{ "best_metric": NaN, "best_model_checkpoint": "miner_id_24/checkpoint-25", "epoch": 0.3831417624521073, "eval_steps": 25, "global_step": 50, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.007662835249042145, "grad_norm": NaN, "learning_rate": 5e-05, "loss": 0.0, "step": 1 }, { "epoch": 0.007662835249042145, "eval_loss": NaN, "eval_runtime": 1.8407, "eval_samples_per_second": 27.164, "eval_steps_per_second": 7.063, "step": 1 }, { "epoch": 0.01532567049808429, "grad_norm": NaN, "learning_rate": 0.0001, "loss": 0.0, "step": 2 }, { "epoch": 0.022988505747126436, "grad_norm": NaN, "learning_rate": 9.990365154573717e-05, "loss": 0.0, "step": 3 }, { "epoch": 0.03065134099616858, "grad_norm": NaN, "learning_rate": 9.961501876182148e-05, "loss": 0.0, "step": 4 }, { "epoch": 0.038314176245210725, "grad_norm": NaN, "learning_rate": 9.913533761814537e-05, "loss": 0.0, "step": 5 }, { "epoch": 0.04597701149425287, "grad_norm": NaN, "learning_rate": 9.846666218300807e-05, "loss": 0.0, "step": 6 }, { "epoch": 0.05363984674329502, "grad_norm": NaN, "learning_rate": 9.761185582727977e-05, "loss": 0.0, "step": 7 }, { "epoch": 0.06130268199233716, "grad_norm": NaN, "learning_rate": 9.657457896300791e-05, "loss": 0.0, "step": 8 }, { "epoch": 0.06896551724137931, "grad_norm": NaN, "learning_rate": 9.535927336897098e-05, "loss": 0.0, "step": 9 }, { "epoch": 0.07662835249042145, "grad_norm": NaN, "learning_rate": 9.397114317029975e-05, "loss": 0.0, "step": 10 }, { "epoch": 0.0842911877394636, "grad_norm": NaN, "learning_rate": 9.241613255361455e-05, "loss": 0.0, "step": 11 }, { "epoch": 0.09195402298850575, "grad_norm": NaN, "learning_rate": 9.070090031310558e-05, "loss": 0.0, "step": 12 }, { "epoch": 0.09961685823754789, "grad_norm": NaN, "learning_rate": 8.883279133655399e-05, "loss": 0.0, "step": 13 }, { "epoch": 0.10727969348659004, "grad_norm": NaN, "learning_rate": 8.681980515339464e-05, "loss": 0.0, "step": 14 }, { "epoch": 0.11494252873563218, "grad_norm": NaN, "learning_rate": 8.467056167950311e-05, "loss": 0.0, "step": 15 }, { "epoch": 0.12260536398467432, "grad_norm": NaN, "learning_rate": 8.239426430539243e-05, "loss": 0.0, "step": 16 }, { "epoch": 0.13026819923371646, "grad_norm": NaN, "learning_rate": 8.000066048588211e-05, "loss": 0.0, "step": 17 }, { "epoch": 0.13793103448275862, "grad_norm": NaN, "learning_rate": 7.75e-05, "loss": 0.0, "step": 18 }, { "epoch": 0.14559386973180077, "grad_norm": NaN, "learning_rate": 7.490299105985507e-05, "loss": 0.0, "step": 19 }, { "epoch": 0.1532567049808429, "grad_norm": NaN, "learning_rate": 7.222075445642904e-05, "loss": 0.0, "step": 20 }, { "epoch": 0.16091954022988506, "grad_norm": NaN, "learning_rate": 6.946477593864228e-05, "loss": 0.0, "step": 21 }, { "epoch": 0.1685823754789272, "grad_norm": NaN, "learning_rate": 6.664685702961344e-05, "loss": 0.0, "step": 22 }, { "epoch": 0.17624521072796934, "grad_norm": NaN, "learning_rate": 6.377906449072578e-05, "loss": 0.0, "step": 23 }, { "epoch": 0.1839080459770115, "grad_norm": NaN, "learning_rate": 6.087367864990233e-05, "loss": 0.0, "step": 24 }, { "epoch": 0.19157088122605365, "grad_norm": NaN, "learning_rate": 5.794314081535644e-05, "loss": 0.0, "step": 25 }, { "epoch": 0.19157088122605365, "eval_loss": NaN, "eval_runtime": 1.3267, "eval_samples_per_second": 37.686, "eval_steps_per_second": 9.798, "step": 25 }, { "epoch": 0.19923371647509577, "grad_norm": NaN, "learning_rate": 5.500000000000001e-05, "loss": 0.0, "step": 26 }, { "epoch": 0.20689655172413793, "grad_norm": NaN, "learning_rate": 5.205685918464356e-05, "loss": 0.0, "step": 27 }, { "epoch": 0.21455938697318008, "grad_norm": NaN, "learning_rate": 4.912632135009769e-05, "loss": 0.0, "step": 28 }, { "epoch": 0.2222222222222222, "grad_norm": NaN, "learning_rate": 4.6220935509274235e-05, "loss": 0.0, "step": 29 }, { "epoch": 0.22988505747126436, "grad_norm": NaN, "learning_rate": 4.3353142970386564e-05, "loss": 0.0, "step": 30 }, { "epoch": 0.23754789272030652, "grad_norm": NaN, "learning_rate": 4.053522406135775e-05, "loss": 0.0, "step": 31 }, { "epoch": 0.24521072796934865, "grad_norm": NaN, "learning_rate": 3.777924554357096e-05, "loss": 0.0, "step": 32 }, { "epoch": 0.25287356321839083, "grad_norm": NaN, "learning_rate": 3.509700894014496e-05, "loss": 0.0, "step": 33 }, { "epoch": 0.26053639846743293, "grad_norm": NaN, "learning_rate": 3.250000000000001e-05, "loss": 0.0, "step": 34 }, { "epoch": 0.2681992337164751, "grad_norm": NaN, "learning_rate": 2.9999339514117912e-05, "loss": 0.0, "step": 35 }, { "epoch": 0.27586206896551724, "grad_norm": NaN, "learning_rate": 2.760573569460757e-05, "loss": 0.0, "step": 36 }, { "epoch": 0.2835249042145594, "grad_norm": NaN, "learning_rate": 2.53294383204969e-05, "loss": 0.0, "step": 37 }, { "epoch": 0.29118773946360155, "grad_norm": NaN, "learning_rate": 2.3180194846605367e-05, "loss": 0.0, "step": 38 }, { "epoch": 0.2988505747126437, "grad_norm": NaN, "learning_rate": 2.1167208663446025e-05, "loss": 0.0, "step": 39 }, { "epoch": 0.3065134099616858, "grad_norm": NaN, "learning_rate": 1.9299099686894423e-05, "loss": 0.0, "step": 40 }, { "epoch": 0.31417624521072796, "grad_norm": NaN, "learning_rate": 1.758386744638546e-05, "loss": 0.0, "step": 41 }, { "epoch": 0.3218390804597701, "grad_norm": NaN, "learning_rate": 1.602885682970026e-05, "loss": 0.0, "step": 42 }, { "epoch": 0.32950191570881227, "grad_norm": NaN, "learning_rate": 1.464072663102903e-05, "loss": 0.0, "step": 43 }, { "epoch": 0.3371647509578544, "grad_norm": NaN, "learning_rate": 1.3425421036992098e-05, "loss": 0.0, "step": 44 }, { "epoch": 0.3448275862068966, "grad_norm": NaN, "learning_rate": 1.2388144172720251e-05, "loss": 0.0, "step": 45 }, { "epoch": 0.3524904214559387, "grad_norm": NaN, "learning_rate": 1.1533337816991932e-05, "loss": 0.0, "step": 46 }, { "epoch": 0.36015325670498083, "grad_norm": NaN, "learning_rate": 1.0864662381854632e-05, "loss": 0.0, "step": 47 }, { "epoch": 0.367816091954023, "grad_norm": NaN, "learning_rate": 1.0384981238178534e-05, "loss": 0.0, "step": 48 }, { "epoch": 0.37547892720306514, "grad_norm": NaN, "learning_rate": 1.0096348454262845e-05, "loss": 0.0, "step": 49 }, { "epoch": 0.3831417624521073, "grad_norm": NaN, "learning_rate": 1e-05, "loss": 0.0, "step": 50 }, { "epoch": 0.3831417624521073, "eval_loss": NaN, "eval_runtime": 1.3111, "eval_samples_per_second": 38.135, "eval_steps_per_second": 9.915, "step": 50 } ], "logging_steps": 1, "max_steps": 50, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 25, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 1, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 1 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.29516617859072e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }