checkpoint-206/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9969751966122202,
  "eval_steps": 100.0,
  "global_step": 206,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04839685420447671,
      "grad_norm": 1.21875,
      "learning_rate": 0.00019883937098250963,
      "loss": 1.4872,
      "step": 10
    },
    {
      "epoch": 0.09679370840895342,
      "grad_norm": 2.359375,
      "learning_rate": 0.00019538442512436328,
      "loss": 0.9191,
      "step": 20
    },
    {
      "epoch": 0.14519056261343014,
      "grad_norm": 2.578125,
      "learning_rate": 0.00018971536063389744,
      "loss": 0.7246,
      "step": 30
    },
    {
      "epoch": 0.19358741681790684,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00018196377112610526,
      "loss": 0.6294,
      "step": 40
    },
    {
      "epoch": 0.24198427102238354,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00017230959099527512,
      "loss": 0.5323,
      "step": 50
    },
    {
      "epoch": 0.29038112522686027,
      "grad_norm": 0.361328125,
      "learning_rate": 0.00016097691867340545,
      "loss": 0.5302,
      "step": 60
    },
    {
      "epoch": 0.338777979431337,
      "grad_norm": 0.287109375,
      "learning_rate": 0.00014822881472734562,
      "loss": 0.4468,
      "step": 70
    },
    {
      "epoch": 0.3871748336358137,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00013436119554425133,
      "loss": 0.4287,
      "step": 80
    },
    {
      "epoch": 0.4355716878402904,
      "grad_norm": 0.35546875,
      "learning_rate": 0.00011969596434867063,
      "loss": 0.4989,
      "step": 90
    },
    {
      "epoch": 0.4839685420447671,
      "grad_norm": 0.408203125,
      "learning_rate": 0.00010457353899807946,
      "loss": 0.4332,
      "step": 100
    },
    {
      "epoch": 0.5323653962492438,
      "grad_norm": 0.283203125,
      "learning_rate": 8.93449500060124e-05,
      "loss": 0.4339,
      "step": 110
    },
    {
      "epoch": 0.5807622504537205,
      "grad_norm": 0.42578125,
      "learning_rate": 7.436369221806201e-05,
      "loss": 0.4827,
      "step": 120
    },
    {
      "epoch": 0.6291591046581972,
      "grad_norm": 0.30078125,
      "learning_rate": 5.9977519284372194e-05,
      "loss": 0.4237,
      "step": 130
    },
    {
      "epoch": 0.677555958862674,
      "grad_norm": 0.2890625,
      "learning_rate": 4.652037140009259e-05,
      "loss": 0.414,
      "step": 140
    },
    {
      "epoch": 0.7259528130671506,
      "grad_norm": 0.275390625,
      "learning_rate": 3.430462369176619e-05,
      "loss": 0.3736,
      "step": 150
    },
    {
      "epoch": 0.7743496672716274,
      "grad_norm": 0.267578125,
      "learning_rate": 2.3613835184605525e-05,
      "loss": 0.3812,
      "step": 160
    },
    {
      "epoch": 0.822746521476104,
      "grad_norm": 0.361328125,
      "learning_rate": 1.4696166665835853e-05,
      "loss": 0.4415,
      "step": 170
    },
    {
      "epoch": 0.8711433756805808,
      "grad_norm": 0.189453125,
      "learning_rate": 7.758620232482084e-06,
      "loss": 0.4423,
      "step": 180
    },
    {
      "epoch": 0.9195402298850575,
      "grad_norm": 0.2890625,
      "learning_rate": 2.9622342385589254e-06,
      "loss": 0.4125,
      "step": 190
    },
    {
      "epoch": 0.9679370840895342,
      "grad_norm": 0.31640625,
      "learning_rate": 4.1834517933907467e-07,
      "loss": 0.4855,
      "step": 200
    }
  ],
  "logging_steps": 10,
  "max_steps": 206,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.488159952764928e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}