| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3639164690398064, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.6e-08, | |
| "loss": 2.8922, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 5.6e-08, | |
| "loss": 2.6572, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 9.199999999999999e-08, | |
| "loss": 2.6816, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.32e-07, | |
| "loss": 2.6259, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.7199999999999998e-07, | |
| "loss": 2.4734, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.12e-07, | |
| "loss": 2.4194, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.52e-07, | |
| "loss": 2.3645, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.9199999999999997e-07, | |
| "loss": 2.287, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.32e-07, | |
| "loss": 2.2451, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.72e-07, | |
| "loss": 2.1339, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.12e-07, | |
| "loss": 2.0578, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.5199999999999997e-07, | |
| "loss": 2.0854, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.92e-07, | |
| "loss": 1.9981, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.32e-07, | |
| "loss": 1.927, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.719999999999999e-07, | |
| "loss": 1.8661, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.119999999999999e-07, | |
| "loss": 1.8546, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.52e-07, | |
| "loss": 1.7949, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 6.919999999999999e-07, | |
| "loss": 1.7926, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 7.319999999999999e-07, | |
| "loss": 1.7737, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.72e-07, | |
| "loss": 1.7178, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.12e-07, | |
| "loss": 1.6872, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.52e-07, | |
| "loss": 1.6504, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 8.92e-07, | |
| "loss": 1.6955, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9.32e-07, | |
| "loss": 1.6822, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.72e-07, | |
| "loss": 1.647, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.012e-06, | |
| "loss": 1.6131, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.052e-06, | |
| "loss": 1.6433, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.092e-06, | |
| "loss": 1.596, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.132e-06, | |
| "loss": 1.5824, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.172e-06, | |
| "loss": 1.5606, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.212e-06, | |
| "loss": 1.5886, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.252e-06, | |
| "loss": 1.5905, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.292e-06, | |
| "loss": 1.5493, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.332e-06, | |
| "loss": 1.5734, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.372e-06, | |
| "loss": 1.5435, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.4119999999999998e-06, | |
| "loss": 1.5359, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.4519999999999998e-06, | |
| "loss": 1.5579, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.4919999999999999e-06, | |
| "loss": 1.5358, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.532e-06, | |
| "loss": 1.5509, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.572e-06, | |
| "loss": 1.5116, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.612e-06, | |
| "loss": 1.5251, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.6519999999999998e-06, | |
| "loss": 1.5107, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.6919999999999999e-06, | |
| "loss": 1.4805, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.7319999999999999e-06, | |
| "loss": 1.4819, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.772e-06, | |
| "loss": 1.4897, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.812e-06, | |
| "loss": 1.4824, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.852e-06, | |
| "loss": 1.4822, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.8919999999999998e-06, | |
| "loss": 1.5, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.932e-06, | |
| "loss": 1.4595, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.972e-06, | |
| "loss": 1.4163, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_wmt2019_zh-en_accuracy": 0.5762586867111048, | |
| "eval_wmt2019_zh-en_loss": 2.037109375, | |
| "eval_wmt2019_zh-en_runtime": 77.5891, | |
| "eval_wmt2019_zh-en_samples_per_second": 12.888, | |
| "eval_wmt2019_zh-en_steps_per_second": 3.222, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_ted_trans_en-ja_accuracy": 0.49833809864188705, | |
| "eval_ted_trans_en-ja_loss": 2.359375, | |
| "eval_ted_trans_en-ja_runtime": 67.0162, | |
| "eval_ted_trans_en-ja_samples_per_second": 11.952, | |
| "eval_ted_trans_en-ja_steps_per_second": 2.999, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_ted_trans_zh-ja_accuracy": 0.41924741924741926, | |
| "eval_ted_trans_zh-ja_loss": 3.099609375, | |
| "eval_ted_trans_zh-ja_runtime": 4.122, | |
| "eval_ted_trans_zh-ja_samples_per_second": 10.189, | |
| "eval_ted_trans_zh-ja_steps_per_second": 2.669, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_sharegpt_accuracy": 0.6638255086604158, | |
| "eval_sharegpt_loss": 1.4677734375, | |
| "eval_sharegpt_runtime": 754.4011, | |
| "eval_sharegpt_samples_per_second": 4.438, | |
| "eval_sharegpt_steps_per_second": 1.109, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_dolly15k_accuracy": 0.5514953610117659, | |
| "eval_dolly15k_loss": 1.9169921875, | |
| "eval_dolly15k_runtime": 63.6309, | |
| "eval_dolly15k_samples_per_second": 11.802, | |
| "eval_dolly15k_steps_per_second": 2.955, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_ikala_accuracy": 0.6825314578991759, | |
| "eval_ikala_loss": 1.2626953125, | |
| "eval_ikala_runtime": 1328.8914, | |
| "eval_ikala_samples_per_second": 10.328, | |
| "eval_ikala_steps_per_second": 2.583, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_oasst_export_accuracy": 0.5725428280687767, | |
| "eval_oasst_export_loss": 2.181640625, | |
| "eval_oasst_export_runtime": 206.1134, | |
| "eval_oasst_export_samples_per_second": 10.184, | |
| "eval_oasst_export_steps_per_second": 2.547, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2.012e-06, | |
| "loss": 1.4116, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2.052e-06, | |
| "loss": 1.4304, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.092e-06, | |
| "loss": 1.4102, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.132e-06, | |
| "loss": 1.3838, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.172e-06, | |
| "loss": 1.4179, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.212e-06, | |
| "loss": 1.4286, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.2519999999999998e-06, | |
| "loss": 1.426, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.292e-06, | |
| "loss": 1.4257, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.332e-06, | |
| "loss": 1.4342, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.372e-06, | |
| "loss": 1.4015, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.412e-06, | |
| "loss": 1.3999, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.452e-06, | |
| "loss": 1.4062, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.492e-06, | |
| "loss": 1.427, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.532e-06, | |
| "loss": 1.4171, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.572e-06, | |
| "loss": 1.4221, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.612e-06, | |
| "loss": 1.3547, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.652e-06, | |
| "loss": 1.3783, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.6920000000000002e-06, | |
| "loss": 1.3574, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.7320000000000003e-06, | |
| "loss": 1.3604, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.7719999999999995e-06, | |
| "loss": 1.3978, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.8119999999999995e-06, | |
| "loss": 1.3879, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.8519999999999995e-06, | |
| "loss": 1.3731, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.8919999999999996e-06, | |
| "loss": 1.3732, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.9319999999999996e-06, | |
| "loss": 1.3787, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2.9719999999999997e-06, | |
| "loss": 1.3842, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 3.0119999999999997e-06, | |
| "loss": 1.3952, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 3.0519999999999997e-06, | |
| "loss": 1.3117, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 3.0919999999999998e-06, | |
| "loss": 1.3531, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 3.132e-06, | |
| "loss": 1.3287, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.172e-06, | |
| "loss": 1.3417, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.212e-06, | |
| "loss": 1.3341, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.2519999999999995e-06, | |
| "loss": 1.364, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.2919999999999996e-06, | |
| "loss": 1.3201, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.3319999999999996e-06, | |
| "loss": 1.3714, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 3.3719999999999996e-06, | |
| "loss": 1.3098, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.4119999999999997e-06, | |
| "loss": 1.3222, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.4519999999999997e-06, | |
| "loss": 1.3451, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.4919999999999998e-06, | |
| "loss": 1.3502, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.532e-06, | |
| "loss": 1.3445, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 3.572e-06, | |
| "loss": 1.3304, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.612e-06, | |
| "loss": 1.3081, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.652e-06, | |
| "loss": 1.3106, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.692e-06, | |
| "loss": 1.3357, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.732e-06, | |
| "loss": 1.3243, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.7719999999999996e-06, | |
| "loss": 1.3164, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.8119999999999997e-06, | |
| "loss": 1.3124, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 3.852e-06, | |
| "loss": 1.3162, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 3.891999999999999e-06, | |
| "loss": 1.3124, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 3.932e-06, | |
| "loss": 1.3347, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 3.971999999999999e-06, | |
| "loss": 1.3174, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_wmt2019_zh-en_accuracy": 0.5819691791089524, | |
| "eval_wmt2019_zh-en_loss": 1.9326171875, | |
| "eval_wmt2019_zh-en_runtime": 76.631, | |
| "eval_wmt2019_zh-en_samples_per_second": 13.05, | |
| "eval_wmt2019_zh-en_steps_per_second": 3.262, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_ted_trans_en-ja_accuracy": 0.5248655214298109, | |
| "eval_ted_trans_en-ja_loss": 2.15625, | |
| "eval_ted_trans_en-ja_runtime": 67.1457, | |
| "eval_ted_trans_en-ja_samples_per_second": 11.929, | |
| "eval_ted_trans_en-ja_steps_per_second": 2.993, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_ted_trans_zh-ja_accuracy": 0.46262002743484226, | |
| "eval_ted_trans_zh-ja_loss": 2.8671875, | |
| "eval_ted_trans_zh-ja_runtime": 3.9271, | |
| "eval_ted_trans_zh-ja_samples_per_second": 10.695, | |
| "eval_ted_trans_zh-ja_steps_per_second": 2.801, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_sharegpt_accuracy": 0.6835767342429696, | |
| "eval_sharegpt_loss": 1.326171875, | |
| "eval_sharegpt_runtime": 751.6419, | |
| "eval_sharegpt_samples_per_second": 4.454, | |
| "eval_sharegpt_steps_per_second": 1.114, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_dolly15k_accuracy": 0.5566094783945817, | |
| "eval_dolly15k_loss": 1.84765625, | |
| "eval_dolly15k_runtime": 63.1655, | |
| "eval_dolly15k_samples_per_second": 11.889, | |
| "eval_dolly15k_steps_per_second": 2.976, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_ikala_accuracy": 0.6979014199391581, | |
| "eval_ikala_loss": 1.1513671875, | |
| "eval_ikala_runtime": 1328.524, | |
| "eval_ikala_samples_per_second": 10.331, | |
| "eval_ikala_steps_per_second": 2.583, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_oasst_export_accuracy": 0.5836786504830943, | |
| "eval_oasst_export_loss": 2.068359375, | |
| "eval_oasst_export_runtime": 206.6429, | |
| "eval_oasst_export_samples_per_second": 10.158, | |
| "eval_oasst_export_steps_per_second": 2.541, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.011999999999999e-06, | |
| "loss": 1.2867, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.0519999999999995e-06, | |
| "loss": 1.2933, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.091999999999999e-06, | |
| "loss": 1.2778, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.1319999999999996e-06, | |
| "loss": 1.3085, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.171999999999999e-06, | |
| "loss": 1.2772, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.212e-06, | |
| "loss": 1.3461, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.251999999999999e-06, | |
| "loss": 1.3247, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.292e-06, | |
| "loss": 1.2988, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.331999999999999e-06, | |
| "loss": 1.3175, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.372e-06, | |
| "loss": 1.3061, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.4119999999999994e-06, | |
| "loss": 1.3001, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.452e-06, | |
| "loss": 1.3325, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.4919999999999995e-06, | |
| "loss": 1.3177, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.532e-06, | |
| "loss": 1.3128, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.572e-06, | |
| "loss": 1.302, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.612e-06, | |
| "loss": 1.3233, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.652e-06, | |
| "loss": 1.3075, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.692e-06, | |
| "loss": 1.3044, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.732e-06, | |
| "loss": 1.2686, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.772e-06, | |
| "loss": 1.3169, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.812e-06, | |
| "loss": 1.3075, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.852e-06, | |
| "loss": 1.2911, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.892e-06, | |
| "loss": 1.289, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.932e-06, | |
| "loss": 1.2944, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.972e-06, | |
| "loss": 1.2753, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.012e-06, | |
| "loss": 1.2949, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.051999999999999e-06, | |
| "loss": 1.2816, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.092e-06, | |
| "loss": 1.3104, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.131999999999999e-06, | |
| "loss": 1.274, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.172e-06, | |
| "loss": 1.296, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.211999999999999e-06, | |
| "loss": 1.29, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.252e-06, | |
| "loss": 1.2643, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.2919999999999995e-06, | |
| "loss": 1.2882, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.332e-06, | |
| "loss": 1.295, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.3719999999999996e-06, | |
| "loss": 1.248, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.412e-06, | |
| "loss": 1.3236, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.452e-06, | |
| "loss": 1.2925, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.492e-06, | |
| "loss": 1.2991, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.532e-06, | |
| "loss": 1.2853, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5.572e-06, | |
| "loss": 1.2835, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.612e-06, | |
| "loss": 1.2687, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.652e-06, | |
| "loss": 1.254, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.692e-06, | |
| "loss": 1.3045, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.732e-06, | |
| "loss": 1.2598, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.772e-06, | |
| "loss": 1.2628, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5.8120000000000004e-06, | |
| "loss": 1.2519, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5.852e-06, | |
| "loss": 1.2902, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5.892e-06, | |
| "loss": 1.2999, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5.932e-06, | |
| "loss": 1.283, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5.972e-06, | |
| "loss": 1.2697, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_wmt2019_zh-en_accuracy": 0.587438249823571, | |
| "eval_wmt2019_zh-en_loss": 1.9453125, | |
| "eval_wmt2019_zh-en_runtime": 77.3466, | |
| "eval_wmt2019_zh-en_samples_per_second": 12.929, | |
| "eval_wmt2019_zh-en_steps_per_second": 3.232, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_ted_trans_en-ja_accuracy": 0.5393802273612374, | |
| "eval_ted_trans_en-ja_loss": 2.05078125, | |
| "eval_ted_trans_en-ja_runtime": 66.8272, | |
| "eval_ted_trans_en-ja_samples_per_second": 11.986, | |
| "eval_ted_trans_en-ja_steps_per_second": 3.008, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_ted_trans_zh-ja_accuracy": 0.46426092990978485, | |
| "eval_ted_trans_zh-ja_loss": 2.8203125, | |
| "eval_ted_trans_zh-ja_runtime": 3.6203, | |
| "eval_ted_trans_zh-ja_samples_per_second": 11.601, | |
| "eval_ted_trans_zh-ja_steps_per_second": 3.038, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_sharegpt_accuracy": 0.6912774549792565, | |
| "eval_sharegpt_loss": 1.2744140625, | |
| "eval_sharegpt_runtime": 755.8288, | |
| "eval_sharegpt_samples_per_second": 4.43, | |
| "eval_sharegpt_steps_per_second": 1.107, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_dolly15k_accuracy": 0.5594429758634393, | |
| "eval_dolly15k_loss": 1.8212890625, | |
| "eval_dolly15k_runtime": 63.0527, | |
| "eval_dolly15k_samples_per_second": 11.911, | |
| "eval_dolly15k_steps_per_second": 2.982, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_ikala_accuracy": 0.7042301907582296, | |
| "eval_ikala_loss": 1.115234375, | |
| "eval_ikala_runtime": 1329.6879, | |
| "eval_ikala_samples_per_second": 10.322, | |
| "eval_ikala_steps_per_second": 2.581, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_oasst_export_accuracy": 0.5868625620259539, | |
| "eval_oasst_export_loss": 2.0390625, | |
| "eval_oasst_export_runtime": 206.437, | |
| "eval_oasst_export_samples_per_second": 10.168, | |
| "eval_oasst_export_steps_per_second": 2.543, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 6.011999999999999e-06, | |
| "loss": 1.3179, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.051999999999999e-06, | |
| "loss": 1.2781, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.0919999999999994e-06, | |
| "loss": 1.2717, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.131999999999999e-06, | |
| "loss": 1.2661, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.1719999999999995e-06, | |
| "loss": 1.287, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 6.211999999999999e-06, | |
| "loss": 1.2784, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.252e-06, | |
| "loss": 1.2767, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.291999999999999e-06, | |
| "loss": 1.2657, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.332e-06, | |
| "loss": 1.2957, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.371999999999999e-06, | |
| "loss": 1.3181, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.412e-06, | |
| "loss": 1.2688, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 6.451999999999999e-06, | |
| "loss": 1.2598, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 6.492e-06, | |
| "loss": 1.2875, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 6.5319999999999995e-06, | |
| "loss": 1.2573, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 6.572e-06, | |
| "loss": 1.2698, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 6.6119999999999995e-06, | |
| "loss": 1.271, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 6.652e-06, | |
| "loss": 1.2724, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.692e-06, | |
| "loss": 1.2527, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.732e-06, | |
| "loss": 1.2659, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.772e-06, | |
| "loss": 1.2298, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.812e-06, | |
| "loss": 1.2217, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.852e-06, | |
| "loss": 1.2695, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 6.892e-06, | |
| "loss": 1.2339, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 6.932e-06, | |
| "loss": 1.2342, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 6.972e-06, | |
| "loss": 1.2652, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 7.011999999999999e-06, | |
| "loss": 1.2411, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 7.0519999999999996e-06, | |
| "loss": 1.2478, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 7.091999999999999e-06, | |
| "loss": 1.2379, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.132e-06, | |
| "loss": 1.2847, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.171999999999999e-06, | |
| "loss": 1.2378, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.212e-06, | |
| "loss": 1.2901, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.251999999999999e-06, | |
| "loss": 1.2662, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.292e-06, | |
| "loss": 1.2622, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.3319999999999994e-06, | |
| "loss": 1.2518, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.372e-06, | |
| "loss": 1.2648, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.4119999999999995e-06, | |
| "loss": 1.2582, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.452e-06, | |
| "loss": 1.2545, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.492e-06, | |
| "loss": 1.2372, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.532e-06, | |
| "loss": 1.2266, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.572e-06, | |
| "loss": 1.2509, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.612e-06, | |
| "loss": 1.2487, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.652e-06, | |
| "loss": 1.2968, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.692e-06, | |
| "loss": 1.2719, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.732e-06, | |
| "loss": 1.2537, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.772e-06, | |
| "loss": 1.2733, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.812e-06, | |
| "loss": 1.2663, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.852e-06, | |
| "loss": 1.2496, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.892e-06, | |
| "loss": 1.2542, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.932e-06, | |
| "loss": 1.2398, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.972e-06, | |
| "loss": 1.2564, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_wmt2019_zh-en_accuracy": 0.5861558940985061, | |
| "eval_wmt2019_zh-en_loss": 1.9638671875, | |
| "eval_wmt2019_zh-en_runtime": 76.6629, | |
| "eval_wmt2019_zh-en_samples_per_second": 13.044, | |
| "eval_wmt2019_zh-en_steps_per_second": 3.261, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_ted_trans_en-ja_accuracy": 0.5495979979766785, | |
| "eval_ted_trans_en-ja_loss": 2.001953125, | |
| "eval_ted_trans_en-ja_runtime": 66.9893, | |
| "eval_ted_trans_en-ja_samples_per_second": 11.957, | |
| "eval_ted_trans_en-ja_steps_per_second": 3.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_ted_trans_zh-ja_accuracy": 0.47411944869831546, | |
| "eval_ted_trans_zh-ja_loss": 2.7109375, | |
| "eval_ted_trans_zh-ja_runtime": 4.1683, | |
| "eval_ted_trans_zh-ja_samples_per_second": 10.076, | |
| "eval_ted_trans_zh-ja_steps_per_second": 2.639, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_sharegpt_accuracy": 0.6977260319800304, | |
| "eval_sharegpt_loss": 1.234375, | |
| "eval_sharegpt_runtime": 751.9973, | |
| "eval_sharegpt_samples_per_second": 4.452, | |
| "eval_sharegpt_steps_per_second": 1.113, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_dolly15k_accuracy": 0.5599267437239759, | |
| "eval_dolly15k_loss": 1.822265625, | |
| "eval_dolly15k_runtime": 63.1103, | |
| "eval_dolly15k_samples_per_second": 11.9, | |
| "eval_dolly15k_steps_per_second": 2.979, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_ikala_accuracy": 0.708770642571952, | |
| "eval_ikala_loss": 1.0927734375, | |
| "eval_ikala_runtime": 1329.0014, | |
| "eval_ikala_samples_per_second": 10.327, | |
| "eval_ikala_steps_per_second": 2.582, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_oasst_export_accuracy": 0.5890918246383349, | |
| "eval_oasst_export_loss": 2.033203125, | |
| "eval_oasst_export_runtime": 207.1646, | |
| "eval_oasst_export_samples_per_second": 10.132, | |
| "eval_oasst_export_steps_per_second": 2.534, | |
| "step": 2000 | |
| } | |
| ], | |
| "max_steps": 43960, | |
| "num_train_epochs": 8, | |
| "total_flos": 952516817190912.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |