trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.08651117724409994,
  "eval_steps": 100000000,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 10.7586,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 5e-05,
      "loss": 6.5292,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.99998214600225e-05,
      "loss": 4.7262,
      "step": 1000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.9999285842640105e-05,
      "loss": 4.1997,
      "step": 1500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.999839315550315e-05,
      "loss": 3.8892,
      "step": 2000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9997143411362066e-05,
      "loss": 3.6618,
      "step": 2500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9995536628067196e-05,
      "loss": 3.5132,
      "step": 3000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9993572828568536e-05,
      "loss": 3.4092,
      "step": 3500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.999125204091544e-05,
      "loss": 3.3295,
      "step": 4000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.998857429825615e-05,
      "loss": 3.2596,
      "step": 4500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9985539638837424e-05,
      "loss": 3.2068,
      "step": 5000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.998214810600389e-05,
      "loss": 3.163,
      "step": 5500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.997839974819749e-05,
      "loss": 3.1256,
      "step": 6000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.997429461895675e-05,
      "loss": 3.0844,
      "step": 6500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.996983277691606e-05,
      "loss": 3.0541,
      "step": 7000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.996501428580478e-05,
      "loss": 3.0293,
      "step": 7500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.995983921444638e-05,
      "loss": 2.999,
      "step": 8000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.9954307636757434e-05,
      "loss": 2.9744,
      "step": 8500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.994841963174656e-05,
      "loss": 2.9525,
      "step": 9000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.9942175283513294e-05,
      "loss": 2.9321,
      "step": 9500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.993557468124691e-05,
      "loss": 2.9237,
      "step": 10000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.9928617919225106e-05,
      "loss": 2.9021,
      "step": 10500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.99213050968127e-05,
      "loss": 2.8766,
      "step": 11000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.991363631846019e-05,
      "loss": 2.8691,
      "step": 11500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.990561169370224e-05,
      "loss": 2.8531,
      "step": 12000
    }
  ],
  "logging_steps": 500,
  "max_steps": 416130,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 3000,
  "total_flos": 4.394858840064e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}