{
  "best_metric": 0.9565775394439697,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-small_alpha0.7_temp5.0_t3/checkpoint-2952",
  "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": 8.189471244812012,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.7718,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 1.7328355312347412,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.5884,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 2.701951026916504,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.5691,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 2.796539545059204,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.5657,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 3.6667864322662354,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.5628,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 1.97449791431427,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.5605,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.26247288503253796,
      "eval_loss": 1.4025862216949463,
      "eval_runtime": 40.6424,
      "eval_samples_per_second": 22.686,
      "eval_steps_per_second": 0.714,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 1.28244948387146,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.565,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 1.9142343997955322,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.5588,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 3.679830312728882,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.5546,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 3.1231205463409424,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.5631,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 3.1345646381378174,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.5591,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 2.961580753326416,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.5356,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 1.47929847240448,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.5309,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3210412147505423,
      "eval_loss": 1.3394988775253296,
      "eval_runtime": 40.4396,
      "eval_samples_per_second": 22.799,
      "eval_steps_per_second": 0.717,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 1.2656458616256714,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.5263,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 3.3095638751983643,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.5215,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 3.9150097370147705,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.4957,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 4.314517021179199,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.5022,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 3.652432441711426,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.5003,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 3.8010597229003906,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.4735,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4967462039045553,
      "eval_loss": 1.1832730770111084,
      "eval_runtime": 40.2778,
      "eval_samples_per_second": 22.891,
      "eval_steps_per_second": 0.72,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 3.1977288722991943,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.4755,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 3.3505699634552,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.475,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 3.71514892578125,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.4479,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 2.5243544578552246,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.4444,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 3.1844165325164795,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.4457,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 3.4737300872802734,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.4342,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 6.344496726989746,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.4298,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5260303687635575,
      "eval_loss": 1.1389656066894531,
      "eval_runtime": 40.6034,
      "eval_samples_per_second": 22.707,
      "eval_steps_per_second": 0.714,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 1.5928798913955688,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.4302,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 3.1686978340148926,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.4258,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 3.4186737537384033,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.4119,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 6.188898086547852,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.4275,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 3.578242063522339,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.4081,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 2.729630470275879,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.4036,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5748373101952278,
      "eval_loss": 1.0784701108932495,
      "eval_runtime": 40.2411,
      "eval_samples_per_second": 22.912,
      "eval_steps_per_second": 0.721,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 4.695616722106934,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.4194,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 3.0300028324127197,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.4021,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 4.037109375,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.3885,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 4.685178279876709,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.396,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 3.1342365741729736,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.3929,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 1.4096190929412842,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.3895,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 2.12451434135437,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.3936,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5921908893709328,
      "eval_loss": 1.0463019609451294,
      "eval_runtime": 40.6746,
      "eval_samples_per_second": 22.668,
      "eval_steps_per_second": 0.713,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 5.905704498291016,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.3867,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.444490909576416,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.3846,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 1.9741724729537964,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.3853,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 3.790639638900757,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.3829,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 4.159935474395752,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.3783,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 2.0675580501556396,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.3746,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6073752711496746,
      "eval_loss": 1.0220353603363037,
      "eval_runtime": 40.7084,
      "eval_samples_per_second": 22.649,
      "eval_steps_per_second": 0.712,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 1.8815311193466187,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.3694,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 2.5251076221466064,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.3674,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 2.4822423458099365,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.3582,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 4.727237224578857,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.3532,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 2.819591999053955,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.3683,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 3.6963562965393066,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.3627,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 3.315181255340576,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.3644,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.631236442516269,
      "eval_loss": 1.002418875694275,
      "eval_runtime": 40.3989,
      "eval_samples_per_second": 22.822,
      "eval_steps_per_second": 0.718,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 3.3769114017486572,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.3556,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 4.508443832397461,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.3449,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 3.271930456161499,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.3605,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 1.3970648050308228,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.3389,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 2.6376571655273438,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.3573,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 3.3899121284484863,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.342,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 3.9985809326171875,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.3514,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6160520607375272,
      "eval_loss": 0.9987125396728516,
      "eval_runtime": 41.3919,
      "eval_samples_per_second": 22.275,
      "eval_steps_per_second": 0.701,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 1.7232149839401245,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.3382,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 3.1317169666290283,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.3499,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 3.785961866378784,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.3392,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 3.2999846935272217,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.3345,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 3.8600687980651855,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.3447,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 2.395810842514038,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.3366,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6247288503253796,
      "eval_loss": 1.0162410736083984,
      "eval_runtime": 41.4003,
      "eval_samples_per_second": 22.27,
      "eval_steps_per_second": 0.7,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 3.0433719158172607,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.3364,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 1.941277027130127,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.3289,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 4.652011394500732,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.34,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 2.4000205993652344,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.33,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 2.2985665798187256,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.325,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 3.168454170227051,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.3345,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 4.306560039520264,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.3377,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6073752711496746,
      "eval_loss": 1.0073492527008057,
      "eval_runtime": 41.3428,
      "eval_samples_per_second": 22.301,
      "eval_steps_per_second": 0.701,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 4.902525424957275,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.3382,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 4.037588596343994,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.3234,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 3.5839176177978516,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.3228,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 4.032567501068115,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.3214,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 2.031604528427124,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.3289,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 3.957519054412842,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.3291,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6323210412147505,
      "eval_loss": 0.9800806641578674,
      "eval_runtime": 41.5692,
      "eval_samples_per_second": 22.18,
      "eval_steps_per_second": 0.698,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 1.7875639200210571,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.3286,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 2.1771202087402344,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.3122,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 4.263763427734375,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.3215,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 2.513144016265869,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.32,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 1.9128166437149048,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.3189,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 3.0551483631134033,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.3102,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 3.003140449523926,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.3156,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.658351409978308,
      "eval_loss": 0.9602269530296326,
      "eval_runtime": 41.7005,
      "eval_samples_per_second": 22.11,
      "eval_steps_per_second": 0.695,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 2.685469388961792,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.3188,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 2.435077428817749,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.3169,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 2.26979398727417,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.3151,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 1.8860948085784912,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.3075,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 4.036872863769531,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.3061,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 1.734885573387146,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.3177,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6442516268980477,
      "eval_loss": 0.974234402179718,
      "eval_runtime": 40.8193,
      "eval_samples_per_second": 22.587,
      "eval_steps_per_second": 0.71,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 4.14552116394043,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.3127,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 3.565781831741333,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.3114,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 3.494729518890381,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.3061,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 1.0175479650497437,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.3065,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 3.574209451675415,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.3118,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 1.8788031339645386,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.3096,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 2.41518497467041,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.31,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.648590021691974,
      "eval_loss": 0.968324601650238,
      "eval_runtime": 41.4772,
      "eval_samples_per_second": 22.229,
      "eval_steps_per_second": 0.699,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 3.1199381351470947,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.2981,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 3.316758871078491,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.3081,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 1.6707791090011597,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.307,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 2.384627103805542,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.2946,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 2.757924795150757,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.3051,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 1.3199679851531982,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.3071,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6507592190889371,
      "eval_loss": 0.9691674709320068,
      "eval_runtime": 41.6947,
      "eval_samples_per_second": 22.113,
      "eval_steps_per_second": 0.696,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 3.2221853733062744,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.3062,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 2.3375253677368164,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.2994,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 4.375495433807373,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.2971,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 2.7444097995758057,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.2964,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 1.3352084159851074,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.3032,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 3.3579742908477783,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.2963,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 2.475299835205078,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.3077,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6572668112798264,
      "eval_loss": 0.9566349983215332,
      "eval_runtime": 41.259,
      "eval_samples_per_second": 22.347,
      "eval_steps_per_second": 0.703,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 2.1953816413879395,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.3015,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 1.230119228363037,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.2907,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 2.223381757736206,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.3046,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 2.9400393962860107,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.304,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 1.3474946022033691,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.301,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 2.508678674697876,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.2949,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 1.5501606464385986,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.2957,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.658351409978308,
      "eval_loss": 0.9565775394439697,
      "eval_runtime": 40.9228,
      "eval_samples_per_second": 22.53,
      "eval_steps_per_second": 0.709,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 1.400895595550537,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.2985,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 3.581841230392456,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.2979,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 1.246849536895752,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.2972,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 1.2201147079467773,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.2955,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 1.5855196714401245,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2927,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 2.1912453174591064,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.3017,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6648590021691974,
      "eval_loss": 0.9574868083000183,
      "eval_runtime": 41.4305,
      "eval_samples_per_second": 22.254,
      "eval_steps_per_second": 0.7,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 2.730165958404541,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.2942,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 1.4807108640670776,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.3009,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 1.0064300298690796,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.2975,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 1.0867955684661865,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.2942,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 2.274994134902954,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.2979,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 3.7017014026641846,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.2917,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 1.147507667541504,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.2911,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6572668112798264,
      "eval_loss": 0.9592885971069336,
      "eval_runtime": 41.3712,
      "eval_samples_per_second": 22.286,
      "eval_steps_per_second": 0.701,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.371750411972767,
      "train_runtime": 14482.7553,
      "train_samples_per_second": 7.214,
      "train_steps_per_second": 0.226
    }
  ],
  "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
}