checkpoint-165/trainer_state.json

{
  "best_global_step": 150,
  "best_metric": 3.6032822132110596,
  "best_model_checkpoint": "./shridhar_multilingual_retrained/checkpoint-150",
  "epoch": 15.0,
  "eval_steps": 50,
  "global_step": 165,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 22.509538650512695,
      "learning_rate": 9.000000000000001e-07,
      "loss": 4.445,
      "step": 10
    },
    {
      "epoch": 1.8470588235294119,
      "grad_norm": 17.885433197021484,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 4.6913,
      "step": 20
    },
    {
      "epoch": 2.7529411764705882,
      "grad_norm": 17.10190200805664,
      "learning_rate": 2.9e-06,
      "loss": 4.5176,
      "step": 30
    },
    {
      "epoch": 3.6588235294117646,
      "grad_norm": 17.561813354492188,
      "learning_rate": 3.900000000000001e-06,
      "loss": 4.2982,
      "step": 40
    },
    {
      "epoch": 4.564705882352941,
      "grad_norm": 16.954017639160156,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 3.9415,
      "step": 50
    },
    {
      "epoch": 4.564705882352941,
      "eval_loss": 4.784646511077881,
      "eval_runtime": 0.6453,
      "eval_samples_per_second": 15.497,
      "eval_steps_per_second": 15.497,
      "step": 50
    },
    {
      "epoch": 5.470588235294118,
      "grad_norm": 18.607324600219727,
      "learning_rate": 5.9e-06,
      "loss": 4.3359,
      "step": 60
    },
    {
      "epoch": 6.376470588235295,
      "grad_norm": 11.686802864074707,
      "learning_rate": 6.9e-06,
      "loss": 4.1427,
      "step": 70
    },
    {
      "epoch": 7.2823529411764705,
      "grad_norm": 13.75390625,
      "learning_rate": 7.9e-06,
      "loss": 3.7839,
      "step": 80
    },
    {
      "epoch": 8.188235294117646,
      "grad_norm": 12.081093788146973,
      "learning_rate": 8.900000000000001e-06,
      "loss": 3.94,
      "step": 90
    },
    {
      "epoch": 9.094117647058823,
      "grad_norm": 14.128493309020996,
      "learning_rate": 9.9e-06,
      "loss": 3.5581,
      "step": 100
    },
    {
      "epoch": 9.094117647058823,
      "eval_loss": 4.216272830963135,
      "eval_runtime": 0.5662,
      "eval_samples_per_second": 17.662,
      "eval_steps_per_second": 17.662,
      "step": 100
    },
    {
      "epoch": 10.0,
      "grad_norm": 9.035210609436035,
      "learning_rate": 1.0900000000000002e-05,
      "loss": 3.4935,
      "step": 110
    },
    {
      "epoch": 10.941176470588236,
      "grad_norm": 6.492658615112305,
      "learning_rate": 1.1900000000000001e-05,
      "loss": 3.4656,
      "step": 120
    },
    {
      "epoch": 11.847058823529412,
      "grad_norm": 7.675276279449463,
      "learning_rate": 1.2900000000000002e-05,
      "loss": 3.2714,
      "step": 130
    },
    {
      "epoch": 12.75294117647059,
      "grad_norm": 7.906505107879639,
      "learning_rate": 1.39e-05,
      "loss": 3.4237,
      "step": 140
    },
    {
      "epoch": 13.658823529411764,
      "grad_norm": 4.878476619720459,
      "learning_rate": 1.4900000000000001e-05,
      "loss": 3.1148,
      "step": 150
    },
    {
      "epoch": 13.658823529411764,
      "eval_loss": 3.6032822132110596,
      "eval_runtime": 0.5923,
      "eval_samples_per_second": 16.885,
      "eval_steps_per_second": 16.885,
      "step": 150
    },
    {
      "epoch": 14.564705882352941,
      "grad_norm": 5.000204563140869,
      "learning_rate": 1.5900000000000004e-05,
      "loss": 3.2597,
      "step": 160
    }
  ],
  "logging_steps": 10,
  "max_steps": 165,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 716754847334400.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}