{
  "best_metric": 1.5028287172317505,
  "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_t5/t5_base_scotus/checkpoint-400",
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 471,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 2.9687695503234863,
      "learning_rate": 0.0004893842887473461,
      "loss": 2.245,
      "step": 10
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.2519373893737793,
      "learning_rate": 0.00047876857749469217,
      "loss": 2.1475,
      "step": 20
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.8483872413635254,
      "learning_rate": 0.0004681528662420382,
      "loss": 2.1533,
      "step": 30
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.697084903717041,
      "learning_rate": 0.00045753715498938433,
      "loss": 2.0035,
      "step": 40
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.669522762298584,
      "learning_rate": 0.0004469214437367304,
      "loss": 2.0135,
      "step": 50
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.3357142857142857,
      "eval_f1_macro": 0.0896297993579241,
      "eval_f1_micro": 0.3357142857142857,
      "eval_loss": 1.9643603563308716,
      "eval_runtime": 3.3846,
      "eval_samples_per_second": 413.638,
      "eval_steps_per_second": 13.0,
      "step": 50
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.081737995147705,
      "learning_rate": 0.00043630573248407644,
      "loss": 1.8487,
      "step": 60
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.5176327228546143,
      "learning_rate": 0.00042569002123142254,
      "loss": 1.7924,
      "step": 70
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.1894452571868896,
      "learning_rate": 0.0004150743099787686,
      "loss": 1.7488,
      "step": 80
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.7363619804382324,
      "learning_rate": 0.0004044585987261147,
      "loss": 1.6172,
      "step": 90
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.9973998069763184,
      "learning_rate": 0.00039384288747346076,
      "loss": 1.6667,
      "step": 100
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.455,
      "eval_f1_macro": 0.22291521072048548,
      "eval_f1_micro": 0.455,
      "eval_loss": 1.757584810256958,
      "eval_runtime": 3.4183,
      "eval_samples_per_second": 409.561,
      "eval_steps_per_second": 12.872,
      "step": 100
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.3672401905059814,
      "learning_rate": 0.0003832271762208068,
      "loss": 1.6553,
      "step": 110
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.7414376735687256,
      "learning_rate": 0.00037261146496815286,
      "loss": 1.6762,
      "step": 120
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.143216133117676,
      "learning_rate": 0.0003619957537154989,
      "loss": 1.5513,
      "step": 130
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.8119118213653564,
      "learning_rate": 0.000351380042462845,
      "loss": 1.5697,
      "step": 140
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.1018612384796143,
      "learning_rate": 0.0003407643312101911,
      "loss": 1.6764,
      "step": 150
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.49142857142857144,
      "eval_f1_macro": 0.24594406581528833,
      "eval_f1_micro": 0.49142857142857144,
      "eval_loss": 1.6189066171646118,
      "eval_runtime": 3.4379,
      "eval_samples_per_second": 407.23,
      "eval_steps_per_second": 12.799,
      "step": 150
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.2958126068115234,
      "learning_rate": 0.00033014861995753713,
      "loss": 1.5329,
      "step": 160
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.6772100925445557,
      "learning_rate": 0.00031953290870488323,
      "loss": 1.3477,
      "step": 170
    },
    {
      "epoch": 1.15,
      "grad_norm": 3.70650315284729,
      "learning_rate": 0.0003089171974522293,
      "loss": 1.499,
      "step": 180
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.109652042388916,
      "learning_rate": 0.00029830148619957534,
      "loss": 1.4567,
      "step": 190
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.5284805297851562,
      "learning_rate": 0.00028768577494692145,
      "loss": 1.3803,
      "step": 200
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 0.48142857142857143,
      "eval_f1_macro": 0.2690374005007927,
      "eval_f1_micro": 0.48142857142857143,
      "eval_loss": 1.6397602558135986,
      "eval_runtime": 3.4426,
      "eval_samples_per_second": 406.667,
      "eval_steps_per_second": 12.781,
      "step": 200
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.480762004852295,
      "learning_rate": 0.0002770700636942675,
      "loss": 1.4702,
      "step": 210
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.323151111602783,
      "learning_rate": 0.0002664543524416136,
      "loss": 1.2652,
      "step": 220
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.6335439682006836,
      "learning_rate": 0.00025583864118895966,
      "loss": 1.4015,
      "step": 230
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.2516725063323975,
      "learning_rate": 0.0002452229299363057,
      "loss": 1.3274,
      "step": 240
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.6842403411865234,
      "learning_rate": 0.00023460721868365182,
      "loss": 1.315,
      "step": 250
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.5471428571428572,
      "eval_f1_macro": 0.3479706465936974,
      "eval_f1_micro": 0.5471428571428572,
      "eval_loss": 1.546648621559143,
      "eval_runtime": 3.4514,
      "eval_samples_per_second": 405.631,
      "eval_steps_per_second": 12.748,
      "step": 250
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.566072940826416,
      "learning_rate": 0.0002239915074309979,
      "loss": 1.3606,
      "step": 260
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.178421974182129,
      "learning_rate": 0.00021337579617834395,
      "loss": 1.396,
      "step": 270
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.773405075073242,
      "learning_rate": 0.00020276008492569003,
      "loss": 1.2954,
      "step": 280
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.5649752616882324,
      "learning_rate": 0.0001921443736730361,
      "loss": 1.2681,
      "step": 290
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.4314775466918945,
      "learning_rate": 0.00018152866242038217,
      "loss": 1.229,
      "step": 300
    },
    {
      "epoch": 1.91,
      "eval_accuracy": 0.5542857142857143,
      "eval_f1_macro": 0.3855062205660112,
      "eval_f1_micro": 0.5542857142857143,
      "eval_loss": 1.5176787376403809,
      "eval_runtime": 3.5144,
      "eval_samples_per_second": 398.358,
      "eval_steps_per_second": 12.52,
      "step": 300
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.010051727294922,
      "learning_rate": 0.00017091295116772822,
      "loss": 1.2579,
      "step": 310
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.9793009757995605,
      "learning_rate": 0.0001602972399150743,
      "loss": 1.2206,
      "step": 320
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.624272108078003,
      "learning_rate": 0.00014968152866242038,
      "loss": 1.0831,
      "step": 330
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.17748761177063,
      "learning_rate": 0.00013906581740976646,
      "loss": 1.2332,
      "step": 340
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.369600296020508,
      "learning_rate": 0.00012845010615711254,
      "loss": 1.0902,
      "step": 350
    },
    {
      "epoch": 2.23,
      "eval_accuracy": 0.5585714285714286,
      "eval_f1_macro": 0.39765198539309016,
      "eval_f1_micro": 0.5585714285714286,
      "eval_loss": 1.5300391912460327,
      "eval_runtime": 3.4532,
      "eval_samples_per_second": 405.425,
      "eval_steps_per_second": 12.742,
      "step": 350
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.4971916675567627,
      "learning_rate": 0.0001178343949044586,
      "loss": 1.1181,
      "step": 360
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.7142956256866455,
      "learning_rate": 0.00010721868365180467,
      "loss": 1.0325,
      "step": 370
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.7627336978912354,
      "learning_rate": 9.660297239915075e-05,
      "loss": 1.1275,
      "step": 380
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.2720165252685547,
      "learning_rate": 8.598726114649682e-05,
      "loss": 1.0764,
      "step": 390
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.69014835357666,
      "learning_rate": 7.537154989384288e-05,
      "loss": 1.0522,
      "step": 400
    },
    {
      "epoch": 2.55,
      "eval_accuracy": 0.5657142857142857,
      "eval_f1_macro": 0.4058731854607469,
      "eval_f1_micro": 0.5657142857142857,
      "eval_loss": 1.5028287172317505,
      "eval_runtime": 3.4556,
      "eval_samples_per_second": 405.134,
      "eval_steps_per_second": 12.733,
      "step": 400
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.6888267993927,
      "learning_rate": 6.475583864118896e-05,
      "loss": 1.1666,
      "step": 410
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.5979206562042236,
      "learning_rate": 5.414012738853504e-05,
      "loss": 1.0935,
      "step": 420
    },
    {
      "epoch": 2.74,
      "grad_norm": 3.747196674346924,
      "learning_rate": 4.35244161358811e-05,
      "loss": 1.09,
      "step": 430
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.983452796936035,
      "learning_rate": 3.2908704883227177e-05,
      "loss": 1.0263,
      "step": 440
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.690558433532715,
      "learning_rate": 2.2292993630573246e-05,
      "loss": 0.9946,
      "step": 450
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.5628571428571428,
      "eval_f1_macro": 0.40603867861430476,
      "eval_f1_micro": 0.5628571428571428,
      "eval_loss": 1.5198277235031128,
      "eval_runtime": 3.4615,
      "eval_samples_per_second": 404.445,
      "eval_steps_per_second": 12.711,
      "step": 450
    },
    {
      "epoch": 2.93,
      "grad_norm": 3.4624648094177246,
      "learning_rate": 1.167728237791932e-05,
      "loss": 1.0296,
      "step": 460
    },
    {
      "epoch": 2.99,
      "grad_norm": 3.968691110610962,
      "learning_rate": 1.0615711252653927e-06,
      "loss": 1.1202,
      "step": 470
    },
    {
      "epoch": 3.0,
      "step": 471,
      "total_flos": 2301505374257152.0,
      "train_loss": 1.4186640820938563,
      "train_runtime": 181.6507,
      "train_samples_per_second": 82.576,
      "train_steps_per_second": 2.593
    }
  ],
  "logging_steps": 10,
  "max_steps": 471,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "total_flos": 2301505374257152.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}