| { | |
| "best_metric": 0.8311111111111111, | |
| "best_model_checkpoint": "train_binary/05-13-2024_12:08:27/checkpoint-160", | |
| "epoch": 5.0, | |
| "eval_steps": 20, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.1047492027282715, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2022, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_acc_product": 0.2339592488915767, | |
| "eval_loss": 0.6791220903396606, | |
| "eval_overall_acc": 0.7155555555555555, | |
| "eval_runtime": 4.3626, | |
| "eval_samples_per_second": 206.3, | |
| "eval_steps_per_second": 1.834, | |
| "eval_type_descriptive_acc": 0.5418502202643172, | |
| "eval_type_expository_acc": 0.9829059829059829, | |
| "eval_type_narrative_acc": 0.6559633027522935, | |
| "eval_type_persuasive_acc": 0.669683257918552, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 9.341121673583984, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.634, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_acc_product": 0.374576132841691, | |
| "eval_loss": 0.5208084583282471, | |
| "eval_overall_acc": 0.79, | |
| "eval_runtime": 4.3469, | |
| "eval_samples_per_second": 207.046, | |
| "eval_steps_per_second": 1.84, | |
| "eval_type_descriptive_acc": 0.6519823788546255, | |
| "eval_type_expository_acc": 0.9358974358974359, | |
| "eval_type_narrative_acc": 0.8027522935779816, | |
| "eval_type_persuasive_acc": 0.7647058823529411, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 12.583885192871094, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.4102, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_acc_product": 0.37668167775444333, | |
| "eval_loss": 0.5568819046020508, | |
| "eval_overall_acc": 0.8033333333333333, | |
| "eval_runtime": 4.3493, | |
| "eval_samples_per_second": 206.93, | |
| "eval_steps_per_second": 1.839, | |
| "eval_type_descriptive_acc": 0.5374449339207048, | |
| "eval_type_expository_acc": 0.9829059829059829, | |
| "eval_type_narrative_acc": 0.8165137614678899, | |
| "eval_type_persuasive_acc": 0.8733031674208145, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 18.182939529418945, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.3891, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_acc_product": 0.3805953902407894, | |
| "eval_loss": 0.528384804725647, | |
| "eval_overall_acc": 0.8, | |
| "eval_runtime": 4.3488, | |
| "eval_samples_per_second": 206.952, | |
| "eval_steps_per_second": 1.84, | |
| "eval_type_descriptive_acc": 0.6035242290748899, | |
| "eval_type_expository_acc": 0.9914529914529915, | |
| "eval_type_narrative_acc": 0.8623853211009175, | |
| "eval_type_persuasive_acc": 0.7375565610859729, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 6.487391471862793, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.2423, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_acc_product": 0.3477393972726718, | |
| "eval_loss": 0.7446629405021667, | |
| "eval_overall_acc": 0.7888888888888889, | |
| "eval_runtime": 4.3484, | |
| "eval_samples_per_second": 206.972, | |
| "eval_steps_per_second": 1.84, | |
| "eval_type_descriptive_acc": 0.6607929515418502, | |
| "eval_type_expository_acc": 0.9358974358974359, | |
| "eval_type_narrative_acc": 0.5779816513761468, | |
| "eval_type_persuasive_acc": 0.9728506787330317, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 10.529717445373535, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.27, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_acc_product": 0.3687367678193366, | |
| "eval_loss": 0.5524727702140808, | |
| "eval_overall_acc": 0.8033333333333333, | |
| "eval_runtime": 4.361, | |
| "eval_samples_per_second": 206.376, | |
| "eval_steps_per_second": 1.834, | |
| "eval_type_descriptive_acc": 0.5154185022026432, | |
| "eval_type_expository_acc": 0.9700854700854701, | |
| "eval_type_narrative_acc": 0.7798165137614679, | |
| "eval_type_persuasive_acc": 0.9457013574660633, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 8.022017478942871, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.116, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_acc_product": 0.42496128327762595, | |
| "eval_loss": 0.666808009147644, | |
| "eval_overall_acc": 0.8177777777777778, | |
| "eval_runtime": 4.3416, | |
| "eval_samples_per_second": 207.296, | |
| "eval_steps_per_second": 1.843, | |
| "eval_type_descriptive_acc": 0.6519823788546255, | |
| "eval_type_expository_acc": 0.9786324786324786, | |
| "eval_type_narrative_acc": 0.7706422018348624, | |
| "eval_type_persuasive_acc": 0.8642533936651584, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 11.408710479736328, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1078, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_acc_product": 0.4513722367715261, | |
| "eval_loss": 0.655240535736084, | |
| "eval_overall_acc": 0.8311111111111111, | |
| "eval_runtime": 4.3513, | |
| "eval_samples_per_second": 206.837, | |
| "eval_steps_per_second": 1.839, | |
| "eval_type_descriptive_acc": 0.6740088105726872, | |
| "eval_type_expository_acc": 0.9871794871794872, | |
| "eval_type_narrative_acc": 0.7385321100917431, | |
| "eval_type_persuasive_acc": 0.918552036199095, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 8.358388900756836, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0381, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_acc_product": 0.4348685552569168, | |
| "eval_loss": 0.7252246141433716, | |
| "eval_overall_acc": 0.8222222222222222, | |
| "eval_runtime": 4.3445, | |
| "eval_samples_per_second": 207.158, | |
| "eval_steps_per_second": 1.841, | |
| "eval_type_descriptive_acc": 0.7224669603524229, | |
| "eval_type_expository_acc": 0.9871794871794872, | |
| "eval_type_narrative_acc": 0.7018348623853211, | |
| "eval_type_persuasive_acc": 0.8687782805429864, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 8.94113540649414, | |
| "learning_rate": 0.0, | |
| "loss": 0.0417, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_acc_product": 0.45515444892534596, | |
| "eval_loss": 0.7260475754737854, | |
| "eval_overall_acc": 0.83, | |
| "eval_runtime": 4.34, | |
| "eval_samples_per_second": 207.373, | |
| "eval_steps_per_second": 1.843, | |
| "eval_type_descriptive_acc": 0.7268722466960352, | |
| "eval_type_expository_acc": 0.9871794871794872, | |
| "eval_type_narrative_acc": 0.7339449541284404, | |
| "eval_type_persuasive_acc": 0.8642533936651584, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 20, | |
| "total_flos": 5941102588416000.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |