| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 28, | |
| "global_step": 280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 3e-05, | |
| "loss": 3.5167, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7034, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3e-05, | |
| "loss": 2.5409, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4931, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_accuracy": 0.5748751060189339, | |
| "eval_loss": 2.390355348587036, | |
| "eval_runtime": 25.7729, | |
| "eval_samples_per_second": 34.261, | |
| "eval_steps_per_second": 2.173, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_bleu": 2.6290615444017718, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4944, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4832, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4336, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3e-05, | |
| "loss": 2.5101, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4699, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6049526162776446, | |
| "eval_loss": 2.101423740386963, | |
| "eval_runtime": 23.9781, | |
| "eval_samples_per_second": 36.825, | |
| "eval_steps_per_second": 2.335, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 2.731416805620766, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2172, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9462, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8014, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8689, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9006, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_accuracy": 0.6392953969397911, | |
| "eval_loss": 1.8079485893249512, | |
| "eval_runtime": 24.4889, | |
| "eval_samples_per_second": 36.057, | |
| "eval_steps_per_second": 2.287, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_bleu": 3.957789284907739, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8975, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8936, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8677, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9317, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6722017129718149, | |
| "eval_loss": 1.5510306358337402, | |
| "eval_runtime": 24.2083, | |
| "eval_samples_per_second": 36.475, | |
| "eval_steps_per_second": 2.313, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 4.403977979732999, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7801, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3602, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3667, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3497, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3984, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_accuracy": 0.707495599907831, | |
| "eval_loss": 1.284958839416504, | |
| "eval_runtime": 24.875, | |
| "eval_samples_per_second": 35.498, | |
| "eval_steps_per_second": 2.251, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_bleu": 7.1028643857654, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3143, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3887, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3563, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3749, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3662, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.737386810999495, | |
| "eval_loss": 1.0899983644485474, | |
| "eval_runtime": 24.1441, | |
| "eval_samples_per_second": 36.572, | |
| "eval_steps_per_second": 2.319, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 8.181972004065782, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.0, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9654, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9704, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9269, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9041, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_accuracy": 0.7670475994371804, | |
| "eval_loss": 0.890863835811615, | |
| "eval_runtime": 24.7388, | |
| "eval_samples_per_second": 35.693, | |
| "eval_steps_per_second": 2.264, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_bleu": 12.34433922960259, | |
| "eval_exact_match": 0.0022650056625141564, | |
| "eval_prefix_exact_match": 0.0022650056625141564, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9277, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8827, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 3e-05, | |
| "loss": 0.99, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9083, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9056, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7902908718310757, | |
| "eval_loss": 0.7502285242080688, | |
| "eval_runtime": 24.8951, | |
| "eval_samples_per_second": 35.469, | |
| "eval_steps_per_second": 2.249, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 14.357948484733997, | |
| "eval_exact_match": 0.0033975084937712344, | |
| "eval_prefix_exact_match": 0.0033975084937712344, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7462, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 3e-05, | |
| "loss": 0.583, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5995, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6094, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6131, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_accuracy": 0.8066557828732234, | |
| "eval_loss": 0.6304216980934143, | |
| "eval_runtime": 25.7927, | |
| "eval_samples_per_second": 34.234, | |
| "eval_steps_per_second": 2.171, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_bleu": 21.42036295397234, | |
| "eval_exact_match": 0.010192525481313703, | |
| "eval_prefix_exact_match": 0.010192525481313703, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6447, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6134, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "learning_rate": 3e-05, | |
| "loss": 0.611, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6101, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.82152049535968, | |
| "eval_loss": 0.5429415106773376, | |
| "eval_runtime": 24.6601, | |
| "eval_samples_per_second": 35.807, | |
| "eval_steps_per_second": 2.271, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 26.261017838942095, | |
| "eval_exact_match": 0.02944507361268403, | |
| "eval_prefix_exact_match": 0.02944507361268403, | |
| "step": 280 | |
| } | |
| ], | |
| "logging_steps": 6, | |
| "max_steps": 560, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 6832039079884800.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |