| { | |
| "best_metric": 1.301220178604126, | |
| "best_model_checkpoint": "roberta-base-CD_baseline/checkpoint-625", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 625, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.637064456939697, | |
| "learning_rate": 1.968e-05, | |
| "loss": 2.2753, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 15.68393325805664, | |
| "learning_rate": 1.936e-05, | |
| "loss": 2.0418, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 12.480018615722656, | |
| "learning_rate": 1.904e-05, | |
| "loss": 1.9195, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 41.00396728515625, | |
| "learning_rate": 1.8720000000000004e-05, | |
| "loss": 1.9629, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.204047203063965, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.7429, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 8.693258285522461, | |
| "learning_rate": 1.8080000000000003e-05, | |
| "loss": 1.9048, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.7029831409454346, | |
| "learning_rate": 1.7760000000000003e-05, | |
| "loss": 1.8083, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.148443698883057, | |
| "learning_rate": 1.7440000000000002e-05, | |
| "loss": 1.6655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 20.234933853149414, | |
| "learning_rate": 1.7120000000000002e-05, | |
| "loss": 1.5108, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 8.840304374694824, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 1.7, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 7.2302350997924805, | |
| "learning_rate": 1.648e-05, | |
| "loss": 1.6971, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.1535422801971436, | |
| "learning_rate": 1.616e-05, | |
| "loss": 1.6077, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.41304347826086957, | |
| "eval_f1": 0.3267868906455863, | |
| "eval_loss": 1.6377873420715332, | |
| "eval_precision": 0.29971778816474115, | |
| "eval_recall": 0.41304347826086957, | |
| "eval_runtime": 1.3051, | |
| "eval_samples_per_second": 176.233, | |
| "eval_steps_per_second": 11.493, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 9.82644271850586, | |
| "learning_rate": 1.584e-05, | |
| "loss": 1.7461, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 7.467548370361328, | |
| "learning_rate": 1.552e-05, | |
| "loss": 1.6311, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 5.879318714141846, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 1.4653, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 12.443910598754883, | |
| "learning_rate": 1.4880000000000002e-05, | |
| "loss": 1.628, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 5.35117244720459, | |
| "learning_rate": 1.4560000000000001e-05, | |
| "loss": 1.5741, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 7.006358623504639, | |
| "learning_rate": 1.4240000000000001e-05, | |
| "loss": 1.8264, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 5.610796928405762, | |
| "learning_rate": 1.392e-05, | |
| "loss": 1.5304, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 23.96417808532715, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 1.4833, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 6.071532249450684, | |
| "learning_rate": 1.3280000000000002e-05, | |
| "loss": 1.4721, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 27.290363311767578, | |
| "learning_rate": 1.2960000000000001e-05, | |
| "loss": 1.6517, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 7.392124176025391, | |
| "learning_rate": 1.2640000000000001e-05, | |
| "loss": 1.5183, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 8.155671119689941, | |
| "learning_rate": 1.232e-05, | |
| "loss": 1.4147, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.289750099182129, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.6016, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.48695652173913045, | |
| "eval_f1": 0.4109191960797593, | |
| "eval_loss": 1.460945963859558, | |
| "eval_precision": 0.39039609511101225, | |
| "eval_recall": 0.48695652173913045, | |
| "eval_runtime": 1.2886, | |
| "eval_samples_per_second": 178.493, | |
| "eval_steps_per_second": 11.641, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 15.545705795288086, | |
| "learning_rate": 1.168e-05, | |
| "loss": 1.3776, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 7.504268646240234, | |
| "learning_rate": 1.136e-05, | |
| "loss": 1.3455, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 8.06009292602539, | |
| "learning_rate": 1.1040000000000001e-05, | |
| "loss": 1.4793, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 18.173667907714844, | |
| "learning_rate": 1.072e-05, | |
| "loss": 1.4226, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 11.790044784545898, | |
| "learning_rate": 1.04e-05, | |
| "loss": 1.386, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 7.876099109649658, | |
| "learning_rate": 1.008e-05, | |
| "loss": 1.3522, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 10.516275405883789, | |
| "learning_rate": 9.760000000000001e-06, | |
| "loss": 1.4296, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 8.990633010864258, | |
| "learning_rate": 9.440000000000001e-06, | |
