426e-06,
+      "loss": 0.4316,
+      "step": 330
+    },
+    {
+      "epoch": 0.5229067930489731,
+      "grad_norm": 0.5326763391494751,
+      "learning_rate": 4.7867298578199055e-06,
+      "loss": 0.389,
+      "step": 331
+    },
+    {
+      "epoch": 0.5244865718799369,
+      "grad_norm": 0.4922950565814972,
+      "learning_rate": 4.770932069510269e-06,
+      "loss": 0.3756,
+      "step": 332
+    },
+    {
+      "epoch": 0.5260663507109005,
+      "grad_norm": 0.4961477518081665,
+      "learning_rate": 4.755134281200632e-06,
+      "loss": 0.4336,
+      "step": 333
+    },
+    {
+      "epoch": 0.5276461295418642,
+      "grad_norm": 0.5258511304855347,
+      "learning_rate": 4.739336492890996e-06,
+      "loss": 0.404,
+      "step": 334
+    },
+    {
+      "epoch": 0.5292259083728278,
+      "grad_norm": 0.5479301810264587,
+      "learning_rate": 4.723538704581359e-06,
+      "loss": 0.3578,
+      "step": 335
+    },
+    {
+      "epoch": 0.5308056872037915,
+      "grad_norm": 0.49883902072906494,
+      "learning_rate": 4.707740916271723e-06,
+      "loss": 0.3809,
+      "step": 336
+    },
+    {
+      "epoch": 0.5323854660347551,
+      "grad_norm": 0.5133053660392761,
+      "learning_rate": 4.691943127962086e-06,
+      "loss": 0.4091,
+      "step": 337
+    },
+    {
+      "epoch": 0.5339652448657188,
+      "grad_norm": 0.6334301829338074,
+      "learning_rate": 4.676145339652449e-06,
+      "loss": 0.4432,
+      "step": 338
+    },
+    {
+      "epoch": 0.5355450236966824,
+      "grad_norm": 0.5124396085739136,
+      "learning_rate": 4.660347551342812e-06,
+      "loss": 0.3557,
+      "step": 339
+    },
+    {
+      "epoch": 0.5371248025276462,
+      "grad_norm": 0.5863746404647827,
+      "learning_rate": 4.644549763033176e-06,
+      "loss": 0.4288,
+      "step": 340
+    },
+    {
+      "epoch": 0.5387045813586098,
+      "grad_norm": 0.6599943041801453,
+      "learning_rate": 4.628751974723539e-06,
+      "loss": 0.398,
+      "step": 341
+    },
+    {
+      "epoch": 0.5402843601895735,
+      "grad_norm": 0.480027437210083,
+      "learning_rate": 4.612954186413903e-06,
+      "loss": 0.4706,
+      "step": 342
+    },
+    {
+      "epoch": 0.5418641390205371,
+      "grad_norm": 0.6601845026016235,
+      "learning_rate": 4.597156398104266e-06,
+      "loss": 0.4092,
+      "step": 343
+    },
+    {
+      "epoch": 0.5434439178515008,
+      "grad_norm": 0.5557224154472351,
+      "learning_rate": 4.581358609794629e-06,
+      "loss": 0.389,
+      "step": 344
+    },
+    {
+      "epoch": 0.5450236966824644,
+      "grad_norm": 0.49160709977149963,
+      "learning_rate": 4.5655608214849925e-06,
+      "loss": 0.4338,
+      "step": 345
+    },
+    {
+      "epoch": 0.5466034755134281,
+      "grad_norm": 0.5284649133682251,
+      "learning_rate": 4.549763033175355e-06,
+      "loss": 0.403,
+      "step": 346
+    },
+    {
+      "epoch": 0.5481832543443917,
+      "grad_norm": 0.5501908659934998,
+      "learning_rate": 4.533965244865719e-06,
+      "loss": 0.4983,
+      "step": 347
+    },
+    {
+      "epoch": 0.5497630331753555,
+      "grad_norm": 0.5585077404975891,
+      "learning_rate": 4.518167456556082e-06,
+      "loss": 0.4219,
+      "step": 348
+    },
+    {
+      "epoch": 0.5513428120063191,
+      "grad_norm": 0.4565962255001068,
+      "learning_rate": 4.502369668246446e-06,
+      "loss": 0.3591,
+      "step": 349
+    },
+    {
+      "epoch": 0.5529225908372828,
+      "grad_norm": 0.5507949590682983,
+      "learning_rate": 4.486571879936809e-06,
+      "loss": 0.4752,
+      "step": 350
+    },
+    {
+      "epoch": 0.5545023696682464,
+      "grad_norm": 0.5490357875823975,
+      "learning_rate": 4.470774091627173e-06,
+      "loss": 0.4291,
+      "step": 351
+    },
+    {
+      "epoch": 0.5560821484992101,
+      "grad_norm": 0.5804268717765808,
+      "learning_rate": 4.4549763033175355e-06,
+      "loss": 0.3113,
+      "step": 352
+    },
+    {
+      "epoch": 0.5576619273301737,
+      "grad_norm": 0.4745613634586334,
+      "learning_rate": 4.439178515007899e-06,
+      "loss": 0.4196,
+      "step": 353
+    },
+    {
+      "epoch": 0.5592417061611374,
+      "grad_norm": 0.6223664283752441,
+      "learning_rate": 4.423380726698262e-06,
+      "loss": 0.4592,
+      "step": 354
+    },
+    {
+      "epoch": 0.5608214849921012,
+      "grad_norm": 0.8797832727432251,
+      "learning_rate": 4.407582938388626e-06,
+      "loss": 0.4448,
+      "step": 355
+    },
+    {
+      "epoch": 0.5624012638230648,
+      "grad_norm": 0.5569826364517212,
+      "learning_rate": 4.391785150078989e-06,
+      "loss": 0.3873,
+      "step": 356
+    },
+    {
+      "epoch": 0.5639810426540285,
+      "grad_norm": 0.4294510781764984,
+      "learning_rate": 4.375987361769353e-06,
+      "loss": 0.3407,
+      "step": 357
+    },
+    {
+      "epoch": 0.5655608214849921,
+      "grad_norm": 0.5657434463500977,
+      "learning_rate": 4.360189573459716e-06,
+      "loss": 0.3345,
+      "step": 358
+    },
+    {
+      "epoch": 0.5671406003159558,
+      "grad_norm": 0.5589077472686768,
+      "learning_rate": 4.3443917851500794e-06,
+      "loss": 0.5237,
+      "step": 359
+    },
+    {
+      "epoch": 0.5687203791469194,
+      "grad_norm": 0.6107128858566284,
+      "learning_rate": 4.328593996840442e-06,
+      "loss": 0.4354,
+      "step": 360
+    },
+    {
+      "epoch": 0.5703001579778831,
+      "grad_norm": 0.5671380758285522,
+      "learning_rate": 4.312796208530806e-06,
+      "loss": 0.3712,
+      "step": 361
+    },
+    {
+      "epoch": 0.5718799368088467,
+      "grad_norm": 0.508173406124115,
+      "learning_rate": 4.29699842022117e-06,
+      "loss": 0.4097,
+      "step": 362
+    },
+    {
+      "epoch": 0.5734597156398105,
+      "grad_norm": 0.6139382719993591,
+      "learning_rate": 4.281200631911533e-06,
+      "loss": 0.2646,
+      "step": 363
+    },
+    {
+      "epoch": 0.5750394944707741,
+      "grad_norm": 0.5677220821380615,
+      "learning_rate": 4.265402843601897e-06,
+      "loss": 0.3748,
+      "step": 364
+    },
+    {
+      "epoch": 0.5766192733017378,
+      "grad_norm": 0.530708372592926,
+      "learning_rate": 4.2496050552922596e-06,
+      "loss": 0.3857,
+      "step": 365
+    },
+    {
+      "epoch": 0.5781990521327014,
+      "grad_norm": 1.176272988319397,
+      "learning_rate": 4.233807266982623e-06,
+      "loss": 0.436,
+      "step": 366
+    },
+    {
+      "epoch": 0.5797788309636651,
+      "grad_norm": 0.6165753602981567,
+      "learning_rate": 4.218009478672986e-06,
+      "loss": 0.3898,
+      "step": 367
+    },
+    {
+      "epoch": 0.5813586097946287,
+      "grad_norm": 0.47574201226234436,
+      "learning_rate": 4.20221169036335e-06,
+      "loss": 0.3685,
+      "step": 368
+    },
+    {
+      "epoch": 0.5829383886255924,
+      "grad_norm": 0.5995083451271057,
+      "learning_rate": 4.186413902053712e-06,
+      "loss": 0.4686,
+      "step": 369
+    },
+    {
+      "epoch": 0.584518167456556,
+      "grad_norm": 0.5809090733528137,
+      "learning_rate": 4.170616113744076e-06,
+      "loss": 0.4514,
+      "step": 370
+    },
+    {
+      "epoch": 0.5860979462875198,
+      "grad_norm": 0.6154018044471741,
+      "learning_rate": 4.15481832543444e-06,
+      "loss": 0.3737,
+      "step": 371
+    },
+    {
+      "epoch": 0.5876777251184834,
+      "grad_norm": 0.5799654126167297,
+      "learning_rate": 4.139020537124803e-06,
+      "loss": 0.4285,
+      "step": 372
+    },
+    {
+      "epoch": 0.5892575039494471,
+      "grad_norm": 0.4476354420185089,
+      "learning_rate": 4.123222748815166e-06,
+      "loss": 0.4362,
+      "step": 373
+    },
+    {
+      "epoch": 0.5908372827804107,
+      "grad_norm": 0.6266714334487915,
+      "learning_rate": 4.107424960505529e-06,
+      "loss": 0.4943,
+      "step": 374
+    },
+    {
+      "epoch": 0.5924170616113744,
+      "grad_norm": 0.5103732347488403,
+      "learning_rate": 4.091627172195893e-06,
+      "loss": 0.4585,
+      "step": 375
+    },
+    {
+      "epoch": 0.593996840442338,
+      "grad_norm": 0.49011877179145813,
+      "learning_rate": 4.075829383886256e-06,
+      "loss": 0.4489,
+      "step": 376
+    },
+    {
+      "epoch": 0.5955766192733017,
+      "grad_norm": 0.5286844372749329,
+      "learning_rate": 4.06003159557662e-06,
+      "loss": 0.4114,
+      "step": 377
+    },
+    {
+      "epoch": 0.5971563981042654,
+      "grad_norm": 0.494807630777359,
+      "learning_rate": 4.044233807266983e-06,
+      "loss": 0.3514,
+      "step": 378
+    },
+    {
+      "epoch": 0.5987361769352291,
+      "grad_norm": 0.46120524406433105,
+      "learning_rate": 4.0284360189573465e-06,
+      "loss": 0.4452,
+      "step": 379
+    },
+    {
+      "epoch": 0.6003159557661928,
+      "grad_norm": 0.6024404764175415,
+      "learning_rate": 4.0126382306477095e-06,
+      "loss": 0.4368,
+      "step": 380
+    },
+    {
+      "epoch": 0.6018957345971564,
+      "grad_norm": 0.8292664885520935,
+      "learning_rate": 3.996840442338073e-06,
+      "loss": 0.4495,
+      "step": 381
+    },
+    {
+      "epoch": 0.6034755134281201,
+      "grad_norm": 0.5312369465827942,
+      "learning_rate": 3.981042654028436e-06,
+      "loss": 0.3642,
+      "step": 382
+    },
+    {
+      "epoch": 0.6050552922590837,
+      "grad_norm": 0.6373758316040039,
+      "learning_rate": 3.9652448657188e-06,
+      "loss": 0.3884,
+      "step": 383
+    },
+    {
+      "epoch": 0.6066350710900474,
+      "grad_norm": 0.5623313188552856,
+      "learning_rate": 3.949447077409163e-06,
+      "loss": 0.3489,
+      "step": 384
+    },
+    {
+      "epoch": 0.608214849921011,
+      "grad_norm": 0.5703821778297424,
+      "learning_rate": 3.933649289099527e-06,
+      "loss": 0.5309,
+      "step": 385
+    },
+    {
+      "epoch": 0.6097946287519748,
+      "grad_norm": 0.5930938720703125,
+      "learning_rate": 3.91785150078989e-06,
+      "loss": 0.4072,
+      "step": 386
+    },
+    {
+      "epoch": 0.6113744075829384,
+      "grad_norm": 0.5636332631111145,
+      "learning_rate": 3.902053712480253e-06,
+      "loss": 0.3938,
+      "step": 387
+    },
+    {
+      "epoch": 0.6129541864139021,
+      "grad_norm": 0.45709583163261414,
+      "learning_rate": 3.886255924170616e-06,
+      "loss": 0.4436,
+      "step": 388
+    },
+    {
+      "epoch": 0.6145339652448657,
+      "grad_norm": 0.5924400687217712,
+      "learning_rate": 3.87045813586098e-06,
+      "loss": 0.2939,
+      "step": 389
+    },
+    {
+      "epoch": 0.6161137440758294,
+      "grad_norm": 0.6232696175575256,
+      "learning_rate": 3.854660347551343e-06,
+      "loss": 0.4183,
+      "step": 390
+    },
+    {
+      "epoch": 0.617693522906793,
+      "grad_norm": 0.5407995581626892,
+      "learning_rate": 3.838862559241707e-06,
+      "loss": 0.3925,
+      "step": 391
+    },
+    {
+      "epoch": 0.6192733017377567,
+      "grad_norm": 0.524691104888916,
+      "learning_rate": 3.82306477093207e-06,
+      "loss": 0.4327,
+      "step": 392
+    },
+    {
+      "epoch": 0.6208530805687204,
+      "grad_norm": 0.5206206440925598,
+      "learning_rate": 3.8072669826224335e-06,
+      "loss": 0.4203,
+      "step": 393
+    },
+    {
+      "epoch": 0.6224328593996841,
+      "grad_norm": 0.6244251132011414,
+      "learning_rate": 3.7914691943127964e-06,
+      "loss": 0.4546,
+      "step": 394
+    },
+    {
+      "epoch": 0.6240126382306477,
+      "grad_norm": 0.707058846950531,
+      "learning_rate": 3.77567140600316e-06,
+      "loss": 0.4015,
+      "step": 395
+    },
+    {
+      "epoch": 0.6255924170616114,
+      "grad_norm": 0.5457757115364075,
+      "learning_rate": 3.759873617693523e-06,
+      "loss": 0.3962,
+      "step": 396
+    },
+    {
+      "epoch": 0.627172195892575,
+      "grad_norm": 0.5757611989974976,
+      "learning_rate": 3.7440758293838865e-06,
+      "loss": 0.4299,
+      "step": 397
+    },
+    {
+      "epoch": 0.6287519747235387,
+      "grad_norm": 0.5844476819038391,
+      "learning_rate": 3.72827804107425e-06,
+      "loss": 0.4674,
+      "step": 398
+    },
+    {
+      "epoch": 0.6303317535545023,
+      "grad_norm": 0.6859634518623352,
+      "learning_rate": 3.7124802527646132e-06,
+      "loss": 0.4253,
+      "step": 399
+    },
+    {
+      "epoch": 0.631911532385466,
+      "grad_norm": 0.5247636437416077,
+      "learning_rate": 3.6966824644549766e-06,
+      "loss": 0.4318,
+      "step": 400
+    },
+    {
+      "epoch": 0.6334913112164297,
+      "grad_norm": 0.6206024885177612,
+      "learning_rate": 3.68088467614534e-06,
+      "loss": 0.3759,
+      "step": 401
+    },
+    {
+      "epoch": 0.6350710900473934,
+      "grad_norm": 0.6237459182739258,
+      "learning_rate": 3.6650868878357033e-06,
+      "loss": 0.3642,
+      "step": 402
+    },
+    {
+      "epoch": 0.636650868878357,
+      "grad_norm": 0.8048799633979797,
+      "learning_rate": 3.6492890995260666e-06,
+      "loss": 0.514,
+      "step": 403
+    },
+    {
+      "epoch": 0.6382306477093207,
+      "grad_norm": 0.4662720561027527,
+      "learning_rate": 3.63349131121643e-06,
+      "loss": 0.3654,
+      "step": 404
+    },
+    {
+      "epoch": 0.6398104265402843,
+      "grad_norm": 0.5561702251434326,
+      "learning_rate": 3.6176935229067934e-06,
+      "loss": 0.3823,
+      "step": 405
+    },
+    {
+      "epoch": 0.641390205371248,
+      "grad_norm": 0.6143206357955933,
+      "learning_rate": 3.6018957345971567e-06,
+      "loss": 0.3938,
+      "step": 406
+    },
+    {
+      "epoch": 0.6429699842022117,
+      "grad_norm": 0.6854034662246704,
+      "learning_rate": 3.58609794628752e-06,
+      "loss": 0.4625,
+      "step": 407
+    },
+    {
+      "epoch": 0.6445497630331753,
+      "grad_norm": 0.5590549111366272,
+      "learning_rate": 3.5703001579778834e-06,
+      "loss": 0.4199,
+      "step": 408
+    },
+    {
+      "epoch": 0.6461295418641391,
+      "grad_norm": 0.642573356628418,
+      "learning_rate": 3.5545023696682468e-06,
+      "loss": 0.4366,
+      "step": 409
+    },
+    {
+      "epoch": 0.6477093206951027,
+      "grad_norm": 0.5898130536079407,
+      "learning_rate": 3.53870458135861e-06,
+      "loss": 0.4691,
+      "step": 410
+    },
+    {
+      "epoch": 0.6492890995260664,
+      "grad_norm": 0.5370688438415527,
+      "learning_rate": 3.5229067930489735e-06,
+      "loss": 0.45,
+      "step": 411
+    },
+    {
+      "epoch": 0.65086887835703,
+      "grad_norm": 0.6769170165061951,
+      "learning_rate": 3.507109004739337e-06,
+      "loss": 0.3962,
+      "step": 412
+    },
+    {
+      "epoch": 0.6524486571879937,
+      "grad_norm": 0.5891703367233276,
+      "learning_rate": 3.4913112164297e-06,
+      "loss": 0.4542,
+      "step": 413
+    },
+    {
+      "epoch": 0.6540284360189573,
+      "grad_norm": 0.42204615473747253,
+      "learning_rate": 3.4755134281200636e-06,
+      "loss": 0.3368,
+      "step": 414
+    },
+    {
+      "epoch": 0.655608214849921,
+      "grad_norm": 0.46033787727355957,
+      "learning_rate": 3.459715639810427e-06,
+      "loss": 0.4357,
+      "step": 415
+    },
+    {
+      "epoch": 0.6571879936808847,
+      "grad_norm": 0.5509577393531799,
+      "learning_rate": 3.4439178515007903e-06,
+      "loss": 0.3939,
+      "step": 416
+    },
+    {
+      "epoch": 0.6587677725118484,
+      "grad_norm": 0.5802867412567139,
+      "learning_rate": 3.4281200631911536e-06,
+      "loss": 0.4073,
+      "step": 417
+    },
+    {
+      "epoch": 0.660347551342812,
+      "grad_norm": 0.6130402684211731,
+      "learning_rate": 3.412322274881517e-06,
+      "loss": 0.3452,
+      "step": 418
+    },
+    {
+      "epoch": 0.6619273301737757,
+      "grad_norm": 0.6854075789451599,
+      "learning_rate": 3.39652448657188e-06,
+      "loss": 0.3551,
+      "step": 419
+    },
+    {
+      "epoch": 0.6635071090047393,
+      "grad_norm": 0.5365926027297974,
+      "learning_rate": 3.3807266982622433e-06,
+      "loss": 0.4011,
+      "step": 420
+    },
+    {
+      "epoch": 0.665086887835703,
+      "grad_norm": 1.0338938236236572,
+      "learning_rate": 3.3649289099526066e-06,
+      "loss": 0.4623,
+      "step": 421
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5612855553627014,
+      "learning_rate": 3.34913112164297e-06,
+      "loss": 0.3738,
+      "step": 422
+    },
+    {
+      "epoch": 0.6682464454976303,
+      "grad_norm": 0.5113286375999451,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3865,
+      "step": 423
+    },
+    {
+      "epoch": 0.669826224328594,
+      "grad_norm": 0.5509905815124512,
+      "learning_rate": 3.3175355450236967e-06,
+      "loss": 0.4093,
+      "step": 424
+    },
+    {
+      "epoch": 0.6714060031595577,
+      "grad_norm": 0.5425525903701782,
+      "learning_rate": 3.30173775671406e-06,
+      "loss": 0.383,
+      "step": 425
+    },
+    {
+      "epoch": 0.6729857819905213,
+      "grad_norm": 0.5866172909736633,
+      "learning_rate": 3.2859399684044234e-06,
+      "loss": 0.4843,
+      "step": 426
+    },
+    {
+      "epoch": 0.674565560821485,
+      "grad_norm": 1.0777703523635864,
+      "learning_rate": 3.2701421800947867e-06,
+      "loss": 0.3748,
+      "step": 427
+    },
+    {
+      "epoch": 0.6761453396524486,
+      "grad_norm": 0.49126845598220825,
+      "learning_rate": 3.25434439178515e-06,
+      "loss": 0.3505,
+      "step": 428
+    },
+    {
+      "epoch": 0.6777251184834123,
+      "grad_norm": 0.5471718311309814,
+      "learning_rate": 3.2385466034755135e-06,
+      "loss": 0.4755,
+      "step": 429
+    },
+    {
+      "epoch": 0.6793048973143759,
+      "grad_norm": 0.5689931511878967,
+      "learning_rate": 3.222748815165877e-06,
+      "loss": 0.3956,
+      "step": 430
+    },
+    {
+      "epoch": 0.6808846761453397,
+      "grad_norm": 0.6496183276176453,
+      "learning_rate": 3.2069510268562406e-06,
+      "loss": 0.4598,
+      "step": 431
+    },
+    {
+      "epoch": 0.6824644549763034,
+      "grad_norm": 0.47042712569236755,
+      "learning_rate": 3.191153238546604e-06,
+      "loss": 0.3756,
+      "step": 432
+    },
+    {
+      "epoch": 0.684044233807267,
+      "grad_norm": 0.5819857120513916,
+      "learning_rate": 3.1753554502369673e-06,
+      "loss": 0.4803,
+      "step": 433
+    },
+    {
+      "epoch": 0.6856240126382307,
+      "grad_norm": 0.5752127766609192,
+      "learning_rate": 3.1595576619273307e-06,
+      "loss": 0.3916,
+      "step": 434
+    },
+    {
+      "epoch": 0.6872037914691943,
+      "grad_norm": 0.6483988761901855,
+      "learning_rate": 3.143759873617694e-06,
+      "loss": 0.4338,
+      "step": 435
+    },
+    {
+      "epoch": 0.688783570300158,
+      "grad_norm": 0.7817516326904297,
+      "learning_rate": 3.1279620853080574e-06,
+      "loss": 0.3645,
+      "step": 436
+    },
+    {
+      "epoch": 0.6903633491311216,
+      "grad_norm": 0.4980696737766266,
+      "learning_rate": 3.1121642969984207e-06,
+      "loss": 0.3962,
+      "step": 437
+    },
+    {
+      "epoch": 0.6919431279620853,
+      "grad_norm": 0.5592882037162781,
+      "learning_rate": 3.096366508688784e-06,
+      "loss": 0.3645,
+      "step": 438
+    },
+    {
+      "epoch": 0.693522906793049,
+      "grad_norm": 0.6228163242340088,
+      "learning_rate": 3.0805687203791474e-06,
+      "loss": 0.3696,
+      "step": 439
+    },
+    {
+      "epoch": 0.6951026856240127,
+      "grad_norm": 0.6718009114265442,
+      "learning_rate": 3.0647709320695108e-06,
+      "loss": 0.4926,
+      "step": 440
+    },
+    {
+      "epoch": 0.6966824644549763,
+      "grad_norm": 0.6085376143455505,
+      "learning_rate": 3.048973143759874e-06,
+      "loss": 0.418,
+      "step": 441
+    },
+    {
+      "epoch": 0.69826224328594,
+      "grad_norm": 0.7716324925422668,
+      "learning_rate": 3.0331753554502375e-06,
+      "loss": 0.4038,
+      "step": 442
+    },
+    {
+      "epoch": 0.6998420221169036,
+      "grad_norm": 0.7239758968353271,
+      "learning_rate": 3.017377567140601e-06,
+      "loss": 0.4596,
+      "step": 443
+    },
+    {
+      "epoch": 0.7014218009478673,
+      "grad_norm": 0.6308011412620544,
+      "learning_rate": 3.001579778830964e-06,
+      "loss": 0.4082,
+      "step": 444
+    },
+    {
+      "epoch": 0.7030015797788309,
+      "grad_norm": 0.515626072883606,
+      "learning_rate": 2.985781990521327e-06,
+      "loss": 0.4688,
+      "step": 445
+    },
+    {
+      "epoch": 0.7045813586097947,
+      "grad_norm": 0.5395441651344299,
+      "learning_rate": 2.9699842022116905e-06,
+      "loss": 0.3448,
+      "step": 446
+    },
+    {
+      "epoch": 0.7061611374407583,
+      "grad_norm": 0.5883680582046509,
+      "learning_rate": 2.954186413902054e-06,
+      "loss": 0.4546,
+      "step": 447
+    },
+    {
+      "epoch": 0.707740916271722,
+      "grad_norm": 0.7300311326980591,
+      "learning_rate": 2.938388625592417e-06,
+      "loss": 0.368,
+      "step": 448
+    },
+    {
+      "epoch": 0.7093206951026856,
+      "grad_norm": 0.5901307463645935,
+      "learning_rate": 2.9225908372827806e-06,
+      "loss": 0.3688,
+      "step": 449
+    },
+    {
+      "epoch": 0.7109004739336493,
+      "grad_norm": 0.6521854996681213,
+      "learning_rate": 2.906793048973144e-06,
+      "loss": 0.3876,
+      "step": 450
+    },
+    {
+      "epoch": 0.7124802527646129,
+      "grad_norm": 0.688450038433075,
+      "learning_rate": 2.8909952606635073e-06,
+      "loss": 0.4298,
+      "step": 451
+    },
+    {
+      "epoch": 0.7140600315955766,
+      "grad_norm": 0.6533556580543518,
+      "learning_rate": 2.8751974723538706e-06,
+      "loss": 0.3589,
+      "step": 452
+    },
+    {
+      "epoch": 0.7156398104265402,
+      "grad_norm": 0.5261491537094116,
+      "learning_rate": 2.859399684044234e-06,
+      "loss": 0.3886,
+      "step": 453
+    },
+    {
+      "epoch": 0.717219589257504,
+      "grad_norm": 0.5488421320915222,
+      "learning_rate": 2.8436018957345973e-06,
+      "loss": 0.411,
+      "step": 454
+    },
+    {
+      "epoch": 0.7187993680884676,
+      "grad_norm": 0.6415657997131348,
+      "learning_rate": 2.8278041074249607e-06,
+      "loss": 0.4581,
+      "step": 455
+    },
+    {
+      "epoch": 0.7203791469194313,
+      "grad_norm": 0.5058445334434509,
+      "learning_rate": 2.812006319115324e-06,
+      "loss": 0.4325,
+      "step": 456
+    },
+    {
+      "epoch": 0.721958925750395,
+      "grad_norm": 0.6409322619438171,
+      "learning_rate": 2.7962085308056874e-06,
+      "loss": 0.3759,
+      "step": 457
+    },
+    {
+      "epoch": 0.7235387045813586,
+      "grad_norm": 0.5578014850616455,
+      "learning_rate": 2.7804107424960508e-06,
+      "loss": 0.3947,
+      "step": 458
+    },
+    {
+      "epoch": 0.7251184834123223,
+      "grad_norm": 0.6064183115959167,
+      "learning_rate": 2.764612954186414e-06,
+      "loss": 0.4766,
+      "step": 459
+    },
+    {
+      "epoch": 0.7266982622432859,
+      "grad_norm": 0.6067904233932495,
+      "learning_rate": 2.7488151658767775e-06,
+      "loss": 0.4698,
+      "step": 460
+    },
+    {
+      "epoch": 0.7282780410742496,
+      "grad_norm": 0.526088297367096,
+      "learning_rate": 2.733017377567141e-06,
+      "loss": 0.3997,
+      "step": 461
+    },
+    {
+      "epoch": 0.7298578199052133,
+      "grad_norm": 0.6290006637573242,
+      "learning_rate": 2.717219589257504e-06,
+      "loss": 0.4393,
+      "step": 462
+    },
+    {
+      "epoch": 0.731437598736177,
+      "grad_norm": 0.5822445154190063,
+      "learning_rate": 2.7014218009478675e-06,
+      "loss": 0.4767,
+      "step": 463
+    },
+    {
+      "epoch": 0.7330173775671406,
+      "grad_norm": 0.5798205733299255,
+      "learning_rate": 2.685624012638231e-06,
+      "loss": 0.4163,
+      "step": 464
+    },
+    {
+      "epoch": 0.7345971563981043,
+      "grad_norm": 0.6234124898910522,
+      "learning_rate": 2.6698262243285942e-06,
+      "loss": 0.387,
+      "step": 465
+    },
+    {
+      "epoch": 0.7361769352290679,
+      "grad_norm": 0.5226984620094299,
+      "learning_rate": 2.6540284360189576e-06,
+      "loss": 0.4144,
+      "step": 466
+    },
+    {
+      "epoch": 0.7377567140600316,
+      "grad_norm": 0.529303789138794,
+      "learning_rate": 2.638230647709321e-06,
+      "loss": 0.4689,
+      "step": 467
+    },
+    {
+      "epoch": 0.7393364928909952,
+      "grad_norm": 0.6620000004768372,
+      "learning_rate": 2.6224328593996843e-06,
+      "loss": 0.4358,
+      "step": 468
+    },
+    {
+      "epoch": 0.740916271721959,
+      "grad_norm": 0.8560294508934021,
+      "learning_rate": 2.606635071090048e-06,
+      "loss": 0.422,
+      "step": 469
+    },
+    {
+      "epoch": 0.7424960505529226,
+      "grad_norm": 0.47033989429473877,
+      "learning_rate": 2.5908372827804106e-06,
+      "loss": 0.4462,
+      "step": 470
+    },
+    {
+      "epoch": 0.7440758293838863,
+      "grad_norm": 0.5476656556129456,
+      "learning_rate": 2.575039494470774e-06,
+      "loss": 0.3818,
+      "step": 471
+    },
+    {
+      "epoch": 0.7456556082148499,
+      "grad_norm": 0.5771902203559875,
+      "learning_rate": 2.5592417061611373e-06,
+      "loss": 0.3835,
+      "step": 472
+    },
+    {
+      "epoch": 0.7472353870458136,
+      "grad_norm": 0.6452733278274536,
+      "learning_rate": 2.5434439178515007e-06,
+      "loss": 0.4224,
+      "step": 473
+    },
+    {
+      "epoch": 0.7488151658767772,
+      "grad_norm": 0.5318686962127686,
+      "learning_rate": 2.527646129541864e-06,
+      "loss": 0.4812,
+      "step": 474
+    },
+    {
+      "epoch": 0.7503949447077409,
+      "grad_norm": 0.6591460108757019,
+      "learning_rate": 2.5118483412322274e-06,
+      "loss": 0.4546,
+      "step": 475
+    },
+    {
+      "epoch": 0.7519747235387045,
+      "grad_norm": 0.5857440829277039,
+      "learning_rate": 2.4960505529225907e-06,
+      "loss": 0.4008,
+      "step": 476
+    },
+    {
+      "epoch": 0.7535545023696683,
+      "grad_norm": 0.6430768370628357,
+      "learning_rate": 2.480252764612954e-06,
+      "loss": 0.3191,
+      "step": 477
+    },
+    {
+      "epoch": 0.7551342812006319,
+      "grad_norm": 0.7442892789840698,
+      "learning_rate": 2.4644549763033174e-06,
+      "loss": 0.4171,
+      "step": 478
+    },
+    {
+      "epoch": 0.7567140600315956,
+      "grad_norm": 0.6390454173088074,
+      "learning_rate": 2.4486571879936812e-06,
+      "loss": 0.5381,
+      "step": 479
+    },
+    {
+      "epoch": 0.7582938388625592,
+      "grad_norm": 0.6277416348457336,
+      "learning_rate": 2.4328593996840446e-06,
+      "loss": 0.4824,
+      "step": 480
+    },
+    {
+      "epoch": 0.7598736176935229,
+      "grad_norm": 0.6043097972869873,
+      "learning_rate": 2.417061611374408e-06,
+      "loss": 0.4266,
+      "step": 481
+    },
+    {
+      "epoch": 0.7614533965244866,
+      "grad_norm": 0.6095964312553406,
+      "learning_rate": 2.4012638230647713e-06,
+      "loss": 0.4258,
+      "step": 482
+    },
+    {
+      "epoch": 0.7630331753554502,
+      "grad_norm": 0.5433639287948608,
+      "learning_rate": 2.3854660347551346e-06,
+      "loss": 0.4873,
+      "step": 483
+    },
+    {
+      "epoch": 0.764612954186414,
+      "grad_norm": 0.49287649989128113,
+      "learning_rate": 2.369668246445498e-06,
+      "loss": 0.4814,
+      "step": 484
+    },
+    {
+      "epoch": 0.7661927330173776,
+      "grad_norm": 0.5905902981758118,
+      "learning_rate": 2.3538704581358613e-06,
+      "loss": 0.4519,
+      "step": 485
+    },
+    {
+      "epoch": 0.7677725118483413,
+      "grad_norm": 0.6697285771369934,
+      "learning_rate": 2.3380726698262247e-06,
+      "loss": 0.4686,
+      "step": 486
+    },
+    {
+      "epoch": 0.7693522906793049,
+      "grad_norm": 0.5338664650917053,
+      "learning_rate": 2.322274881516588e-06,
+      "loss": 0.401,
+      "step": 487
+    },
+    {
+      "epoch": 0.7709320695102686,
+      "grad_norm": 0.5338428616523743,
+      "learning_rate": 2.3064770932069514e-06,
+      "loss": 0.4045,
+      "step": 488
+    },
+    {
+      "epoch": 0.7725118483412322,
+      "grad_norm": 0.6102830171585083,
+      "learning_rate": 2.2906793048973143e-06,
+      "loss": 0.3785,
+      "step": 489
+    },
+    {
+      "epoch": 0.7740916271721959,
+      "grad_norm": 0.5787335634231567,
+      "learning_rate": 2.2748815165876777e-06,
+      "loss": 0.42,
+      "step": 490
+    },
+    {
+      "epoch": 0.7756714060031595,
+      "grad_norm": 0.7426438331604004,
+      "learning_rate": 2.259083728278041e-06,
+      "loss": 0.4676,
+      "step": 491
+    },
+    {
+      "epoch": 0.7772511848341233,
+      "grad_norm": 0.5988475680351257,
+      "learning_rate": 2.2432859399684044e-06,
+      "loss": 0.5404,
+      "step": 492
+    },
+    {
+      "epoch": 0.7788309636650869,
+      "grad_norm": 0.6289830803871155,
+      "learning_rate": 2.2274881516587678e-06,
+      "loss": 0.396,
+      "step": 493
+    },
+    {
+      "epoch": 0.7804107424960506,
+      "grad_norm": 0.6077900528907776,
+      "learning_rate": 2.211690363349131e-06,
+      "loss": 0.4016,
+      "step": 494
+    },
+    {
+      "epoch": 0.7819905213270142,
+      "grad_norm": 0.8171889781951904,
+      "learning_rate": 2.1958925750394945e-06,
+      "loss": 0.3638,
+      "step": 495
+    },
+    {
+      "epoch": 0.7835703001579779,
+      "grad_norm": 0.6225026845932007,
+      "learning_rate": 2.180094786729858e-06,
+      "loss": 0.4088,
+      "step": 496
+    },
+    {
+      "epoch": 0.7851500789889415,
+      "grad_norm": 0.6262929439544678,
+      "learning_rate": 2.164296998420221e-06,
+      "loss": 0.3311,
+      "step": 497
+    },
+    {
+      "epoch": 0.7867298578199052,
+      "grad_norm": 0.662129282951355,
+      "learning_rate": 2.148499210110585e-06,
+      "loss": 0.4434,
+      "step": 498
+    },
+    {
+      "epoch": 0.7883096366508688,
+      "grad_norm": 0.5046777725219727,
+      "learning_rate": 2.1327014218009483e-06,
+      "loss": 0.5042,
+      "step": 499
+    },
+    {
+      "epoch": 0.7898894154818326,
+      "grad_norm": 0.6273382306098938,
+      "learning_rate": 2.1169036334913117e-06,
+      "loss": 0.345,
+      "step": 500
+    },
+    {
+      "epoch": 0.7914691943127962,
+      "grad_norm": 0.5484871864318848,
+      "learning_rate": 2.101105845181675e-06,
+      "loss": 0.3476,
+      "step": 501
+    },
+    {
+      "epoch": 0.7930489731437599,
+      "grad_norm": 0.6779518723487854,
+      "learning_rate": 2.085308056872038e-06,
+      "loss": 0.4062,
+      "step": 502
+    },
+    {
+      "epoch": 0.7946287519747235,
+      "grad_norm": 0.4969736635684967,
+      "learning_rate": 2.0695102685624013e-06,
+      "loss": 0.3615,
+      "step": 503
+    },
+    {
+      "epoch": 0.7962085308056872,
+      "grad_norm": 0.5542388558387756,
+      "learning_rate": 2.0537124802527647e-06,
+      "loss": 0.39,
+      "step": 504
+    },
+    {
+      "epoch": 0.7977883096366508,
+      "grad_norm": 0.8587651252746582,
+      "learning_rate": 2.037914691943128e-06,
+      "loss": 0.423,
+      "step": 505
+    },
+    {
+      "epoch": 0.7993680884676145,
+      "grad_norm": 0.6399357318878174,
+      "learning_rate": 2.0221169036334914e-06,
+      "loss": 0.4645,
+      "step": 506
+    },
+    {
+      "epoch": 0.8009478672985783,
+      "grad_norm": 0.5677849650382996,
+      "learning_rate": 2.0063191153238547e-06,
+      "loss": 0.3749,
+      "step": 507
+    },
+    {
+      "epoch": 0.8025276461295419,
+      "grad_norm": 0.5609621405601501,
+      "learning_rate": 1.990521327014218e-06,
+      "loss": 0.4727,
+      "step": 508
+    },
+    {
+      "epoch": 0.8041074249605056,
+      "grad_norm": 0.615185558795929,
+      "learning_rate": 1.9747235387045814e-06,
+      "loss": 0.4349,
+      "step": 509
+    },
+    {
+      "epoch": 0.8056872037914692,
+      "grad_norm": 0.5093739032745361,
+      "learning_rate": 1.958925750394945e-06,
+      "loss": 0.3502,
+      "step": 510
+    },
+    {
+      "epoch": 0.8072669826224329,
+      "grad_norm": 0.8513323068618774,
+      "learning_rate": 1.943127962085308e-06,
+      "loss": 0.3902,
+      "step": 511
+    },
+    {
+      "epoch": 0.8088467614533965,
+      "grad_norm": 0.6797610521316528,
+      "learning_rate": 1.9273301737756715e-06,
+      "loss": 0.4987,
+      "step": 512
+    },
+    {
+      "epoch": 0.8104265402843602,
+      "grad_norm": 0.5715585947036743,
+      "learning_rate": 1.911532385466035e-06,
+      "loss": 0.3965,
+      "step": 513
+    },
+    {
+      "epoch": 0.8120063191153238,
+      "grad_norm": 0.5537532567977905,
+      "learning_rate": 1.8957345971563982e-06,
+      "loss": 0.3832,
+      "step": 514
+    },
+    {
+      "epoch": 0.8135860979462876,
+      "grad_norm": 0.5337470173835754,
+      "learning_rate": 1.8799368088467616e-06,
+      "loss": 0.4136,
+      "step": 515
+    },
+    {
+      "epoch": 0.8151658767772512,
+      "grad_norm": 0.5929555892944336,
+      "learning_rate": 1.864139020537125e-06,
+      "loss": 0.3901,
+      "step": 516
+    },
+    {
+      "epoch": 0.8167456556082149,
+      "grad_norm": 0.6738921403884888,
+      "learning_rate": 1.8483412322274883e-06,
+      "loss": 0.4128,
+      "step": 517
+    },
+    {
+      "epoch": 0.8183254344391785,
+      "grad_norm": 0.598659098148346,
+      "learning_rate": 1.8325434439178516e-06,
+      "loss": 0.3707,
+      "step": 518
+    },
+    {
+      "epoch": 0.8199052132701422,
+      "grad_norm": 0.5679790377616882,
+      "learning_rate": 1.816745655608215e-06,
+      "loss": 0.457,
+      "step": 519
+    },
+    {
+      "epoch": 0.8214849921011058,
+      "grad_norm": 0.5459115505218506,
+      "learning_rate": 1.8009478672985784e-06,
+      "loss": 0.3613,
+      "step": 520
+    },
+    {
+      "epoch": 0.8230647709320695,
+      "grad_norm": 0.5752125978469849,
+      "learning_rate": 1.7851500789889417e-06,
+      "loss": 0.479,
+      "step": 521
+    },
+    {
+      "epoch": 0.8246445497630331,
+      "grad_norm": 0.5184637904167175,
+      "learning_rate": 1.769352290679305e-06,
+      "loss": 0.4126,
+      "step": 522
+    },
+    {
+      "epoch": 0.8262243285939969,
+      "grad_norm": 0.6329041123390198,
+      "learning_rate": 1.7535545023696684e-06,
+      "loss": 0.4221,
+      "step": 523
+    },
+    {
+      "epoch": 0.8278041074249605,
+      "grad_norm": 0.5233784317970276,
+      "learning_rate": 1.7377567140600318e-06,
+      "loss": 0.4375,
+      "step": 524
+    },
+    {
+      "epoch": 0.8293838862559242,
+      "grad_norm": 0.5424541234970093,
+      "learning_rate": 1.7219589257503951e-06,
+      "loss": 0.4447,
+      "step": 525
+    },
+    {
+      "epoch": 0.8309636650868878,
+      "grad_norm": 0.5534167885780334,
+      "learning_rate": 1.7061611374407585e-06,
+      "loss": 0.3672,
+      "step": 526
+    },
+    {
+      "epoch": 0.8325434439178515,
+      "grad_norm": 0.605102002620697,
+      "learning_rate": 1.6903633491311216e-06,
+      "loss": 0.4319,
+      "step": 527
+    },
+    {
+      "epoch": 0.8341232227488151,
+      "grad_norm": 0.5609396696090698,
+      "learning_rate": 1.674565560821485e-06,
+      "loss": 0.3984,
+      "step": 528
+    },
+    {
+      "epoch": 0.8357030015797788,
+      "grad_norm": 0.7964479923248291,
+      "learning_rate": 1.6587677725118483e-06,
+      "loss": 0.407,
+      "step": 529
+    },
+    {
+      "epoch": 0.8372827804107424,
+      "grad_norm": 0.4886048436164856,
+      "learning_rate": 1.6429699842022117e-06,
+      "loss": 0.4506,
+      "step": 530
+    },
+    {
+      "epoch": 0.8388625592417062,
+      "grad_norm": 0.543812096118927,
+      "learning_rate": 1.627172195892575e-06,
+      "loss": 0.3141,
+      "step": 531
+    },
+    {
+      "epoch": 0.8404423380726699,
+      "grad_norm": 0.5370059609413147,
+      "learning_rate": 1.6113744075829384e-06,
+      "loss": 0.3712,
+      "step": 532
+    },
+    {
+      "epoch": 0.8420221169036335,
+      "grad_norm": 0.7402203679084778,
+      "learning_rate": 1.595576619273302e-06,
+      "loss": 0.4136,
+      "step": 533
+    },
+    {
+      "epoch": 0.8436018957345972,
+      "grad_norm": 0.6814244985580444,
+      "learning_rate": 1.5797788309636653e-06,
+      "loss": 0.4634,
+      "step": 534
+    },
+    {
+      "epoch": 0.8451816745655608,
+      "grad_norm": 0.5919080972671509,
+      "learning_rate": 1.5639810426540287e-06,
+      "loss": 0.4238,
+      "step": 535
+    },
+    {
+      "epoch": 0.8467614533965245,
+      "grad_norm": 0.617522120475769,
+      "learning_rate": 1.548183254344392e-06,
+      "loss": 0.3431,
+      "step": 536
+    },
+    {
+      "epoch": 0.8483412322274881,
+      "grad_norm": 0.49482643604278564,
+      "learning_rate": 1.5323854660347554e-06,
+      "loss": 0.3882,
+      "step": 537
+    },
+    {
+      "epoch": 0.8499210110584519,
+      "grad_norm": 0.5525531768798828,
+      "learning_rate": 1.5165876777251187e-06,
+      "loss": 0.4053,
+      "step": 538
+    },
+    {
+      "epoch": 0.8515007898894155,
+      "grad_norm": 0.6634103655815125,
+      "learning_rate": 1.500789889415482e-06,
+      "loss": 0.4624,
+      "step": 539
+    },
+    {
+      "epoch": 0.8530805687203792,
+      "grad_norm": 0.45309382677078247,
+      "learning_rate": 1.4849921011058452e-06,
+      "loss": 0.3486,
+      "step": 540
+    },
+    {
+      "epoch": 0.8546603475513428,
+      "grad_norm": 0.778338611125946,
+      "learning_rate": 1.4691943127962086e-06,
+      "loss": 0.3984,
+      "step": 541
+    },
+    {
+      "epoch": 0.8562401263823065,
+      "grad_norm": 0.6093356609344482,
+      "learning_rate": 1.453396524486572e-06,
+      "loss": 0.333,
+      "step": 542
+    },
+    {
+      "epoch": 0.8578199052132701,
+      "grad_norm": 0.49551188945770264,
+      "learning_rate": 1.4375987361769353e-06,
+      "loss": 0.3915,
+      "step": 543
+    },
+    {
+      "epoch": 0.8593996840442338,
+      "grad_norm": 0.5423188209533691,
+      "learning_rate": 1.4218009478672987e-06,
+      "loss": 0.4192,
+      "step": 544
+    },
+    {
+      "epoch": 0.8609794628751974,
+      "grad_norm": 0.8111097812652588,
+      "learning_rate": 1.406003159557662e-06,
+      "loss": 0.473,
+      "step": 545
+    },
+    {
+      "epoch": 0.8625592417061612,
+      "grad_norm": 0.6064862012863159,
+      "learning_rate": 1.3902053712480254e-06,
+      "loss": 0.4164,
+      "step": 546
+    },
+    {
+      "epoch": 0.8641390205371248,
+      "grad_norm": 0.6180470585823059,
+      "learning_rate": 1.3744075829383887e-06,
+      "loss": 0.4351,
+      "step": 547
+    },
+    {
+      "epoch": 0.8657187993680885,
+      "grad_norm": 0.5101069808006287,
+      "learning_rate": 1.358609794628752e-06,
+      "loss": 0.3806,
+      "step": 548
+    },
+    {
+      "epoch": 0.8672985781990521,
+      "grad_norm": 0.6269749402999878,
+      "learning_rate": 1.3428120063191154e-06,
+      "loss": 0.4028,
+      "step": 549
+    },
+    {
+      "epoch": 0.8688783570300158,
+      "grad_norm": 0.6344918608665466,
+      "learning_rate": 1.3270142180094788e-06,
+      "loss": 0.3206,
+      "step": 550
+    },
+    {
+      "epoch": 0.8704581358609794,
+      "grad_norm": 0.7053835988044739,
+      "learning_rate": 1.3112164296998422e-06,
+      "loss": 0.4404,
+      "step": 551
+    },
+    {
+      "epoch": 0.8720379146919431,
+      "grad_norm": 0.4780917465686798,
+      "learning_rate": 1.2954186413902053e-06,
+      "loss": 0.4089,
+      "step": 552
+    },
+    {
+      "epoch": 0.8736176935229067,
+      "grad_norm": 0.5235942006111145,
+      "learning_rate": 1.2796208530805687e-06,
+      "loss": 0.3992,
+      "step": 553
+    },
+    {
+      "epoch": 0.8751974723538705,
+      "grad_norm": 0.5037370324134827,
+      "learning_rate": 1.263823064770932e-06,
+      "loss": 0.3727,
+      "step": 554
+    },
+    {
+      "epoch": 0.8767772511848341,
+      "grad_norm": 0.5422868132591248,
+      "learning_rate": 1.2480252764612954e-06,
+      "loss": 0.4524,
+      "step": 555
+    },
+    {
+      "epoch": 0.8783570300157978,
+      "grad_norm": 0.5287191271781921,
+      "learning_rate": 1.2322274881516587e-06,
+      "loss": 0.3445,
+      "step": 556
+    },
+    {
+      "epoch": 0.8799368088467614,
+      "grad_norm": 0.49679964780807495,
+      "learning_rate": 1.2164296998420223e-06,
+      "loss": 0.3357,
+      "step": 557
+    },
+    {
+      "epoch": 0.8815165876777251,
+      "grad_norm": 0.5391539931297302,
+      "learning_rate": 1.2006319115323856e-06,
+      "loss": 0.4645,
+      "step": 558
+    },
+    {
+      "epoch": 0.8830963665086888,
+      "grad_norm": 0.5474575757980347,
+      "learning_rate": 1.184834123222749e-06,
+      "loss": 0.4109,
+      "step": 559
+    },
+    {
+      "epoch": 0.8846761453396524,
+      "grad_norm": 0.5920886993408203,
+      "learning_rate": 1.1690363349131124e-06,
+      "loss": 0.4034,
+      "step": 560
+    },
+    {
+      "epoch": 0.8862559241706162,
+      "grad_norm": 0.5637263655662537,
+      "learning_rate": 1.1532385466034757e-06,
+      "loss": 0.392,
+      "step": 561
+    },
+    {
+      "epoch": 0.8878357030015798,
+      "grad_norm": 0.6719076037406921,
+      "learning_rate": 1.1374407582938388e-06,
+      "loss": 0.3798,
+      "step": 562
+    },
+    {
+      "epoch": 0.8894154818325435,
+      "grad_norm": 0.5554001927375793,
+      "learning_rate": 1.1216429699842022e-06,
+      "loss": 0.3901,
+      "step": 563
+    },
+    {
+      "epoch": 0.8909952606635071,
+      "grad_norm": 0.6078475713729858,
+      "learning_rate": 1.1058451816745656e-06,
+      "loss": 0.3574,
+      "step": 564
+    },
+    {
+      "epoch": 0.8925750394944708,
+      "grad_norm": 0.9478325843811035,
+      "learning_rate": 1.090047393364929e-06,
+      "loss": 0.3831,
+      "step": 565
+    },
+    {
+      "epoch": 0.8941548183254344,
+      "grad_norm": 0.5259877443313599,
+      "learning_rate": 1.0742496050552925e-06,
+      "loss": 0.4003,
+      "step": 566
+    },
+    {
+      "epoch": 0.8957345971563981,
+      "grad_norm": 0.5395880937576294,
+      "learning_rate": 1.0584518167456558e-06,
+      "loss": 0.3513,
+      "step": 567
+    },
+    {
+      "epoch": 0.8973143759873617,
+      "grad_norm": 0.5458592772483826,
+      "learning_rate": 1.042654028436019e-06,
+      "loss": 0.49,
+      "step": 568
+    },
+    {
+      "epoch": 0.8988941548183255,
+      "grad_norm": 0.5552616715431213,
+      "learning_rate": 1.0268562401263823e-06,
+      "loss": 0.3905,
+      "step": 569
+    },
+    {
+      "epoch": 0.9004739336492891,
+      "grad_norm": 0.551466166973114,
+      "learning_rate": 1.0110584518167457e-06,
+      "loss": 0.4241,
+      "step": 570
+    },
+    {
+      "epoch": 0.9020537124802528,
+      "grad_norm": 0.7195900082588196,
+      "learning_rate": 9.95260663507109e-07,
+      "loss": 0.3912,
+      "step": 571
+    },
+    {
+      "epoch": 0.9036334913112164,
+      "grad_norm": 0.5951517820358276,
+      "learning_rate": 9.794628751974724e-07,
+      "loss": 0.4267,
+      "step": 572
+    },
+    {
+      "epoch": 0.9052132701421801,
+      "grad_norm": 0.7582541108131409,
+      "learning_rate": 9.636650868878358e-07,
+      "loss": 0.4024,
+      "step": 573
+    },
+    {
+      "epoch": 0.9067930489731437,
+      "grad_norm": 0.6346389651298523,
+      "learning_rate": 9.478672985781991e-07,
+      "loss": 0.4677,
+      "step": 574
+    },
+    {
+      "epoch": 0.9083728278041074,
+      "grad_norm": 0.7323048710823059,
+      "learning_rate": 9.320695102685625e-07,
+      "loss": 0.4332,
+      "step": 575
+    },
+    {
+      "epoch": 0.909952606635071,
+      "grad_norm": 0.5796726942062378,
+      "learning_rate": 9.162717219589258e-07,
+      "loss": 0.3514,
+      "step": 576
+    },
+    {
+      "epoch": 0.9115323854660348,
+      "grad_norm": 0.7424004673957825,
+      "learning_rate": 9.004739336492892e-07,
+      "loss": 0.4178,
+      "step": 577
+    },
+    {
+      "epoch": 0.9131121642969984,
+      "grad_norm": 0.525142252445221,
+      "learning_rate": 8.846761453396525e-07,
+      "loss": 0.4498,
+      "step": 578
+    },
+    {
+      "epoch": 0.9146919431279621,
+      "grad_norm": 0.5565955638885498,
+      "learning_rate": 8.688783570300159e-07,
+      "loss": 0.4532,
+      "step": 579
+    },
+    {
+      "epoch": 0.9162717219589257,
+      "grad_norm": 0.540267288684845,
+      "learning_rate": 8.530805687203792e-07,
+      "loss": 0.4828,
+      "step": 580
+    },
+    {
+      "epoch": 0.9178515007898894,
+      "grad_norm": 0.5061677694320679,
+      "learning_rate": 8.372827804107425e-07,
+      "loss": 0.3505,
+      "step": 581
+    },
+    {
+      "epoch": 0.919431279620853,
+      "grad_norm": 0.5490908622741699,
+      "learning_rate": 8.214849921011058e-07,
+      "loss": 0.4402,
+      "step": 582
+    },
+    {
+      "epoch": 0.9210110584518167,
+      "grad_norm": 0.5788997411727905,
+      "learning_rate": 8.056872037914692e-07,
+      "loss": 0.3256,
+      "step": 583
+    },
+    {
+      "epoch": 0.9225908372827805,
+      "grad_norm": 0.5741492509841919,
+      "learning_rate": 7.898894154818327e-07,
+      "loss": 0.451,
+      "step": 584
+    },
+    {
+      "epoch": 0.9241706161137441,
+      "grad_norm": 0.5012090802192688,
+      "learning_rate": 7.74091627172196e-07,
+      "loss": 0.3513,
+      "step": 585
+    },
+    {
+      "epoch": 0.9257503949447078,
+      "grad_norm": 0.5613192915916443,
+      "learning_rate": 7.582938388625594e-07,
+      "loss": 0.3499,
+      "step": 586
+    },
+    {
+      "epoch": 0.9273301737756714,
+      "grad_norm": 0.5941815376281738,
+      "learning_rate": 7.424960505529226e-07,
+      "loss": 0.4133,
+      "step": 587
+    },
+    {
+      "epoch": 0.9289099526066351,
+      "grad_norm": 0.7772453427314758,
+      "learning_rate": 7.26698262243286e-07,
+      "loss": 0.3818,
+      "step": 588
+    },
+    {
+      "epoch": 0.9304897314375987,
+      "grad_norm": 0.5977700352668762,
+      "learning_rate": 7.109004739336493e-07,
+      "loss": 0.4099,
+      "step": 589
+    },
+    {
+      "epoch": 0.9320695102685624,
+      "grad_norm": 0.7777069807052612,
+      "learning_rate": 6.951026856240127e-07,
+      "loss": 0.4341,
+      "step": 590
+    },
+    {
+      "epoch": 0.933649289099526,
+      "grad_norm": 0.5362728834152222,
+      "learning_rate": 6.79304897314376e-07,
+      "loss": 0.4431,
+      "step": 591
+    },
+    {
+      "epoch": 0.9352290679304898,
+      "grad_norm": 0.5126134157180786,
+      "learning_rate": 6.635071090047394e-07,
+      "loss": 0.3713,
+      "step": 592
+    },
+    {
+      "epoch": 0.9368088467614534,
+      "grad_norm": 0.5886785984039307,
+      "learning_rate": 6.477093206951026e-07,
+      "loss": 0.405,
+      "step": 593
+    },
+    {
+      "epoch": 0.9383886255924171,
+      "grad_norm": 0.5328089594841003,
+      "learning_rate": 6.31911532385466e-07,
+      "loss": 0.3952,
+      "step": 594
+    },
+    {
+      "epoch": 0.9399684044233807,
+      "grad_norm": 0.7170501351356506,
+      "learning_rate": 6.161137440758294e-07,
+      "loss": 0.3979,
+      "step": 595
+    },
+    {
+      "epoch": 0.9415481832543444,
+      "grad_norm": 0.6048548817634583,
+      "learning_rate": 6.003159557661928e-07,
+      "loss": 0.3425,
+      "step": 596
+    },
+    {
+      "epoch": 0.943127962085308,
+      "grad_norm": 0.5635291337966919,
+      "learning_rate": 5.845181674565562e-07,
+      "loss": 0.3008,
+      "step": 597
+    },
+    {
+      "epoch": 0.9447077409162717,
+      "grad_norm": 0.6890112161636353,
+      "learning_rate": 5.687203791469194e-07,
+      "loss": 0.4205,
+      "step": 598
+    },
+    {
+      "epoch": 0.9462875197472354,
+      "grad_norm": 0.5197014212608337,
+      "learning_rate": 5.529225908372828e-07,
+      "loss": 0.4589,
+      "step": 599
+    },
+    {
+      "epoch": 0.9478672985781991,
+      "grad_norm": 0.5197718143463135,
+      "learning_rate": 5.371248025276462e-07,
+      "loss": 0.2678,
+      "step": 600
+    },
+    {
+      "epoch": 0.9494470774091627,
+      "grad_norm": 0.44931474328041077,
+      "learning_rate": 5.213270142180095e-07,
+      "loss": 0.4351,
+      "step": 601
+    },
+    {
+      "epoch": 0.9510268562401264,
+      "grad_norm": 0.47795984148979187,
+      "learning_rate": 5.055292259083728e-07,
+      "loss": 0.4392,
+      "step": 602
+    },
+    {
+      "epoch": 0.95260663507109,
+      "grad_norm": 0.6027578115463257,
+      "learning_rate": 4.897314375987362e-07,
+      "loss": 0.4499,
+      "step": 603
+    },
+    {
+      "epoch": 0.9541864139020537,
+      "grad_norm": 0.6160722374916077,
+      "learning_rate": 4.7393364928909956e-07,
+      "loss": 0.434,
+      "step": 604
+    },
+    {
+      "epoch": 0.9557661927330173,
+      "grad_norm": 0.8371343612670898,
+      "learning_rate": 4.581358609794629e-07,
+      "loss": 0.3911,
+      "step": 605
+    },
+    {
+      "epoch": 0.957345971563981,
+      "grad_norm": 0.5282484292984009,
+      "learning_rate": 4.4233807266982627e-07,
+      "loss": 0.4445,
+      "step": 606
+    },
+    {
+      "epoch": 0.9589257503949447,
+      "grad_norm": 0.5557743310928345,
+      "learning_rate": 4.265402843601896e-07,
+      "loss": 0.4103,
+      "step": 607
+    },
+    {
+      "epoch": 0.9605055292259084,
+      "grad_norm": 0.6362637281417847,
+      "learning_rate": 4.107424960505529e-07,
+      "loss": 0.3856,
+      "step": 608
+    },
+    {
+      "epoch": 0.9620853080568721,
+      "grad_norm": 0.745617151260376,
+      "learning_rate": 3.9494470774091633e-07,
+      "loss": 0.4179,
+      "step": 609
+    },
+    {
+      "epoch": 0.9636650868878357,
+      "grad_norm": 0.659038782119751,
+      "learning_rate": 3.791469194312797e-07,
+      "loss": 0.4027,
+      "step": 610
+    },
+    {
+      "epoch": 0.9652448657187994,
+      "grad_norm": 0.645199716091156,
+      "learning_rate": 3.63349131121643e-07,
+      "loss": 0.3501,
+      "step": 611
+    },
+    {
+      "epoch": 0.966824644549763,
+      "grad_norm": 0.4868941605091095,
+      "learning_rate": 3.4755134281200634e-07,
+      "loss": 0.3385,
+      "step": 612
+    },
+    {
+      "epoch": 0.9684044233807267,
+      "grad_norm": 0.5993934273719788,
+      "learning_rate": 3.317535545023697e-07,
+      "loss": 0.369,
+      "step": 613
+    },
+    {
+      "epoch": 0.9699842022116903,
+      "grad_norm": 0.6094574928283691,
+      "learning_rate": 3.15955766192733e-07,
+      "loss": 0.4899,
+      "step": 614
+    },
+    {
+      "epoch": 0.9715639810426541,
+      "grad_norm": 0.6989656686782837,
+      "learning_rate": 3.001579778830964e-07,
+      "loss": 0.4346,
+      "step": 615
+    },
+    {
+      "epoch": 0.9731437598736177,
+      "grad_norm": 0.5412940382957458,
+      "learning_rate": 2.843601895734597e-07,
+      "loss": 0.4515,
+      "step": 616
+    },
+    {
+      "epoch": 0.9747235387045814,
+      "grad_norm": 0.507622241973877,
+      "learning_rate": 2.685624012638231e-07,
+      "loss": 0.4171,
+      "step": 617
+    },
+    {
+      "epoch": 0.976303317535545,
+      "grad_norm": 0.4564089775085449,
+      "learning_rate": 2.527646129541864e-07,
+      "loss": 0.3452,
+      "step": 618
+    },
+    {
+      "epoch": 0.9778830963665087,
+      "grad_norm": 0.48170286417007446,
+      "learning_rate": 2.3696682464454978e-07,
+      "loss": 0.3866,
+      "step": 619
+    },
+    {
+      "epoch": 0.9794628751974723,
+      "grad_norm": 0.47774481773376465,
+      "learning_rate": 2.2116903633491313e-07,
+      "loss": 0.4425,
+      "step": 620
+    },
+    {
+      "epoch": 0.981042654028436,
+      "grad_norm": 0.4460739493370056,
+      "learning_rate": 2.0537124802527646e-07,
+      "loss": 0.3991,
+      "step": 621
+    },
+    {
+      "epoch": 0.9826224328593997,
+      "grad_norm": 0.536359965801239,
+      "learning_rate": 1.8957345971563984e-07,
+      "loss": 0.327,
+      "step": 622
+    },
+    {
+      "epoch": 0.9842022116903634,
+      "grad_norm": 0.5439571738243103,
+      "learning_rate": 1.7377567140600317e-07,
+      "loss": 0.408,
+      "step": 623
+    },
+    {
+      "epoch": 0.985781990521327,
+      "grad_norm": 0.8827345967292786,
+      "learning_rate": 1.579778830963665e-07,
+      "loss": 0.4924,
+      "step": 624
+    },
+    {
+      "epoch": 0.9873617693522907,
+      "grad_norm": 0.4992835521697998,
+      "learning_rate": 1.4218009478672986e-07,
+      "loss": 0.3921,
+      "step": 625
+    },
+    {
+      "epoch": 0.9889415481832543,
+      "grad_norm": 0.7306237816810608,
+      "learning_rate": 1.263823064770932e-07,
+      "loss": 0.5063,
+      "step": 626
+    },
+    {
+      "epoch": 0.990521327014218,
+      "grad_norm": 0.5200903415679932,
+      "learning_rate": 1.1058451816745657e-07,
+      "loss": 0.358,
+      "step": 627
+    },
+    {
+      "epoch": 0.9921011058451816,
+      "grad_norm": 0.42708104848861694,
+      "learning_rate": 9.478672985781992e-08,
+      "loss": 0.3361,
+      "step": 628
+    },
+    {
+      "epoch": 0.9936808846761453,
+      "grad_norm": 0.5993225574493408,
+      "learning_rate": 7.898894154818325e-08,
+      "loss": 0.3625,
+      "step": 629
+    },
+    {
+      "epoch": 0.995260663507109,
+      "grad_norm": 0.49995774030685425,
+      "learning_rate": 6.31911532385466e-08,
+      "loss": 0.3746,
+      "step": 630
+    },
+    {
+      "epoch": 0.9968404423380727,
+      "grad_norm": 0.5806180238723755,
+      "learning_rate": 4.739336492890996e-08,
+      "loss": 0.3727,
+      "step": 631
+    },
+    {
+      "epoch": 0.9984202211690363,
+      "grad_norm": 0.5514349341392517,
+      "learning_rate": 3.15955766192733e-08,
+      "loss": 0.4634,
+      "step": 632
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4094119668006897,
+      "learning_rate": 1.579778830963665e-08,
+      "loss": 0.2044,
+      "step": 633
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 633,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.9805266972408545e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}