| { | |
| "best_metric": 0.06340890377759933, | |
| "best_model_checkpoint": "./beans_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": 2.198349952697754, | |
| "learning_rate": 1.9692307692307696e-05, | |
| "loss": 1.0245, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.8462625741958618, | |
| "learning_rate": 1.9384615384615386e-05, | |
| "loss": 0.9453, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 2.1076345443725586, | |
| "learning_rate": 1.907692307692308e-05, | |
| "loss": 0.8404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.612368106842041, | |
| "learning_rate": 1.876923076923077e-05, | |
| "loss": 0.6951, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 4.421859264373779, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.6618, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.2416698932647705, | |
| "learning_rate": 1.8153846153846155e-05, | |
| "loss": 0.581, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.765328049659729, | |
| "learning_rate": 1.784615384615385e-05, | |
| "loss": 0.5209, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.254733681678772, | |
| "learning_rate": 1.753846153846154e-05, | |
| "loss": 0.3972, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 2.554185152053833, | |
| "learning_rate": 1.7230769230769234e-05, | |
| "loss": 0.3536, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 2.9679086208343506, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.3718, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.998707890510559, | |
| "learning_rate": 1.6615384615384618e-05, | |
| "loss": 0.3356, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 2.4654996395111084, | |
| "learning_rate": 1.630769230769231e-05, | |
| "loss": 0.246, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.244895935058594, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.2823, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.21853642165660858, | |
| "eval_runtime": 0.8655, | |
| "eval_samples_per_second": 153.664, | |
| "eval_steps_per_second": 19.641, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.4838427305221558, | |
| "learning_rate": 1.5692307692307693e-05, | |
| "loss": 0.1994, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 3.1009955406188965, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2254, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.4830702543258667, | |
| "learning_rate": 1.5076923076923078e-05, | |
| "loss": 0.2016, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.7268257141113281, | |
| "learning_rate": 1.4769230769230772e-05, | |
| "loss": 0.2199, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.6102820634841919, | |
| "learning_rate": 1.4461538461538462e-05, | |
| "loss": 0.2105, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 4.0108184814453125, | |
| "learning_rate": 1.4153846153846156e-05, | |
| "loss": 0.1635, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.453694224357605, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.2531, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.6146100759506226, | |
| "learning_rate": 1.353846153846154e-05, | |
| "loss": 0.1778, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 2.1759166717529297, | |
| "learning_rate": 1.3230769230769231e-05, | |
| "loss": 0.2135, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 1.7484408617019653, | |
| "learning_rate": 1.2923076923076925e-05, | |
| "loss": 0.1679, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 4.208325386047363, | |
| "learning_rate": 1.2615384615384616e-05, | |
| "loss": 0.2595, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3881552517414093, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.1382, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.348543643951416, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.132, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.1254686862230301, | |
| "eval_runtime": 1.0009, | |
| "eval_samples_per_second": 132.881, | |
| "eval_steps_per_second": 16.985, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.48037123680114746, | |
| "learning_rate": 1.1692307692307694e-05, | |
| "loss": 0.0898, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.33670032024383545, | |
| "learning_rate": 1.1384615384615385e-05, | |
| "loss": 0.1446, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 7.312185287475586, | |
| "learning_rate": 1.1076923076923079e-05, | |
| "loss": 0.1248, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 7.81497049331665, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.1396, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 7.189246654510498, | |
| "learning_rate": 1.0461538461538463e-05, | |
| "loss": 0.1793, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.31857696175575256, | |
| "learning_rate": 1.0153846153846154e-05, | |
| "loss": 0.1576, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 7.1352620124816895, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 0.1798, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 1.186557412147522, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 0.0741, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.33909401297569275, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.1346, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 12.247209548950195, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 0.1237, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.3360549509525299, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 0.151, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 1.2044156789779663, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 0.1285, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.2921423614025116, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.1448, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.09480106830596924, | |
| "eval_runtime": 1.0292, | |
| "eval_samples_per_second": 129.23, | |
| "eval_steps_per_second": 16.518, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 2.2833054065704346, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.1362, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 6.499863624572754, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 0.1572, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 8.841352462768555, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 0.1412, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.23157350718975067, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 0.0745, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 6.295877933502197, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 0.1145, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 12.639660835266113, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.087, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 4.014486312866211, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.0995, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 8.148310661315918, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 0.0774, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.22080053389072418, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 0.103, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.20995458960533142, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 0.0743, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.8845565319061279, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.1458, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.2121868133544922, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.1277, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3017064034938812, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0873, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9924812030075187, | |
| "eval_loss": 0.06340890377759933, | |
| "eval_runtime": 0.7533, | |
| "eval_samples_per_second": 176.547, | |
| "eval_steps_per_second": 22.566, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 6.891432285308838, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 0.098, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 1.4789180755615234, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 0.1661, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.4021252989768982, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.1548, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 1.2701594829559326, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.0505, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.3795933723449707, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.135, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.27298393845558167, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 0.0819, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 8.698312759399414, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 0.0776, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 2.449575901031494, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.0949, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 5.441997051239014, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.0871, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.25316253304481506, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 0.0819, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.21708577871322632, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 0.0719, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.23087237775325775, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.0773, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.589409589767456, | |
| "learning_rate": 0.0, | |
| "loss": 0.1172, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.08086416125297546, | |
| "eval_runtime": 0.7469, | |
| "eval_samples_per_second": 178.064, | |
| "eval_steps_per_second": 22.76, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 650, | |
| "total_flos": 4.006371770595533e+17, | |
| "train_loss": 0.22014901995658875, | |
| "train_runtime": 93.7188, | |
| "train_samples_per_second": 55.165, | |
| "train_steps_per_second": 6.936 | |
| } | |
| ], | |
| "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": 4.006371770595533e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |