trainer_state.json

{
  "best_global_step": 100,
  "best_metric": 0.5687752366065979,
  "best_model_checkpoint": "/home/ubuntu/training-data/outputs/einstein_prefix_v3/einstein_20251223_023941/checkpoint-100",
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 153,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.19900497512437812,
      "grad_norm": 3.298532724380493,
      "learning_rate": 0.00011250000000000001,
      "loss": 1.7049,
      "step": 10
    },
    {
      "epoch": 0.39800995024875624,
      "grad_norm": 1.7755048274993896,
      "learning_rate": 0.0001956204379562044,
      "loss": 1.3648,
      "step": 20
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.9607742428779602,
      "learning_rate": 0.00018102189781021898,
      "loss": 0.9361,
      "step": 30
    },
    {
      "epoch": 0.7960199004975125,
      "grad_norm": 0.4132033586502075,
      "learning_rate": 0.0001664233576642336,
      "loss": 0.6577,
      "step": 40
    },
    {
      "epoch": 0.9950248756218906,
      "grad_norm": 0.3440336585044861,
      "learning_rate": 0.00015182481751824817,
      "loss": 0.6169,
      "step": 50
    },
    {
      "epoch": 1.1791044776119404,
      "grad_norm": 0.22007885575294495,
      "learning_rate": 0.0001372262773722628,
      "loss": 0.6068,
      "step": 60
    },
    {
      "epoch": 1.3781094527363185,
      "grad_norm": 0.280690461397171,
      "learning_rate": 0.00012262773722627738,
      "loss": 0.5801,
      "step": 70
    },
    {
      "epoch": 1.5771144278606966,
      "grad_norm": 0.3772982954978943,
      "learning_rate": 0.00010802919708029198,
      "loss": 0.5776,
      "step": 80
    },
    {
      "epoch": 1.7761194029850746,
      "grad_norm": 0.1831785887479782,
      "learning_rate": 9.343065693430657e-05,
      "loss": 0.5752,
      "step": 90
    },
    {
      "epoch": 1.9751243781094527,
      "grad_norm": 0.23180606961250305,
      "learning_rate": 7.883211678832118e-05,
      "loss": 0.5574,
      "step": 100
    },
    {
      "epoch": 1.9751243781094527,
      "eval_loss": 0.5687752366065979,
      "eval_runtime": 110.6819,
      "eval_samples_per_second": 0.407,
      "eval_steps_per_second": 0.407,
      "step": 100
    },
    {
      "epoch": 2.1592039800995027,
      "grad_norm": 0.19315272569656372,
      "learning_rate": 6.423357664233576e-05,
      "loss": 0.5573,
      "step": 110
    },
    {
      "epoch": 2.3582089552238807,
      "grad_norm": 0.20070014894008636,
      "learning_rate": 4.963503649635037e-05,
      "loss": 0.5427,
      "step": 120
    },
    {
      "epoch": 2.557213930348259,
      "grad_norm": 0.20601622760295868,
      "learning_rate": 3.5036496350364965e-05,
      "loss": 0.5611,
      "step": 130
    },
    {
      "epoch": 2.756218905472637,
      "grad_norm": 0.2782047390937805,
      "learning_rate": 2.0437956204379563e-05,
      "loss": 0.5398,
      "step": 140
    },
    {
      "epoch": 2.955223880597015,
      "grad_norm": 0.22529636323451996,
      "learning_rate": 5.839416058394161e-06,
      "loss": 0.5597,
      "step": 150
    }
  ],
  "logging_steps": 10,
  "max_steps": 153,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.478542845711483e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}