{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 11.971223021582734,
  "eval_steps": 25,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.64,
      "grad_norm": 494.05772193159083,
      "learning_rate": 9.999874838141888e-05,
      "loss": 1.7591,
      "step": 25
    },
    {
      "epoch": 0.64,
      "eval_loss": 4.927945613861084,
      "eval_runtime": 213.4346,
      "eval_samples_per_second": 46.857,
      "eval_steps_per_second": 0.736,
      "step": 25
    },
    {
      "epoch": 1.28,
      "grad_norm": 41.85279295369776,
      "learning_rate": 9.915628588978522e-05,
      "loss": 2.0299,
      "step": 50
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.818238377571106,
      "eval_runtime": 211.9322,
      "eval_samples_per_second": 47.19,
      "eval_steps_per_second": 0.741,
      "step": 50
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.844358526437997,
      "learning_rate": 9.67797005288181e-05,
      "loss": 0.6558,
      "step": 75
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.5749660134315491,
      "eval_runtime": 211.6596,
      "eval_samples_per_second": 47.25,
      "eval_steps_per_second": 0.742,
      "step": 75
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.085279119700608,
      "learning_rate": 9.294316336102132e-05,
      "loss": 0.4785,
      "step": 100
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.38230499625205994,
      "eval_runtime": 213.3859,
      "eval_samples_per_second": 46.868,
      "eval_steps_per_second": 0.736,
      "step": 100
    },
    {
      "epoch": 3.2,
      "grad_norm": 3.033775912026443,
      "learning_rate": 8.776640921382584e-05,
      "loss": 0.3837,
      "step": 125
    },
    {
      "epoch": 3.2,
      "eval_loss": 0.29413464665412903,
      "eval_runtime": 213.545,
      "eval_samples_per_second": 46.833,
      "eval_steps_per_second": 0.735,
      "step": 125
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.1383110535612262,
      "learning_rate": 8.141099986478212e-05,
      "loss": 0.3073,
      "step": 150
    },
    {
      "epoch": 3.84,
      "eval_loss": 0.23183606564998627,
      "eval_runtime": 212.4796,
      "eval_samples_per_second": 47.068,
      "eval_steps_per_second": 0.739,
      "step": 150
    },
    {
      "epoch": 4.48,
      "grad_norm": 1.081127753088496,
      "learning_rate": 7.407528184577019e-05,
      "loss": 0.2119,
      "step": 175
    },
    {
      "epoch": 4.48,
      "eval_loss": 0.18712490797042847,
      "eval_runtime": 213.5483,
      "eval_samples_per_second": 46.833,
      "eval_steps_per_second": 0.735,
      "step": 175
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.36145826033813083,
      "learning_rate": 6.598819622856227e-05,
      "loss": 0.1632,
      "step": 200
    },
    {
      "epoch": 5.12,
      "eval_loss": 0.15954019129276276,
      "eval_runtime": 211.2408,
      "eval_samples_per_second": 47.344,
      "eval_steps_per_second": 0.743,
      "step": 200
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.3381871309692255,
      "learning_rate": 5.7402133582686576e-05,
      "loss": 0.1297,
      "step": 225
    },
    {
      "epoch": 5.76,
      "eval_loss": 0.14867956936359406,
      "eval_runtime": 214.0779,
      "eval_samples_per_second": 46.717,
      "eval_steps_per_second": 0.733,
      "step": 225
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.3059190090918364,
      "learning_rate": 4.85850570958441e-05,
      "loss": 0.1035,
      "step": 250
    },
    {
      "epoch": 6.39,
      "eval_loss": 0.14759863913059235,
      "eval_runtime": 212.6859,
      "eval_samples_per_second": 47.022,
      "eval_steps_per_second": 0.738,
      "step": 250
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.28211565488590556,
      "learning_rate": 3.9812139687108815e-05,
      "loss": 0.0856,
      "step": 275
    },
    {
      "epoch": 7.03,
      "eval_loss": 0.14268410205841064,
      "eval_runtime": 210.2857,
      "eval_samples_per_second": 47.559,
      "eval_steps_per_second": 0.747,
      "step": 275
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.3745125634810898,
      "learning_rate": 3.135717611098458e-05,
      "loss": 0.0574,
      "step": 300
    },
    {
      "epoch": 7.67,
      "eval_loss": 0.14823941886425018,
      "eval_runtime": 214.2104,
      "eval_samples_per_second": 46.688,
      "eval_steps_per_second": 0.733,
      "step": 300
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.3453636602728437,
      "learning_rate": 2.3484038072721758e-05,
      "loss": 0.0448,
      "step": 325
    },
    {
      "epoch": 8.31,
      "eval_loss": 0.1552451252937317,
      "eval_runtime": 210.5548,
      "eval_samples_per_second": 47.498,
      "eval_steps_per_second": 0.746,
      "step": 325
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.28973521029613436,
      "learning_rate": 1.6438439032954855e-05,
      "loss": 0.0318,
      "step": 350
    },
    {
      "epoch": 8.95,
      "eval_loss": 0.15622578561306,
      "eval_runtime": 210.6816,
      "eval_samples_per_second": 47.47,
      "eval_steps_per_second": 0.745,
      "step": 350
    },
    {
      "epoch": 9.59,
      "grad_norm": 0.20130206793775604,
      "learning_rate": 1.0440265714600572e-05,
      "loss": 0.0196,
      "step": 375
    },
    {
      "epoch": 9.59,
      "eval_loss": 0.1708817183971405,
      "eval_runtime": 213.4033,
      "eval_samples_per_second": 46.864,
      "eval_steps_per_second": 0.736,
      "step": 375
    },
    {
      "epoch": 10.23,
      "grad_norm": 0.12286721385196743,
      "learning_rate": 5.676715638695063e-06,
      "loss": 0.0146,
      "step": 400
    },
    {
      "epoch": 10.23,
      "eval_loss": 0.1793455183506012,
      "eval_runtime": 212.2341,
      "eval_samples_per_second": 47.122,
      "eval_steps_per_second": 0.74,
      "step": 400
    },
    {
      "epoch": 10.87,
      "grad_norm": 0.07993244910369399,
      "learning_rate": 2.2964548604209213e-06,
      "loss": 0.0084,
      "step": 425
    },
    {
      "epoch": 10.87,
      "eval_loss": 0.1853875368833542,
      "eval_runtime": 213.6886,
      "eval_samples_per_second": 46.802,
      "eval_steps_per_second": 0.735,
      "step": 425
    },
    {
      "epoch": 11.51,
      "grad_norm": 0.05627841974218994,
      "learning_rate": 4.049782370561583e-07,
      "loss": 0.0058,
      "step": 450
    },
    {
      "epoch": 11.51,
      "eval_loss": 0.1918596774339676,
      "eval_runtime": 211.3759,
      "eval_samples_per_second": 47.314,
      "eval_steps_per_second": 0.743,
      "step": 450
    },
    {
      "epoch": 11.97,
      "step": 468,
      "total_flos": 3135460343808000.0,
      "train_loss": 0.346917372992915,
      "train_runtime": 79535.2013,
      "train_samples_per_second": 12.071,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 25,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 2000,
  "total_flos": 3135460343808000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}