last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.6237152218818665,
      "learning_rate": 4e-05,
      "loss": 2.4096,
      "step": 100
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.4872891902923584,
      "learning_rate": 8e-05,
      "loss": 1.5214,
      "step": 200
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.8721675872802734,
      "learning_rate": 0.00012,
      "loss": 0.9786,
      "step": 300
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5907843708992004,
      "eval_runtime": 331.9763,
      "eval_samples_per_second": 2.109,
      "eval_steps_per_second": 0.527,
      "step": 350
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 1.8639676570892334,
      "learning_rate": 0.00016,
      "loss": 0.5839,
      "step": 400
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.6524343490600586,
      "learning_rate": 0.0002,
      "loss": 0.4143,
      "step": 500
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 1.1418366432189941,
      "learning_rate": 0.00019333333333333333,
      "loss": 0.3445,
      "step": 600
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0962567329406738,
      "learning_rate": 0.0001866666666666667,
      "loss": 0.3012,
      "step": 700
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2971023917198181,
      "eval_runtime": 331.597,
      "eval_samples_per_second": 2.111,
      "eval_steps_per_second": 0.528,
      "step": 700
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.3611701726913452,
      "learning_rate": 0.00018,
      "loss": 0.2614,
      "step": 800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.40023404359817505,
      "learning_rate": 0.00017333333333333334,
      "loss": 0.2425,
      "step": 900
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.463142991065979,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.2351,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.24983520805835724,
      "eval_runtime": 331.8028,
      "eval_samples_per_second": 2.11,
      "eval_steps_per_second": 0.527,
      "step": 1050
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.3725013732910156,
      "learning_rate": 0.00016,
      "loss": 0.2173,
      "step": 1100
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.3499898612499237,
      "learning_rate": 0.00015333333333333334,
      "loss": 0.216,
      "step": 1200
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.369825541973114,
      "learning_rate": 0.00014666666666666666,
      "loss": 0.2049,
      "step": 1300
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.28506389260292053,
      "learning_rate": 0.00014,
      "loss": 0.2089,
      "step": 1400
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.24056243896484375,
      "eval_runtime": 331.7171,
      "eval_samples_per_second": 2.11,
      "eval_steps_per_second": 0.528,
      "step": 1400
    }
  ],
  "logging_steps": 100,
  "max_steps": 3500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.435572847915827e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}