max_seq_length_128_experiments/distilbert/distilbert_base_uncased_scotus/trainer_state.json

{
  "best_metric": 1.5984479188919067,
  "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/distilbert/distilbert_base_uncased_scotus/checkpoint-200",
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 237,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "grad_norm": 2.7217512130737305,
      "learning_rate": 1.9156118143459917e-05,
      "loss": 2.4003,
      "step": 10
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.063310384750366,
      "learning_rate": 1.8312236286919833e-05,
      "loss": 2.2082,
      "step": 20
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5437555313110352,
      "learning_rate": 1.746835443037975e-05,
      "loss": 2.1492,
      "step": 30
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.9715864658355713,
      "learning_rate": 1.662447257383966e-05,
      "loss": 2.076,
      "step": 40
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.068974018096924,
      "learning_rate": 1.578059071729958e-05,
      "loss": 1.9785,
      "step": 50
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 0.39571428571428574,
      "eval_f1_macro": 0.1007228498239773,
      "eval_f1_micro": 0.39571428571428574,
      "eval_loss": 1.9737688302993774,
      "eval_runtime": 0.7432,
      "eval_samples_per_second": 1883.664,
      "eval_steps_per_second": 29.6,
      "step": 50
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.165163516998291,
      "learning_rate": 1.4936708860759495e-05,
      "loss": 1.9311,
      "step": 60
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.0208511352539062,
      "learning_rate": 1.4092827004219412e-05,
      "loss": 1.7853,
      "step": 70
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.3129401206970215,
      "learning_rate": 1.3248945147679326e-05,
      "loss": 1.8448,
      "step": 80
    },
    {
      "epoch": 1.14,
      "grad_norm": 3.225861072540283,
      "learning_rate": 1.240506329113924e-05,
      "loss": 1.6705,
      "step": 90
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.632293939590454,
      "learning_rate": 1.1561181434599158e-05,
      "loss": 1.6121,
      "step": 100
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 0.48714285714285716,
      "eval_f1_macro": 0.18153434278456468,
      "eval_f1_micro": 0.48714285714285716,
      "eval_loss": 1.7287405729293823,
      "eval_runtime": 0.7378,
      "eval_samples_per_second": 1897.481,
      "eval_steps_per_second": 29.818,
      "step": 100
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.7373692989349365,
      "learning_rate": 1.0717299578059072e-05,
      "loss": 1.694,
      "step": 110
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.147618293762207,
      "learning_rate": 9.87341772151899e-06,
      "loss": 1.6588,
      "step": 120
    },
    {
      "epoch": 1.65,
      "grad_norm": 2.533884286880493,
      "learning_rate": 9.029535864978903e-06,
      "loss": 1.6191,
      "step": 130
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.6445531845092773,
      "learning_rate": 8.18565400843882e-06,
      "loss": 1.5258,
      "step": 140
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.774038314819336,
      "learning_rate": 7.341772151898735e-06,
      "loss": 1.6415,
      "step": 150
    },
    {
      "epoch": 1.9,
      "eval_accuracy": 0.5057142857142857,
      "eval_f1_macro": 0.19889010501840287,
      "eval_f1_micro": 0.5057142857142857,
      "eval_loss": 1.6443278789520264,
      "eval_runtime": 0.7425,
      "eval_samples_per_second": 1885.397,
      "eval_steps_per_second": 29.628,
      "step": 150
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.8844122886657715,
      "learning_rate": 6.49789029535865e-06,
      "loss": 1.5591,
      "step": 160
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.7135651111602783,
      "learning_rate": 5.654008438818566e-06,
      "loss": 1.4955,
      "step": 170
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.7151761054992676,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 1.544,
      "step": 180
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.708754301071167,
      "learning_rate": 3.9662447257383965e-06,
      "loss": 1.4917,
      "step": 190
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.7288765907287598,
      "learning_rate": 3.1223628691983127e-06,
      "loss": 1.4786,
      "step": 200
    },
    {
      "epoch": 2.53,
      "eval_accuracy": 0.5242857142857142,
      "eval_f1_macro": 0.2442080274860101,
      "eval_f1_micro": 0.5242857142857142,
      "eval_loss": 1.5984479188919067,
      "eval_runtime": 0.7477,
      "eval_samples_per_second": 1872.513,
      "eval_steps_per_second": 29.425,
      "step": 200
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.4662201404571533,
      "learning_rate": 2.278481012658228e-06,
      "loss": 1.4878,
      "step": 210
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.5271451473236084,
      "learning_rate": 1.4345991561181436e-06,
      "loss": 1.4375,
      "step": 220
    },
    {
      "epoch": 2.91,
      "grad_norm": 3.117112159729004,
      "learning_rate": 5.907172995780591e-07,
      "loss": 1.4802,
      "step": 230
    },
    {
      "epoch": 3.0,
      "step": 237,
      "total_flos": 502414912258048.0,
      "train_loss": 1.721195534814762,
      "train_runtime": 30.5391,
      "train_samples_per_second": 491.173,
      "train_steps_per_second": 7.761
    }
  ],
  "logging_steps": 10,
  "max_steps": 237,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "total_flos": 502414912258048.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}