{
  "best_global_step": 2090,
  "best_metric": 0.26646581292152405,
  "best_model_checkpoint": "/media/user/Expansion1/multilingual-e5-small-aligned-v2-conversation-refusal/checkpoint-2090",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 5225,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 14.723045349121094,
      "learning_rate": 4.522488038277512e-05,
      "loss": 0.3758,
      "num_input_tokens_seen": 512000,
      "step": 500,
      "train_runtime": 13.5893,
      "train_tokens_per_second": 37676.655
    },
    {
      "epoch": 0.9569377990430622,
      "grad_norm": 1.9174224138259888,
      "learning_rate": 4.044019138755981e-05,
      "loss": 0.308,
      "num_input_tokens_seen": 1024000,
      "step": 1000,
      "train_runtime": 26.8861,
      "train_tokens_per_second": 38086.661
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9023456199138343,
      "eval_loss": 0.29579150676727295,
      "eval_runtime": 1.1525,
      "eval_samples_per_second": 1812.63,
      "eval_steps_per_second": 227.338,
      "num_input_tokens_seen": 1069440,
      "step": 1045
    },
    {
      "epoch": 1.4354066985645932,
      "grad_norm": 12.026004791259766,
      "learning_rate": 3.56555023923445e-05,
      "loss": 0.2384,
      "num_input_tokens_seen": 1535360,
      "step": 1500,
      "train_runtime": 42.6862,
      "train_tokens_per_second": 35968.498
    },
    {
      "epoch": 1.9138755980861244,
      "grad_norm": 3.3293282985687256,
      "learning_rate": 3.0870813397129186e-05,
      "loss": 0.2253,
      "num_input_tokens_seen": 2047360,
      "step": 2000,
      "train_runtime": 56.0083,
      "train_tokens_per_second": 36554.566
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9152704643370033,
      "eval_loss": 0.26646581292152405,
      "eval_runtime": 1.1367,
      "eval_samples_per_second": 1837.79,
      "eval_steps_per_second": 230.493,
      "num_input_tokens_seen": 2138880,
      "step": 2090
    },
    {
      "epoch": 2.3923444976076556,
      "grad_norm": 1.1071521043777466,
      "learning_rate": 2.6086124401913874e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 2558720,
      "step": 2500,
      "train_runtime": 71.8393,
      "train_tokens_per_second": 35617.282
    },
    {
      "epoch": 2.8708133971291865,
      "grad_norm": 2.7163774967193604,
      "learning_rate": 2.1301435406698568e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 3070720,
      "step": 3000,
      "train_runtime": 85.1841,
      "train_tokens_per_second": 36048.061
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9128769746290091,
      "eval_loss": 0.3344113528728485,
      "eval_runtime": 1.1569,
      "eval_samples_per_second": 1805.706,
      "eval_steps_per_second": 226.47,
      "num_input_tokens_seen": 3208320,
      "step": 3135
    },
    {
      "epoch": 3.349282296650718,
      "grad_norm": 0.12201690673828125,
      "learning_rate": 1.6516746411483256e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 3582080,
      "step": 3500,
      "train_runtime": 101.233,
      "train_tokens_per_second": 35384.512
    },
    {
      "epoch": 3.827751196172249,
      "grad_norm": 0.03180946409702301,
      "learning_rate": 1.1732057416267943e-05,
      "loss": 0.1058,
      "num_input_tokens_seen": 4094080,
      "step": 4000,
      "train_runtime": 114.5749,
      "train_tokens_per_second": 35732.797
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9009095260890378,
      "eval_loss": 0.4442259669303894,
      "eval_runtime": 1.1377,
      "eval_samples_per_second": 1836.223,
      "eval_steps_per_second": 230.297,
      "num_input_tokens_seen": 4277760,
      "step": 4180
    },
    {
      "epoch": 4.30622009569378,
      "grad_norm": 0.1291642189025879,
      "learning_rate": 6.9473684210526315e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 4605440,
      "step": 4500,
      "train_runtime": 130.6244,
      "train_tokens_per_second": 35257.12
    },
    {
      "epoch": 4.784688995215311,
      "grad_norm": 0.04361966624855995,
      "learning_rate": 2.1626794258373207e-06,
      "loss": 0.0575,
      "num_input_tokens_seen": 5117440,
      "step": 5000,
      "train_runtime": 143.9296,
      "train_tokens_per_second": 35555.165
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8985160363810436,
      "eval_loss": 0.5183091163635254,
      "eval_runtime": 1.1353,
      "eval_samples_per_second": 1840.123,
      "eval_steps_per_second": 230.786,
      "num_input_tokens_seen": 5347200,
      "step": 5225
    },
    {
      "epoch": 5.0,
      "num_input_tokens_seen": 5347200,
      "step": 5225,
      "total_flos": 687958325107200.0,
      "train_loss": 0.17576850672087602,
      "train_runtime": 152.9681,
      "train_samples_per_second": 273.096,
      "train_steps_per_second": 34.157
    }
  ],
  "logging_steps": 500,
  "max_steps": 5225,
  "num_input_tokens_seen": 5347200,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 687958325107200.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}