{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.2597631645250693,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "learning_rate": 0.000975,
      "loss": 1.0238,
      "step": 500
    },
    {
      "epoch": 0.06,
      "eval_loss": 0.8929525017738342,
      "eval_runtime": 4.0089,
      "eval_samples_per_second": 74.833,
      "eval_steps_per_second": 37.417,
      "step": 500
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00095,
      "loss": 0.9971,
      "step": 1000
    },
    {
      "epoch": 0.13,
      "eval_loss": 0.8510345220565796,
      "eval_runtime": 4.0089,
      "eval_samples_per_second": 74.833,
      "eval_steps_per_second": 37.417,
      "step": 1000
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.000925,
      "loss": 0.9522,
      "step": 1500
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.8565580248832703,
      "eval_runtime": 4.0278,
      "eval_samples_per_second": 74.483,
      "eval_steps_per_second": 37.241,
      "step": 1500
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0009000000000000001,
      "loss": 0.9143,
      "step": 2000
    },
    {
      "epoch": 0.25,
      "eval_loss": 0.8810315132141113,
      "eval_runtime": 4.0097,
      "eval_samples_per_second": 74.819,
      "eval_steps_per_second": 37.409,
      "step": 2000
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.000875,
      "loss": 0.9151,
      "step": 2500
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.8461073040962219,
      "eval_runtime": 4.0108,
      "eval_samples_per_second": 74.798,
      "eval_steps_per_second": 37.399,
      "step": 2500
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00085,
      "loss": 0.9263,
      "step": 3000
    },
    {
      "epoch": 0.38,
      "eval_loss": 0.8521466255187988,
      "eval_runtime": 4.0059,
      "eval_samples_per_second": 74.89,
      "eval_steps_per_second": 37.445,
      "step": 3000
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.000825,
      "loss": 0.9102,
      "step": 3500
    },
    {
      "epoch": 0.44,
      "eval_loss": 0.8521981835365295,
      "eval_runtime": 4.0413,
      "eval_samples_per_second": 74.234,
      "eval_steps_per_second": 37.117,
      "step": 3500
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0008,
      "loss": 0.8866,
      "step": 4000
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.8259674906730652,
      "eval_runtime": 4.0316,
      "eval_samples_per_second": 74.412,
      "eval_steps_per_second": 37.206,
      "step": 4000
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0007750000000000001,
      "loss": 0.911,
      "step": 4500
    },
    {
      "epoch": 0.57,
      "eval_loss": 0.8008900284767151,
      "eval_runtime": 4.0224,
      "eval_samples_per_second": 74.583,
      "eval_steps_per_second": 37.292,
      "step": 4500
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00075,
      "loss": 0.8536,
      "step": 5000
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.808663547039032,
      "eval_runtime": 5.0615,
      "eval_samples_per_second": 59.271,
      "eval_steps_per_second": 29.635,
      "step": 5000
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.000725,
      "loss": 0.8277,
      "step": 5500
    },
    {
      "epoch": 0.69,
      "eval_loss": 0.7903586626052856,
      "eval_runtime": 4.0081,
      "eval_samples_per_second": 74.849,
      "eval_steps_per_second": 37.425,
      "step": 5500
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0007,
      "loss": 0.8413,
      "step": 6000
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.8013813495635986,
      "eval_runtime": 3.994,
      "eval_samples_per_second": 75.113,
      "eval_steps_per_second": 37.556,
      "step": 6000
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.000675,
      "loss": 0.8491,
      "step": 6500
    },
    {
      "epoch": 0.82,
      "eval_loss": 0.7867017984390259,
      "eval_runtime": 4.0072,
      "eval_samples_per_second": 74.864,
      "eval_steps_per_second": 37.432,
      "step": 6500
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0006500000000000001,
      "loss": 0.8077,
      "step": 7000
    },
    {
      "epoch": 0.88,
      "eval_loss": 0.759165346622467,
      "eval_runtime": 4.0166,
      "eval_samples_per_second": 74.691,
      "eval_steps_per_second": 37.345,
      "step": 7000
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.000625,
      "loss": 0.796,
      "step": 7500
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.7527943849563599,
      "eval_runtime": 3.998,
      "eval_samples_per_second": 75.037,
      "eval_steps_per_second": 37.518,
      "step": 7500
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0006,
      "loss": 0.7952,
      "step": 8000
    },
    {
      "epoch": 1.01,
      "eval_loss": 0.7407393455505371,
      "eval_runtime": 3.9421,
      "eval_samples_per_second": 76.102,
      "eval_steps_per_second": 38.051,
      "step": 8000
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.000575,
      "loss": 0.5761,
      "step": 8500
    },
    {
      "epoch": 1.07,
      "eval_loss": 0.7574229836463928,
      "eval_runtime": 3.9855,
      "eval_samples_per_second": 75.273,
      "eval_steps_per_second": 37.636,
      "step": 8500
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00055,
      "loss": 0.6084,
      "step": 9000
    },
    {
      "epoch": 1.13,
      "eval_loss": 0.7349967956542969,
      "eval_runtime": 3.995,
      "eval_samples_per_second": 75.093,
      "eval_steps_per_second": 37.547,
      "step": 9000
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0005250000000000001,
      "loss": 0.5863,
      "step": 9500
    },
    {
      "epoch": 1.2,
      "eval_loss": 0.7407773733139038,
      "eval_runtime": 3.9874,
      "eval_samples_per_second": 75.238,
      "eval_steps_per_second": 37.619,
      "step": 9500
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0005,
      "loss": 0.5835,
      "step": 10000
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.719527006149292,
      "eval_runtime": 3.9471,
      "eval_samples_per_second": 76.006,
      "eval_steps_per_second": 38.003,
      "step": 10000
    }
  ],
  "logging_steps": 500,
  "max_steps": 20000,
  "num_train_epochs": 3,
  "save_steps": 5000,
  "total_flos": 1.773098919161856e+16,
  "trial_name": null,
  "trial_params": null
}