| { | |
| "best_metric": 0.06572757661342621, | |
| "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": 2.1983089447021484, | |
| "learning_rate": 1.9692307692307696e-05, | |
| "loss": 1.0245, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.8456082344055176, | |
| "learning_rate": 1.9384615384615386e-05, | |
| "loss": 0.9453, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 2.1001663208007812, | |
| "learning_rate": 1.907692307692308e-05, | |
| "loss": 0.8404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.6200923919677734, | |
| "learning_rate": 1.876923076923077e-05, | |
| "loss": 0.695, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 3.9886152744293213, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.661, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.0928611755371094, | |
| "learning_rate": 1.8153846153846155e-05, | |
| "loss": 0.5811, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.8000515699386597, | |
| "learning_rate": 1.784615384615385e-05, | |
| "loss": 0.5187, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.279348373413086, | |
| "learning_rate": 1.753846153846154e-05, | |
| "loss": 0.3915, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 2.4376962184906006, | |
| "learning_rate": 1.7230769230769234e-05, | |
| "loss": 0.3521, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 2.9530019760131836, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.3709, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.8786017894744873, | |
| "learning_rate": 1.6615384615384618e-05, | |
| "loss": 0.3359, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 2.5607118606567383, | |
| "learning_rate": 1.630769230769231e-05, | |
| "loss": 0.2476, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.597635269165039, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.284, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.2164773941040039, | |
| "eval_runtime": 15.625, | |
| "eval_samples_per_second": 8.512, | |
| "eval_steps_per_second": 1.088, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.7280329465866089, | |
| "learning_rate": 1.5692307692307693e-05, | |
| "loss": 0.1987, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 3.229095220565796, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2205, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.469906210899353, | |
| "learning_rate": 1.5076923076923078e-05, | |
| "loss": 0.2047, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.524368941783905, | |
| "learning_rate": 1.4769230769230772e-05, | |
| "loss": 0.2219, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.5878937840461731, | |
| "learning_rate": 1.4461538461538462e-05, | |
| "loss": 0.2116, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 4.049538612365723, | |
| "learning_rate": 1.4153846153846156e-05, | |
| "loss": 0.1605, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.47798386216163635, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.2582, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.6020946502685547, | |
| "learning_rate": 1.353846153846154e-05, | |
| "loss": 0.1857, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 2.6440248489379883, | |
| "learning_rate": 1.3230769230769231e-05, | |
| "loss": 0.2161, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 1.9017517566680908, | |
| "learning_rate": 1.2923076923076925e-05, | |
| "loss": 0.1619, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 10.503934860229492, | |
| "learning_rate": 1.2615384615384616e-05, | |
| "loss": 0.2592, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3896147906780243, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.1432, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.34977179765701294, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.1316, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.13314025104045868, | |
| "eval_runtime": 15.4736, | |
| "eval_samples_per_second": 8.595, | |
| "eval_steps_per_second": 1.099, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.4498082995414734, | |
| "learning_rate": 1.1692307692307694e-05, | |
| "loss": 0.0908, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.33902597427368164, | |
| "learning_rate": 1.1384615384615385e-05, | |
| "loss": 0.1483, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 6.657054424285889, | |
| "learning_rate": 1.1076923076923079e-05, | |
| "loss": 0.1296, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 7.5347466468811035, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.1387, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 5.809398651123047, | |
| "learning_rate": 1.0461538461538463e-05, | |
| "loss": 0.1809, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.3026523292064667, | |
| "learning_rate": 1.0153846153846154e-05, | |
| "loss": 0.1444, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 6.505568981170654, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 0.1819, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 1.1542584896087646, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 0.0727, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.5524181127548218, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.1315, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 10.899353981018066, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 0.1265, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.2956291139125824, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 0.1525, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 1.2137246131896973, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 0.1344, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.29744699597358704, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.1429, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.09923107177019119, | |
| "eval_runtime": 15.5311, | |
| "eval_samples_per_second": 8.563, | |
| "eval_steps_per_second": 1.095, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 2.0318715572357178, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.1347, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 6.056100845336914, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 0.1599, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 9.551461219787598, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 0.142, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.2325032651424408, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 0.0717, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 5.357461452484131, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 0.1151, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 12.689959526062012, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.0865, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 4.816635608673096, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.0974, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 8.048978805541992, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 0.079, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.2229355424642563, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 0.0944, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.2118719518184662, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 0.0788, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.8754830956459045, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.1364, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.20934270322322845, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.1295, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3156948685646057, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0775, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9924812030075187, | |
| "eval_loss": 0.06572757661342621, | |
| "eval_runtime": 15.8319, | |
| "eval_samples_per_second": 8.401, | |
| "eval_steps_per_second": 1.074, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 9.230475425720215, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 0.1011, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 1.6108723878860474, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 0.1655, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.3194853961467743, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.1479, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 1.1996067762374878, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.0508, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.2994942367076874, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.1303, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.267814040184021, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 0.0836, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 9.885746002197266, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 0.0725, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 4.1704487800598145, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.0976, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 5.211251735687256, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.0857, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.26706379652023315, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 0.0802, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.219241201877594, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 0.0675, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.2313096821308136, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.0776, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6467680931091309, | |
| "learning_rate": 0.0, | |
| "loss": 0.1142, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.07825341820716858, | |
| "eval_runtime": 19.0701, | |
| "eval_samples_per_second": 6.974, | |
| "eval_steps_per_second": 0.891, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 650, | |
| "total_flos": 4.006371770595533e+17, | |
| "train_loss": 0.21960290945493258, | |
| "train_runtime": 2197.69, | |
| "train_samples_per_second": 2.352, | |
| "train_steps_per_second": 0.296 | |
| } | |
| ], | |
| "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 | |
| } | |