{
  "best_metric": 0.47813913226127625,
  "best_model_checkpoint": "/mnt/beegfs/farid/mlora/outputs/xnli/aya-101/ar/rank4_lr5e-5/checkpoint-6000",
  "epoch": 0.24445893089960888,
  "eval_steps": 500,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020371577574967405,
      "grad_norm": 3.1687538623809814,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1228,
      "step": 500
    },
    {
      "epoch": 0.020371577574967405,
      "eval_accuracy": 0.4465863453815261,
      "eval_f1": 0.4398462893270641,
      "eval_loss": 1.0515310764312744,
      "eval_runtime": 413.6686,
      "eval_samples_per_second": 6.019,
      "eval_steps_per_second": 0.377,
      "step": 500
    },
    {
      "epoch": 0.04074315514993481,
      "grad_norm": 10.303964614868164,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.901,
      "step": 1000
    },
    {
      "epoch": 0.04074315514993481,
      "eval_accuracy": 0.7469879518072289,
      "eval_f1": 0.7498668828149658,
      "eval_loss": 0.6499696969985962,
      "eval_runtime": 411.8975,
      "eval_samples_per_second": 6.045,
      "eval_steps_per_second": 0.379,
      "step": 1000
    },
    {
      "epoch": 0.06111473272490222,
      "grad_norm": 27.607500076293945,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6865,
      "step": 1500
    },
    {
      "epoch": 0.06111473272490222,
      "eval_accuracy": 0.7763052208835342,
      "eval_f1": 0.7760746295957563,
      "eval_loss": 0.6157358884811401,
      "eval_runtime": 411.7973,
      "eval_samples_per_second": 6.047,
      "eval_steps_per_second": 0.379,
      "step": 1500
    },
    {
      "epoch": 0.08148631029986962,
      "grad_norm": 10.66073989868164,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6189,
      "step": 2000
    },
    {
      "epoch": 0.08148631029986962,
      "eval_accuracy": 0.7799196787148595,
      "eval_f1": 0.7829493062150646,
      "eval_loss": 0.5638271570205688,
      "eval_runtime": 411.2214,
      "eval_samples_per_second": 6.055,
      "eval_steps_per_second": 0.379,
      "step": 2000
    },
    {
      "epoch": 0.10185788787483703,
      "grad_norm": 4.5490522384643555,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.621,
      "step": 2500
    },
    {
      "epoch": 0.10185788787483703,
      "eval_accuracy": 0.8012048192771084,
      "eval_f1": 0.8020131355543914,
      "eval_loss": 0.5373384952545166,
      "eval_runtime": 411.4783,
      "eval_samples_per_second": 6.051,
      "eval_steps_per_second": 0.379,
      "step": 2500
    },
    {
      "epoch": 0.12222946544980444,
      "grad_norm": 9.11120319366455,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.5901,
      "step": 3000
    },
    {
      "epoch": 0.12222946544980444,
      "eval_accuracy": 0.8132530120481928,
      "eval_f1": 0.813184387788095,
      "eval_loss": 0.5145502090454102,
      "eval_runtime": 411.6491,
      "eval_samples_per_second": 6.049,
      "eval_steps_per_second": 0.379,
      "step": 3000
    },
    {
      "epoch": 0.14260104302477183,
      "grad_norm": 7.9284515380859375,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.5831,
      "step": 3500
    },
    {
      "epoch": 0.14260104302477183,
      "eval_accuracy": 0.8068273092369478,
      "eval_f1": 0.807843707897922,
      "eval_loss": 0.5040690302848816,
      "eval_runtime": 411.4454,
      "eval_samples_per_second": 6.052,
      "eval_steps_per_second": 0.379,
      "step": 3500
    },
    {
      "epoch": 0.16297262059973924,
      "grad_norm": 5.931895732879639,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.5816,
      "step": 4000
    },
    {
      "epoch": 0.16297262059973924,
      "eval_accuracy": 0.8164658634538152,
      "eval_f1": 0.8168966682189732,
      "eval_loss": 0.49493011832237244,
      "eval_runtime": 411.9765,
      "eval_samples_per_second": 6.044,
      "eval_steps_per_second": 0.379,
      "step": 4000
    },
    {
      "epoch": 0.18334419817470665,
      "grad_norm": 5.388001918792725,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.5606,
      "step": 4500
    },
    {
      "epoch": 0.18334419817470665,
      "eval_accuracy": 0.8220883534136546,
      "eval_f1": 0.8223408197075474,
      "eval_loss": 0.48176905512809753,
      "eval_runtime": 411.5334,
      "eval_samples_per_second": 6.051,
      "eval_steps_per_second": 0.379,
      "step": 4500
    },
    {
      "epoch": 0.20371577574967406,
      "grad_norm": 8.442779541015625,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.5742,
      "step": 5000
    },
    {
      "epoch": 0.20371577574967406,
      "eval_accuracy": 0.8152610441767069,
      "eval_f1": 0.8151921057938069,
      "eval_loss": 0.49218347668647766,
      "eval_runtime": 411.5662,
      "eval_samples_per_second": 6.05,
      "eval_steps_per_second": 0.379,
      "step": 5000
    },
    {
      "epoch": 0.22408735332464147,
      "grad_norm": 4.897560119628906,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.5648,
      "step": 5500
    },
    {
      "epoch": 0.22408735332464147,
      "eval_accuracy": 0.8236947791164658,
      "eval_f1": 0.823621391191149,
      "eval_loss": 0.47885623574256897,
      "eval_runtime": 431.6447,
      "eval_samples_per_second": 5.769,
      "eval_steps_per_second": 0.361,
      "step": 5500
    },
    {
      "epoch": 0.24445893089960888,
      "grad_norm": 6.240907669067383,
      "learning_rate": 0.0,
      "loss": 0.5461,
      "step": 6000
    },
    {
      "epoch": 0.24445893089960888,
      "eval_accuracy": 0.8253012048192772,
      "eval_f1": 0.8254650982786048,
      "eval_loss": 0.47813913226127625,
      "eval_runtime": 412.3144,
      "eval_samples_per_second": 6.039,
      "eval_steps_per_second": 0.378,
      "step": 6000
    }
  ],
  "logging_steps": 500,
  "max_steps": 6000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 8.03166870528e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}