trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 10005,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 1.5337328910827637,
      "learning_rate": 1.900049975012494e-05,
      "loss": 0.4281,
      "step": 500
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.992353081703186,
      "learning_rate": 1.8000999500249876e-05,
      "loss": 0.2656,
      "step": 1000
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4741365909576416,
      "learning_rate": 1.7001499250374815e-05,
      "loss": 0.263,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5607998967170715,
      "learning_rate": 1.600199900049975e-05,
      "loss": 0.2228,
      "step": 2000
    },
    {
      "epoch": 1.25,
      "grad_norm": 56.73757553100586,
      "learning_rate": 1.500249875062469e-05,
      "loss": 0.1552,
      "step": 2500
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.09034252166748,
      "learning_rate": 1.4002998500749626e-05,
      "loss": 0.1395,
      "step": 3000
    },
    {
      "epoch": 1.75,
      "grad_norm": 49.66099166870117,
      "learning_rate": 1.3003498250874565e-05,
      "loss": 0.1396,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.33147394657135,
      "learning_rate": 1.2003998000999502e-05,
      "loss": 0.1468,
      "step": 4000
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.00538645638152957,
      "learning_rate": 1.100449775112444e-05,
      "loss": 0.0595,
      "step": 4500
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.2792365550994873,
      "learning_rate": 1.0004997501249377e-05,
      "loss": 0.0529,
      "step": 5000
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.09408382326364517,
      "learning_rate": 9.005497251374314e-06,
      "loss": 0.0578,
      "step": 5500
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.005479688756167889,
      "learning_rate": 8.005997001499251e-06,
      "loss": 0.0523,
      "step": 6000
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.0009208513656631112,
      "learning_rate": 7.006496751624189e-06,
      "loss": 0.0145,
      "step": 6500
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.2281578779220581,
      "learning_rate": 6.006996501749126e-06,
      "loss": 0.0129,
      "step": 7000
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.005618613678961992,
      "learning_rate": 5.0074962518740634e-06,
      "loss": 0.0162,
      "step": 7500
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0017482700059190392,
      "learning_rate": 4.0079960019990015e-06,
      "loss": 0.0212,
      "step": 8000
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.0020746004302054644,
      "learning_rate": 3.008495752123938e-06,
      "loss": 0.0045,
      "step": 8500
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.0015063141472637653,
      "learning_rate": 2.008995502248876e-06,
      "loss": 0.006,
      "step": 9000
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.006012303754687309,
      "learning_rate": 1.0094952523738132e-06,
      "loss": 0.0064,
      "step": 9500
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0025774994865059853,
      "learning_rate": 9.995002498750625e-09,
      "loss": 0.0067,
      "step": 10000
    },
    {
      "epoch": 5.0,
      "step": 10005,
      "total_flos": 1.863979218076416e+16,
      "train_loss": 0.10352551692517793,
      "train_runtime": 13226.2571,
      "train_samples_per_second": 6.049,
      "train_steps_per_second": 0.756
    }
  ],
  "logging_steps": 500,
  "max_steps": 10005,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.863979218076416e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}