{
  "best_metric": 0.7371716499328613,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.7_temp3.0_t3/checkpoint-1968",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 3280,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1524390243902439,
      "grad_norm": 4.130408763885498,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.6029,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 4.129245281219482,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.5746,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 6.233711242675781,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.5797,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 2.5758605003356934,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.5691,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 2.7497589588165283,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.5691,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 4.016096591949463,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.5666,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2559652928416486,
      "eval_loss": 1.4206671714782715,
      "eval_runtime": 42.5074,
      "eval_samples_per_second": 21.69,
      "eval_steps_per_second": 0.682,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 2.849764585494995,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.5721,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 4.841067314147949,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.5721,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 5.506284236907959,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.5574,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 3.889922618865967,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.5672,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 2.9700119495391846,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.5597,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 3.7748219966888428,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.5416,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 2.7717812061309814,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.5216,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3774403470715835,
      "eval_loss": 1.3016068935394287,
      "eval_runtime": 41.9522,
      "eval_samples_per_second": 21.977,
      "eval_steps_per_second": 0.691,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 3.9956743717193604,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.4796,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 4.206533432006836,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.4553,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 4.580026149749756,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.4393,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 6.77695894241333,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.4157,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 3.3758962154388428,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.4111,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 3.101966619491577,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.3815,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6778741865509761,
      "eval_loss": 0.9241301417350769,
      "eval_runtime": 41.9124,
      "eval_samples_per_second": 21.998,
      "eval_steps_per_second": 0.692,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 1.5353446006774902,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.3763,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 2.3866560459136963,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.3331,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 2.588350534439087,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.3266,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 3.552384376525879,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.336,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 2.6808509826660156,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.3151,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 1.148443579673767,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.3283,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 3.3858516216278076,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.324,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6995661605206074,
      "eval_loss": 0.8860035538673401,
      "eval_runtime": 41.7503,
      "eval_samples_per_second": 22.084,
      "eval_steps_per_second": 0.695,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 1.106536865234375,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.3084,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 3.1932008266448975,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.2866,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 3.1825125217437744,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.2863,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 1.2984044551849365,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.2852,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 5.454014301300049,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.281,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 1.7663265466690063,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.2839,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7830802603036876,
      "eval_loss": 0.7562471628189087,
      "eval_runtime": 42.1558,
      "eval_samples_per_second": 21.871,
      "eval_steps_per_second": 0.688,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 1.5635110139846802,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.2849,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 1.5313557386398315,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.2657,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 1.4946922063827515,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.2579,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 1.3229933977127075,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.2669,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 1.4292865991592407,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.2617,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 2.3933095932006836,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.2663,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 0.8884098529815674,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.263,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7874186550976139,
      "eval_loss": 0.7732966542243958,
      "eval_runtime": 40.9777,
      "eval_samples_per_second": 22.5,
      "eval_steps_per_second": 0.708,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 1.6093999147415161,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.2549,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.152940034866333,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.2572,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 1.8871915340423584,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.2528,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 1.3599646091461182,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.2575,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 1.6081894636154175,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.2535,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 1.464537501335144,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.2514,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7939262472885033,
      "eval_loss": 0.7519813776016235,
      "eval_runtime": 42.1977,
      "eval_samples_per_second": 21.85,
      "eval_steps_per_second": 0.687,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 0.8137891292572021,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.247,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 1.0123687982559204,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.2486,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 0.910114586353302,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.2439,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 1.0693368911743164,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.2409,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 1.2416306734085083,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.2425,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 1.1681289672851562,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.2521,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 0.911630392074585,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.2459,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7960954446854663,
      "eval_loss": 0.7617538571357727,
      "eval_runtime": 41.1882,
      "eval_samples_per_second": 22.385,
      "eval_steps_per_second": 0.704,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 1.316224217414856,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.2408,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 0.7942056655883789,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.2341,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 0.9957849383354187,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.2492,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 1.141812801361084,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.2357,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 0.5211377739906311,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.238,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 1.0767700672149658,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.2369,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 0.5546083450317383,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.2449,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7982646420824295,
      "eval_loss": 0.7674908638000488,
      "eval_runtime": 41.6918,
      "eval_samples_per_second": 22.115,
      "eval_steps_per_second": 0.696,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 0.7693997025489807,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.2326,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 0.9971864223480225,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.2423,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 1.401846170425415,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.238,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 0.9956872463226318,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.2388,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 0.7329416871070862,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.2379,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 1.5872021913528442,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.2312,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8145336225596529,
      "eval_loss": 0.7584421634674072,
      "eval_runtime": 41.5972,
      "eval_samples_per_second": 22.165,
      "eval_steps_per_second": 0.697,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 1.1239310503005981,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.2322,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 0.5998251438140869,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.2301,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 0.7929293513298035,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.2367,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 0.8035590648651123,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.2338,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 1.1374906301498413,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.2309,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 1.0718984603881836,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.2345,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 1.033934473991394,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.2297,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8036876355748374,
      "eval_loss": 0.7757138013839722,
      "eval_runtime": 41.1766,
      "eval_samples_per_second": 22.391,
      "eval_steps_per_second": 0.704,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 0.7480637431144714,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.2343,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 0.845191240310669,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.2265,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 0.9163962006568909,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.2291,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 0.6352323889732361,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.2268,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 0.691192090511322,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.2382,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 0.6980682611465454,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.2322,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8221258134490239,
      "eval_loss": 0.7371716499328613,
      "eval_runtime": 41.6322,
      "eval_samples_per_second": 22.146,
      "eval_steps_per_second": 0.697,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 0.6022236943244934,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.2281,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 0.5455940365791321,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.2282,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 0.7416942119598389,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.2331,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.473399817943573,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.2296,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 0.7426900863647461,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.2307,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 0.7116826772689819,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.2248,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 0.8426291346549988,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.2239,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8134490238611713,
      "eval_loss": 0.7653362154960632,
      "eval_runtime": 41.4508,
      "eval_samples_per_second": 22.243,
      "eval_steps_per_second": 0.7,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 0.3719276785850525,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.2301,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 0.4737173020839691,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.232,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 0.699379026889801,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.2257,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 0.6058977246284485,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.2216,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 1.1857759952545166,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.2239,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 1.0432237386703491,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.2282,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8232104121475055,
      "eval_loss": 0.7396612167358398,
      "eval_runtime": 41.9046,
      "eval_samples_per_second": 22.002,
      "eval_steps_per_second": 0.692,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 0.44644126296043396,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.2261,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 0.6213334798812866,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.2292,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 1.1576259136199951,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.227,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 0.4390643239021301,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.2252,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 0.7716153860092163,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.2221,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 0.38759616017341614,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.2228,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 0.2993668019771576,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.2291,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8047722342733189,
      "eval_loss": 0.7517526745796204,
      "eval_runtime": 41.4988,
      "eval_samples_per_second": 22.218,
      "eval_steps_per_second": 0.699,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 0.8219616413116455,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.2161,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 0.58041912317276,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.2323,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 0.34345969557762146,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.2265,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 0.5889475345611572,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.2233,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 0.3325055241584778,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.2214,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 0.6006044745445251,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.2243,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8134490238611713,
      "eval_loss": 0.7498494386672974,
      "eval_runtime": 41.9082,
      "eval_samples_per_second": 22.0,
      "eval_steps_per_second": 0.692,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 0.9776254296302795,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.2271,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 0.5019322633743286,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.2231,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 0.5922698378562927,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.2215,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 0.47184258699417114,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.2193,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 0.703269898891449,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.2242,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 0.4298715591430664,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.2183,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 0.5792725682258606,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.2285,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.824295010845987,
      "eval_loss": 0.7430695295333862,
      "eval_runtime": 40.5765,
      "eval_samples_per_second": 22.722,
      "eval_steps_per_second": 0.715,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 0.3882294297218323,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.2248,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 0.4255586862564087,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.2218,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 0.915451169013977,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.228,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 0.29044169187545776,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.2275,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 0.5202724933624268,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.2207,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 0.47653186321258545,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.2167,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 0.4258553683757782,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.2191,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.824295010845987,
      "eval_loss": 0.7445759177207947,
      "eval_runtime": 40.8119,
      "eval_samples_per_second": 22.591,
      "eval_steps_per_second": 0.711,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 0.2472049742937088,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.222,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 1.0352897644042969,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.2224,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 0.3187510669231415,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.2192,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 0.4212016463279724,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.2228,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.3100457489490509,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2198,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 0.26730892062187195,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.2238,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8232104121475055,
      "eval_loss": 0.7432267069816589,
      "eval_runtime": 40.2321,
      "eval_samples_per_second": 22.917,
      "eval_steps_per_second": 0.721,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 0.31010693311691284,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.222,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 0.45770835876464844,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.2281,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 0.32277268171310425,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.2238,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 0.31574589014053345,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.2219,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 0.3392852246761322,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.2226,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 0.3520653247833252,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.2177,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 0.21607498824596405,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.2187,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.824295010845987,
      "eval_loss": 0.7467628717422485,
      "eval_runtime": 39.8668,
      "eval_samples_per_second": 23.127,
      "eval_steps_per_second": 0.727,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.28329727395278653,
      "train_runtime": 14840.8024,
      "train_samples_per_second": 7.04,
      "train_steps_per_second": 0.221
    }
  ],
  "logging_steps": 25,
  "max_steps": 3280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}