| "loss": 1.4941, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 8.365500450134277, | |
| "learning_rate": 9.12e-06, | |
| "loss": 1.4153, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 9.450441360473633, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.522, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 18.165035247802734, | |
| "learning_rate": 8.48e-06, | |
| "loss": 1.3895, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 10.70251750946045, | |
| "learning_rate": 8.16e-06, | |
| "loss": 1.2479, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5043478260869565, | |
| "eval_f1": 0.44849074152912793, | |
| "eval_loss": 1.4184553623199463, | |
| "eval_precision": 0.4235901359719885, | |
| "eval_recall": 0.5043478260869565, | |
| "eval_runtime": 1.303, | |
| "eval_samples_per_second": 176.51, | |
| "eval_steps_per_second": 11.511, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 23.491443634033203, | |
| "learning_rate": 7.840000000000001e-06, | |
| "loss": 1.2631, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 8.99567985534668, | |
| "learning_rate": 7.520000000000001e-06, | |
| "loss": 1.0625, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 12.7035493850708, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.3823, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 15.28470230102539, | |
| "learning_rate": 6.88e-06, | |
| "loss": 1.2633, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 10.78918170928955, | |
| "learning_rate": 6.560000000000001e-06, | |
| "loss": 1.1007, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 14.376765251159668, | |
| "learning_rate": 6.24e-06, | |
| "loss": 1.2674, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 11.458234786987305, | |
| "learning_rate": 5.92e-06, | |
| "loss": 1.0922, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 15.360867500305176, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.2077, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 20.423187255859375, | |
| "learning_rate": 5.28e-06, | |
| "loss": 1.302, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 13.024535179138184, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 1.1218, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 14.865490913391113, | |
| "learning_rate": 4.6400000000000005e-06, | |
| "loss": 1.1858, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 17.874656677246094, | |
| "learning_rate": 4.32e-06, | |
| "loss": 1.1319, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 16.873672485351562, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.1542, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5434782608695652, | |
| "eval_f1": 0.5141449277787993, | |
| "eval_loss": 1.3071645498275757, | |
| "eval_precision": 0.5397252083665127, | |
| "eval_recall": 0.5434782608695652, | |
| "eval_runtime": 2.7726, | |
| "eval_samples_per_second": 82.954, | |
| "eval_steps_per_second": 5.41, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 16.543500900268555, | |
| "learning_rate": 3.6800000000000003e-06, | |
| "loss": 1.0007, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 13.177757263183594, | |
| "learning_rate": 3.3600000000000004e-06, | |
| "loss": 1.0294, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 10.053601264953613, | |
| "learning_rate": 3.04e-06, | |
| "loss": 1.1163, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 20.3472957611084, | |
| "learning_rate": 2.7200000000000002e-06, | |
| "loss": 1.1222, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 10.218711853027344, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.964, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 10.360238075256348, | |
| "learning_rate": 2.08e-06, | |
| "loss": 1.0094, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 16.11949348449707, | |
| "learning_rate": 1.76e-06, | |
| "loss": 1.0102, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 14.713994979858398, | |
| "learning_rate": 1.44e-06, | |
| "loss": 1.0562, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 12.219269752502441, | |
| "learning_rate": 1.12e-06, | |
| "loss": 0.9808, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 27.00701141357422, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.1036, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 18.196247100830078, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 1.1907, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 18.10931968688965, | |
| "learning_rate": 1.6e-07, | |
| "loss": 1.1302, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5434782608695652, | |
| "eval_f1": 0.5062253924435693, | |
| "eval_loss": 1.301220178604126, | |
| "eval_precision": 0.511256656273167, | |
| "eval_recall": 0.5434782608695652, | |
| "eval_runtime": 2.7886, | |
| "eval_samples_per_second": 82.478, | |
| "eval_steps_per_second": 5.379, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 625, | |
| "total_flos": 1434186246250944.0, | |
| "train_loss": 1.4087707061767578, | |
| "train_runtime": 222.2422, | |
| "train_samples_per_second": 44.906, | |
| "train_steps_per_second": 2.812 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 625, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 1434186246250944.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |