diff --git "a/checkpoint-34364/trainer_state.json" "b/checkpoint-34364/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-34364/trainer_state.json"
@@ -0,0 +1,4195 @@
+{
+  "best_metric": 0.8927125930786133,
+  "best_model_checkpoint": "/workspace/previous_works/MedBLIP/output/MedBLIP-0005/checkpoint-34364",
+  "epoch": 1.80001047614059,
+  "eval_steps": 17182,
+  "global_step": 34364,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0030380807710439473,
+      "grad_norm": 119.46666717529297,
+      "learning_rate": 1.3268156424581008e-07,
+      "loss": 5.0724,
+      "step": 58
+    },
+    {
+      "epoch": 0.006076161542087895,
+      "grad_norm": 34.45277786254883,
+      "learning_rate": 3.3519553072625703e-07,
+      "loss": 3.6479,
+      "step": 116
+    },
+    {
+      "epoch": 0.009114242313131841,
+      "grad_norm": 14.232598304748535,
+      "learning_rate": 5.37709497206704e-07,
+      "loss": 2.2356,
+      "step": 174
+    },
+    {
+      "epoch": 0.01215232308417579,
+      "grad_norm": 10.709288597106934,
+      "learning_rate": 7.402234636871509e-07,
+      "loss": 1.9953,
+      "step": 232
+    },
+    {
+      "epoch": 0.015190403855219737,
+      "grad_norm": 7.259119987487793,
+      "learning_rate": 9.427374301675979e-07,
+      "loss": 1.8188,
+      "step": 290
+    },
+    {
+      "epoch": 0.018228484626263683,
+      "grad_norm": 6.452203273773193,
+      "learning_rate": 1.1452513966480447e-06,
+      "loss": 1.5639,
+      "step": 348
+    },
+    {
+      "epoch": 0.021266565397307632,
+      "grad_norm": 9.458000183105469,
+      "learning_rate": 1.3477653631284918e-06,
+      "loss": 1.504,
+      "step": 406
+    },
+    {
+      "epoch": 0.02430464616835158,
+      "grad_norm": 6.162489414215088,
+      "learning_rate": 1.5502793296089386e-06,
+      "loss": 1.2581,
+      "step": 464
+    },
+    {
+      "epoch": 0.027342726939395528,
+      "grad_norm": 6.684648513793945,
+      "learning_rate": 1.7527932960893857e-06,
+      "loss": 1.1713,
+      "step": 522
+    },
+    {
+      "epoch": 0.030380807710439474,
+      "grad_norm": 6.966240882873535,
+      "learning_rate": 1.9553072625698325e-06,
+      "loss": 1.0504,
+      "step": 580
+    },
+    {
+      "epoch": 0.03341888848148342,
+      "grad_norm": 6.379108905792236,
+      "learning_rate": 2.1578212290502796e-06,
+      "loss": 0.968,
+      "step": 638
+    },
+    {
+      "epoch": 0.036456969252527366,
+      "grad_norm": 5.3792619705200195,
+      "learning_rate": 2.3603351955307262e-06,
+      "loss": 0.9209,
+      "step": 696
+    },
+    {
+      "epoch": 0.039495050023571315,
+      "grad_norm": 6.413719177246094,
+      "learning_rate": 2.5628491620111733e-06,
+      "loss": 0.8723,
+      "step": 754
+    },
+    {
+      "epoch": 0.042533130794615265,
+      "grad_norm": 8.826435089111328,
+      "learning_rate": 2.7653631284916204e-06,
+      "loss": 0.8652,
+      "step": 812
+    },
+    {
+      "epoch": 0.045571211565659214,
+      "grad_norm": 6.294381618499756,
+      "learning_rate": 2.9678770949720674e-06,
+      "loss": 0.8099,
+      "step": 870
+    },
+    {
+      "epoch": 0.04860929233670316,
+      "grad_norm": 7.355430603027344,
+      "learning_rate": 3.170391061452514e-06,
+      "loss": 0.7396,
+      "step": 928
+    },
+    {
+      "epoch": 0.051647373107747106,
+      "grad_norm": 10.120753288269043,
+      "learning_rate": 3.372905027932961e-06,
+      "loss": 0.7511,
+      "step": 986
+    },
+    {
+      "epoch": 0.054685453878791056,
+      "grad_norm": 5.653336048126221,
+      "learning_rate": 3.575418994413408e-06,
+      "loss": 0.6688,
+      "step": 1044
+    },
+    {
+      "epoch": 0.057723534649835,
+      "grad_norm": 5.749114990234375,
+      "learning_rate": 3.7779329608938552e-06,
+      "loss": 0.677,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06076161542087895,
+      "grad_norm": 7.65744686126709,
+      "learning_rate": 3.980446927374302e-06,
+      "loss": 0.7195,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0637996961919229,
+      "grad_norm": 9.296794891357422,
+      "learning_rate": 4.1829608938547485e-06,
+      "loss": 0.7598,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06683777696296685,
+      "grad_norm": 6.070080757141113,
+      "learning_rate": 4.385474860335196e-06,
+      "loss": 0.7015,
+      "step": 1276
+    },
+    {
+      "epoch": 0.0698758577340108,
+      "grad_norm": 6.638489723205566,
+      "learning_rate": 4.5879888268156435e-06,
+      "loss": 0.6778,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07291393850505473,
+      "grad_norm": 9.076967239379883,
+      "learning_rate": 4.790502793296089e-06,
+      "loss": 0.5551,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07595201927609868,
+      "grad_norm": 6.929805278778076,
+      "learning_rate": 4.993016759776537e-06,
+      "loss": 0.6792,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07899010004714263,
+      "grad_norm": 8.316506385803223,
+      "learning_rate": 5.195530726256983e-06,
+      "loss": 0.6379,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08202818081818658,
+      "grad_norm": 9.471745491027832,
+      "learning_rate": 5.398044692737431e-06,
+      "loss": 0.6242,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08506626158923053,
+      "grad_norm": 6.022659778594971,
+      "learning_rate": 5.6005586592178775e-06,
+      "loss": 0.6368,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08810434236027448,
+      "grad_norm": 7.15187406539917,
+      "learning_rate": 5.803072625698325e-06,
+      "loss": 0.5381,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09114242313131843,
+      "grad_norm": 7.10537052154541,
+      "learning_rate": 6.005586592178772e-06,
+      "loss": 0.6885,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09418050390236236,
+      "grad_norm": 5.685272216796875,
+      "learning_rate": 6.208100558659218e-06,
+      "loss": 0.6066,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09721858467340631,
+      "grad_norm": 6.733754634857178,
+      "learning_rate": 6.410614525139666e-06,
+      "loss": 0.6124,
+      "step": 1856
+    },
+    {
+      "epoch": 0.10025666544445026,
+      "grad_norm": 6.112730026245117,
+      "learning_rate": 6.613128491620112e-06,
+      "loss": 0.6224,
+      "step": 1914
+    },
+    {
+      "epoch": 0.10329474621549421,
+      "grad_norm": 5.784328460693359,
+      "learning_rate": 6.815642458100559e-06,
+      "loss": 0.6289,
+      "step": 1972
+    },
+    {
+      "epoch": 0.10633282698653816,
+      "grad_norm": 9.69115924835205,
+      "learning_rate": 7.0181564245810065e-06,
+      "loss": 0.5952,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10937090775758211,
+      "grad_norm": 5.509926795959473,
+      "learning_rate": 7.220670391061453e-06,
+      "loss": 0.6288,
+      "step": 2088
+    },
+    {
+      "epoch": 0.11240898852862606,
+      "grad_norm": 6.545931339263916,
+      "learning_rate": 7.423184357541901e-06,
+      "loss": 0.5243,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11544706929967,
+      "grad_norm": 7.0921173095703125,
+      "learning_rate": 7.625698324022347e-06,
+      "loss": 0.5527,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11848515007071395,
+      "grad_norm": 5.278844833374023,
+      "learning_rate": 7.828212290502794e-06,
+      "loss": 0.6739,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1215232308417579,
+      "grad_norm": 5.799619197845459,
+      "learning_rate": 8.03072625698324e-06,
+      "loss": 0.5913,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12456131161280184,
+      "grad_norm": 7.148493766784668,
+      "learning_rate": 8.233240223463687e-06,
+      "loss": 0.6117,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1275993923838458,
+      "grad_norm": 5.538400650024414,
+      "learning_rate": 8.435754189944135e-06,
+      "loss": 0.5512,
+      "step": 2436
+    },
+    {
+      "epoch": 0.13063747315488974,
+      "grad_norm": 5.364485740661621,
+      "learning_rate": 8.638268156424582e-06,
+      "loss": 0.5458,
+      "step": 2494
+    },
+    {
+      "epoch": 0.1336755539259337,
+      "grad_norm": 7.3765549659729,
+      "learning_rate": 8.840782122905029e-06,
+      "loss": 0.5847,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13671363469697764,
+      "grad_norm": 5.364510536193848,
+      "learning_rate": 9.043296089385475e-06,
+      "loss": 0.542,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1397517154680216,
+      "grad_norm": 3.656923770904541,
+      "learning_rate": 9.245810055865922e-06,
+      "loss": 0.5528,
+      "step": 2668
+    },
+    {
+      "epoch": 0.14278979623906554,
+      "grad_norm": 3.117631435394287,
+      "learning_rate": 9.448324022346369e-06,
+      "loss": 0.5793,
+      "step": 2726
+    },
+    {
+      "epoch": 0.14582787701010946,
+      "grad_norm": 6.822358131408691,
+      "learning_rate": 9.650837988826817e-06,
+      "loss": 0.5967,
+      "step": 2784
+    },
+    {
+      "epoch": 0.1488659577811534,
+      "grad_norm": 4.629315376281738,
+      "learning_rate": 9.853351955307264e-06,
+      "loss": 0.5945,
+      "step": 2842
+    },
+    {
+      "epoch": 0.15190403855219736,
+      "grad_norm": 3.6398866176605225,
+      "learning_rate": 1.005586592178771e-05,
+      "loss": 0.6005,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1549421193232413,
+      "grad_norm": 5.811204433441162,
+      "learning_rate": 1.0258379888268157e-05,
+      "loss": 0.5065,
+      "step": 2958
+    },
+    {
+      "epoch": 0.15798020009428526,
+      "grad_norm": 3.8507301807403564,
+      "learning_rate": 1.0460893854748604e-05,
+      "loss": 0.5802,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1610182808653292,
+      "grad_norm": 5.666468143463135,
+      "learning_rate": 1.066340782122905e-05,
+      "loss": 0.5342,
+      "step": 3074
+    },
+    {
+      "epoch": 0.16405636163637316,
+      "grad_norm": 2.025376558303833,
+      "learning_rate": 1.0865921787709498e-05,
+      "loss": 0.5271,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1670944424074171,
+      "grad_norm": 5.698912143707275,
+      "learning_rate": 1.1068435754189945e-05,
+      "loss": 0.538,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17013252317846106,
+      "grad_norm": 4.067931652069092,
+      "learning_rate": 1.1270949720670392e-05,
+      "loss": 0.5601,
+      "step": 3248
+    },
+    {
+      "epoch": 0.173170603949505,
+      "grad_norm": 5.068817138671875,
+      "learning_rate": 1.1473463687150838e-05,
+      "loss": 0.5528,
+      "step": 3306
+    },
+    {
+      "epoch": 0.17620868472054896,
+      "grad_norm": 7.116920471191406,
+      "learning_rate": 1.1675977653631285e-05,
+      "loss": 0.5418,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1792467654915929,
+      "grad_norm": 8.113608360290527,
+      "learning_rate": 1.1878491620111732e-05,
+      "loss": 0.5196,
+      "step": 3422
+    },
+    {
+      "epoch": 0.18228484626263686,
+      "grad_norm": 4.820245265960693,
+      "learning_rate": 1.208100558659218e-05,
+      "loss": 0.6029,
+      "step": 3480
+    },
+    {
+      "epoch": 0.18532292703368078,
+      "grad_norm": 1.1077276468276978,
+      "learning_rate": 1.2283519553072627e-05,
+      "loss": 0.5779,
+      "step": 3538
+    },
+    {
+      "epoch": 0.18836100780472473,
+      "grad_norm": 3.024624824523926,
+      "learning_rate": 1.2486033519553073e-05,
+      "loss": 0.5091,
+      "step": 3596
+    },
+    {
+      "epoch": 0.19139908857576868,
+      "grad_norm": 8.059369087219238,
+      "learning_rate": 1.268854748603352e-05,
+      "loss": 0.5012,
+      "step": 3654
+    },
+    {
+      "epoch": 0.19443716934681263,
+      "grad_norm": 3.9895098209381104,
+      "learning_rate": 1.2891061452513967e-05,
+      "loss": 0.578,
+      "step": 3712
+    },
+    {
+      "epoch": 0.19747525011785658,
+      "grad_norm": 7.111061096191406,
+      "learning_rate": 1.3093575418994415e-05,
+      "loss": 0.4958,
+      "step": 3770
+    },
+    {
+      "epoch": 0.20051333088890053,
+      "grad_norm": 5.541796684265137,
+      "learning_rate": 1.3296089385474861e-05,
+      "loss": 0.589,
+      "step": 3828
+    },
+    {
+      "epoch": 0.20355141165994448,
+      "grad_norm": 4.365527629852295,
+      "learning_rate": 1.3498603351955308e-05,
+      "loss": 0.6003,
+      "step": 3886
+    },
+    {
+      "epoch": 0.20658949243098843,
+      "grad_norm": 4.486824035644531,
+      "learning_rate": 1.3701117318435755e-05,
+      "loss": 0.6135,
+      "step": 3944
+    },
+    {
+      "epoch": 0.20962757320203237,
+      "grad_norm": 5.487951278686523,
+      "learning_rate": 1.3903631284916201e-05,
+      "loss": 0.6234,
+      "step": 4002
+    },
+    {
+      "epoch": 0.21266565397307632,
+      "grad_norm": 2.3850884437561035,
+      "learning_rate": 1.410614525139665e-05,
+      "loss": 0.5536,
+      "step": 4060
+    },
+    {
+      "epoch": 0.21570373474412027,
+      "grad_norm": 3.7957749366760254,
+      "learning_rate": 1.4305167597765364e-05,
+      "loss": 0.5211,
+      "step": 4118
+    },
+    {
+      "epoch": 0.21874181551516422,
+      "grad_norm": 5.272162437438965,
+      "learning_rate": 1.450768156424581e-05,
+      "loss": 0.4992,
+      "step": 4176
+    },
+    {
+      "epoch": 0.22177989628620817,
+      "grad_norm": 4.292142868041992,
+      "learning_rate": 1.4710195530726259e-05,
