| { | |
| "best_metric": 0.04662612825632095, | |
| "best_model_checkpoint": "./beans_lora_outputs/checkpoint-520", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 650, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.4020828008651733, | |
| "learning_rate": 0.0004923076923076924, | |
| "loss": 1.0336, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.1858539581298828, | |
| "learning_rate": 0.0004846153846153846, | |
| "loss": 0.9408, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 1.314568281173706, | |
| "learning_rate": 0.000476923076923077, | |
| "loss": 0.7742, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.8166171312332153, | |
| "learning_rate": 0.00046923076923076926, | |
| "loss": 0.5557, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.892727792263031, | |
| "learning_rate": 0.0004615384615384616, | |
| "loss": 0.5459, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.4540048837661743, | |
| "learning_rate": 0.00045384615384615385, | |
| "loss": 0.4776, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.7993093729019165, | |
| "learning_rate": 0.0004461538461538462, | |
| "loss": 0.4206, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.4129832088947296, | |
| "learning_rate": 0.00043846153846153845, | |
| "loss": 0.2951, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 1.1914254426956177, | |
| "learning_rate": 0.00043076923076923083, | |
| "loss": 0.2664, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 2.3125526905059814, | |
| "learning_rate": 0.0004230769230769231, | |
| "loss": 0.3194, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.2275660037994385, | |
| "learning_rate": 0.0004153846153846154, | |
| "loss": 0.2889, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.6353504657745361, | |
| "learning_rate": 0.0004076923076923077, | |
| "loss": 0.1575, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.177952527999878, | |
| "learning_rate": 0.0004, | |
| "loss": 0.2167, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9473684210526315, | |
| "eval_loss": 0.17138947546482086, | |
| "eval_runtime": 9.2462, | |
| "eval_samples_per_second": 14.384, | |
| "eval_steps_per_second": 1.839, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.3688104748725891, | |
| "learning_rate": 0.0003923076923076923, | |
| "loss": 0.1432, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 1.1805006265640259, | |
| "learning_rate": 0.00038461538461538467, | |
| "loss": 0.2113, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.634524405002594, | |
| "learning_rate": 0.00037692307692307694, | |
| "loss": 0.1851, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.15019208192825317, | |
| "learning_rate": 0.00036923076923076927, | |
| "loss": 0.1933, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.3095666766166687, | |
| "learning_rate": 0.00036153846153846154, | |
| "loss": 0.1281, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 1.1930674314498901, | |
| "learning_rate": 0.00035384615384615386, | |
| "loss": 0.1348, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.1358138620853424, | |
| "learning_rate": 0.00034615384615384613, | |
| "loss": 0.2277, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 2.295457363128662, | |
| "learning_rate": 0.0003384615384615385, | |
| "loss": 0.1802, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.1559237241744995, | |
| "learning_rate": 0.0003307692307692308, | |
| "loss": 0.1972, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.8495343923568726, | |
| "learning_rate": 0.0003230769230769231, | |
| "loss": 0.176, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 1.684859037399292, | |
| "learning_rate": 0.0003153846153846154, | |
| "loss": 0.2912, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.2597804367542267, | |
| "learning_rate": 0.0003076923076923077, | |
| "loss": 0.1231, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.09072303026914597, | |
| "learning_rate": 0.0003, | |
| "loss": 0.1288, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9548872180451128, | |
| "eval_loss": 0.13066871464252472, | |
| "eval_runtime": 7.0796, | |
| "eval_samples_per_second": 18.786, | |
| "eval_steps_per_second": 2.401, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.09045865386724472, | |
| "learning_rate": 0.00029230769230769235, | |
| "loss": 0.0485, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.0802936926484108, | |
| "learning_rate": 0.0002846153846153846, | |
| "loss": 0.1315, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 2.443058967590332, | |
| "learning_rate": 0.00027692307692307695, | |
| "loss": 0.1178, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 2.5117151737213135, | |
| "learning_rate": 0.0002692307692307692, | |
| "loss": 0.1197, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.6276033520698547, | |
| "learning_rate": 0.00026153846153846154, | |
| "loss": 0.1327, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.07618015259504318, | |
| "learning_rate": 0.0002538461538461538, | |
| "loss": 0.1751, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 1.933332920074463, | |
| "learning_rate": 0.0002461538461538462, | |
| "loss": 0.1409, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 1.7085862159729004, | |
| "learning_rate": 0.0002384615384615385, | |
| "loss": 0.076, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.08163443952798843, | |
| "learning_rate": 0.0002307692307692308, | |
| "loss": 0.1995, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 3.548895835876465, | |
| "learning_rate": 0.0002230769230769231, | |
| "loss": 0.1168, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.2048652619123459, | |
| "learning_rate": 0.00021538461538461541, | |
| "loss": 0.1259, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.46744510531425476, | |
| "learning_rate": 0.0002076923076923077, | |
| "loss": 0.0786, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.05968974158167839, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1056, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.06689254194498062, | |
| "eval_runtime": 7.9823, | |
| "eval_samples_per_second": 16.662, | |
| "eval_steps_per_second": 2.13, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 1.8612762689590454, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.1471, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 3.9264113903045654, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.1647, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 1.6901596784591675, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.0862, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.046002600342035294, | |
| "learning_rate": 0.00016923076923076926, | |
| "loss": 0.0178, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 1.302810549736023, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.1012, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 1.406225323677063, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.0446, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 3.55471134185791, | |
| "learning_rate": 0.00014615384615384618, | |
| "loss": 0.0553, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 3.552142858505249, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0779, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.5659548044204712, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0464, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.04047946259379387, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0512, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.4618072211742401, | |
| "learning_rate": 0.0001153846153846154, | |
| "loss": 0.1398, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.030209071934223175, | |
| "learning_rate": 0.00010769230769230771, | |
| "loss": 0.113, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0347941592335701, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0772, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.04662612825632095, | |
| "eval_runtime": 6.9929, | |
| "eval_samples_per_second": 19.019, | |
| "eval_steps_per_second": 2.431, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.3986341953277588, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0794, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.538861870765686, | |
| "learning_rate": 8.461538461538463e-05, | |
| "loss": 0.1024, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.10060963779687881, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.1196, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.4288579225540161, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0233, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.06229620799422264, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0887, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.06911199539899826, | |
| "learning_rate": 5.3846153846153853e-05, | |
| "loss": 0.055, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.31500548124313354, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0385, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.1616271734237671, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.1278, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 2.4755680561065674, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0742, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.2346724271774292, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0398, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.09390363842248917, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0235, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.0435369536280632, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0719, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.09755070507526398, | |
| "learning_rate": 0.0, | |
| "loss": 0.0941, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.08044340461492538, | |
| "eval_runtime": 9.1805, | |
| "eval_samples_per_second": 14.487, | |
| "eval_steps_per_second": 1.852, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 650, | |
| "total_flos": 4.0477911662555136e+17, | |
| "train_loss": 0.18834018239608177, | |
| "train_runtime": 471.6062, | |
| "train_samples_per_second": 10.963, | |
| "train_steps_per_second": 1.378 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 650, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.0477911662555136e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |