model/gpt2-355M/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.662756598240469,
  "global_step": 20000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.37,
      "learning_rate": 5e-05,
      "loss": 3.6227,
      "step": 500
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.8747494989979966e-05,
      "loss": 3.3016,
      "step": 1000
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.7494989979959916e-05,
      "loss": 3.1452,
      "step": 1500
    },
    {
      "epoch": 1.47,
      "learning_rate": 4.624248496993988e-05,
      "loss": 2.932,
      "step": 2000
    },
    {
      "epoch": 1.83,
      "learning_rate": 4.4989979959919844e-05,
      "loss": 2.925,
      "step": 2500
    },
    {
      "epoch": 2.2,
      "learning_rate": 4.37374749498998e-05,
      "loss": 2.7771,
      "step": 3000
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.248496993987976e-05,
      "loss": 2.6615,
      "step": 3500
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.123246492985972e-05,
      "loss": 2.6681,
      "step": 4000
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.997995991983968e-05,
      "loss": 2.4825,
      "step": 4500
    },
    {
      "epoch": 3.67,
      "learning_rate": 3.872745490981964e-05,
      "loss": 2.4556,
      "step": 5000
    },
    {
      "epoch": 4.03,
      "learning_rate": 3.74749498997996e-05,
      "loss": 2.4436,
      "step": 5500
    },
    {
      "epoch": 4.4,
      "learning_rate": 3.6222444889779564e-05,
      "loss": 2.2667,
      "step": 6000
    },
    {
      "epoch": 4.77,
      "learning_rate": 3.496993987975952e-05,
      "loss": 2.2716,
      "step": 6500
    },
    {
      "epoch": 5.13,
      "learning_rate": 3.371743486973948e-05,
      "loss": 2.2282,
      "step": 7000
    },
    {
      "epoch": 5.5,
      "learning_rate": 3.246492985971944e-05,
      "loss": 2.1086,
      "step": 7500
    },
    {
      "epoch": 5.87,
      "learning_rate": 3.12124248496994e-05,
      "loss": 2.1339,
      "step": 8000
    },
    {
      "epoch": 6.23,
      "learning_rate": 2.9959919839679363e-05,
      "loss": 2.0177,
      "step": 8500
    },
    {
      "epoch": 6.6,
      "learning_rate": 2.870741482965932e-05,
      "loss": 1.9853,
      "step": 9000
    },
    {
      "epoch": 6.96,
      "learning_rate": 2.745490981963928e-05,
      "loss": 1.9993,
      "step": 9500
    },
    {
      "epoch": 7.33,
      "learning_rate": 2.620240480961924e-05,
      "loss": 1.859,
      "step": 10000
    },
    {
      "epoch": 7.7,
      "learning_rate": 2.49498997995992e-05,
      "loss": 1.8696,
      "step": 10500
    },
    {
      "epoch": 8.06,
      "learning_rate": 2.3697394789579158e-05,
      "loss": 1.858,
      "step": 11000
    },
    {
      "epoch": 8.43,
      "learning_rate": 2.244488977955912e-05,
      "loss": 1.7593,
      "step": 11500
    },
    {
      "epoch": 8.8,
      "learning_rate": 2.119238476953908e-05,
      "loss": 1.7654,
      "step": 12000
    },
    {
      "epoch": 9.16,
      "learning_rate": 1.993987975951904e-05,
      "loss": 1.7326,
      "step": 12500
    },
    {
      "epoch": 9.53,
      "learning_rate": 1.8687374749499e-05,
      "loss": 1.6704,
      "step": 13000
    },
    {
      "epoch": 9.9,
      "learning_rate": 1.743486973947896e-05,
      "loss": 1.689,
      "step": 13500
    },
    {
      "epoch": 10.26,
      "learning_rate": 1.6182364729458917e-05,
      "loss": 1.6195,
      "step": 14000
    },
    {
      "epoch": 10.63,
      "learning_rate": 1.492985971943888e-05,
      "loss": 1.6111,
      "step": 14500
    },
    {
      "epoch": 11.0,
      "learning_rate": 1.3677354709418838e-05,
      "loss": 1.6198,
      "step": 15000
    },
    {
      "epoch": 11.36,
      "learning_rate": 1.2424849699398798e-05,
      "loss": 1.5393,
      "step": 15500
    },
    {
      "epoch": 11.73,
      "learning_rate": 1.1172344689378759e-05,
      "loss": 1.5502,
      "step": 16000
    },
    {
      "epoch": 12.1,
      "learning_rate": 9.919839679358718e-06,
      "loss": 1.5391,
      "step": 16500
    },
    {
      "epoch": 12.46,
      "learning_rate": 8.667334669338678e-06,
      "loss": 1.4915,
      "step": 17000
    },
    {
      "epoch": 12.83,
      "learning_rate": 7.414829659318638e-06,
      "loss": 1.507,
      "step": 17500
    },
    {
      "epoch": 13.2,
      "learning_rate": 6.162324649298597e-06,
      "loss": 1.4794,
      "step": 18000
    },
    {
      "epoch": 13.56,
      "learning_rate": 4.9098196392785576e-06,
      "loss": 1.4626,
      "step": 18500
    },
    {
      "epoch": 13.93,
      "learning_rate": 3.6573146292585176e-06,
      "loss": 1.464,
      "step": 19000
    },
    {
      "epoch": 14.3,
      "learning_rate": 2.404809619238477e-06,
      "loss": 1.4426,
      "step": 19500
    },
    {
      "epoch": 14.66,
      "learning_rate": 1.152304609218437e-06,
      "loss": 1.4393,
      "step": 20000
    }
  ],
  "max_steps": 20460,
  "num_train_epochs": 15,
  "total_flos": 43573965472923648,
  "trial_name": null,
  "trial_params": null
}