+      "loss": 0.5124,
+      "step": 4234
+    },
+    {
+      "epoch": 0.22481797705725212,
+      "grad_norm": 4.948460102081299,
+      "learning_rate": 1.4912709497206705e-05,
+      "loss": 0.5575,
+      "step": 4292
+    },
+    {
+      "epoch": 0.22785605782829604,
+      "grad_norm": 3.456590414047241,
+      "learning_rate": 1.5115223463687152e-05,
+      "loss": 0.5672,
+      "step": 4350
+    },
+    {
+      "epoch": 0.23089413859934,
+      "grad_norm": 5.341044902801514,
+      "learning_rate": 1.53177374301676e-05,
+      "loss": 0.5106,
+      "step": 4408
+    },
+    {
+      "epoch": 0.23393221937038394,
+      "grad_norm": 5.5106987953186035,
+      "learning_rate": 1.5520251396648043e-05,
+      "loss": 0.6037,
+      "step": 4466
+    },
+    {
+      "epoch": 0.2369703001414279,
+      "grad_norm": 4.941389560699463,
+      "learning_rate": 1.5722765363128495e-05,
+      "loss": 0.5995,
+      "step": 4524
+    },
+    {
+      "epoch": 0.24000838091247184,
+      "grad_norm": 3.719957113265991,
+      "learning_rate": 1.592527932960894e-05,
+      "loss": 0.5531,
+      "step": 4582
+    },
+    {
+      "epoch": 0.2430464616835158,
+      "grad_norm": 4.435623645782471,
+      "learning_rate": 1.612779329608939e-05,
+      "loss": 0.5502,
+      "step": 4640
+    },
+    {
+      "epoch": 0.24608454245455974,
+      "grad_norm": 4.688556671142578,
+      "learning_rate": 1.6330307262569833e-05,
+      "loss": 0.5056,
+      "step": 4698
+    },
+    {
+      "epoch": 0.2491226232256037,
+      "grad_norm": 5.511931896209717,
+      "learning_rate": 1.653282122905028e-05,
+      "loss": 0.5242,
+      "step": 4756
+    },
+    {
+      "epoch": 0.25216070399664764,
+      "grad_norm": 4.933206558227539,
+      "learning_rate": 1.673533519553073e-05,
+      "loss": 0.5115,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2551987847676916,
+      "grad_norm": 2.942838191986084,
+      "learning_rate": 1.6937849162011175e-05,
+      "loss": 0.5941,
+      "step": 4872
+    },
+    {
+      "epoch": 0.25823686553873554,
+      "grad_norm": 4.0710625648498535,
+      "learning_rate": 1.7140363128491623e-05,
+      "loss": 0.4887,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2612749463097795,
+      "grad_norm": 3.212920665740967,
+      "learning_rate": 1.7342877094972068e-05,
+      "loss": 0.5051,
+      "step": 4988
+    },
+    {
+      "epoch": 0.26431302708082344,
+      "grad_norm": 4.390661716461182,
+      "learning_rate": 1.7545391061452513e-05,
+      "loss": 0.5735,
+      "step": 5046
+    },
+    {
+      "epoch": 0.2673511078518674,
+      "grad_norm": 3.784395217895508,
+      "learning_rate": 1.7747905027932965e-05,
+      "loss": 0.4847,
+      "step": 5104
+    },
+    {
+      "epoch": 0.27038918862291134,
+      "grad_norm": 4.238777160644531,
+      "learning_rate": 1.795041899441341e-05,
+      "loss": 0.4303,
+      "step": 5162
+    },
+    {
+      "epoch": 0.2734272693939553,
+      "grad_norm": 4.616554260253906,
+      "learning_rate": 1.8152932960893855e-05,
+      "loss": 0.53,
+      "step": 5220
+    },
+    {
+      "epoch": 0.27646535016499924,
+      "grad_norm": 3.0670206546783447,
+      "learning_rate": 1.8355446927374303e-05,
+      "loss": 0.5089,
+      "step": 5278
+    },
+    {
+      "epoch": 0.2795034309360432,
+      "grad_norm": 5.144998550415039,
+      "learning_rate": 1.8557960893854748e-05,
+      "loss": 0.5352,
+      "step": 5336
+    },
+    {
+      "epoch": 0.28254151170708713,
+      "grad_norm": 3.9276976585388184,
+      "learning_rate": 1.8760474860335196e-05,
+      "loss": 0.5001,
+      "step": 5394
+    },
+    {
+      "epoch": 0.2855795924781311,
+      "grad_norm": 4.67507266998291,
+      "learning_rate": 1.8962988826815645e-05,
+      "loss": 0.5106,
+      "step": 5452
+    },
+    {
+      "epoch": 0.288617673249175,
+      "grad_norm": 11.027061462402344,
+      "learning_rate": 1.916550279329609e-05,
+      "loss": 0.5311,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2916557540202189,
+      "grad_norm": 4.6279802322387695,
+      "learning_rate": 1.9368016759776538e-05,
+      "loss": 0.4722,
+      "step": 5568
+    },
+    {
+      "epoch": 0.2946938347912629,
+      "grad_norm": 3.2744059562683105,
+      "learning_rate": 1.9570530726256983e-05,
+      "loss": 0.5435,
+      "step": 5626
+    },
+    {
+      "epoch": 0.2977319155623068,
+      "grad_norm": 4.361588478088379,
+      "learning_rate": 1.977304469273743e-05,
+      "loss": 0.5182,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3007699963333508,
+      "grad_norm": 4.725919246673584,
+      "learning_rate": 1.997555865921788e-05,
+      "loss": 0.5108,
+      "step": 5742
+    },
+    {
+      "epoch": 0.3038080771043947,
+      "grad_norm": 4.126678943634033,
+      "learning_rate": 1.999995169004151e-05,
+      "loss": 0.4113,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3068461578754387,
+      "grad_norm": 4.924627780914307,
+      "learning_rate": 1.999977932757864e-05,
+      "loss": 0.5911,
+      "step": 5858
+    },
+    {
+      "epoch": 0.3098842386464826,
+      "grad_norm": 3.836568832397461,
+      "learning_rate": 1.9999482004657697e-05,
+      "loss": 0.5589,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3129223194175266,
+      "grad_norm": 2.1909499168395996,
+      "learning_rate": 1.999905972499412e-05,
+      "loss": 0.5321,
+      "step": 5974
+    },
+    {
+      "epoch": 0.3159604001885705,
+      "grad_norm": 4.15761661529541,
+      "learning_rate": 1.9998512493864858e-05,
+      "loss": 0.4898,
+      "step": 6032
+    },
+    {
+      "epoch": 0.31899848095961447,
+      "grad_norm": 4.483209133148193,
+      "learning_rate": 1.9997840318108285e-05,
+      "loss": 0.5339,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3220365617306584,
+      "grad_norm": 4.208775997161865,
+      "learning_rate": 1.9997058007847493e-05,
+      "loss": 0.4381,
+      "step": 6148
+    },
+    {
+      "epoch": 0.32507464250170237,
+      "grad_norm": 4.7916951179504395,
+      "learning_rate": 1.999613812340473e-05,
+      "loss": 0.5622,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3281127232727463,
+      "grad_norm": 3.041353702545166,
+      "learning_rate": 1.999509332400555e-05,
+      "loss": 0.5797,
+      "step": 6264
+    },
+    {
+      "epoch": 0.33115080404379027,
+      "grad_norm": 2.720505714416504,
+      "learning_rate": 1.999392362270611e-05,
+      "loss": 0.5213,
+      "step": 6322
+    },
+    {
+      "epoch": 0.3341888848148342,
+      "grad_norm": 4.793097496032715,
+      "learning_rate": 1.999262903412336e-05,
+      "loss": 0.4593,
+      "step": 6380
+    },
+    {
+      "epoch": 0.33722696558587817,
+      "grad_norm": 3.890002489089966,
+      "learning_rate": 1.999120957443491e-05,
+      "loss": 0.4486,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3402650463569221,
+      "grad_norm": 3.537182569503784,
+      "learning_rate": 1.9989665261378772e-05,
+      "loss": 0.4879,
+      "step": 6496
+    },
+    {
+      "epoch": 0.34330312712796607,
+      "grad_norm": 1.6273483037948608,
+      "learning_rate": 1.998799611425319e-05,
+      "loss": 0.483,
+      "step": 6554
+    },
+    {
+      "epoch": 0.34634120789901,
+      "grad_norm": 1.544161081314087,
+      "learning_rate": 1.9986202153916356e-05,
+      "loss": 0.5295,
+      "step": 6612
+    },
+    {
+      "epoch": 0.34937928867005397,
+      "grad_norm": 4.516360759735107,
+      "learning_rate": 1.9984283402786177e-05,
+      "loss": 0.5544,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3524173694410979,
+      "grad_norm": 3.9603912830352783,
+      "learning_rate": 1.998223988483998e-05,
+      "loss": 0.5005,
+      "step": 6728
+    },
+    {
+      "epoch": 0.35545545021214187,
+      "grad_norm": 4.063785076141357,
+      "learning_rate": 1.998007162561423e-05,
+      "loss": 0.5339,
+      "step": 6786
+    },
+    {
+      "epoch": 0.3584935309831858,
+      "grad_norm": 4.825593948364258,
+      "learning_rate": 1.9977778652204192e-05,
+      "loss": 0.4702,
+      "step": 6844
+    },
+    {
+      "epoch": 0.36153161175422976,
+      "grad_norm": 0.5379557013511658,
+      "learning_rate": 1.997536099326359e-05,
+      "loss": 0.5397,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3645696925252737,
+      "grad_norm": 3.922156810760498,
+      "learning_rate": 1.9972818679004273e-05,
+      "loss": 0.5663,
+      "step": 6960
+    },
+    {
+      "epoch": 0.36760777329631766,
+      "grad_norm": 3.376941442489624,
+      "learning_rate": 1.9970198778515604e-05,
+      "loss": 0.5321,
+      "step": 7018
+    },
+    {
+      "epoch": 0.37064585406736156,
+      "grad_norm": 4.569897174835205,
+      "learning_rate": 1.9967409398301135e-05,
+      "loss": 0.4517,
+      "step": 7076
+    },
+    {
+      "epoch": 0.3736839348384055,
+      "grad_norm": 4.267284393310547,
+      "learning_rate": 1.9964495462133642e-05,
+      "loss": 0.5225,
+      "step": 7134
+    },
+    {
+      "epoch": 0.37672201560944946,
+      "grad_norm": 4.1275506019592285,
+      "learning_rate": 1.9961457006426603e-05,
+      "loss": 0.5007,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3797600963804934,
+      "grad_norm": 4.481261253356934,
+      "learning_rate": 1.995829406914954e-05,
+      "loss": 0.4754,
+      "step": 7250
+    },
+    {
+      "epoch": 0.38279817715153736,
+      "grad_norm": 2.876922845840454,
+      "learning_rate": 1.995500668982753e-05,
+      "loss": 0.4729,
+      "step": 7308
+    },
+    {
+      "epoch": 0.3858362579225813,
+      "grad_norm": 1.0541763305664062,
+      "learning_rate": 1.9951594909540727e-05,
+      "loss": 0.5697,
+      "step": 7366
+    },
+    {
+      "epoch": 0.38887433869362525,
+      "grad_norm": 3.462268114089966,
+      "learning_rate": 1.9948058770923837e-05,
+      "loss": 0.4803,
+      "step": 7424
+    },
+    {
+      "epoch": 0.3919124194646692,
+      "grad_norm": 3.209782123565674,
+      "learning_rate": 1.9944398318165578e-05,
+      "loss": 0.5239,
+      "step": 7482
+    },
+    {
+      "epoch": 0.39495050023571315,
+      "grad_norm": 4.3836445808410645,
+      "learning_rate": 1.994061359700815e-05,
+      "loss": 0.5096,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3979885810067571,
+      "grad_norm": 2.8133575916290283,
+      "learning_rate": 1.9936704654746642e-05,
+      "loss": 0.4546,
+      "step": 7598
+    },
+    {
+      "epoch": 0.40102666177780105,
+      "grad_norm": 3.709463596343994,
+      "learning_rate": 1.9932671540228456e-05,
+      "loss": 0.4882,
+      "step": 7656
+    },
+    {
+      "epoch": 0.404064742548845,
+      "grad_norm": 3.9174060821533203,
+      "learning_rate": 1.992851430385269e-05,
+      "loss": 0.4311,
+      "step": 7714
+    },
+    {
+      "epoch": 0.40710282331988895,
+      "grad_norm": 2.9282238483428955,
+      "learning_rate": 1.99242329975695e-05,
+      "loss": 0.5204,
+      "step": 7772
+    },
+    {
+      "epoch": 0.4101409040909329,
+      "grad_norm": 4.139567852020264,
+      "learning_rate": 1.9919827674879473e-05,
+      "loss": 0.4739,
+      "step": 7830
+    },
+    {
+      "epoch": 0.41317898486197685,
+      "grad_norm": 3.436636447906494,
+      "learning_rate": 1.9915298390832935e-05,
+      "loss": 0.4838,
+      "step": 7888
+    },
+    {
+      "epoch": 0.4162170656330208,
+      "grad_norm": 3.512646198272705,
+      "learning_rate": 1.9910645202029272e-05,
+      "loss": 0.4594,
+      "step": 7946
+    },
+    {
+      "epoch": 0.41925514640406475,
+      "grad_norm": 3.1627018451690674,
+      "learning_rate": 1.9905868166616234e-05,
+      "loss": 0.5628,
+      "step": 8004
+    },
+    {
+      "epoch": 0.4222932271751087,
+      "grad_norm": 1.1955090761184692,
+      "learning_rate": 1.990096734428919e-05,
+      "loss": 0.4587,
+      "step": 8062
+    },
+    {
+      "epoch": 0.42533130794615265,
+      "grad_norm": 3.0422959327697754,
+      "learning_rate": 1.989594279629039e-05,
+      "loss": 0.5523,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4283693887171966,
+      "grad_norm": 3.0934972763061523,
+      "learning_rate": 1.98907945854082e-05,
+      "loss": 0.4855,
+      "step": 8178
+    },
+    {
+      "epoch": 0.43140746948824055,
+      "grad_norm": 3.9694907665252686,
+      "learning_rate": 1.9885522775976324e-05,
+      "loss": 0.543,
+      "step": 8236
+    },
+    {
+      "epoch": 0.4344455502592845,
+      "grad_norm": 4.713873386383057,
+      "learning_rate": 1.9880127433872983e-05,
+      "loss": 0.4901,
+      "step": 8294
+    },
+    {
+      "epoch": 0.43748363103032845,
+      "grad_norm": 2.3840503692626953,
+      "learning_rate": 1.987460862652011e-05,
+      "loss": 0.4265,
+      "step": 8352
+    },
+    {
+      "epoch": 0.4405217118013724,
+      "grad_norm": 4.123522758483887,
+      "learning_rate": 1.9868966422882496e-05,
+      "loss": 0.4237,
+      "step": 8410
+    },
+    {
+      "epoch": 0.44355979257241634,
+      "grad_norm": 3.1017978191375732,
+      "learning_rate": 1.986320089346693e-05,
+      "loss": 0.4106,
+      "step": 8468
+    },
+    {
+      "epoch": 0.4465978733434603,
+      "grad_norm": 2.8059699535369873,
+      "learning_rate": 1.9857414684867994e-05,
+      "loss": 0.4641,
+      "step": 8526
+    },
+    {
+      "epoch": 0.44963595411450424,
+      "grad_norm": 4.327667236328125,
+      "learning_rate": 1.985140484474396e-05,
+      "loss": 0.4337,
+      "step": 8584
+    },
+    {
+      "epoch": 0.45267403488554814,
+      "grad_norm": 0.9626501798629761,
+      "learning_rate": 1.9845271898297104e-05,
+      "loss": 0.4932,
+      "step": 8642
+    },
+    {
+      "epoch": 0.4557121156565921,
+      "grad_norm": 1.3852657079696655,
+      "learning_rate": 1.9839015922166693e-05,
+      "loss": 0.4866,
+      "step": 8700
+    },
+    {
+      "epoch": 0.45875019642763604,
+      "grad_norm": 3.2711095809936523,
+      "learning_rate": 1.983263699452942e-05,
+      "loss": 0.4825,
+      "step": 8758
+    },
+    {
+      "epoch": 0.46178827719868,
+      "grad_norm": 4.84442138671875,
+      "learning_rate": 1.9826135195098416e-05,
+      "loss": 0.4559,
+      "step": 8816
+    },
+    {
+      "epoch": 0.46482635796972394,
+      "grad_norm": 1.2177191972732544,
+      "learning_rate": 1.9819510605122255e-05,
+      "loss": 0.4795,
+      "step": 8874
+    },
+    {
+      "epoch": 0.4678644387407679,
+      "grad_norm": 3.1849379539489746,
+      "learning_rate": 1.981276330738395e-05,
+      "loss": 0.474,
+      "step": 8932
+    },
+    {
+      "epoch": 0.47090251951181183,
+      "grad_norm": 4.420878887176514,
+      "learning_rate": 1.9805893386199892e-05,
+      "loss": 0.4876,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4739406002828558,
+      "grad_norm": 2.714984893798828,
+      "learning_rate": 1.9798900927418835e-05,
+      "loss": 0.4491,
+      "step": 9048
+    },
+    {
+      "epoch": 0.47697868105389973,
+      "grad_norm": 2.185593843460083,
+      "learning_rate": 1.9791786018420792e-05,
+      "loss": 0.4808,
+      "step": 9106
+    },
+    {
+      "epoch": 0.4800167618249437,
+      "grad_norm": 3.3326094150543213,
+      "learning_rate": 1.9784548748115946e-05,
+      "loss": 0.4502,
+      "step": 9164
+    },
+    {
+      "epoch": 0.48305484259598763,
+      "grad_norm": 3.4437661170959473,
+      "learning_rate": 1.977718920694356e-05,
+      "loss": 0.5049,
+      "step": 9222
+    },
+    {
+      "epoch": 0.4860929233670316,
+      "grad_norm": 5.456835746765137,
+      "learning_rate": 1.9769707486870825e-05,
+      "loss": 0.4791,
+      "step": 9280
+    },
+    {
+      "epoch": 0.48913100413807553,
+      "grad_norm": 5.109498023986816,
+      "learning_rate": 1.9762103681391724e-05,
+      "loss": 0.5396,
+      "step": 9338
+    },
+    {
+      "epoch": 0.4921690849091195,
+      "grad_norm": 4.347654342651367,
+      "learning_rate": 1.9754377885525854e-05,
+      "loss": 0.4433,
+      "step": 9396
+    },
+    {
+      "epoch": 0.49520716568016343,
+      "grad_norm": 3.837158203125,
+      "learning_rate": 1.9746530195817243e-05,
+      "loss": 0.4791,
+      "step": 9454
+    },
+    {
+      "epoch": 0.4982452464512074,
+      "grad_norm": 3.8552966117858887,
+      "learning_rate": 1.9738699146560578e-05,
+      "loss": 0.4979,
+      "step": 9512
+    },
+    {
+      "epoch": 0.5012833272222513,
+      "grad_norm": 1.76126229763031,
+      "learning_rate": 1.973061006224811e-05,
+      "loss": 0.4716,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5043214079932953,
+      "grad_norm": 5.198726654052734,
+      "learning_rate": 1.9722399381103267e-05,
+      "loss": 0.4801,
+      "step": 9628
+    },
+    {
+      "epoch": 0.5073594887643392,
+      "grad_norm": 3.2313361167907715,
+      "learning_rate": 1.9714067205729356e-05,
+      "loss": 0.4592,
+      "step": 9686
+    },
+    {
+      "epoch": 0.5103975695353832,
+      "grad_norm": 0.9486598968505859,
+      "learning_rate": 1.9705613640247928e-05,
+      "loss": 0.4399,
+      "step": 9744
+    },
+    {
+      "epoch": 0.5134356503064271,
+      "grad_norm": 3.271669864654541,
+      "learning_rate": 1.9697038790297442e-05,
+      "loss": 0.4722,
+      "step": 9802
+    },
+    {
+      "epoch": 0.5164737310774711,
+      "grad_norm": 5.1848039627075195,
+      "learning_rate": 1.9688342763031993e-05,
+      "loss": 0.4336,
+      "step": 9860
+    },
+    {
+      "epoch": 0.519511811848515,
+      "grad_norm": 4.134024620056152,
+      "learning_rate": 1.967952566711993e-05,
+      "loss": 0.4534,
+      "step": 9918
+    },
+    {
+      "epoch": 0.522549892619559,
+      "grad_norm": 3.0904159545898438,
+      "learning_rate": 1.9670587612742515e-05,
+      "loss": 0.4461,
+      "step": 9976
+    },
+    {
+      "epoch": 0.5255879733906029,
+      "grad_norm": 3.3785481452941895,
+      "learning_rate": 1.9661528711592553e-05,
+      "loss": 0.4906,
+      "step": 10034
+    },
+    {
+      "epoch": 0.5286260541616469,
+      "grad_norm": 4.755141258239746,
+      "learning_rate": 1.9652349076872986e-05,
+      "loss": 0.4519,
+      "step": 10092
+    },
+    {
+      "epoch": 0.5316641349326908,
+      "grad_norm": 4.502477645874023,
+      "learning_rate": 1.9643048823295482e-05,
+      "loss": 0.5454,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5347022157037348,
+      "grad_norm": 2.3361642360687256,
+      "learning_rate": 1.9633628067078997e-05,
+      "loss": 0.5069,
+      "step": 10208
+    },
+    {
+      "epoch": 0.5377402964747787,
+      "grad_norm": 3.6974456310272217,
+      "learning_rate": 1.9624086925948333e-05,
+      "loss": 0.4604,
+      "step": 10266
+    },
+    {
+      "epoch": 0.5407783772458227,
+      "grad_norm": 3.7012462615966797,
+      "learning_rate": 1.9614425519132654e-05,
+      "loss": 0.5368,
+      "step": 10324
+    },
+    {
+      "epoch": 0.5438164580168666,
+      "grad_norm": 0.9825100898742676,
+      "learning_rate": 1.9604643967364013e-05,
+      "loss": 0.4917,
+      "step": 10382
+    },
+    {
+      "epoch": 0.5468545387879106,
+      "grad_norm": 2.8980348110198975,
+      "learning_rate": 1.959474239287582e-05,
+      "loss": 0.4571,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5498926195589545,
+      "grad_norm": 6.615330696105957,
+      "learning_rate": 1.9584720919401342e-05,
+      "loss": 0.4949,
+      "step": 10498
+    },
+    {
+      "epoch": 0.5529307003299985,
+      "grad_norm": 4.613067626953125,
+      "learning_rate": 1.9574579672172126e-05,
+      "loss": 0.4072,
+      "step": 10556
+    },
+    {
+      "epoch": 0.5559687811010424,
+      "grad_norm": 3.3984858989715576,
+      "learning_rate": 1.9564318777916456e-05,
+      "loss": 0.412,
+      "step": 10614
+    },
+    {
+      "epoch": 0.5590068618720864,
+      "grad_norm": 5.624422550201416,
+      "learning_rate": 1.9553938364857775e-05,
+      "loss": 0.4781,
+      "step": 10672
+    },
+    {
+      "epoch": 0.5620449426431303,
+      "grad_norm": 4.486995697021484,
+      "learning_rate": 1.954343856271306e-05,
+      "loss": 0.4426,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5650830234141743,
+      "grad_norm": 2.862964391708374,
+      "learning_rate": 1.953281950269121e-05,
+      "loss": 0.506,
+      "step": 10788
+    },
+    {
+      "epoch": 0.5681211041852182,
+      "grad_norm": 0.556151807308197,
+      "learning_rate": 1.9522267467101615e-05,
+      "loss": 0.4095,
+      "step": 10846
+    },
+    {
+      "epoch": 0.5711591849562622,
+      "grad_norm": 3.035536527633667,
+      "learning_rate": 1.9511412341335318e-05,
+      "loss": 0.517,
+      "step": 10904
+    },
+    {
+      "epoch": 0.574197265727306,
+      "grad_norm": 3.8603086471557617,
+      "learning_rate": 1.950043835790185e-05,
+      "loss": 0.4344,
+      "step": 10962
+    },
+    {
+      "epoch": 0.57723534649835,
+      "grad_norm": 4.4469499588012695,
+      "learning_rate": 1.9489345653935635e-05,
+      "loss": 0.4774,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5802734272693939,
+      "grad_norm": 3.4457666873931885,
+      "learning_rate": 1.9478134368054676e-05,
+      "loss": 0.4274,
+      "step": 11078
+    },
+    {
+      "epoch": 0.5833115080404379,
+      "grad_norm": 3.056290864944458,
+      "learning_rate": 1.9466804640358798e-05,
+      "loss": 0.4432,
+      "step": 11136
+    },
+    {
+      "epoch": 0.5863495888114818,
+      "grad_norm": 4.071867942810059,
+      "learning_rate": 1.9455356612427928e-05,
+      "loss": 0.4344,
+      "step": 11194
+    },
+    {
+      "epoch": 0.5893876695825258,
+      "grad_norm": 2.0395846366882324,
+      "learning_rate": 1.9443790427320303e-05,
+      "loss": 0.4714,
+      "step": 11252
+    },
+    {
+      "epoch": 0.5924257503535697,
+      "grad_norm": 4.563007354736328,
+      "learning_rate": 1.9432106229570685e-05,
+      "loss": 0.5157,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5954638311246137,
+      "grad_norm": 3.7986621856689453,
+      "learning_rate": 1.9420304165188574e-05,
+      "loss": 0.4977,
+      "step": 11368
+    },
+    {
+      "epoch": 0.5985019118956576,
+      "grad_norm": 5.301217555999756,
+      "learning_rate": 1.9408384381656358e-05,
+      "loss": 0.4662,
+      "step": 11426
+    },
+    {
+      "epoch": 0.6015399926667016,
+      "grad_norm": 2.3288731575012207,
+      "learning_rate": 1.939634702792749e-05,
+      "loss": 0.4493,
+      "step": 11484
+    },
+    {
+      "epoch": 0.6045780734377455,
+      "grad_norm": 3.7128169536590576,
+      "learning_rate": 1.9384192254424606e-05,
+      "loss": 0.4865,
+      "step": 11542
+    },
+    {
+      "epoch": 0.6076161542087894,
+      "grad_norm": 4.314477920532227,
+      "learning_rate": 1.9371920213037665e-05,
+      "loss": 0.4715,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6106542349798334,
+      "grad_norm": 2.6989047527313232,
+      "learning_rate": 1.935953105712205e-05,
+      "loss": 0.4345,
+      "step": 11658
+    },
+    {
+      "epoch": 0.6136923157508773,
+      "grad_norm": 3.5463671684265137,
+      "learning_rate": 1.9347024941496628e-05,
+      "loss": 0.4611,
+      "step": 11716
+    },
+    {
+      "epoch": 0.6167303965219213,
+      "grad_norm": 4.914857387542725,
+      "learning_rate": 1.9334402022441848e-05,
+      "loss": 0.4952,
+      "step": 11774
+    },
+    {
+      "epoch": 0.6197684772929652,
+      "grad_norm": 1.5133031606674194,
+      "learning_rate": 1.932188309270537e-05,
+      "loss": 0.4401,
+      "step": 11832
+    },
+    {
+      "epoch": 0.6228065580640092,
+      "grad_norm": 2.30916428565979,
+      "learning_rate": 1.9309029048500578e-05,
+      "loss": 0.4177,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6258446388350531,
+      "grad_norm": 2.825598955154419,
+      "learning_rate": 1.929605867567532e-05,
+      "loss": 0.4529,
+      "step": 11948
+    },
+    {
+      "epoch": 0.6288827196060971,
+      "grad_norm": 5.285458087921143,
+      "learning_rate": 1.9282972136311554e-05,
+      "loss": 0.4806,
+      "step": 12006
+    },
+    {
+      "epoch": 0.631920800377141,
+      "grad_norm": 2.597923755645752,
+      "learning_rate": 1.9269769593942872e-05,
+      "loss": 0.4566,
+      "step": 12064
+    },
+    {
+      "epoch": 0.634958881148185,
+      "grad_norm": 2.780212640762329,
+      "learning_rate": 1.9256451213552497e-05,
+      "loss": 0.4725,
+      "step": 12122
+    },
+    {
+      "epoch": 0.6379969619192289,
+      "grad_norm": 4.1638031005859375,
+      "learning_rate": 1.9243017161571194e-05,
+      "loss": 0.463,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6410350426902729,
+      "grad_norm": 4.174670219421387,
+      "learning_rate": 1.9229467605875196e-05,
+      "loss": 0.5236,
+      "step": 12238
+    },
+    {
+      "epoch": 0.6440731234613168,
+      "grad_norm": 1.9128369092941284,
+      "learning_rate": 1.9215802715784096e-05,
+      "loss": 0.4621,
+      "step": 12296
+    },
+    {
+      "epoch": 0.6471112042323608,
+      "grad_norm": 4.490901947021484,
+      "learning_rate": 1.9202022662058773e-05,
+      "loss": 0.4517,
+      "step": 12354
+    },
+    {
+      "epoch": 0.6501492850034047,
+      "grad_norm": 4.426553726196289,
+      "learning_rate": 1.9188127616899202e-05,
+      "loss": 0.488,
+      "step": 12412
+    },
+    {
+      "epoch": 0.6531873657744487,
+      "grad_norm": 3.694254159927368,
+      "learning_rate": 1.917411775394233e-05,
+      "loss": 0.4705,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6562254465454926,
+      "grad_norm": 2.3134400844573975,
+      "learning_rate": 1.9159993248259916e-05,
+      "loss": 0.4402,
+      "step": 12528
+    },
+    {
+      "epoch": 0.6592635273165366,
+      "grad_norm": 2.868987798690796,
+      "learning_rate": 1.9145754276356323e-05,
+      "loss": 0.4085,
+      "step": 12586
+    },
+    {
+      "epoch": 0.6623016080875805,
+      "grad_norm": 3.815828323364258,
+      "learning_rate": 1.9131401016166326e-05,
+      "loss": 0.5569,
+      "step": 12644
+    },
+    {
+      "epoch": 0.6653396888586245,
+      "grad_norm": 2.865863800048828,
+      "learning_rate": 1.911693364705287e-05,
+      "loss": 0.4515,
+      "step": 12702
+    },
+    {
+      "epoch": 0.6683777696296684,
+      "grad_norm": 3.7862603664398193,
+      "learning_rate": 1.9102352349804865e-05,
+      "loss": 0.4685,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6714158504007124,
+      "grad_norm": 2.3399360179901123,
+      "learning_rate": 1.9087657306634884e-05,
+      "loss": 0.5087,
+      "step": 12818
+    },
+    {
+      "epoch": 0.6744539311717563,
+      "grad_norm": 3.208674430847168,
+      "learning_rate": 1.9072848701176905e-05,
+      "loss": 0.4322,
+      "step": 12876
+    },
+    {
+      "epoch": 0.6774920119428003,
+      "grad_norm": 1.2508207559585571,
+      "learning_rate": 1.9057926718484036e-05,
+      "loss": 0.39,
+      "step": 12934
+    },
+    {
+      "epoch": 0.6805300927138442,
+      "grad_norm": 3.029885768890381,
+      "learning_rate": 1.9042891545026164e-05,
+      "loss": 0.4881,
+      "step": 12992
+    },
+    {
+      "epoch": 0.6835681734848882,
+      "grad_norm": 1.723024606704712,
+      "learning_rate": 1.9028005500450692e-05,
+      "loss": 0.4016,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6866062542559321,
+      "grad_norm": 3.0587244033813477,
+      "learning_rate": 1.9012746453978195e-05,
+      "loss": 0.451,
+      "step": 13108
+    },
+    {
+      "epoch": 0.6896443350269761,
+      "grad_norm": 3.979196548461914,
+      "learning_rate": 1.899737478132781e-05,
+      "loss": 0.4584,
+      "step": 13166
+    },
+    {
+      "epoch": 0.69268241579802,
+      "grad_norm": 3.2428181171417236,
+      "learning_rate": 1.8981890674588902e-05,
+      "loss": 0.4419,
+      "step": 13224
+    },
+    {
+      "epoch": 0.695720496569064,
+      "grad_norm": 1.9672743082046509,
+      "learning_rate": 1.8966294327255843e-05,
+      "loss": 0.4463,
+      "step": 13282
+    },
+    {
+      "epoch": 0.6987585773401079,
+      "grad_norm": 3.543287754058838,
+      "learning_rate": 1.895058593422561e-05,
+      "loss": 0.5232,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7017966581111519,
+      "grad_norm": 2.751725435256958,
+      "learning_rate": 1.8934765691795337e-05,
+      "loss": 0.4627,
+      "step": 13398
+    },
+    {
+      "epoch": 0.7048347388821958,
+      "grad_norm": 3.9089314937591553,
+      "learning_rate": 1.8918833797659854e-05,
+      "loss": 0.4701,
+      "step": 13456
+    },
+    {
+      "epoch": 0.7078728196532398,
+      "grad_norm": 2.623382806777954,
+      "learning_rate": 1.890279045090924e-05,
+      "loss": 0.4627,
+      "step": 13514
+    },
+    {
+      "epoch": 0.7109109004242837,
+      "grad_norm": 3.44734263420105,
+      "learning_rate": 1.8886635852026307e-05,
+      "loss": 0.5063,
+      "step": 13572
+    },
+    {
+      "epoch": 0.7139489811953277,
+      "grad_norm": 4.096603870391846,
+      "learning_rate": 1.887037020288412e-05,
+      "loss": 0.4205,
+      "step": 13630
+    },
+    {
+      "epoch": 0.7169870619663716,
+      "grad_norm": 3.9694747924804688,
+      "learning_rate": 1.8853993706743465e-05,
+      "loss": 0.479,
+      "step": 13688
+    },
+    {
+      "epoch": 0.7200251427374156,
+      "grad_norm": 2.2461936473846436,
+      "learning_rate": 1.88375065682503e-05,
+      "loss": 0.4222,
+      "step": 13746
+    },
+    {
+      "epoch": 0.7230632235084595,
+      "grad_norm": 4.268979549407959,
+      "learning_rate": 1.882090899343321e-05,
+      "loss": 0.4013,
+      "step": 13804
+    },
+    {
+      "epoch": 0.7261013042795035,
+      "grad_norm": 2.9464776515960693,
+      "learning_rate": 1.8804201189700833e-05,
+      "loss": 0.5184,
+      "step": 13862
+    },
+    {
+      "epoch": 0.7291393850505474,
+      "grad_norm": 3.1404519081115723,
+      "learning_rate": 1.8787383365839248e-05,
+      "loss": 0.4451,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7321774658215914,
+      "grad_norm": 3.048670530319214,
+      "learning_rate": 1.8770455732009393e-05,
+      "loss": 0.457,
+      "step": 13978
+    },
+    {
+      "epoch": 0.7352155465926353,
+      "grad_norm": 3.074151039123535,
+      "learning_rate": 1.8753418499744426e-05,
+      "loss": 0.4711,
+      "step": 14036
+    },
+    {
+      "epoch": 0.7382536273636792,
+      "grad_norm": 4.1698150634765625,
+      "learning_rate": 1.873627188194708e-05,
+      "loss": 0.4281,
+      "step": 14094
+    },
+    {
+      "epoch": 0.7412917081347231,
+      "grad_norm": 2.6520071029663086,
+      "learning_rate": 1.8719016092887e-05,
+      "loss": 0.497,
+      "step": 14152
+    },
+    {
+      "epoch": 0.7443297889057671,
+      "grad_norm": 1.3818339109420776,
+      "learning_rate": 1.870165134819808e-05,
+      "loss": 0.4234,
+      "step": 14210
+    },
+    {
+      "epoch": 0.747367869676811,
+      "grad_norm": 3.5460736751556396,
+      "learning_rate": 1.868417786487575e-05,
+      "loss": 0.4444,
+      "step": 14268
+    },
+    {
+      "epoch": 0.750405950447855,
+      "grad_norm": 2.8102331161499023,
+      "learning_rate": 1.8666595861274283e-05,
+      "loss": 0.4159,
+      "step": 14326
+    },
+    {
+      "epoch": 0.7534440312188989,
+      "grad_norm": 3.3770508766174316,
+      "learning_rate": 1.8648905557104046e-05,
+      "loss": 0.4357,
+      "step": 14384
+    },
+    {
+      "epoch": 0.7564821119899429,
+      "grad_norm": 2.95613169670105,
+      "learning_rate": 1.863110717342876e-05,
+      "loss": 0.4627,
+      "step": 14442
+    },
+    {
+      "epoch": 0.7595201927609868,
+      "grad_norm": 2.80786395072937,
+      "learning_rate": 1.8613200932662764e-05,
+      "loss": 0.4331,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7625582735320308,
+      "grad_norm": 2.9433181285858154,
+      "learning_rate": 1.8595187058568197e-05,
+      "loss": 0.5087,
+      "step": 14558
+    },
+    {
+      "epoch": 0.7655963543030747,
+      "grad_norm": 2.6625008583068848,
+      "learning_rate": 1.8577065776252218e-05,
+      "loss": 0.5018,
+      "step": 14616
+    },
+    {
+      "epoch": 0.7686344350741187,
+      "grad_norm": 3.8713533878326416,
+      "learning_rate": 1.8558837312164198e-05,
+      "loss": 0.4454,
+      "step": 14674
+    },
+    {
+      "epoch": 0.7716725158451626,
+      "grad_norm": 3.236130475997925,
+      "learning_rate": 1.8540501894092894e-05,
+      "loss": 0.4463,
+      "step": 14732
+    },
+    {
+      "epoch": 0.7747105966162066,
+      "grad_norm": 1.9471006393432617,
+      "learning_rate": 1.8522059751163578e-05,
+      "loss": 0.4615,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7777486773872505,
+      "grad_norm": 1.1234129667282104,
+      "learning_rate": 1.85035111138352e-05,
+      "loss": 0.3841,
+      "step": 14848
+    },
+    {
+      "epoch": 0.7807867581582945,
+      "grad_norm": 3.155194044113159,
+      "learning_rate": 1.8484856213897496e-05,
+      "loss": 0.4932,
+      "step": 14906
+    },
+    {
+      "epoch": 0.7838248389293384,
+      "grad_norm": 1.2532273530960083,
+      "learning_rate": 1.8466095284468103e-05,
+      "loss": 0.427,
+      "step": 14964
+    },
+    {
+      "epoch": 0.7868629197003824,
+      "grad_norm": 3.315812349319458,
+      "learning_rate": 1.8447228559989618e-05,
+      "loss": 0.4406,
+      "step": 15022
+    },
+    {
+      "epoch": 0.7899010004714263,
+      "grad_norm": 2.3999452590942383,
+      "learning_rate": 1.842858427754608e-05,
+      "loss": 0.4413,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7929390812424703,
+      "grad_norm": 2.896650791168213,
+      "learning_rate": 1.8409508485466538e-05,
+      "loss": 0.4068,
+      "step": 15138
+    },
+    {
+      "epoch": 0.7959771620135142,
+      "grad_norm": 3.3152272701263428,
+      "learning_rate": 1.8390327605464747e-05,
+      "loss": 0.4708,
+      "step": 15196
+    },
+    {
+      "epoch": 0.7990152427845582,
+      "grad_norm": 2.573716163635254,
+      "learning_rate": 1.8371041877231145e-05,
+      "loss": 0.4506,
+      "step": 15254
+    },
+    {
+      "epoch": 0.8020533235556021,
+      "grad_norm": 1.0098395347595215,
+      "learning_rate": 1.8351651541766398e-05,
+      "loss": 0.4614,
+      "step": 15312
+    },
+    {
+      "epoch": 0.805091404326646,
+      "grad_norm": 2.7257494926452637,
+      "learning_rate": 1.8332156841378376e-05,
+      "loss": 0.481,
+      "step": 15370
+    },
+    {
+      "epoch": 0.80812948509769,
+      "grad_norm": 3.291948080062866,
+      "learning_rate": 1.8312558019679113e-05,
+      "loss": 0.4872,
+      "step": 15428
+    },
+    {
+      "epoch": 0.811167565868734,
+      "grad_norm": 0.6372181177139282,
+      "learning_rate": 1.82928553215818e-05,
+      "loss": 0.4664,
+      "step": 15486
+    },
+    {
+      "epoch": 0.8142056466397779,
+      "grad_norm": 2.14487361907959,
+      "learning_rate": 1.8273048993297682e-05,
+      "loss": 0.4443,
+      "step": 15544
+    },
+    {
+      "epoch": 0.8172437274108219,
+      "grad_norm": 1.6703099012374878,
+      "learning_rate": 1.8253139282333005e-05,
+      "loss": 0.4683,
+      "step": 15602
+    },
+    {
+      "epoch": 0.8202818081818658,
+      "grad_norm": 3.7610647678375244,
+      "learning_rate": 1.8233126437485925e-05,
+      "loss": 0.4299,
+      "step": 15660
+    },
+    {
+      "epoch": 0.8233198889529098,
+      "grad_norm": 3.429608106613159,
+      "learning_rate": 1.821301070884338e-05,
+      "loss": 0.3976,
+      "step": 15718
+    },
+    {
+      "epoch": 0.8263579697239537,
+      "grad_norm": 3.0198211669921875,
+      "learning_rate": 1.819279234777799e-05,
+      "loss": 0.407,
+      "step": 15776
+    },
+    {
+      "epoch": 0.8293960504949976,
+      "grad_norm": 3.1796703338623047,
+      "learning_rate": 1.817247160694489e-05,
+      "loss": 0.4235,
+      "step": 15834
+    },
+    {
+      "epoch": 0.8324341312660416,
+      "grad_norm": 2.235328197479248,
+      "learning_rate": 1.81520487402786e-05,
+      "loss": 0.403,
+      "step": 15892
+    },
+    {
+      "epoch": 0.8354722120370855,
+      "grad_norm": 3.7409324645996094,
+      "learning_rate": 1.8131524002989816e-05,
+      "loss": 0.4325,
+      "step": 15950
+    },
+    {
+      "epoch": 0.8385102928081295,
+      "grad_norm": 2.7824206352233887,
+      "learning_rate": 1.811089765156227e-05,
+      "loss": 0.432,
+      "step": 16008
+    },
+    {
+      "epoch": 0.8415483735791734,
+      "grad_norm": 1.6856441497802734,
+      "learning_rate": 1.8090528175270648e-05,
+      "loss": 0.4156,
+      "step": 16066
+    },
+    {
+      "epoch": 0.8445864543502174,
+      "grad_norm": 2.9528396129608154,
+      "learning_rate": 1.8069701110949214e-05,
+      "loss": 0.4486,
+      "step": 16124
+    },
+    {
+      "epoch": 0.8476245351212613,
+      "grad_norm": 1.5881803035736084,
+      "learning_rate": 1.8048773205047752e-05,
+      "loss": 0.4133,
+      "step": 16182
+    },
+    {
+      "epoch": 0.8506626158923053,
+      "grad_norm": 5.50068473815918,
+      "learning_rate": 1.8027744719088103e-05,
+      "loss": 0.4553,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8537006966633492,
+      "grad_norm": 3.2943098545074463,
+      "learning_rate": 1.800661591584899e-05,
+      "loss": 0.4428,
+      "step": 16298
+    },
+    {
+      "epoch": 0.8567387774343932,
+      "grad_norm": 1.3706376552581787,
+      "learning_rate": 1.798538705936273e-05,
+      "loss": 0.4779,
+      "step": 16356
+    },
+    {
+      "epoch": 0.8597768582054371,
+      "grad_norm": 2.18271541595459,
+      "learning_rate": 1.796405841491194e-05,
+      "loss": 0.4687,
+      "step": 16414
+    },
+    {
+      "epoch": 0.8628149389764811,
+      "grad_norm": 2.5106441974639893,
+      "learning_rate": 1.794263024902622e-05,
+      "loss": 0.4016,
+      "step": 16472
+    },
+    {
+      "epoch": 0.865853019747525,
+      "grad_norm": 2.757732629776001,
+      "learning_rate": 1.7921102829478832e-05,
+      "loss": 0.4948,
+      "step": 16530
+    },
+    {
+      "epoch": 0.868891100518569,
+      "grad_norm": 0.37621229887008667,
+      "learning_rate": 1.7899476425283318e-05,
+      "loss": 0.4304,
+      "step": 16588
+    },
+    {
+      "epoch": 0.8719291812896129,
+      "grad_norm": 4.135168552398682,
+      "learning_rate": 1.787775130669019e-05,
+      "loss": 0.4195,
+      "step": 16646
+    },
+    {
+      "epoch": 0.8749672620606569,
+      "grad_norm": 2.2052392959594727,
+      "learning_rate": 1.7855927745183504e-05,
+      "loss": 0.4449,
+      "step": 16704
+    },
+    {
+      "epoch": 0.8780053428317008,
+      "grad_norm": 2.8733346462249756,
+      "learning_rate": 1.7834006013477513e-05,
+      "loss": 0.5016,
+      "step": 16762
+    },
+    {
+      "epoch": 0.8810434236027448,
+      "grad_norm": 1.8927271366119385,
+      "learning_rate": 1.7811986385513226e-05,
+      "loss": 0.3793,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8840815043737887,
+      "grad_norm": 3.7612531185150146,
+      "learning_rate": 1.7789869136454988e-05,
+      "loss": 0.3601,
+      "step": 16878
+    },
+    {
+      "epoch": 0.8871195851448327,
+      "grad_norm": 1.6613848209381104,
+      "learning_rate": 1.7767654542687057e-05,
+      "loss": 0.4772,
+      "step": 16936
+    },
+    {
+      "epoch": 0.8901576659158766,
+      "grad_norm": 2.5755159854888916,
+      "learning_rate": 1.7745342881810144e-05,
+      "loss": 0.4475,
+      "step": 16994
+    },
+    {
+      "epoch": 0.8931957466869206,
+      "grad_norm": 2.7520928382873535,
+      "learning_rate": 1.7722934432637937e-05,
+      "loss": 0.3942,
+      "step": 17052
+    },
+    {
+      "epoch": 0.8962338274579645,
+      "grad_norm": 4.439705848693848,
+      "learning_rate": 1.770042947519362e-05,
+      "loss": 0.4361,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8992719082290085,
+      "grad_norm": 2.091926097869873,
+      "learning_rate": 1.7677828290706382e-05,
+      "loss": 0.42,
+      "step": 17168
+    },
+    {
+      "epoch": 0.9000052380702949,
+      "eval_accuracy": 0.8844256401062012,
+      "eval_loss": 0.44025254249572754,
+      "eval_runtime": 5730.5358,
+      "eval_samples_per_second": 0.835,
+      "eval_steps_per_second": 0.835,
+      "step": 17182
+    },
+    {
+      "epoch": 0.9023099890000523,
+      "grad_norm": 0.736885666847229,
+      "learning_rate": 1.7655131161607887e-05,
+      "loss": 0.4806,
+      "step": 17226
+    },
+    {
+      "epoch": 0.9053480697710963,
+      "grad_norm": 2.1999900341033936,
+      "learning_rate": 1.7632732159520203e-05,
+      "loss": 0.4541,
+      "step": 17284
+    },
+    {
+      "epoch": 0.9083861505421402,
+      "grad_norm": 2.4895384311676025,
+      "learning_rate": 1.76098456352832e-05,
+      "loss": 0.4643,
+      "step": 17342
+    },
+    {
+      "epoch": 0.9114242313131842,
+      "grad_norm": 2.773494005203247,
+      "learning_rate": 1.7586864015968063e-05,
+      "loss": 0.4031,
+      "step": 17400
+    },
+    {
+      "epoch": 0.9144623120842281,
+      "grad_norm": 3.805938720703125,
+      "learning_rate": 1.7563787588760503e-05,
+      "loss": 0.4756,
+      "step": 17458
+    },
+    {
+      "epoch": 0.9175003928552721,
+      "grad_norm": 2.412860631942749,
+      "learning_rate": 1.7540616642030974e-05,
+      "loss": 0.4453,
+      "step": 17516
+    },
+    {
+      "epoch": 0.920538473626316,
+      "grad_norm": 3.0736114978790283,
+      "learning_rate": 1.751735146533107e-05,
+      "loss": 0.4374,
+      "step": 17574
+    },
+    {
+      "epoch": 0.92357655439736,
+      "grad_norm": 3.2423524856567383,
+      "learning_rate": 1.7493992349389927e-05,
+      "loss": 0.3971,
+      "step": 17632
+    },
+    {
+      "epoch": 0.9266146351684039,
+      "grad_norm": 2.4448719024658203,
+      "learning_rate": 1.7470539586110572e-05,
+      "loss": 0.4407,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9296527159394479,
+      "grad_norm": 4.234783172607422,
+      "learning_rate": 1.7446993468566268e-05,
+      "loss": 0.4136,
+      "step": 17748
+    },
+    {
+      "epoch": 0.9326907967104918,
+      "grad_norm": 2.7946712970733643,
+      "learning_rate": 1.742335429099688e-05,
+      "loss": 0.4021,
+      "step": 17806
+    },
+    {
+      "epoch": 0.9357288774815358,
+      "grad_norm": 0.8968492746353149,
+      "learning_rate": 1.7399622348805165e-05,
+      "loss": 0.4591,
+      "step": 17864
+    },
+    {
+      "epoch": 0.9387669582525797,
+      "grad_norm": 2.746527671813965,
+      "learning_rate": 1.7375797938553108e-05,
+      "loss": 0.3938,
+      "step": 17922
+    },
+    {
+      "epoch": 0.9418050390236237,
+      "grad_norm": 1.0526750087738037,
+      "learning_rate": 1.73518813579582e-05,
+      "loss": 0.4577,
+      "step": 17980
+    },
+    {
+      "epoch": 0.9448431197946676,
+      "grad_norm": 2.459176778793335,
+      "learning_rate": 1.7327872905889727e-05,
+      "loss": 0.395,
+      "step": 18038
+    },
+    {
+      "epoch": 0.9478812005657116,
+      "grad_norm": 3.1182572841644287,
+      "learning_rate": 1.7303772882365018e-05,
+      "loss": 0.4536,
+      "step": 18096
+    },
+    {
+      "epoch": 0.9509192813367555,
+      "grad_norm": 2.8542633056640625,
+      "learning_rate": 1.7279581588545723e-05,
+      "loss": 0.4448,
+      "step": 18154
+    },
+    {
+      "epoch": 0.9539573621077995,
+      "grad_norm": 2.421351432800293,
+      "learning_rate": 1.7255299326734026e-05,
+      "loss": 0.4568,
+      "step": 18212
+    },
+    {
+      "epoch": 0.9569954428788434,
+      "grad_norm": 1.9418818950653076,
+      "learning_rate": 1.7230926400368878e-05,
+      "loss": 0.4509,
+      "step": 18270
+    },
+    {
+      "epoch": 0.9600335236498874,
+      "grad_norm": 2.297189950942993,
+      "learning_rate": 1.720688565864609e-05,
+      "loss": 0.4185,
+      "step": 18328
+    },
+    {
+      "epoch": 0.9630716044209313,
+      "grad_norm": 3.155817747116089,
+      "learning_rate": 1.7182333868082773e-05,
+      "loss": 0.4724,
+      "step": 18386
+    },
+    {
+      "epoch": 0.9661096851919753,
+      "grad_norm": 2.3270270824432373,
+      "learning_rate": 1.715769232476584e-05,
+      "loss": 0.4434,
+      "step": 18444
+    },
+    {
+      "epoch": 0.9691477659630192,
+      "grad_norm": 2.4003474712371826,
+      "learning_rate": 1.7132961336623944e-05,
+      "loss": 0.4325,
+      "step": 18502
+    },
+    {
+      "epoch": 0.9721858467340632,
+      "grad_norm": 4.449118614196777,
+      "learning_rate": 1.710814121270346e-05,
+      "loss": 0.4497,
+      "step": 18560
+    },
+    {
+      "epoch": 0.9752239275051071,
+      "grad_norm": 4.001181125640869,
+      "learning_rate": 1.7083232263164643e-05,
+      "loss": 0.4133,
+      "step": 18618
+    },
+    {
+      "epoch": 0.9782620082761511,
+      "grad_norm": 1.3908356428146362,
+      "learning_rate": 1.7058234799277733e-05,
+      "loss": 0.436,
+      "step": 18676
+    },
+    {
+      "epoch": 0.981300089047195,
+      "grad_norm": 2.0721793174743652,
+      "learning_rate": 1.703314913341908e-05,
+      "loss": 0.327,
+      "step": 18734
+    },
+    {
+      "epoch": 0.984338169818239,
+      "grad_norm": 1.8085274696350098,
+      "learning_rate": 1.700797557906723e-05,
+      "loss": 0.4782,
+      "step": 18792
+    },
+    {
+      "epoch": 0.9873762505892829,
+      "grad_norm": 2.877991199493408,
+      "learning_rate": 1.6982714450799006e-05,
+      "loss": 0.4804,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9904143313603269,
+      "grad_norm": 2.9189906120300293,
+      "learning_rate": 1.6957366064285604e-05,
+      "loss": 0.4473,
+      "step": 18908
+    },
+    {
+      "epoch": 0.9934524121313708,
+      "grad_norm": 2.6251885890960693,
+      "learning_rate": 1.6931930736288605e-05,
+      "loss": 0.4665,
+      "step": 18966
+    },
+    {
+      "epoch": 0.9964904929024148,
+      "grad_norm": 2.457298517227173,
+      "learning_rate": 1.6906408784656045e-05,
+      "loss": 0.4931,
+      "step": 19024
+    },
+    {
+      "epoch": 0.9995285736734587,
+      "grad_norm": 4.4407057762146,
+      "learning_rate": 1.6880800528318443e-05,
+      "loss": 0.4835,
+      "step": 19082
+    },
+    {
+      "epoch": 1.0025666544445027,
+      "grad_norm": 0.4853041172027588,
+      "learning_rate": 1.68551062872848e-05,
+      "loss": 0.4143,
+      "step": 19140
+    },
+    {
+      "epoch": 1.0056047352155466,
+      "grad_norm": 3.372084379196167,
+      "learning_rate": 1.682932638263862e-05,
+      "loss": 0.4104,
+      "step": 19198
+    },
+    {
+      "epoch": 1.0086428159865906,
+      "grad_norm": 3.5660102367401123,
+      "learning_rate": 1.6803461136533877e-05,
+      "loss": 0.3998,
+      "step": 19256
+    },
+    {
+      "epoch": 1.0116808967576345,
+      "grad_norm": 3.154710531234741,
+      "learning_rate": 1.6777510872191012e-05,
+      "loss": 0.4479,
+      "step": 19314
+    },
+    {
+      "epoch": 1.0147189775286785,
+      "grad_norm": 2.6014859676361084,
+      "learning_rate": 1.675147591389286e-05,
+      "loss": 0.4407,
+      "step": 19372
+    },
+    {
+      "epoch": 1.0177570582997224,
+      "grad_norm": 2.906419515609741,
+      "learning_rate": 1.672535658698064e-05,
+      "loss": 0.3687,
+      "step": 19430
+    },
+    {
+      "epoch": 1.0207951390707664,
+      "grad_norm": 4.82797908782959,
+      "learning_rate": 1.669915321784986e-05,
+      "loss": 0.4467,
+      "step": 19488
+    },
+    {
+      "epoch": 1.0238332198418103,
+      "grad_norm": 2.7285468578338623,
+      "learning_rate": 1.6673320066929267e-05,
+      "loss": 0.4124,
+      "step": 19546
+    },
+    {
+      "epoch": 1.0268713006128543,
+      "grad_norm": 3.0213897228240967,
+      "learning_rate": 1.664695103165033e-05,
+      "loss": 0.3955,
+      "step": 19604
+    },
+    {
+      "epoch": 1.0299093813838982,
+      "grad_norm": 1.75594961643219,
+      "learning_rate": 1.662049893393386e-05,
+      "loss": 0.3951,
+      "step": 19662
+    },
+    {
+      "epoch": 1.0329474621549422,
+      "grad_norm": 4.119887828826904,
+      "learning_rate": 1.659396410433378e-05,
+      "loss": 0.405,
+      "step": 19720
+    },
+    {
+      "epoch": 1.035985542925986,
+      "grad_norm": 2.607999801635742,
+      "learning_rate": 1.6567346874437857e-05,
+      "loss": 0.4512,
+      "step": 19778
+    },
+    {
+      "epoch": 1.03902362369703,
+      "grad_norm": 2.761012077331543,
+      "learning_rate": 1.6540647576863546e-05,
+      "loss": 0.3692,
+      "step": 19836
+    },
+    {
+      "epoch": 1.042061704468074,
+      "grad_norm": 1.8808788061141968,
+      "learning_rate": 1.6513866545253866e-05,
+      "loss": 0.4663,
+      "step": 19894
+    },
+    {
+      "epoch": 1.045099785239118,
+      "grad_norm": 3.0192790031433105,
+      "learning_rate": 1.648700411427319e-05,
+      "loss": 0.4485,
+      "step": 19952
+    },
+    {
+      "epoch": 1.048137866010162,
+      "grad_norm": 3.0104329586029053,
+      "learning_rate": 1.64600606196031e-05,
+      "loss": 0.4003,
+      "step": 20010
+    },
+    {
+      "epoch": 1.0511759467812058,
+      "grad_norm": 2.610039472579956,
+      "learning_rate": 1.6433036397938168e-05,
+      "loss": 0.3967,
+      "step": 20068
+    },
+    {
+      "epoch": 1.0542140275522498,
+      "grad_norm": 2.601706027984619,
+      "learning_rate": 1.6405931786981753e-05,
+      "loss": 0.3656,
+      "step": 20126
+    },
+    {
+      "epoch": 1.0572521083232937,
+      "grad_norm": 1.4956645965576172,
+      "learning_rate": 1.63787471254418e-05,
+      "loss": 0.3829,
+      "step": 20184
+    },
+    {
+      "epoch": 1.0602901890943377,
+      "grad_norm": 2.5245773792266846,
+      "learning_rate": 1.635148275302657e-05,
+      "loss": 0.4309,
+      "step": 20242
+    },
+    {
+      "epoch": 1.0633282698653816,
+      "grad_norm": 2.844923734664917,
+      "learning_rate": 1.6324139010440435e-05,
+      "loss": 0.4478,
+      "step": 20300
+    },
+    {
+      "epoch": 1.0663663506364256,
+      "grad_norm": 2.980348587036133,
+      "learning_rate": 1.629671623937959e-05,
+      "loss": 0.3524,
+      "step": 20358
+    },
+    {
+      "epoch": 1.0694044314074695,
+      "grad_norm": 2.8273379802703857,
+      "learning_rate": 1.626921478252781e-05,
+      "loss": 0.4499,
+      "step": 20416
+    },
+    {
+      "epoch": 1.0724425121785135,
+      "grad_norm": 3.1932373046875,
+      "learning_rate": 1.624163498355213e-05,
+      "loss": 0.3989,
+      "step": 20474
+    },
+    {
+      "epoch": 1.0754805929495574,
+      "grad_norm": 3.9423575401306152,
+      "learning_rate": 1.62139771870986e-05,
+      "loss": 0.4338,
+      "step": 20532
+    },
+    {
+      "epoch": 1.0785186737206014,
+      "grad_norm": 3.1419034004211426,
+      "learning_rate": 1.618624173878793e-05,
+      "loss": 0.4207,
+      "step": 20590
+    },
+    {
+      "epoch": 1.0815567544916453,
+      "grad_norm": 0.7265225052833557,
+      "learning_rate": 1.61584289852112e-05,
+      "loss": 0.3271,
+      "step": 20648
+    },
+    {
+      "epoch": 1.0845948352626893,
+      "grad_norm": 1.7413750886917114,
+      "learning_rate": 1.613053927392553e-05,
+      "loss": 0.3501,
+      "step": 20706
+    },
+    {
+      "epoch": 1.0876329160337332,
+      "grad_norm": 4.131454944610596,
+      "learning_rate": 1.6102572953449715e-05,
+      "loss": 0.5085,
+      "step": 20764
+    },
+    {
+      "epoch": 1.0906709968047772,
+      "grad_norm": 3.0570194721221924,
+      "learning_rate": 1.6074530373259887e-05,
+      "loss": 0.4154,
+      "step": 20822
+    },
+    {
+      "epoch": 1.0937090775758211,
+      "grad_norm": 4.058136463165283,
+      "learning_rate": 1.6046897326469475e-05,
+      "loss": 0.4007,
+      "step": 20880
+    },
+    {
+      "epoch": 1.096747158346865,
+      "grad_norm": 2.809772491455078,
+      "learning_rate": 1.601870457882787e-05,
+      "loss": 0.4188,
+      "step": 20938
+    },
+    {
+      "epoch": 1.099785239117909,
+      "grad_norm": 2.9016387462615967,
+      "learning_rate": 1.5990436619518428e-05,
+      "loss": 0.3567,
+      "step": 20996
+    },
+    {
+      "epoch": 1.102823319888953,
+      "grad_norm": 3.3889174461364746,
+      "learning_rate": 1.5962093801786668e-05,
+      "loss": 0.3956,
+      "step": 21054
+    },
+    {
+      "epoch": 1.105861400659997,
+      "grad_norm": 1.6921156644821167,
+      "learning_rate": 1.5933676479813547e-05,
+      "loss": 0.352,
+      "step": 21112
+    },
+    {
+      "epoch": 1.108899481431041,
+      "grad_norm": 3.156926393508911,
+      "learning_rate": 1.5905185008711063e-05,
+      "loss": 0.4026,
+      "step": 21170
+    },
+    {
+      "epoch": 1.1119375622020848,
+      "grad_norm": 1.8040690422058105,
+      "learning_rate": 1.58766197445178e-05,
+      "loss": 0.4248,
+      "step": 21228
+    },
+    {
+      "epoch": 1.1149756429731288,
+      "grad_norm": 3.1559841632843018,
+      "learning_rate": 1.5848475435042218e-05,
+      "loss": 0.3691,
+      "step": 21286
+    },
+    {
+      "epoch": 1.1180137237441727,
+      "grad_norm": 2.4738497734069824,
+      "learning_rate": 1.5820260539561704e-05,
+      "loss": 0.367,
+      "step": 21344
+    },
+    {
+      "epoch": 1.1210518045152167,
+      "grad_norm": 3.917677402496338,
+      "learning_rate": 1.579147854311163e-05,
+      "loss": 0.426,
+      "step": 21402
+    },
+    {
+      "epoch": 1.1240898852862606,
+      "grad_norm": 2.7536489963531494,
+      "learning_rate": 1.576262417448334e-05,
+      "loss": 0.4269,
+      "step": 21460
+    },
+    {
+      "epoch": 1.1271279660573046,
+      "grad_norm": 3.678410768508911,
+      "learning_rate": 1.5733697794250292e-05,
+      "loss": 0.3977,
+      "step": 21518
+    },
+    {
+      "epoch": 1.1301660468283483,
+      "grad_norm": 3.2154643535614014,
+      "learning_rate": 1.5704699763885845e-05,
+      "loss": 0.3647,
+      "step": 21576
+    },
+    {
+      "epoch": 1.1332041275993925,
+      "grad_norm": 2.71803617477417,
+      "learning_rate": 1.5675630445758707e-05,
+      "loss": 0.3774,
+      "step": 21634
+    },
+    {
+      "epoch": 1.1362422083704362,
+      "grad_norm": 2.9909651279449463,
+      "learning_rate": 1.5646490203128424e-05,
+      "loss": 0.3707,
+      "step": 21692
+    },
+    {
+      "epoch": 1.1392802891414804,
+      "grad_norm": 3.9027791023254395,
+      "learning_rate": 1.561727940014084e-05,
+      "loss": 0.3925,
+      "step": 21750
+    },
+    {
+      "epoch": 1.1423183699125241,
+      "grad_norm": 2.427164316177368,
+      "learning_rate": 1.558799840182354e-05,
+      "loss": 0.3666,
+      "step": 21808
+    },
+    {
+      "epoch": 1.1453564506835683,
+      "grad_norm": 4.000829219818115,
+      "learning_rate": 1.55586475740813e-05,
+      "loss": 0.4233,
+      "step": 21866
+    },
+    {
+      "epoch": 1.148394531454612,
+      "grad_norm": 2.324054718017578,
+      "learning_rate": 1.5529227283691498e-05,
+      "loss": 0.4315,
+      "step": 21924
+    },
+    {
+      "epoch": 1.1514326122256562,
+      "grad_norm": 2.127596378326416,
+      "learning_rate": 1.549973789829954e-05,
+      "loss": 0.3987,
+      "step": 21982
+    },
+    {
+      "epoch": 1.1544706929967,
+      "grad_norm": 2.658895254135132,
+      "learning_rate": 1.5470179786414278e-05,
+      "loss": 0.3766,
+      "step": 22040
+    },
+    {
+      "epoch": 1.157508773767744,
+      "grad_norm": 3.48286509513855,
+      "learning_rate": 1.5440553317403375e-05,
+      "loss": 0.414,
+      "step": 22098
+    },
+    {
+      "epoch": 1.1605468545387878,
+      "grad_norm": 2.305206298828125,
+      "learning_rate": 1.5410858861488717e-05,
+      "loss": 0.3961,
+      "step": 22156
+    },
+    {
+      "epoch": 1.163584935309832,
+      "grad_norm": 2.7347195148468018,
+      "learning_rate": 1.5381096789741777e-05,
+      "loss": 0.3038,
+      "step": 22214
+    },
+    {
+      "epoch": 1.1666230160808757,
+      "grad_norm": 3.5154871940612793,
+      "learning_rate": 1.5351267474078967e-05,
+      "loss": 0.3442,
+      "step": 22272
+    },
+    {
+      "epoch": 1.1696610968519197,
+      "grad_norm": 3.8937666416168213,
+      "learning_rate": 1.532137128725701e-05,
+      "loss": 0.396,
+      "step": 22330
+    },
+    {
+      "epoch": 1.1726991776229636,
+      "grad_norm": 3.284895896911621,
+      "learning_rate": 1.529140860286828e-05,
+      "loss": 0.3605,
+      "step": 22388
+    },
+    {
+      "epoch": 1.1757372583940076,
+      "grad_norm": 2.9780056476593018,
+      "learning_rate": 1.5261379795336102e-05,
+      "loss": 0.3908,
+      "step": 22446
+    },
+    {
+      "epoch": 1.1787753391650515,
+      "grad_norm": 3.4775991439819336,
+      "learning_rate": 1.5231285239910119e-05,
+      "loss": 0.4244,
+      "step": 22504
+    },
+    {
+      "epoch": 1.1818134199360955,
+      "grad_norm": 3.3921220302581787,
+      "learning_rate": 1.520112531266157e-05,
+      "loss": 0.4234,
+      "step": 22562
+    },
+    {
+      "epoch": 1.1848515007071394,
+      "grad_norm": 4.370563507080078,
+      "learning_rate": 1.5170900390478605e-05,
+      "loss": 0.3605,
+      "step": 22620
+    },
+    {
+      "epoch": 1.1878895814781834,
+      "grad_norm": 2.411207675933838,
+      "learning_rate": 1.5140610851061573e-05,
+      "loss": 0.4109,
+      "step": 22678
+    },
+    {
+      "epoch": 1.1909276622492273,
+      "grad_norm": 1.8209956884384155,
+      "learning_rate": 1.5110257072918297e-05,
+      "loss": 0.3973,
+      "step": 22736
+    },
+    {
+      "epoch": 1.1939657430202713,
+      "grad_norm": 1.9300421476364136,
+      "learning_rate": 1.5079839435359347e-05,
+      "loss": 0.385,
+      "step": 22794
+    },
+    {
+      "epoch": 1.1970038237913152,
+      "grad_norm": 2.628262519836426,
+      "learning_rate": 1.504935831849331e-05,
+      "loss": 0.3439,
+      "step": 22852
+    },
+    {
+      "epoch": 1.2000419045623592,
+      "grad_norm": 1.3332035541534424,
+      "learning_rate": 1.5018814103222013e-05,
+      "loss": 0.3466,
+      "step": 22910
+    },
+    {
+      "epoch": 1.203079985333403,
+      "grad_norm": 0.541492223739624,
+      "learning_rate": 1.4988207171235807e-05,
+      "loss": 0.4193,
+      "step": 22968
+    },
+    {
+      "epoch": 1.206118066104447,
+      "grad_norm": 3.0223543643951416,
+      "learning_rate": 1.4957537905008744e-05,
+      "loss": 0.3866,
+      "step": 23026
+    },
+    {
+      "epoch": 1.209156146875491,
+      "grad_norm": 2.3658430576324463,
+      "learning_rate": 1.492680668779384e-05,
+      "loss": 0.4248,
+      "step": 23084
+    },
+    {
+      "epoch": 1.212194227646535,
+      "grad_norm": 2.964507579803467,
+      "learning_rate": 1.4896013903618272e-05,
+      "loss": 0.3904,
+      "step": 23142
+    },
+    {
+      "epoch": 1.215232308417579,
+      "grad_norm": 2.325568675994873,
+      "learning_rate": 1.4865159937278566e-05,
+      "loss": 0.3901,
+      "step": 23200
+    },
+    {
+      "epoch": 1.2182703891886228,
+      "grad_norm": 2.5834431648254395,
+      "learning_rate": 1.4834245174335812e-05,
+      "loss": 0.386,
+      "step": 23258
+    },
+    {
+      "epoch": 1.2213084699596668,
+      "grad_norm": 3.9499247074127197,
+      "learning_rate": 1.480327000111083e-05,
+      "loss": 0.4012,
+      "step": 23316
+    },
+    {
+      "epoch": 1.2243465507307107,
+      "grad_norm": 3.858048439025879,
+      "learning_rate": 1.477223480467934e-05,
+      "loss": 0.4277,
+      "step": 23374
+    },
+    {
+      "epoch": 1.2273846315017547,
+      "grad_norm": 2.3730807304382324,
+      "learning_rate": 1.4741139972867137e-05,
+      "loss": 0.3963,
+      "step": 23432
+    },
+    {
+      "epoch": 1.2304227122727986,
+      "grad_norm": 3.197197914123535,
+      "learning_rate": 1.4709985894245246e-05,
+      "loss": 0.4269,
+      "step": 23490
+    },
+    {
+      "epoch": 1.2334607930438426,
+      "grad_norm": 1.0639739036560059,
+      "learning_rate": 1.4678772958125043e-05,
+      "loss": 0.486,
+      "step": 23548
+    },
+    {
+      "epoch": 1.2364988738148865,
+      "grad_norm": 4.4826202392578125,
+      "learning_rate": 1.4647501554553417e-05,
+      "loss": 0.3983,
+      "step": 23606
+    },
+    {
+      "epoch": 1.2395369545859305,
+      "grad_norm": 3.2974660396575928,
+      "learning_rate": 1.4616172074307886e-05,
+      "loss": 0.3893,
+      "step": 23664
+    },
+    {
+      "epoch": 1.2425750353569744,
+      "grad_norm": 4.817580223083496,
+      "learning_rate": 1.4584784908891705e-05,
+      "loss": 0.4044,
+      "step": 23722
+    },
+    {
+      "epoch": 1.2456131161280184,
+      "grad_norm": 4.333406925201416,
+      "learning_rate": 1.455334045052899e-05,
+      "loss": 0.3787,
+      "step": 23780
+    },
+    {
+      "epoch": 1.2486511968990623,
+      "grad_norm": 4.401366233825684,
+      "learning_rate": 1.4521839092159802e-05,
+      "loss": 0.3653,
+      "step": 23838
+    },
+    {
+      "epoch": 1.2516892776701063,
+      "grad_norm": 1.6555136442184448,
+      "learning_rate": 1.4490281227435248e-05,
+      "loss": 0.3371,
+      "step": 23896
+    },
+    {
+      "epoch": 1.2547273584411502,
+      "grad_norm": 3.3965320587158203,
+      "learning_rate": 1.4459212792449709e-05,
+      "loss": 0.3923,
+      "step": 23954
+    },
+    {
+      "epoch": 1.2577654392121942,
+      "grad_norm": 1.8068640232086182,
+      "learning_rate": 1.4427544056073314e-05,
+      "loss": 0.4196,
+      "step": 24012
+    },
+    {
+      "epoch": 1.2608035199832381,
+      "grad_norm": 2.8050827980041504,
+      "learning_rate": 1.4395819991682645e-05,
+      "loss": 0.448,
+      "step": 24070
+    },
+    {
+      "epoch": 1.263841600754282,
+      "grad_norm": 3.339679002761841,
+      "learning_rate": 1.4364040995711812e-05,
+      "loss": 0.4015,
+      "step": 24128
+    },
+    {
+      "epoch": 1.266879681525326,
+      "grad_norm": 1.8822154998779297,
+      "learning_rate": 1.4332207465281365e-05,
+      "loss": 0.3494,
+      "step": 24186
+    },
+    {
+      "epoch": 1.26991776229637,
+      "grad_norm": 2.0527522563934326,
+      "learning_rate": 1.4300319798193339e-05,
+      "loss": 0.4453,
+      "step": 24244
+    },
+    {
+      "epoch": 1.272955843067414,
+      "grad_norm": 2.963618516921997,
+      "learning_rate": 1.4268378392926277e-05,
+      "loss": 0.451,
+      "step": 24302
+    },
+    {
+      "epoch": 1.2759939238384579,
+      "grad_norm": 2.991400957107544,
+      "learning_rate": 1.4236383648630245e-05,
+      "loss": 0.3719,
+      "step": 24360
+    },
+    {
+      "epoch": 1.2790320046095018,
+      "grad_norm": 1.891254186630249,
+      "learning_rate": 1.4204335965121862e-05,
+      "loss": 0.334,
+      "step": 24418
+    },
+    {
+      "epoch": 1.2820700853805458,
+      "grad_norm": 3.0473315715789795,
+      "learning_rate": 1.4172235742879283e-05,
+      "loss": 0.4188,
+      "step": 24476
+    },
+    {
+      "epoch": 1.2851081661515897,
+      "grad_norm": 1.8583441972732544,
+      "learning_rate": 1.414008338303721e-05,
+      "loss": 0.4133,
+      "step": 24534
+    },
+    {
+      "epoch": 1.2881462469226337,
+      "grad_norm": 2.5623059272766113,
+      "learning_rate": 1.4107879287381872e-05,
+      "loss": 0.4317,
+      "step": 24592
+    },
+    {
+      "epoch": 1.2911843276936776,
+      "grad_norm": 1.2033942937850952,
+      "learning_rate": 1.4075623858346e-05,
+      "loss": 0.3494,
+      "step": 24650
+    },
+    {
+      "epoch": 1.2942224084647216,
+      "grad_norm": 1.8575024604797363,
+      "learning_rate": 1.404331749900381e-05,
+      "loss": 0.4533,
+      "step": 24708
+    },
+    {
+      "epoch": 1.2972604892357655,
+      "grad_norm": 2.617656946182251,
+      "learning_rate": 1.4010960613065956e-05,
+      "loss": 0.2822,
+      "step": 24766
+    },
+    {
+      "epoch": 1.3002985700068095,
+      "grad_norm": 1.466433048248291,
+      "learning_rate": 1.397855360487449e-05,
+      "loss": 0.3965,
+      "step": 24824
+    },
+    {
+      "epoch": 1.3033366507778534,
+      "grad_norm": 2.6398956775665283,
+      "learning_rate": 1.3946096879397808e-05,
+      "loss": 0.4247,
+      "step": 24882
+    },
+    {
+      "epoch": 1.3063747315488974,
+      "grad_norm": 3.450256109237671,
+      "learning_rate": 1.3913590842225589e-05,
+      "loss": 0.3854,
+      "step": 24940
+    },
+    {
+      "epoch": 1.3094128123199413,
+      "grad_norm": 3.451451063156128,
+      "learning_rate": 1.388103589956372e-05,
+      "loss": 0.3852,
+      "step": 24998
+    },
+    {
+      "epoch": 1.3124508930909853,
+      "grad_norm": 3.278472900390625,
+      "learning_rate": 1.3848432458229241e-05,
+      "loss": 0.3999,
+      "step": 25056
+    },
+    {
+      "epoch": 1.3154889738620292,
+      "grad_norm": 0.6535636782646179,
+      "learning_rate": 1.381578092564524e-05,
+      "loss": 0.4151,
+      "step": 25114
+    },
+    {
+      "epoch": 1.3185270546330732,
+      "grad_norm": 3.161670207977295,
+      "learning_rate": 1.378308170983576e-05,
+      "loss": 0.3413,
+      "step": 25172
+    },
+    {
+      "epoch": 1.3215651354041171,
+      "grad_norm": 3.4061434268951416,
+      "learning_rate": 1.375033521942072e-05,
+      "loss": 0.3902,
+      "step": 25230
+    },
+    {
+      "epoch": 1.324603216175161,
+      "grad_norm": 3.568188428878784,
+      "learning_rate": 1.3717541863610799e-05,
+      "loss": 0.3391,
+      "step": 25288
+    },
+    {
+      "epoch": 1.327641296946205,
+      "grad_norm": 1.4888482093811035,
+      "learning_rate": 1.368470205220231e-05,
+      "loss": 0.4018,
+      "step": 25346
+    },
+    {
+      "epoch": 1.330679377717249,
+      "grad_norm": 3.4848194122314453,
+      "learning_rate": 1.36518161955721e-05,
+      "loss": 0.4045,
+      "step": 25404
+    },
+    {
+      "epoch": 1.333717458488293,
+      "grad_norm": 2.341862916946411,
+      "learning_rate": 1.3618884704672413e-05,
+      "loss": 0.3615,
+      "step": 25462
+    },
+    {
+      "epoch": 1.3367555392593369,
+      "grad_norm": 4.151137351989746,
+      "learning_rate": 1.3585907991025737e-05,
+      "loss": 0.4122,
+      "step": 25520
+    },
+    {
+      "epoch": 1.3397936200303808,
+      "grad_norm": 3.4976377487182617,
+      "learning_rate": 1.3552886466719696e-05,
+      "loss": 0.3674,
+      "step": 25578
+    },
+    {
+      "epoch": 1.3428317008014248,
+      "grad_norm": 3.3959484100341797,
+      "learning_rate": 1.3519820544401882e-05,
+      "loss": 0.3742,
+      "step": 25636
+    },
+    {
+      "epoch": 1.3458697815724687,
+      "grad_norm": 1.3133649826049805,
+      "learning_rate": 1.3486710637274687e-05,
+      "loss": 0.3388,
+      "step": 25694
+    },
+    {
+      "epoch": 1.3489078623435127,
+      "grad_norm": 3.167113780975342,
+      "learning_rate": 1.3453557159090159e-05,
+      "loss": 0.3886,
+      "step": 25752
+    },
+    {
+      "epoch": 1.3519459431145566,
+      "grad_norm": 2.9960269927978516,
+      "learning_rate": 1.342036052414482e-05,
+      "loss": 0.3896,
+      "step": 25810
+    },
+    {
+      "epoch": 1.3549840238856006,
+      "grad_norm": 3.482513427734375,
+      "learning_rate": 1.3387121147274498e-05,
+      "loss": 0.4203,
+      "step": 25868
+    },
+    {
+      "epoch": 1.3580221046566445,
+      "grad_norm": 2.4168648719787598,
+      "learning_rate": 1.3353839443849134e-05,
+      "loss": 0.343,
+      "step": 25926
+    },
+    {
+      "epoch": 1.3610601854276885,
+      "grad_norm": 0.6455244421958923,
+      "learning_rate": 1.33205158297676e-05,
+      "loss": 0.3872,
+      "step": 25984
+    },
+    {
+      "epoch": 1.3640982661987324,
+      "grad_norm": 4.288425445556641,
+      "learning_rate": 1.3287150721452488e-05,
+      "loss": 0.3655,
+      "step": 26042
+    },
+    {
+      "epoch": 1.3671363469697764,
+      "grad_norm": 2.454954147338867,
+      "learning_rate": 1.3254320850234712e-05,
+      "loss": 0.4089,
+      "step": 26100
+    },
+    {
+      "epoch": 1.3701744277408203,
+      "grad_norm": 3.7248048782348633,
+      "learning_rate": 1.322087470228127e-05,
+      "loss": 0.3784,
+      "step": 26158
+    },
+    {
+      "epoch": 1.3732125085118643,
+      "grad_norm": 1.3180829286575317,
+      "learning_rate": 1.3187388305241823e-05,
+      "loss": 0.3933,
+      "step": 26216
+    },
+    {
+      "epoch": 1.3762505892829082,
+      "grad_norm": 1.625256061553955,
+      "learning_rate": 1.3153862077573157e-05,
+      "loss": 0.4252,
+      "step": 26274
+    },
+    {
+      "epoch": 1.3792886700539522,
+      "grad_norm": 2.6077065467834473,
+      "learning_rate": 1.312029643822979e-05,
+      "loss": 0.385,
+      "step": 26332
+    },
+    {
+      "epoch": 1.3823267508249961,
+      "grad_norm": 3.2308919429779053,
+      "learning_rate": 1.3086691806658749e-05,
+      "loss": 0.4137,
+      "step": 26390
+    },
+    {
+      "epoch": 1.38536483159604,
+      "grad_norm": 3.302467107772827,
+      "learning_rate": 1.3053048602794315e-05,
+      "loss": 0.3842,
+      "step": 26448
+    },
+    {
+      "epoch": 1.388402912367084,
+      "grad_norm": 2.4501612186431885,
+      "learning_rate": 1.3019367247052781e-05,
+      "loss": 0.3889,
+      "step": 26506
+    },
+    {
+      "epoch": 1.391440993138128,
+      "grad_norm": 3.628817319869995,
+      "learning_rate": 1.29856481603272e-05,
+      "loss": 0.3531,
+      "step": 26564
+    },
+    {
+      "epoch": 1.394479073909172,
+      "grad_norm": 2.43445086479187,
+      "learning_rate": 1.2951891763982125e-05,
+      "loss": 0.402,
+      "step": 26622
+    },
+    {
+      "epoch": 1.3975171546802159,
+      "grad_norm": 0.9383435845375061,
+      "learning_rate": 1.2918098479848336e-05,
+      "loss": 0.445,
+      "step": 26680
+    },
+    {
+      "epoch": 1.4005552354512598,
+      "grad_norm": 3.4569339752197266,
+      "learning_rate": 1.2884268730217577e-05,
+      "loss": 0.3929,
+      "step": 26738
+    },
+    {
+      "epoch": 1.4035933162223038,
+      "grad_norm": 1.1176038980484009,
+      "learning_rate": 1.2850402937837283e-05,
+      "loss": 0.4028,
+      "step": 26796
+    },
+    {
+      "epoch": 1.4066313969933477,
+      "grad_norm": 3.2979259490966797,
+      "learning_rate": 1.2816501525905282e-05,
+      "loss": 0.4184,
+      "step": 26854
+    },
+    {
+      "epoch": 1.4096694777643917,
+      "grad_norm": 3.3147165775299072,
+      "learning_rate": 1.2782564918064522e-05,
+      "loss": 0.4289,
+      "step": 26912
+    },
+    {
+      "epoch": 1.4127075585354356,
+      "grad_norm": 3.0880119800567627,
+      "learning_rate": 1.2748593538397764e-05,
+      "loss": 0.4247,
+      "step": 26970
+    },
+    {
+      "epoch": 1.4157456393064796,
+      "grad_norm": 2.7575788497924805,
+      "learning_rate": 1.27145878114223e-05,
+      "loss": 0.3868,
+      "step": 27028
+    },
+    {
+      "epoch": 1.4187837200775235,
+      "grad_norm": 2.9284110069274902,
+      "learning_rate": 1.2680548162084614e-05,
+      "loss": 0.4337,
+      "step": 27086
+    },
+    {
+      "epoch": 1.4218218008485675,
+      "grad_norm": 2.6792969703674316,
+      "learning_rate": 1.2646475015755124e-05,
+      "loss": 0.4215,
+      "step": 27144
+    },
+    {
+      "epoch": 1.4248598816196114,
+      "grad_norm": 2.336975574493408,
+      "learning_rate": 1.261236879822282e-05,
+      "loss": 0.407,
+      "step": 27202
+    },
+    {
+      "epoch": 1.4278979623906554,
+      "grad_norm": 2.287140369415283,
+      "learning_rate": 1.2578818810950262e-05,
+      "loss": 0.3581,
+      "step": 27260
+    },
+    {
+      "epoch": 1.4309360431616993,
+      "grad_norm": 3.0004403591156006,
+      "learning_rate": 1.2544648281900015e-05,
+      "loss": 0.412,
+      "step": 27318
+    },
+    {
+      "epoch": 1.4339741239327433,
+      "grad_norm": 0.6953569650650024,
+      "learning_rate": 1.2510445954106563e-05,
+      "loss": 0.4086,
+      "step": 27376
+    },
+    {
+      "epoch": 1.437012204703787,
+      "grad_norm": 1.8512517213821411,
+      "learning_rate": 1.2476212254973198e-05,
+      "loss": 0.367,
+      "step": 27434
+    },
+    {
+      "epoch": 1.4400502854748312,
+      "grad_norm": 3.226573944091797,
+      "learning_rate": 1.2441947612295222e-05,
+      "loss": 0.4576,
+      "step": 27492
+    },
+    {
+      "epoch": 1.4430883662458749,
+      "grad_norm": 2.18180251121521,
+      "learning_rate": 1.2407652454254632e-05,
+      "loss": 0.406,
+      "step": 27550
+    },
+    {
+      "epoch": 1.446126447016919,
+      "grad_norm": 2.2790727615356445,
+      "learning_rate": 1.2373327209414759e-05,
+      "loss": 0.3834,
+      "step": 27608
+    },
+    {
+      "epoch": 1.4491645277879628,
+      "grad_norm": 3.7362709045410156,
+      "learning_rate": 1.2338972306714889e-05,
+      "loss": 0.3668,
+      "step": 27666
+    },
+    {
+      "epoch": 1.452202608559007,
+      "grad_norm": 3.6803033351898193,
+      "learning_rate": 1.2304588175464941e-05,
+      "loss": 0.3552,
+      "step": 27724
+    },
+    {
+      "epoch": 1.4552406893300507,
+      "grad_norm": 4.011587142944336,
+      "learning_rate": 1.2270175245340074e-05,
+      "loss": 0.3629,
+      "step": 27782
+    },
+    {
+      "epoch": 1.4582787701010949,
+      "grad_norm": 2.53393292427063,
+      "learning_rate": 1.223573394637533e-05,
+      "loss": 0.4304,
+      "step": 27840
+    },
+    {
+      "epoch": 1.4613168508721386,
+      "grad_norm": 1.7455615997314453,
+      "learning_rate": 1.2201264708960252e-05,
+      "loss": 0.3834,
+      "step": 27898
+    },
+    {
+      "epoch": 1.4643549316431828,
+      "grad_norm": 3.8155839443206787,
+      "learning_rate": 1.2166767963833519e-05,
+      "loss": 0.3604,
+      "step": 27956
+    },
+    {
+      "epoch": 1.4673930124142265,
+      "grad_norm": 2.2630105018615723,
+      "learning_rate": 1.213224414207755e-05,
+      "loss": 0.3833,
+      "step": 28014
+    },
+    {
+      "epoch": 1.4704310931852707,
+      "grad_norm": 2.6760213375091553,
+      "learning_rate": 1.209769367511312e-05,
+      "loss": 0.3574,
+      "step": 28072
+    },
+    {
+      "epoch": 1.4734691739563144,
+      "grad_norm": 3.230642318725586,
+      "learning_rate": 1.206311699469398e-05,
+      "loss": 0.4048,
+      "step": 28130
+    },
+    {
+      "epoch": 1.4765072547273586,
+      "grad_norm": 4.475136756896973,
+      "learning_rate": 1.2028514532901445e-05,
+      "loss": 0.3787,
+      "step": 28188
+    },
+    {
+      "epoch": 1.4795453354984023,
+      "grad_norm": 3.1134748458862305,
+      "learning_rate": 1.1993886722139004e-05,
+      "loss": 0.4555,
+      "step": 28246
+    },
+    {
+      "epoch": 1.4825834162694465,
+      "grad_norm": 1.694558024406433,
+      "learning_rate": 1.19598316646051e-05,
+      "loss": 0.3722,
+      "step": 28304
+    },
+    {
+      "epoch": 1.4856214970404902,
+      "grad_norm": 2.7863929271698,
+      "learning_rate": 1.1925154872829044e-05,
+      "loss": 0.3436,
+      "step": 28362
+    },
+    {
+      "epoch": 1.4886595778115344,
+      "grad_norm": 1.5842992067337036,
+      "learning_rate": 1.189045402369863e-05,
+      "loss": 0.4055,
+      "step": 28420
+    },
+    {
+      "epoch": 1.491697658582578,
+      "grad_norm": 3.305001974105835,
+      "learning_rate": 1.185572955084683e-05,
+      "loss": 0.4177,
+      "step": 28478
+    },
+    {
+      "epoch": 1.4947357393536223,
+      "grad_norm": 1.0446484088897705,
+      "learning_rate": 1.1820981888201819e-05,
+      "loss": 0.3333,
+      "step": 28536
+    },
+    {
+      "epoch": 1.497773820124666,
+      "grad_norm": 2.8945727348327637,
+      "learning_rate": 1.178621146998157e-05,
+      "loss": 0.4184,
+      "step": 28594
+    },
+    {
+      "epoch": 1.5008119008957101,
+      "grad_norm": 2.099456548690796,
+      "learning_rate": 1.1751418730688405e-05,
+      "loss": 0.3384,
+      "step": 28652
+    },
+    {
+      "epoch": 1.5038499816667539,
+      "grad_norm": 2.770231246948242,
+      "learning_rate": 1.1716604105103582e-05,
+      "loss": 0.4045,
+      "step": 28710
+    },
+    {
+      "epoch": 1.506888062437798,
+      "grad_norm": 2.9615721702575684,
+      "learning_rate": 1.1681768028281859e-05,
+      "loss": 0.358,
+      "step": 28768
+    },
+    {
+      "epoch": 1.5099261432088418,
+      "grad_norm": 2.6768476963043213,
+      "learning_rate": 1.1646910935546055e-05,
+      "loss": 0.3421,
+      "step": 28826
+    },
+    {
+      "epoch": 1.512964223979886,
+      "grad_norm": 2.7448065280914307,
+      "learning_rate": 1.1612033262481607e-05,
+      "loss": 0.3749,
+      "step": 28884
+    },
+    {
+      "epoch": 1.5160023047509297,
+      "grad_norm": 2.4983344078063965,
+      "learning_rate": 1.1577135444931136e-05,
+      "loss": 0.2883,
+      "step": 28942
+    },
+    {
+      "epoch": 1.5190403855219738,
+      "grad_norm": 3.426910400390625,
+      "learning_rate": 1.1542217918988993e-05,
+      "loss": 0.4035,
+      "step": 29000
+    },
+    {
+      "epoch": 1.5220784662930176,
+      "grad_norm": 2.1255111694335938,
+      "learning_rate": 1.1507281120995808e-05,
+      "loss": 0.3872,
+      "step": 29058
+    },
+    {
+      "epoch": 1.5251165470640617,
+      "grad_norm": 2.398613691329956,
+      "learning_rate": 1.147232548753304e-05,
+      "loss": 0.3485,
+      "step": 29116
+    },
+    {
+      "epoch": 1.5281546278351055,
+      "grad_norm": 1.043369174003601,
+      "learning_rate": 1.1437351455417533e-05,
+      "loss": 0.3921,
+      "step": 29174
+    },
+    {
+      "epoch": 1.5311927086061496,
+      "grad_norm": 0.4801720976829529,
+      "learning_rate": 1.1402359461696034e-05,
+      "loss": 0.3838,
+      "step": 29232
+    },
+    {
+      "epoch": 1.5342307893771934,
+      "grad_norm": 1.7236027717590332,
+      "learning_rate": 1.1367349943639748e-05,
+      "loss": 0.4114,
+      "step": 29290
+    },
+    {
+      "epoch": 1.5372688701482375,
+      "grad_norm": 2.06709885597229,
+      "learning_rate": 1.1332323338738873e-05,
+      "loss": 0.3896,
+      "step": 29348
+    },
+    {
+      "epoch": 1.5403069509192813,
+      "grad_norm": 2.651384115219116,
+      "learning_rate": 1.1297280084697126e-05,
+      "loss": 0.3873,
+      "step": 29406
+    },
+    {
+      "epoch": 1.5433450316903254,
+      "grad_norm": 1.6569814682006836,
+      "learning_rate": 1.1262825227855019e-05,
+      "loss": 0.3591,
+      "step": 29464
+    },
+    {
+      "epoch": 1.5463831124613692,
+      "grad_norm": 1.6780743598937988,
+      "learning_rate": 1.1227750257706836e-05,
+      "loss": 0.314,
+      "step": 29522
+    },
+    {
+      "epoch": 1.5494211932324133,
+      "grad_norm": 2.9294652938842773,
+      "learning_rate": 1.1192659945196629e-05,
+      "loss": 0.3737,
+      "step": 29580
+    },
+    {
+      "epoch": 1.552459274003457,
+      "grad_norm": 2.5051159858703613,
+      "learning_rate": 1.115755472882423e-05,
+      "loss": 0.3653,
+      "step": 29638
+    },
+    {
+      "epoch": 1.5554973547745012,
+      "grad_norm": 0.6437486410140991,
+      "learning_rate": 1.1122435047275705e-05,
+      "loss": 0.3389,
+      "step": 29696
+    },
+    {
+      "epoch": 1.558535435545545,
+      "grad_norm": 2.644324541091919,
+      "learning_rate": 1.1087301339417893e-05,
+      "loss": 0.3918,
+      "step": 29754
+    },
+    {
+      "epoch": 1.561573516316589,
+      "grad_norm": 2.6001169681549072,
+      "learning_rate": 1.1052154044292904e-05,
+      "loss": 0.3319,
+      "step": 29812
+    },
+    {
+      "epoch": 1.5646115970876329,
+      "grad_norm": 1.9335983991622925,
+      "learning_rate": 1.101699360111264e-05,
+      "loss": 0.4825,
+      "step": 29870
+    },
+    {
+      "epoch": 1.5676496778586768,
+      "grad_norm": 2.186579704284668,
+      "learning_rate": 1.0981820449253304e-05,
+      "loss": 0.3636,
+      "step": 29928
+    },
+    {
+      "epoch": 1.5706877586297208,
+      "grad_norm": 1.987005591392517,
+      "learning_rate": 1.0946635028249916e-05,
+      "loss": 0.4214,
+      "step": 29986
+    },
+    {
+      "epoch": 1.5737258394007647,
+      "grad_norm": 2.18890380859375,
+      "learning_rate": 1.0911437777790807e-05,
+      "loss": 0.3795,
+      "step": 30044
+    },
+    {
+      "epoch": 1.5767639201718087,
+      "grad_norm": 2.2321784496307373,
+      "learning_rate": 1.0876229137712135e-05,
+      "loss": 0.3954,
+      "step": 30102
+    },
+    {
+      "epoch": 1.5798020009428526,
+      "grad_norm": 3.391721725463867,
+      "learning_rate": 1.0841009547992398e-05,
+      "loss": 0.4151,
+      "step": 30160
+    },
+    {
+      "epoch": 1.5828400817138966,
+      "grad_norm": 4.004073619842529,
+      "learning_rate": 1.0805779448746907e-05,
+      "loss": 0.3318,
+      "step": 30218
+    },
+    {
+      "epoch": 1.5858781624849405,
+      "grad_norm": 2.895787477493286,
+      "learning_rate": 1.0770539280222312e-05,
+      "loss": 0.3688,
+      "step": 30276
+    },
+    {
+      "epoch": 1.5889162432559845,
+      "grad_norm": 2.4140095710754395,
+      "learning_rate": 1.073528948279109e-05,
+      "loss": 0.4059,
+      "step": 30334
+    },
+    {
+      "epoch": 1.5919543240270284,
+      "grad_norm": 3.6965787410736084,
+      "learning_rate": 1.070003049694605e-05,
+      "loss": 0.359,
+      "step": 30392
+    },
+    {
+      "epoch": 1.5949924047980724,
+      "grad_norm": 2.7172248363494873,
+      "learning_rate": 1.0664762763294812e-05,
+      "loss": 0.4282,
+      "step": 30450
+    },
+    {
+      "epoch": 1.5980304855691163,
+      "grad_norm": 3.0504825115203857,
+      "learning_rate": 1.0629486722554316e-05,
+      "loss": 0.3838,
+      "step": 30508
+    },
+    {
+      "epoch": 1.6010685663401603,
+      "grad_norm": 1.9200332164764404,
+      "learning_rate": 1.0594202815545319e-05,
+      "loss": 0.3814,
+      "step": 30566
+    },
+    {
+      "epoch": 1.6041066471112042,
+      "grad_norm": 3.9770891666412354,
+      "learning_rate": 1.0558911483186856e-05,
+      "loss": 0.3785,
+      "step": 30624
+    },
+    {
+      "epoch": 1.6071447278822482,
+      "grad_norm": 2.1418545246124268,
+      "learning_rate": 1.0523613166490776e-05,
+      "loss": 0.3282,
+      "step": 30682
+    },
+    {
+      "epoch": 1.610182808653292,
+      "grad_norm": 2.4927940368652344,
+      "learning_rate": 1.0488308306556192e-05,
+      "loss": 0.3979,
+      "step": 30740
+    },
+    {
+      "epoch": 1.613220889424336,
+      "grad_norm": 1.7132062911987305,
+      "learning_rate": 1.0452997344563982e-05,
+      "loss": 0.371,
+      "step": 30798
+    },
+    {
+      "epoch": 1.61625897019538,
+      "grad_norm": 2.9751667976379395,
+      "learning_rate": 1.0417680721771288e-05,
+      "loss": 0.3782,
+      "step": 30856
+    },
+    {
+      "epoch": 1.619297050966424,
+      "grad_norm": 3.013737678527832,
+      "learning_rate": 1.0382358879505982e-05,
+      "loss": 0.3899,
+      "step": 30914
+    },
+    {
+      "epoch": 1.622335131737468,
+      "grad_norm": 3.8316636085510254,
+      "learning_rate": 1.0347032259161162e-05,
+      "loss": 0.3942,
+      "step": 30972
+    },
+    {
+      "epoch": 1.6253732125085119,
+      "grad_norm": 2.597729444503784,
+      "learning_rate": 1.031170130218964e-05,
+      "loss": 0.3941,
+      "step": 31030
+    },
+    {
+      "epoch": 1.6284112932795558,
+      "grad_norm": 3.033374547958374,
+      "learning_rate": 1.0276975702213507e-05,
+      "loss": 0.3367,
+      "step": 31088
+    },
+    {
+      "epoch": 1.6314493740505998,
+      "grad_norm": 3.6567564010620117,
+      "learning_rate": 1.0241637452361323e-05,
+      "loss": 0.3943,
+      "step": 31146
+    },
+    {
+      "epoch": 1.6344874548216437,
+      "grad_norm": 2.558410167694092,
+      "learning_rate": 1.0206296182929831e-05,
+      "loss": 0.4043,
+      "step": 31204
+    },
+    {
+      "epoch": 1.6375255355926877,
+      "grad_norm": 1.9596946239471436,
+      "learning_rate": 1.01709523355549e-05,
+      "loss": 0.3589,
+      "step": 31262
+    },
+    {
+      "epoch": 1.6405636163637316,
+      "grad_norm": 3.5439186096191406,
+      "learning_rate": 1.013560635190461e-05,
+      "loss": 0.3769,
+      "step": 31320
+    },
+    {
+      "epoch": 1.6436016971347756,
+      "grad_norm": 2.5455784797668457,
+      "learning_rate": 1.010025867367374e-05,
+      "loss": 0.3099,
+      "step": 31378
+    },
+    {
+      "epoch": 1.6466397779058195,
+      "grad_norm": 2.672985792160034,
+      "learning_rate": 1.0064909742578242e-05,
+      "loss": 0.3384,
+      "step": 31436
+    },
+    {
+      "epoch": 1.6496778586768635,
+      "grad_norm": 3.4179306030273438,
+      "learning_rate": 1.002956000034973e-05,
+      "loss": 0.418,
+      "step": 31494
+    },
+    {
+      "epoch": 1.6527159394479074,
+      "grad_norm": 3.8588337898254395,
+      "learning_rate": 9.99420988872995e-06,
+      "loss": 0.4403,
+      "step": 31552
+    },
+    {
+      "epoch": 1.6557540202189514,
+      "grad_norm": 2.2216336727142334,
+      "learning_rate": 9.958859849465258e-06,
+      "loss": 0.3876,
+      "step": 31610
+    },
+    {
+      "epoch": 1.6587921009899953,
+      "grad_norm": 5.450603485107422,
+      "learning_rate": 9.923510324301119e-06,
+      "loss": 0.3786,
+      "step": 31668
+    },
+    {
+      "epoch": 1.6618301817610392,
+      "grad_norm": 3.075505256652832,
+      "learning_rate": 9.888161754976566e-06,
+      "loss": 0.3599,
+      "step": 31726
+    },
+    {
+      "epoch": 1.6648682625320832,
+      "grad_norm": 0.3762458562850952,
+      "learning_rate": 9.852814583218681e-06,
+      "loss": 0.3758,
+      "step": 31784
+    },
+    {
+      "epoch": 1.6679063433031271,
+      "grad_norm": 4.629846572875977,
+      "learning_rate": 9.817469250737098e-06,
+      "loss": 0.3707,
+      "step": 31842
+    },
+    {
+      "epoch": 1.670944424074171,
+      "grad_norm": 2.84089732170105,
+      "learning_rate": 9.782126199218453e-06,
+      "loss": 0.385,
+      "step": 31900
+    },
+    {
+      "epoch": 1.673982504845215,
+      "grad_norm": 3.581580400466919,
+      "learning_rate": 9.74678587032088e-06,
+      "loss": 0.346,
+      "step": 31958
+    },
+    {
+      "epoch": 1.677020585616259,
+      "grad_norm": 2.798215866088867,
+      "learning_rate": 9.711448705668503e-06,
+      "loss": 0.4142,
+      "step": 32016
+    },
+    {
+      "epoch": 1.680058666387303,
+      "grad_norm": 2.810606002807617,
+      "learning_rate": 9.676115146845887e-06,
+      "loss": 0.3711,
+      "step": 32074
+    },
+    {
+      "epoch": 1.683096747158347,
+      "grad_norm": 0.6601367592811584,
+      "learning_rate": 9.640785635392543e-06,
+      "loss": 0.4049,
+      "step": 32132
+    },
+    {
+      "epoch": 1.6861348279293908,
+      "grad_norm": 2.879519462585449,
+      "learning_rate": 9.60546061279741e-06,
+      "loss": 0.385,
+      "step": 32190
+    },
+    {
+      "epoch": 1.6891729087004348,
+      "grad_norm": 4.493093013763428,
+      "learning_rate": 9.570140520493336e-06,
+      "loss": 0.385,
+      "step": 32248
+    },
+    {
+      "epoch": 1.6922109894714787,
+      "grad_norm": 1.8380403518676758,
+      "learning_rate": 9.53482579985154e-06,
+      "loss": 0.3804,
+      "step": 32306
+    },
+    {
+      "epoch": 1.6952490702425227,
+      "grad_norm": 0.3858829736709595,
+      "learning_rate": 9.499516892176139e-06,
+      "loss": 0.3626,
+      "step": 32364
+    },
+    {
+      "epoch": 1.6982871510135666,
+      "grad_norm": 3.132283926010132,
+      "learning_rate": 9.464214238698589e-06,
+      "loss": 0.3205,
+      "step": 32422
+    },
+    {
+      "epoch": 1.7013252317846106,
+      "grad_norm": 2.6827683448791504,
+      "learning_rate": 9.428918280572203e-06,
+      "loss": 0.3605,
+      "step": 32480
+    },
+    {
+      "epoch": 1.7043633125556545,
+      "grad_norm": 3.2188720703125,
+      "learning_rate": 9.394237823897566e-06,
+      "loss": 0.4137,
+      "step": 32538
+    },
+    {
+      "epoch": 1.7074013933266985,
+      "grad_norm": 2.3144068717956543,
+      "learning_rate": 9.35895644521292e-06,
+      "loss": 0.4068,
+      "step": 32596
+    },
+    {
+      "epoch": 1.7104394740977424,
+      "grad_norm": 2.736628293991089,
+      "learning_rate": 9.323683077214672e-06,
+      "loss": 0.3743,
+      "step": 32654
+    },
+    {
+      "epoch": 1.7134775548687864,
+      "grad_norm": 3.6730895042419434,
+      "learning_rate": 9.28841816069017e-06,
+      "loss": 0.3719,
+      "step": 32712
+    },
+    {
+      "epoch": 1.7165156356398303,
+      "grad_norm": 4.749345779418945,
+      "learning_rate": 9.253162136321158e-06,
+      "loss": 0.3911,
+      "step": 32770
+    },
+    {
+      "epoch": 1.7195537164108743,
+      "grad_norm": 1.4852901697158813,
+      "learning_rate": 9.217915444678246e-06,
+      "loss": 0.4266,
+      "step": 32828
+    },
+    {
+      "epoch": 1.7225917971819182,
+      "grad_norm": 2.9961495399475098,
+      "learning_rate": 9.182678526215428e-06,
+      "loss": 0.3772,
+      "step": 32886
+    },
+    {
+      "epoch": 1.725629877952962,
+      "grad_norm": 2.647796154022217,
+      "learning_rate": 9.147451821264571e-06,
+      "loss": 0.3617,
+      "step": 32944
+    },
+    {
+      "epoch": 1.7286679587240061,
+      "grad_norm": 1.1617308855056763,
+      "learning_rate": 9.112235770029908e-06,
+      "loss": 0.2908,
+      "step": 33002
+    },
+    {
+      "epoch": 1.7317060394950499,
+      "grad_norm": 2.2398462295532227,
+      "learning_rate": 9.077030812582535e-06,
+      "loss": 0.3906,
+      "step": 33060
+    },
+    {
+      "epoch": 1.734744120266094,
+      "grad_norm": 3.6835782527923584,
+      "learning_rate": 9.041837388854928e-06,
+      "loss": 0.4513,
+      "step": 33118
+    },
+    {
+      "epoch": 1.7377822010371378,
+      "grad_norm": 2.344299077987671,
+      "learning_rate": 9.006655938635422e-06,
+      "loss": 0.4084,
+      "step": 33176
+    },
+    {
+      "epoch": 1.740820281808182,
+      "grad_norm": 3.1176178455352783,
+      "learning_rate": 8.971486901562728e-06,
+      "loss": 0.3877,
+      "step": 33234
+    },
+    {
+      "epoch": 1.7438583625792257,
+      "grad_norm": 1.6773039102554321,
+      "learning_rate": 8.936330717120455e-06,
+      "loss": 0.3969,
+      "step": 33292
+    },
+    {
+      "epoch": 1.7468964433502698,
+      "grad_norm": 3.7852532863616943,
+      "learning_rate": 8.901187824631575e-06,
+      "loss": 0.3343,
+      "step": 33350
+    },
+    {
+      "epoch": 1.7499345241213136,
+      "grad_norm": 2.9691007137298584,
+      "learning_rate": 8.866058663252984e-06,
+      "loss": 0.3555,
+      "step": 33408
+    },
+    {
+      "epoch": 1.7529726048923577,
+      "grad_norm": 2.2175984382629395,
+      "learning_rate": 8.830943671969973e-06,
+      "loss": 0.3689,
+      "step": 33466
+    },
+    {
+      "epoch": 1.7560106856634015,
+      "grad_norm": 4.200769901275635,
+      "learning_rate": 8.795843289590765e-06,
+      "loss": 0.3527,
+      "step": 33524
+    },
+    {
+      "epoch": 1.7590487664344456,
+      "grad_norm": 0.7551002502441406,
+      "learning_rate": 8.760757954741032e-06,
+      "loss": 0.3522,
+      "step": 33582
+    },
+    {
+      "epoch": 1.7620868472054894,
+      "grad_norm": 2.0418272018432617,
+      "learning_rate": 8.725688105858394e-06,
+      "loss": 0.4124,
+      "step": 33640
+    },
+    {
+      "epoch": 1.7651249279765335,
+      "grad_norm": 2.172964096069336,
+      "learning_rate": 8.690634181186958e-06,
+      "loss": 0.4303,
+      "step": 33698
+    },
+    {
+      "epoch": 1.7681630087475773,
+      "grad_norm": 1.0999021530151367,
+      "learning_rate": 8.655596618771844e-06,
+      "loss": 0.3028,
+      "step": 33756
+    },
+    {
+      "epoch": 1.7712010895186214,
+      "grad_norm": 5.170861721038818,
+      "learning_rate": 8.620575856453699e-06,
+      "loss": 0.3741,
+      "step": 33814
+    },
+    {
+      "epoch": 1.7742391702896652,
+      "grad_norm": 2.6932246685028076,
+      "learning_rate": 8.585572331863224e-06,
+      "loss": 0.4294,
+      "step": 33872
+    },
+    {
+      "epoch": 1.7772772510607093,
+      "grad_norm": 0.71751469373703,
+      "learning_rate": 8.551189534523404e-06,
+      "loss": 0.338,
+      "step": 33930
+    },
+    {
+      "epoch": 1.780315331831753,
+      "grad_norm": 2.928600788116455,
+      "learning_rate": 8.516221481428949e-06,
+      "loss": 0.4693,
+      "step": 33988
+    },
+    {
+      "epoch": 1.7833534126027972,
+      "grad_norm": 3.5148239135742188,
+      "learning_rate": 8.481271970107997e-06,
+      "loss": 0.4074,
+      "step": 34046
+    },
+    {
+      "epoch": 1.786391493373841,
+      "grad_norm": 2.133580207824707,
+      "learning_rate": 8.446341437300874e-06,
+      "loss": 0.4129,
+      "step": 34104
+    },
+    {
+      "epoch": 1.7894295741448851,
+      "grad_norm": 2.5686025619506836,
+      "learning_rate": 8.411430319510761e-06,
+      "loss": 0.3526,
+      "step": 34162
+    },
+    {
+      "epoch": 1.7924676549159289,
+      "grad_norm": 4.3522562980651855,
+      "learning_rate": 8.376539052998205e-06,
+      "loss": 0.3587,
+      "step": 34220
+    },
+    {
+      "epoch": 1.795505735686973,
+      "grad_norm": 3.373617649078369,
+      "learning_rate": 8.34166807377569e-06,
+      "loss": 0.4069,
+      "step": 34278
+    },
+    {
+      "epoch": 1.7985438164580168,
+      "grad_norm": 2.4383530616760254,
+      "learning_rate": 8.306817817602193e-06,
+      "loss": 0.4002,
+      "step": 34336
+    },
+    {
+      "epoch": 1.80001047614059,
+      "eval_accuracy": 0.8927125930786133,
+      "eval_loss": 0.4064957797527313,
+      "eval_runtime": 5473.958,
+      "eval_samples_per_second": 0.874,
+      "eval_steps_per_second": 0.874,
+      "step": 34364
+    }
+  ],
+  "logging_steps": 58,
+  "max_steps": 57273,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 17182,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8389182403659694e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}