trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9645776566757496,
  "eval_steps": 500,
  "global_step": 135,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21798365122615804,
      "grad_norm": 2.1800439383266883,
      "learning_rate": 5e-06,
      "loss": 1.0138,
      "step": 10
    },
    {
      "epoch": 0.4359673024523161,
      "grad_norm": 2.321066726127001,
      "learning_rate": 5e-06,
      "loss": 0.9311,
      "step": 20
    },
    {
      "epoch": 0.6539509536784741,
      "grad_norm": 1.2528230223771615,
      "learning_rate": 5e-06,
      "loss": 0.902,
      "step": 30
    },
    {
      "epoch": 0.8719346049046321,
      "grad_norm": 1.3255206188515178,
      "learning_rate": 5e-06,
      "loss": 0.8817,
      "step": 40
    },
    {
      "epoch": 0.9809264305177112,
      "eval_loss": 0.8728658556938171,
      "eval_runtime": 32.8266,
      "eval_samples_per_second": 37.652,
      "eval_steps_per_second": 0.609,
      "step": 45
    },
    {
      "epoch": 1.1008174386920981,
      "grad_norm": 1.1084771353250091,
      "learning_rate": 5e-06,
      "loss": 0.9119,
      "step": 50
    },
    {
      "epoch": 1.318801089918256,
      "grad_norm": 1.1331211841837239,
      "learning_rate": 5e-06,
      "loss": 0.8201,
      "step": 60
    },
    {
      "epoch": 1.5367847411444142,
      "grad_norm": 0.7209847897858535,
      "learning_rate": 5e-06,
      "loss": 0.8167,
      "step": 70
    },
    {
      "epoch": 1.7547683923705724,
      "grad_norm": 0.6424432226183309,
      "learning_rate": 5e-06,
      "loss": 0.8136,
      "step": 80
    },
    {
      "epoch": 1.9727520435967303,
      "grad_norm": 0.6477980910912198,
      "learning_rate": 5e-06,
      "loss": 0.8115,
      "step": 90
    },
    {
      "epoch": 1.9727520435967303,
      "eval_loss": 0.8485425710678101,
      "eval_runtime": 32.0874,
      "eval_samples_per_second": 38.52,
      "eval_steps_per_second": 0.623,
      "step": 90
    },
    {
      "epoch": 2.2016348773841963,
      "grad_norm": 0.9029436230736683,
      "learning_rate": 5e-06,
      "loss": 0.8127,
      "step": 100
    },
    {
      "epoch": 2.4196185286103544,
      "grad_norm": 0.6308547054658898,
      "learning_rate": 5e-06,
      "loss": 0.746,
      "step": 110
    },
    {
      "epoch": 2.637602179836512,
      "grad_norm": 0.6233413271157902,
      "learning_rate": 5e-06,
      "loss": 0.7542,
      "step": 120
    },
    {
      "epoch": 2.8555858310626703,
      "grad_norm": 0.6392219564241265,
      "learning_rate": 5e-06,
      "loss": 0.7547,
      "step": 130
    },
    {
      "epoch": 2.9645776566757496,
      "eval_loss": 0.8474076986312866,
      "eval_runtime": 32.1945,
      "eval_samples_per_second": 38.392,
      "eval_steps_per_second": 0.621,
      "step": 135
    },
    {
      "epoch": 2.9645776566757496,
      "step": 135,
      "total_flos": 225920648478720.0,
      "train_loss": 0.840225980899952,
      "train_runtime": 4572.4428,
      "train_samples_per_second": 15.402,
      "train_steps_per_second": 0.03
    }
  ],
  "logging_steps": 10,
  "max_steps": 135,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 225920648478720.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}