{
  "best_metric": 0.6019545197486877,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.5_temp3.0_t3/checkpoint-2624",
  "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": 4.68709135055542,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.8426,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 5.267493724822998,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.822,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 6.456117630004883,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.8236,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 3.630889415740967,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.8086,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 3.2471823692321777,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.8139,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 5.173975944519043,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.8129,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.25488069414316705,
      "eval_loss": 1.4044036865234375,
      "eval_runtime": 42.4407,
      "eval_samples_per_second": 21.724,
      "eval_steps_per_second": 0.683,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 3.622982978820801,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.8198,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 4.415459156036377,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.8064,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 7.020495891571045,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.799,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 2.6109766960144043,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.8132,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 6.117157459259033,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.8,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 4.821333885192871,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.753,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 2.979583740234375,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.7193,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5065075921908894,
      "eval_loss": 1.1586508750915527,
      "eval_runtime": 40.0726,
      "eval_samples_per_second": 23.008,
      "eval_steps_per_second": 0.724,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 2.5219600200653076,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.6487,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 4.031715393066406,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.6221,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 5.4756550788879395,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.6056,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 7.903572082519531,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.5807,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 5.278177738189697,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.5794,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 5.145104885101318,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.5431,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7418655097613883,
      "eval_loss": 0.7721553444862366,
      "eval_runtime": 40.3437,
      "eval_samples_per_second": 22.854,
      "eval_steps_per_second": 0.719,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 4.718493938446045,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.5102,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 3.389207363128662,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.4575,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 3.910080909729004,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.4499,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 2.741947889328003,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.4714,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 4.898369789123535,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.4257,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 3.3292341232299805,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.4285,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 3.275188446044922,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.4342,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7581344902386117,
      "eval_loss": 0.760450005531311,
      "eval_runtime": 40.3605,
      "eval_samples_per_second": 22.844,
      "eval_steps_per_second": 0.719,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 2.784367322921753,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.4119,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 3.02673077583313,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.3809,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 3.911879062652588,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.381,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 1.1639741659164429,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.3795,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 2.106140375137329,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.3655,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 1.7877476215362549,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.3763,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8080260303687635,
      "eval_loss": 0.6501278281211853,
      "eval_runtime": 40.4912,
      "eval_samples_per_second": 22.77,
      "eval_steps_per_second": 0.716,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 2.096224784851074,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.3742,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 1.554193377494812,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.3406,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 2.2673614025115967,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.3343,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 3.702033042907715,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.3588,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 1.59395432472229,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.3453,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 1.8020344972610474,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.3375,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 4.047776699066162,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.3418,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7906724511930586,
      "eval_loss": 0.6815962791442871,
      "eval_runtime": 40.4542,
      "eval_samples_per_second": 22.791,
      "eval_steps_per_second": 0.717,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 1.2697885036468506,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.3351,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.248262882232666,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.3314,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 1.2794231176376343,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.3203,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 1.554882287979126,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.3298,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 2.1057960987091064,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.3256,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 3.269249439239502,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.323,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8101952277657267,
      "eval_loss": 0.6434170603752136,
      "eval_runtime": 40.6177,
      "eval_samples_per_second": 22.699,
      "eval_steps_per_second": 0.714,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 2.0187323093414307,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.3111,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 1.4456573724746704,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.3194,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 1.4459925889968872,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.3148,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 0.9452431201934814,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.3111,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 1.2967967987060547,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.3086,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 1.7521177530288696,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.3199,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 1.7025812864303589,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.3123,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8080260303687635,
      "eval_loss": 0.6708961725234985,
      "eval_runtime": 40.4633,
      "eval_samples_per_second": 22.786,
      "eval_steps_per_second": 0.717,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 0.8077727556228638,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.3065,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 1.584625244140625,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.3031,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 0.5091081857681274,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.316,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 1.150061845779419,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.3019,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 0.7301451563835144,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.3035,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 1.6500414609909058,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.3043,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 1.2119061946868896,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.3109,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8123644251626898,
      "eval_loss": 0.6393240690231323,
      "eval_runtime": 40.4639,
      "eval_samples_per_second": 22.786,
      "eval_steps_per_second": 0.717,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 0.7057688236236572,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.2964,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 1.1362090110778809,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.3089,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 1.779889464378357,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.302,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 2.1354031562805176,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.3065,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 0.4051612913608551,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.2999,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 0.9028952717781067,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.2944,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8329718004338394,
      "eval_loss": 0.6249269247055054,
      "eval_runtime": 40.2234,
      "eval_samples_per_second": 22.922,
      "eval_steps_per_second": 0.721,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 1.136090636253357,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.2958,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 0.6200639605522156,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.2954,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 1.4423284530639648,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.2998,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 0.7923824787139893,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.2995,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 0.7103340029716492,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.295,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 0.8508581519126892,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.2998,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 0.4668588638305664,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.2928,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.834056399132321,
      "eval_loss": 0.6533224582672119,
      "eval_runtime": 40.121,
      "eval_samples_per_second": 22.98,
      "eval_steps_per_second": 0.723,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 0.9473902583122253,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.296,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 0.9022158980369568,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.2893,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 0.8550190329551697,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.2924,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 1.9592097997665405,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.2915,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 1.4749720096588135,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.3016,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 1.2009644508361816,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.2966,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8275488069414316,
      "eval_loss": 0.6237086057662964,
      "eval_runtime": 40.4159,
      "eval_samples_per_second": 22.813,
      "eval_steps_per_second": 0.718,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 0.3396455943584442,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.2915,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 0.7679861783981323,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.2923,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 0.7669865489006042,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.2994,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.6513790488243103,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.2927,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 0.6599563956260681,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.293,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 0.6201032400131226,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.2865,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 0.9056888222694397,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.2845,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8373101952277657,
      "eval_loss": 0.6263235807418823,
      "eval_runtime": 40.5187,
      "eval_samples_per_second": 22.755,
      "eval_steps_per_second": 0.716,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 0.5516000986099243,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.2922,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 0.49217310547828674,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.2963,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 0.6513165235519409,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.2884,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 0.5991350412368774,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.2841,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 1.1869876384735107,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.2855,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 0.799327552318573,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.2933,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8503253796095445,
      "eval_loss": 0.6041625142097473,
      "eval_runtime": 40.6934,
      "eval_samples_per_second": 22.657,
      "eval_steps_per_second": 0.713,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 0.41806021332740784,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.2891,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 0.6208969950675964,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.2906,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 0.5198073387145996,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.2897,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 0.4373537302017212,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.2883,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 0.8531773090362549,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.2839,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 0.49251678586006165,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.2848,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 0.300030380487442,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.2917,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8449023861171366,
      "eval_loss": 0.6151757836341858,
      "eval_runtime": 39.8717,
      "eval_samples_per_second": 23.124,
      "eval_steps_per_second": 0.727,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 1.0790830850601196,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.2778,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 0.44671347737312317,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.2961,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 0.4193664491176605,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.2874,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 0.4943961501121521,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.287,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 0.39981332421302795,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.2829,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 0.4732746481895447,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.2865,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8546637744034707,
      "eval_loss": 0.6019545197486877,
      "eval_runtime": 39.1986,
      "eval_samples_per_second": 23.521,
      "eval_steps_per_second": 0.74,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 0.38862547278404236,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.2925,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 0.5937957763671875,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.2854,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 0.6209492087364197,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.2838,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 0.9960609674453735,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.2808,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 0.48445960879325867,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.2857,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 0.7934386134147644,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.2806,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 1.530712366104126,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.2923,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8568329718004338,
      "eval_loss": 0.6054571270942688,
      "eval_runtime": 40.7766,
      "eval_samples_per_second": 22.611,
      "eval_steps_per_second": 0.711,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 0.42646315693855286,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.2867,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 0.440929651260376,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.2835,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 0.4175315201282501,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.2908,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 0.3104482591152191,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.2908,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 0.582192599773407,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.2835,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 0.26515841484069824,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.2775,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 0.6080241799354553,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.281,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8438177874186551,
      "eval_loss": 0.6052865982055664,
      "eval_runtime": 40.1833,
      "eval_samples_per_second": 22.945,
      "eval_steps_per_second": 0.722,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 0.4831451177597046,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.2845,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 0.5719716548919678,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.2838,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 0.3154873847961426,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.2811,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 0.23891954123973846,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.2851,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.4182262122631073,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2809,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 0.21054212749004364,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.2865,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8503253796095445,
      "eval_loss": 0.6038812398910522,
      "eval_runtime": 40.5586,
      "eval_samples_per_second": 22.733,
      "eval_steps_per_second": 0.715,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 0.3340577781200409,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.2835,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 0.2887159585952759,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.291,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 0.31478652358055115,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.2865,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 0.397370845079422,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.2845,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 0.48177850246429443,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.2836,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 0.3912806808948517,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.279,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 0.20423896610736847,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.28,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8481561822125814,
      "eval_loss": 0.6056901216506958,
      "eval_runtime": 41.085,
      "eval_samples_per_second": 22.441,
      "eval_steps_per_second": 0.706,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.3744926492252001,
      "train_runtime": 14758.7315,
      "train_samples_per_second": 7.079,
      "train_steps_per_second": 0.222
    }
  ],
  "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
}