{
  "best_global_step": 900,
  "best_metric": 0.8228043143297381,
  "best_model_checkpoint": "/content/drive/MyDrive/expA_realonly/checkpoint-900",
  "epoch": 15.0,
  "eval_steps": 50,
  "global_step": 1290,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011627906976744186,
      "grad_norm": 4.9740142822265625,
      "learning_rate": 0.0,
      "loss": 0.317932665348053,
      "step": 1
    },
    {
      "epoch": 0.29069767441860467,
      "grad_norm": 1.5566426515579224,
      "learning_rate": 6.282722513089005e-06,
      "loss": 0.3199571371078491,
      "step": 25
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 1.6379677057266235,
      "learning_rate": 1.2827225130890053e-05,
      "loss": 0.24125898361206055,
      "step": 50
    },
    {
      "epoch": 0.5813953488372093,
      "eval_accuracy": 0.8287126584998925,
      "eval_f1": 0.08611410118406888,
      "eval_loss": 0.10772959887981415,
      "eval_precision": 0.15810276679841898,
      "eval_recall": 0.05917159763313609,
      "eval_runtime": 4.2273,
      "eval_samples_per_second": 80.667,
      "eval_steps_per_second": 1.419,
      "step": 50
    },
    {
      "epoch": 0.872093023255814,
      "grad_norm": 8.703161239624023,
      "learning_rate": 1.93717277486911e-05,
      "loss": 0.20946041107177735,
      "step": 75
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.9212121963500977,
      "learning_rate": 2.591623036649215e-05,
      "loss": 0.19275867462158203,
      "step": 100
    },
    {
      "epoch": 1.1627906976744187,
      "eval_accuracy": 0.7635933806146572,
      "eval_f1": 0.3981945837512538,
      "eval_loss": 0.07719173282384872,
      "eval_precision": 0.30121396054628224,
      "eval_recall": 0.5872781065088757,
      "eval_runtime": 1.6016,
      "eval_samples_per_second": 212.909,
      "eval_steps_per_second": 3.746,
      "step": 100
    },
    {
      "epoch": 1.4534883720930232,
      "grad_norm": 1.46477210521698,
      "learning_rate": 3.246073298429319e-05,
      "loss": 0.14603865623474122,
      "step": 125
    },
    {
      "epoch": 1.744186046511628,
      "grad_norm": 3.729416608810425,
      "learning_rate": 3.900523560209424e-05,
      "loss": 0.13855588912963868,
      "step": 150
    },
    {
      "epoch": 1.744186046511628,
      "eval_accuracy": 0.81130453470879,
      "eval_f1": 0.46785521339816316,
      "eval_loss": 0.06223325803875923,
      "eval_precision": 0.36851063829787234,
      "eval_recall": 0.6405325443786982,
      "eval_runtime": 1.5906,
      "eval_samples_per_second": 214.385,
      "eval_steps_per_second": 3.772,
      "step": 150
    },
    {
      "epoch": 2.0348837209302326,
      "grad_norm": 1.2140589952468872,
      "learning_rate": 4.554973821989529e-05,
      "loss": 0.09624839782714843,
      "step": 175
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.875888466835022,
      "learning_rate": 4.999346304613061e-05,
      "loss": 0.0821054744720459,
      "step": 200
    },
    {
      "epoch": 2.3255813953488373,
      "eval_accuracy": 0.7872340425531915,
      "eval_f1": 0.4847619047619048,
      "eval_loss": 0.04723217338323593,
      "eval_precision": 0.3574438202247191,
      "eval_recall": 0.7529585798816568,
      "eval_runtime": 1.6255,
      "eval_samples_per_second": 209.777,
      "eval_steps_per_second": 3.691,
      "step": 200
    },
    {
      "epoch": 2.616279069767442,
      "grad_norm": 1.0979472398757935,
      "learning_rate": 4.98888472605987e-05,
      "loss": 0.08414460182189941,
      "step": 225
    },
    {
      "epoch": 2.9069767441860463,
      "grad_norm": 0.49558770656585693,
      "learning_rate": 4.965717280180432e-05,
      "loss": 0.08369583129882813,
      "step": 250
    },
    {
      "epoch": 2.9069767441860463,
      "eval_accuracy": 0.6211046636578551,
      "eval_f1": 0.3483915126625599,
      "eval_loss": 0.045315902680158615,
      "eval_precision": 0.22662511130899377,
      "eval_recall": 0.7529585798816568,
      "eval_runtime": 1.6274,
      "eval_samples_per_second": 209.539,
      "eval_steps_per_second": 3.687,
      "step": 250
    },
    {
      "epoch": 3.197674418604651,
      "grad_norm": 0.8114362955093384,
      "learning_rate": 4.9299622378173246e-05,
      "loss": 0.05607881069183349,
      "step": 275
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.6278374195098877,
      "learning_rate": 4.881802130053548e-05,
      "loss": 0.043469696044921874,
      "step": 300
    },
    {
      "epoch": 3.488372093023256,
      "eval_accuracy": 0.8082957231893402,
      "eval_f1": 0.5285784074255008,
      "eval_loss": 0.04308881610631943,
      "eval_precision": 0.39460247994164843,
      "eval_recall": 0.8002958579881657,
      "eval_runtime": 1.6193,
      "eval_samples_per_second": 210.585,
      "eval_steps_per_second": 3.705,
      "step": 300
    },
    {
      "epoch": 3.7790697674418605,
      "grad_norm": 1.0219486951828003,
      "learning_rate": 4.821482816383218e-05,
      "loss": 0.047760300636291504,
      "step": 325
    },
    {
      "epoch": 4.069767441860465,
      "grad_norm": 0.509274959564209,
      "learning_rate": 4.749312229587908e-05,
      "loss": 0.03177810907363891,
      "step": 350
    },
    {
      "epoch": 4.069767441860465,
      "eval_accuracy": 0.9196217494089834,
      "eval_f1": 0.7152317880794702,
      "eval_loss": 0.04378104954957962,
      "eval_precision": 0.6474820143884892,
      "eval_recall": 0.7988165680473372,
      "eval_runtime": 1.6177,
      "eval_samples_per_second": 210.792,
      "eval_steps_per_second": 3.709,
      "step": 350
    },
    {
      "epoch": 4.3604651162790695,
      "grad_norm": 0.47098585963249207,
      "learning_rate": 4.6656588037260834e-05,
      "loss": 0.020146708488464355,
      "step": 375
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 0.7534321546554565,
      "learning_rate": 4.570949593260837e-05,
      "loss": 0.018921481370925905,
      "step": 400
    },
    {
      "epoch": 4.651162790697675,
      "eval_accuracy": 0.9097356544165055,
      "eval_f1": 0.6987341772151898,
      "eval_loss": 0.05498046055436134,
      "eval_precision": 0.6106194690265486,
      "eval_recall": 0.8165680473372781,
      "eval_runtime": 1.6157,
      "eval_samples_per_second": 211.05,
      "eval_steps_per_second": 3.713,
      "step": 400
    },
    {
      "epoch": 4.941860465116279,
      "grad_norm": 0.7499828934669495,
      "learning_rate": 4.465668092927841e-05,
      "loss": 0.01998784065246582,
      "step": 425
    },
    {
      "epoch": 5.232558139534884,
      "grad_norm": 0.24809418618679047,
      "learning_rate": 4.350351769473198e-05,
      "loss": 0.01218403697013855,
      "step": 450
    },
    {
      "epoch": 5.232558139534884,
      "eval_accuracy": 0.9084461637653127,
      "eval_f1": 0.6831746031746032,
      "eval_loss": 0.05535344406962395,
      "eval_precision": 0.5984427141268076,
      "eval_recall": 0.7958579881656804,
      "eval_runtime": 1.6123,
      "eval_samples_per_second": 211.498,
      "eval_steps_per_second": 3.721,
      "step": 450
    },
    {
      "epoch": 5.523255813953488,
      "grad_norm": 0.3144132196903229,
      "learning_rate": 4.2255893178617745e-05,
      "loss": 0.012306832075119019,
      "step": 475
    },
    {
      "epoch": 5.813953488372093,
      "grad_norm": 0.1607646942138672,
      "learning_rate": 4.092017655963198e-05,
      "loss": 0.010539753437042236,
      "step": 500
    },
    {
      "epoch": 5.813953488372093,
      "eval_accuracy": 0.9527186761229315,
      "eval_f1": 0.8059040590405904,
      "eval_loss": 0.07191809266805649,
      "eval_precision": 0.8041237113402062,
      "eval_recall": 0.8076923076923077,
      "eval_runtime": 1.6569,
      "eval_samples_per_second": 205.806,
      "eval_steps_per_second": 3.621,
      "step": 500
    },
    {
      "epoch": 6.104651162790698,
      "grad_norm": 0.19801200926303864,
      "learning_rate": 3.9503186730577816e-05,
      "loss": 0.008691226243972778,
      "step": 525
    },
    {
      "epoch": 6.395348837209302,
      "grad_norm": 0.4177681505680084,
      "learning_rate": 3.801215748761385e-05,
      "loss": 0.004750225245952606,
      "step": 550
    },
    {
      "epoch": 6.395348837209302,
      "eval_accuracy": 0.9458413926499033,
      "eval_f1": 0.7863866763215062,
      "eval_loss": 0.07378304749727249,
      "eval_precision": 0.7702127659574468,
      "eval_recall": 0.8032544378698225,
      "eval_runtime": 1.6214,
      "eval_samples_per_second": 210.313,
      "eval_steps_per_second": 3.701,
      "step": 550
    },
    {
      "epoch": 6.686046511627907,
      "grad_norm": 0.30753573775291443,
      "learning_rate": 3.645470060140278e-05,
      "loss": 0.005743271708488464,
      "step": 575
    },
    {
      "epoch": 6.976744186046512,
      "grad_norm": 0.13237181305885315,
      "learning_rate": 3.4838766958683304e-05,
      "loss": 0.006215065121650696,
      "step": 600
    },
    {
      "epoch": 6.976744186046512,
      "eval_accuracy": 0.9447668171072426,
      "eval_f1": 0.7789934354485777,
      "eval_loss": 0.062308527529239655,
      "eval_precision": 0.7683453237410072,
      "eval_recall": 0.7899408284023669,
      "eval_runtime": 1.6197,
      "eval_samples_per_second": 210.528,
      "eval_steps_per_second": 3.704,
      "step": 600
    },
    {
      "epoch": 7.267441860465116,
      "grad_norm": 0.023809686303138733,
      "learning_rate": 3.317260597263932e-05,
      "loss": 0.002318141460418701,
      "step": 625
    },
    {
      "epoch": 7.558139534883721,
      "grad_norm": 0.03782212734222412,
      "learning_rate": 3.146472346927845e-05,
      "loss": 0.003790220320224762,
      "step": 650
    },
    {
      "epoch": 7.558139534883721,
      "eval_accuracy": 0.9529335912314636,
      "eval_f1": 0.8062827225130891,
      "eval_loss": 0.07683796435594559,
      "eval_precision": 0.8154311649016641,
      "eval_recall": 0.7973372781065089,
      "eval_runtime": 1.6235,
      "eval_samples_per_second": 210.043,
      "eval_steps_per_second": 3.696,
      "step": 650
    },
    {
      "epoch": 7.848837209302325,
      "grad_norm": 0.037971507757902145,
      "learning_rate": 2.9723838264811545e-05,
      "loss": 0.0018679146468639374,
      "step": 675
    },
    {
      "epoch": 8.13953488372093,
      "grad_norm": 0.6083823442459106,
      "learning_rate": 2.7958837655707814e-05,
      "loss": 0.0019237904250621796,
      "step": 700
    },
    {
      "epoch": 8.13953488372093,
      "eval_accuracy": 0.9565871480765098,
      "eval_f1": 0.8152091254752851,
      "eval_loss": 0.08732757717370987,
      "eval_precision": 0.838810641627543,
      "eval_recall": 0.7928994082840237,
      "eval_runtime": 1.6124,
      "eval_samples_per_second": 211.484,
      "eval_steps_per_second": 3.721,
      "step": 700
    },
    {
      "epoch": 8.430232558139535,
      "grad_norm": 0.022007286548614502,
      "learning_rate": 2.6178732048650694e-05,
      "loss": 0.0010710173845291137,
      "step": 725
    },
    {
      "epoch": 8.720930232558139,
      "grad_norm": 0.09439104795455933,
      "learning_rate": 2.4392608962010652e-05,
      "loss": 0.0006907544285058976,
      "step": 750
    },
    {
      "epoch": 8.720930232558139,
      "eval_accuracy": 0.9449817322157748,
      "eval_f1": 0.7829848594087959,
      "eval_loss": 0.08308640867471695,
      "eval_precision": 0.7637130801687764,
      "eval_recall": 0.8032544378698225,
      "eval_runtime": 1.6359,
      "eval_samples_per_second": 208.445,
      "eval_steps_per_second": 3.668,
      "step": 750
    },
    {
      "epoch": 9.011627906976743,
      "grad_norm": 0.018290311098098755,
      "learning_rate": 2.2609586633659256e-05,
      "loss": 0.0005397907271981239,
      "step": 775
    },
    {
      "epoch": 9.30232558139535,
      "grad_norm": 0.002839206252247095,
      "learning_rate": 2.0838767471959014e-05,
      "loss": 0.0002632186934351921,
      "step": 800
    },
    {
      "epoch": 9.30232558139535,
      "eval_accuracy": 0.9544379969911885,
      "eval_f1": 0.8081880212282032,
      "eval_loss": 0.09754534065723419,
      "eval_precision": 0.8289269051321928,
      "eval_recall": 0.7884615384615384,
      "eval_runtime": 1.6188,
      "eval_samples_per_second": 210.653,
      "eval_steps_per_second": 3.706,
      "step": 800
    },
    {
      "epoch": 9.593023255813954,
      "grad_norm": 0.04134448245167732,
      "learning_rate": 1.9089191587563417e-05,
      "loss": 0.0005076020956039428,
      "step": 825
    },
    {
      "epoch": 9.883720930232558,
      "grad_norm": 0.08968573063611984,
      "learning_rate": 1.7369790643249573e-05,
      "loss": 0.0005874237418174743,
      "step": 850
    },
    {
      "epoch": 9.883720930232558,
      "eval_accuracy": 0.9548678272082527,
      "eval_f1": 0.8093797276853252,
      "eval_loss": 0.09803693741559982,
      "eval_precision": 0.8281733746130031,
      "eval_recall": 0.7914201183431953,
      "eval_runtime": 1.6209,
      "eval_samples_per_second": 210.379,
      "eval_steps_per_second": 3.702,
      "step": 850
    },
    {
      "epoch": 10.174418604651162,
      "grad_norm": 0.02889181859791279,
      "learning_rate": 1.5689342257382207e-05,
      "loss": 0.0004838050529360771,
      "step": 875
    },
    {
      "epoch": 10.465116279069768,
      "grad_norm": 0.020970497280359268,
      "learning_rate": 1.4056425193781048e-05,
      "loss": 0.0001906752586364746,
      "step": 900
    },
    {
      "epoch": 10.465116279069768,
      "eval_accuracy": 0.9587362991618311,
      "eval_f1": 0.8228043143297381,
      "eval_loss": 0.10382715612649918,
      "eval_precision": 0.8585209003215434,
      "eval_recall": 0.7899408284023669,
      "eval_runtime": 1.6341,
      "eval_samples_per_second": 208.675,
      "eval_steps_per_second": 3.672,
      "step": 900
    },
    {
      "epoch": 10.755813953488373,
      "grad_norm": 0.028520096093416214,
      "learning_rate": 1.2479375566749694e-05,
      "loss": 0.00012492083944380284,
      "step": 925
    },
    {
      "epoch": 11.046511627906977,
      "grad_norm": 0.03838531672954559,
      "learning_rate": 1.0966244284840926e-05,
      "loss": 0.00028255857527256014,
      "step": 950
    },
    {
      "epoch": 11.046511627906977,
      "eval_accuracy": 0.9580915538362347,
      "eval_f1": 0.819369715603382,
      "eval_loss": 0.10540532320737839,
      "eval_precision": 0.8528,
      "eval_recall": 0.7884615384615384,
      "eval_runtime": 1.6495,
      "eval_samples_per_second": 206.73,
      "eval_steps_per_second": 3.637,
      "step": 950
    },
    {
      "epoch": 11.337209302325581,
      "grad_norm": 0.05398673936724663,
      "learning_rate": 9.524755950610204e-06,
      "loss": 0.00014415644109249114,
      "step": 975
    },
    {
      "epoch": 11.627906976744185,
      "grad_norm": 0.026561638340353966,
      "learning_rate": 8.162269426175681e-06,
      "loss": 9.150288999080658e-05,
      "step": 1000
    },
    {
      "epoch": 11.627906976744185,
      "eval_accuracy": 0.9574468085106383,
      "eval_f1": 0.8180428134556575,
      "eval_loss": 0.10544126480817795,
      "eval_precision": 0.8465189873417721,
      "eval_recall": 0.7914201183431953,
      "eval_runtime": 1.6274,
      "eval_samples_per_second": 209.539,
      "eval_steps_per_second": 3.687,
      "step": 1000
    },
    {
      "epoch": 11.918604651162791,
      "grad_norm": 0.010734460316598415,
      "learning_rate": 6.885740265899526e-06,
      "loss": 4.033456556499005e-05,
      "step": 1025
    },
    {
      "epoch": 12.209302325581396,
      "grad_norm": 0.02024712599813938,
      "learning_rate": 5.701685207973243e-06,
      "loss": 6.52561942115426e-05,
      "step": 1050
    },
    {
      "epoch": 12.209302325581396,
      "eval_accuracy": 0.9576617236191705,
      "eval_f1": 0.8190184049079755,
      "eval_loss": 0.10666228085756302,
      "eval_precision": 0.8503184713375797,
      "eval_recall": 0.7899408284023669,
      "eval_runtime": 1.6269,
      "eval_samples_per_second": 209.598,
      "eval_steps_per_second": 3.688,
      "step": 1050
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.034469809383153915,
      "learning_rate": 4.616148906179082e-06,
      "loss": 7.551711052656174e-05,
      "step": 1075
    },
    {
      "epoch": 12.790697674418604,
      "grad_norm": 0.004050907213240862,
      "learning_rate": 3.6346730716634025e-06,
      "loss": 7.49318953603506e-05,
      "step": 1100
    },
    {
      "epoch": 12.790697674418604,
      "eval_accuracy": 0.9565871480765098,
      "eval_f1": 0.8146453089244853,
      "eval_loss": 0.10619472712278366,
      "eval_precision": 0.8409448818897638,
      "eval_recall": 0.7899408284023669,
      "eval_runtime": 1.6196,
      "eval_samples_per_second": 210.546,
      "eval_steps_per_second": 3.705,
      "step": 1100
    },
    {
      "epoch": 13.081395348837209,
      "grad_norm": 0.11917345970869064,
      "learning_rate": 2.7622681822545764e-06,
      "loss": 8.815748617053032e-05,
      "step": 1125
    },
    {
      "epoch": 13.372093023255815,
      "grad_norm": 0.02682509273290634,
      "learning_rate": 2.0033879037506004e-06,
      "loss": 6.015756167471409e-05,
      "step": 1150
    },
    {
      "epoch": 13.372093023255815,
      "eval_accuracy": 0.9583064689447668,
      "eval_f1": 0.8205521472392638,
      "eval_loss": 0.10740578919649124,
      "eval_precision": 0.8519108280254777,
      "eval_recall": 0.7914201183431953,
      "eval_runtime": 1.6217,
      "eval_samples_per_second": 210.273,
      "eval_steps_per_second": 3.7,
      "step": 1150
    },
    {
      "epoch": 13.662790697674419,
      "grad_norm": 0.004100531339645386,
      "learning_rate": 1.361906353756917e-06,
      "loss": 4.690625239163637e-05,
      "step": 1175
    },
    {
      "epoch": 13.953488372093023,
      "grad_norm": 0.008385015651583672,
      "learning_rate": 8.410983241436132e-07,
      "loss": 6.560925394296646e-05,
      "step": 1200
    },
    {
      "epoch": 13.953488372093023,
      "eval_accuracy": 0.9578766387277026,
      "eval_f1": 0.8181120491174213,
      "eval_loss": 0.10830119252204895,
      "eval_precision": 0.8500797448165869,
      "eval_recall": 0.7884615384615384,
      "eval_runtime": 1.6315,
      "eval_samples_per_second": 209.012,
      "eval_steps_per_second": 3.678,
      "step": 1200
    },
    {
      "epoch": 14.244186046511627,
      "grad_norm": 0.003520221682265401,
      "learning_rate": 4.436225630870927e-07,
      "loss": 0.00012058223597705364,
      "step": 1225
    },
    {
      "epoch": 14.534883720930232,
      "grad_norm": 0.00905795767903328,
      "learning_rate": 1.7150820204230865e-07,
      "loss": 8.818465285003186e-05,
      "step": 1250
    },
    {
      "epoch": 14.534883720930232,
      "eval_accuracy": 0.9578766387277026,
      "eval_f1": 0.8187403993855606,
      "eval_loss": 0.10815286636352539,
      "eval_precision": 0.8514376996805112,
      "eval_recall": 0.7884615384615384,
      "eval_runtime": 1.6307,
      "eval_samples_per_second": 209.118,
      "eval_steps_per_second": 3.68,
      "step": 1250
    },
    {
      "epoch": 14.825581395348838,
      "grad_norm": 0.006475365720689297,
      "learning_rate": 2.6144396936325645e-08,
      "loss": 4.936248995363712e-05,
      "step": 1275
    }
  ],
  "logging_steps": 25,
  "max_steps": 1290,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 10,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 7
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6995052164620800.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}