checkpoint-648/trainer_state.json

{
  "best_metric": 4.273556709289551,
  "best_model_checkpoint": "/opt/ml/model/checkpoint-400",
  "epoch": 0.9988439306358381,
  "eval_steps": 200,
  "global_step": 648,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07707129094412331,
      "grad_norm": 2.1371634006500244,
      "learning_rate": 4.6219135802469136e-05,
      "loss": 7.0635,
      "step": 50
    },
    {
      "epoch": 0.15414258188824662,
      "grad_norm": 11.696390151977539,
      "learning_rate": 4.236111111111111e-05,
      "loss": 4.747,
      "step": 100
    },
    {
      "epoch": 0.23121387283236994,
      "grad_norm": 5.832604885101318,
      "learning_rate": 3.850308641975309e-05,
      "loss": 4.4816,
      "step": 150
    },
    {
      "epoch": 0.30828516377649323,
      "grad_norm": 2.8921449184417725,
      "learning_rate": 3.4645061728395064e-05,
      "loss": 4.444,
      "step": 200
    },
    {
      "epoch": 0.30828516377649323,
      "eval_loss": 4.3647613525390625,
      "eval_runtime": 1152.4316,
      "eval_samples_per_second": 4.504,
      "eval_steps_per_second": 0.563,
      "step": 200
    },
    {
      "epoch": 0.3853564547206166,
      "grad_norm": 4.978172302246094,
      "learning_rate": 3.0787037037037034e-05,
      "loss": 4.3124,
      "step": 250
    },
    {
      "epoch": 0.4624277456647399,
      "grad_norm": 2.087311029434204,
      "learning_rate": 2.692901234567901e-05,
      "loss": 4.309,
      "step": 300
    },
    {
      "epoch": 0.5394990366088632,
      "grad_norm": 4.583073616027832,
      "learning_rate": 2.307098765432099e-05,
      "loss": 4.3329,
      "step": 350
    },
    {
      "epoch": 0.6165703275529865,
      "grad_norm": 2.860360622406006,
      "learning_rate": 1.9212962962962962e-05,
      "loss": 4.3039,
      "step": 400
    },
    {
      "epoch": 0.6165703275529865,
      "eval_loss": 4.273556709289551,
      "eval_runtime": 1152.4004,
      "eval_samples_per_second": 4.505,
      "eval_steps_per_second": 0.563,
      "step": 400
    },
    {
      "epoch": 0.6936416184971098,
      "grad_norm": 3.887697219848633,
      "learning_rate": 1.5354938271604938e-05,
      "loss": 4.332,
      "step": 450
    },
    {
      "epoch": 0.7707129094412332,
      "grad_norm": 2.878767728805542,
      "learning_rate": 1.1496913580246914e-05,
      "loss": 4.2523,
      "step": 500
    },
    {
      "epoch": 0.8477842003853564,
      "grad_norm": 4.9191155433654785,
      "learning_rate": 7.63888888888889e-06,
      "loss": 4.2528,
      "step": 550
    },
    {
      "epoch": 0.9248554913294798,
      "grad_norm": 6.933360576629639,
      "learning_rate": 3.7808641975308645e-06,
      "loss": 4.2671,
      "step": 600
    },
    {
      "epoch": 0.9248554913294798,
      "eval_loss": 4.235105037689209,
      "eval_runtime": 1152.402,
      "eval_samples_per_second": 4.505,
      "eval_steps_per_second": 0.563,
      "step": 600
    }
  ],
  "logging_steps": 50,
  "max_steps": 648,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.785053161456927e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}