trainer_state.json

{
  "best_metric": 70.29279085912718,
  "best_model_checkpoint": "/data/jcanete/all_results/mlqa/albeto_base/epochs_4_bs_64_lr_5e-5/checkpoint-4200",
  "epoch": 3.9996101364522416,
  "global_step": 5128,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "eval_exact_match": 36.4,
      "eval_f1": 61.84101912660733,
      "step": 300
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.517355694227769e-05,
      "loss": 2.3283,
      "step": 500
    },
    {
      "epoch": 0.47,
      "eval_exact_match": 38.6,
      "eval_f1": 64.05444583625557,
      "step": 600
    },
    {
      "epoch": 0.7,
      "eval_exact_match": 39.8,
      "eval_f1": 64.38318427781859,
      "step": 900
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.030811232449298e-05,
      "loss": 1.8958,
      "step": 1000
    },
    {
      "epoch": 0.94,
      "eval_exact_match": 42.6,
      "eval_f1": 67.53463785062704,
      "step": 1200
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.5432917316692674e-05,
      "loss": 1.7304,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "eval_exact_match": 41.6,
      "eval_f1": 67.26857117036066,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "eval_exact_match": 40.8,
      "eval_f1": 68.19818328108937,
      "step": 1800
    },
    {
      "epoch": 1.56,
      "learning_rate": 3.055772230889236e-05,
      "loss": 1.606,
      "step": 2000
    },
    {
      "epoch": 1.64,
      "eval_exact_match": 42.6,
      "eval_f1": 68.13227861379248,
      "step": 2100
    },
    {
      "epoch": 1.87,
      "eval_exact_match": 45.2,
      "eval_f1": 69.37715961143962,
      "step": 2400
    },
    {
      "epoch": 1.95,
      "learning_rate": 2.5682527301092048e-05,
      "loss": 1.5968,
      "step": 2500
    },
    {
      "epoch": 2.11,
      "eval_exact_match": 42.4,
      "eval_f1": 68.23125120786263,
      "step": 2700
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.080733229329173e-05,
      "loss": 1.4139,
      "step": 3000
    },
    {
      "epoch": 2.34,
      "eval_exact_match": 43.2,
      "eval_f1": 69.28256897389873,
      "step": 3000
    },
    {
      "epoch": 2.57,
      "eval_exact_match": 43.8,
      "eval_f1": 69.8210897722378,
      "step": 3300
    },
    {
      "epoch": 2.73,
      "learning_rate": 1.593213728549142e-05,
      "loss": 1.3802,
      "step": 3500
    },
    {
      "epoch": 2.81,
      "eval_exact_match": 44.0,
      "eval_f1": 69.9704820053898,
      "step": 3600
    },
    {
      "epoch": 3.04,
      "eval_exact_match": 43.6,
      "eval_f1": 70.21665964298872,
      "step": 3900
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.1056942277691108e-05,
      "loss": 1.3413,
      "step": 4000
    },
    {
      "epoch": 3.28,
      "eval_exact_match": 44.0,
      "eval_f1": 70.29279085912718,
      "step": 4200
    },
    {
      "epoch": 3.51,
      "learning_rate": 6.1817472698907965e-06,
      "loss": 1.2214,
      "step": 4500
    },
    {
      "epoch": 3.51,
      "eval_exact_match": 43.2,
      "eval_f1": 69.84257510948594,
      "step": 4500
    },
    {
      "epoch": 3.74,
      "eval_exact_match": 44.2,
      "eval_f1": 69.60819358399327,
      "step": 4800
    },
    {
      "epoch": 3.9,
      "learning_rate": 1.3065522620904836e-06,
      "loss": 1.2215,
      "step": 5000
    },
    {
      "epoch": 3.98,
      "eval_exact_match": 43.4,
      "eval_f1": 69.19469528834327,
      "step": 5100
    },
    {
      "epoch": 4.0,
      "step": 5128,
      "total_flos": 5575305579646752.0,
      "train_loss": 1.5647673971381462,
      "train_runtime": 3554.3046,
      "train_samples_per_second": 92.34,
      "train_steps_per_second": 1.443
    }
  ],
  "max_steps": 5128,
  "num_train_epochs": 4,
  "total_flos": 5575305579646752.0,
  "trial_name": null,
  "trial_params": null
}