last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.03626562117128155,
  "eval_steps": 5,
  "global_step": 37,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00098015192354815,
      "eval_loss": 1.3752902746200562,
      "eval_runtime": 32.5898,
      "eval_samples_per_second": 13.194,
      "eval_steps_per_second": 6.597,
      "step": 1
    },
    {
      "epoch": 0.00294045577064445,
      "grad_norm": 2.109035015106201,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.4108,
      "step": 3
    },
    {
      "epoch": 0.00490075961774075,
      "eval_loss": 1.2923061847686768,
      "eval_runtime": 33.1287,
      "eval_samples_per_second": 12.98,
      "eval_steps_per_second": 6.49,
      "step": 5
    },
    {
      "epoch": 0.0058809115412889,
      "grad_norm": 1.7424509525299072,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.234,
      "step": 6
    },
    {
      "epoch": 0.008821367311933349,
      "grad_norm": 1.0848296880722046,
      "learning_rate": 0.0001,
      "loss": 1.0658,
      "step": 9
    },
    {
      "epoch": 0.0098015192354815,
      "eval_loss": 0.8554880023002625,
      "eval_runtime": 33.1587,
      "eval_samples_per_second": 12.968,
      "eval_steps_per_second": 6.484,
      "step": 10
    },
    {
      "epoch": 0.0117618230825778,
      "grad_norm": 1.0243185758590698,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.8279,
      "step": 12
    },
    {
      "epoch": 0.014702278853222249,
      "grad_norm": 0.9655507206916809,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.6354,
      "step": 15
    },
    {
      "epoch": 0.014702278853222249,
      "eval_loss": 0.5227863192558289,
      "eval_runtime": 33.2549,
      "eval_samples_per_second": 12.93,
      "eval_steps_per_second": 6.465,
      "step": 15
    },
    {
      "epoch": 0.017642734623866698,
      "grad_norm": 1.0472897291183472,
      "learning_rate": 0.0002,
      "loss": 0.4133,
      "step": 18
    },
    {
      "epoch": 0.019603038470963,
      "eval_loss": 0.3974114656448364,
      "eval_runtime": 33.2711,
      "eval_samples_per_second": 12.924,
      "eval_steps_per_second": 6.462,
      "step": 20
    },
    {
      "epoch": 0.02058319039451115,
      "grad_norm": 0.9422470331192017,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.4808,
      "step": 21
    },
    {
      "epoch": 0.0235236461651556,
      "grad_norm": 0.7602350115776062,
      "learning_rate": 0.00015469481581224272,
      "loss": 0.3845,
      "step": 24
    },
    {
      "epoch": 0.02450379808870375,
      "eval_loss": 0.3662199378013611,
      "eval_runtime": 33.3018,
      "eval_samples_per_second": 12.912,
      "eval_steps_per_second": 6.456,
      "step": 25
    },
    {
      "epoch": 0.02646410193580005,
      "grad_norm": 0.7190614342689514,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.3378,
      "step": 27
    },
    {
      "epoch": 0.029404557706444498,
      "grad_norm": 0.6798431277275085,
      "learning_rate": 5.983045753470308e-05,
      "loss": 0.3959,
      "step": 30
    },
    {
      "epoch": 0.029404557706444498,
      "eval_loss": 0.3478722870349884,
      "eval_runtime": 33.3253,
      "eval_samples_per_second": 12.903,
      "eval_steps_per_second": 6.452,
      "step": 30
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 0.7356574535369873,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.3745,
      "step": 33
    },
    {
      "epoch": 0.03430531732418525,
      "eval_loss": 0.3429311215877533,
      "eval_runtime": 33.3237,
      "eval_samples_per_second": 12.904,
      "eval_steps_per_second": 6.452,
      "step": 35
    },
    {
      "epoch": 0.035285469247733396,
      "grad_norm": 0.6864279508590698,
      "learning_rate": 1.3638696597277679e-06,
      "loss": 0.3072,
      "step": 36
    }
  ],
  "logging_steps": 3,
  "max_steps": 37,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 18,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2895506741067776e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}