checkpoint-200/trainer_state.json

{
  "best_global_step": 200,
  "best_metric": 0.0776444673538208,
  "best_model_checkpoint": "./paligemma-clevr-finetuned/checkpoint-200",
  "epoch": 0.25396825396825395,
  "eval_steps": 100,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012698412698412698,
      "grad_norm": 6.424576759338379,
      "learning_rate": 9.898477157360407e-05,
      "loss": 7.0764,
      "step": 10
    },
    {
      "epoch": 0.025396825396825397,
      "grad_norm": 2.169109344482422,
      "learning_rate": 9.771573604060914e-05,
      "loss": 0.5921,
      "step": 20
    },
    {
      "epoch": 0.0380952380952381,
      "grad_norm": 1.0041604042053223,
      "learning_rate": 9.644670050761421e-05,
      "loss": 0.2141,
      "step": 30
    },
    {
      "epoch": 0.050793650793650794,
      "grad_norm": 1.1403199434280396,
      "learning_rate": 9.51776649746193e-05,
      "loss": 0.1424,
      "step": 40
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 1.898563265800476,
      "learning_rate": 9.390862944162437e-05,
      "loss": 0.1758,
      "step": 50
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 3.6021909713745117,
      "learning_rate": 9.263959390862943e-05,
      "loss": 0.1356,
      "step": 60
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 7.687479496002197,
      "learning_rate": 9.137055837563452e-05,
      "loss": 0.1585,
      "step": 70
    },
    {
      "epoch": 0.10158730158730159,
      "grad_norm": 1.7669498920440674,
      "learning_rate": 9.01015228426396e-05,
      "loss": 0.1537,
      "step": 80
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 0.5084421634674072,
      "learning_rate": 8.883248730964467e-05,
      "loss": 0.118,
      "step": 90
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 0.9338592886924744,
      "learning_rate": 8.756345177664976e-05,
      "loss": 0.0885,
      "step": 100
    },
    {
      "epoch": 0.12698412698412698,
      "eval_loss": 0.08724059909582138,
      "eval_runtime": 408.8329,
      "eval_samples_per_second": 3.424,
      "eval_steps_per_second": 0.856,
      "step": 100
    },
    {
      "epoch": 0.13968253968253969,
      "grad_norm": 2.7368178367614746,
      "learning_rate": 8.629441624365483e-05,
      "loss": 0.1042,
      "step": 110
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 2.3904881477355957,
      "learning_rate": 8.50253807106599e-05,
      "loss": 0.076,
      "step": 120
    },
    {
      "epoch": 0.16507936507936508,
      "grad_norm": 1.2015010118484497,
      "learning_rate": 8.375634517766498e-05,
      "loss": 0.1791,
      "step": 130
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.8636962175369263,
      "learning_rate": 8.248730964467005e-05,
      "loss": 0.0762,
      "step": 140
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 3.2262871265411377,
      "learning_rate": 8.134517766497463e-05,
      "loss": 0.1217,
      "step": 150
    },
    {
      "epoch": 0.20317460317460317,
      "grad_norm": 4.662909030914307,
      "learning_rate": 8.00761421319797e-05,
      "loss": 0.1213,
      "step": 160
    },
    {
      "epoch": 0.21587301587301588,
      "grad_norm": 0.6849185824394226,
      "learning_rate": 7.880710659898477e-05,
      "loss": 0.1106,
      "step": 170
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.5793686509132385,
      "learning_rate": 7.753807106598985e-05,
      "loss": 0.1171,
      "step": 180
    },
    {
      "epoch": 0.24126984126984127,
      "grad_norm": 3.4240646362304688,
      "learning_rate": 7.626903553299492e-05,
      "loss": 0.0528,
      "step": 190
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 1.2842432260513306,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.1149,
      "step": 200
    },
    {
      "epoch": 0.25396825396825395,
      "eval_loss": 0.0776444673538208,
      "eval_runtime": 428.189,
      "eval_samples_per_second": 3.27,
      "eval_steps_per_second": 0.817,
      "step": 200
    }
  ],
  "logging_steps": 10,
  "max_steps": 788,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.432387427328e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}