|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.2416626389560174, |
|
"eval_steps": 500, |
|
"global_step": 1000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": NaN, |
|
"learning_rate": 2.0000000000000002e-07, |
|
"loss": 2.0414, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 53.515953063964844, |
|
"learning_rate": 6.5e-07, |
|
"loss": 1.8355, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 27.920907974243164, |
|
"learning_rate": 1.15e-06, |
|
"loss": 1.5172, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 20.47172737121582, |
|
"learning_rate": 1.65e-06, |
|
"loss": 1.2654, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 8.903529167175293, |
|
"learning_rate": 2.1499999999999997e-06, |
|
"loss": 1.1095, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 5.902841567993164, |
|
"learning_rate": 2.65e-06, |
|
"loss": 1.0017, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 2.9340908527374268, |
|
"learning_rate": 3.1500000000000003e-06, |
|
"loss": 0.8846, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 8.701367378234863, |
|
"learning_rate": 3.6499999999999998e-06, |
|
"loss": 0.8572, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 1.3205362558364868, |
|
"learning_rate": 4.15e-06, |
|
"loss": 0.839, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 3.0334558486938477, |
|
"learning_rate": 4.65e-06, |
|
"loss": 0.7838, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 1.8086520433425903, |
|
"learning_rate": 5.15e-06, |
|
"loss": 0.7782, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 29.50135040283203, |
|
"learning_rate": 5.65e-06, |
|
"loss": 0.7883, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 3.2708683013916016, |
|
"learning_rate": 6.15e-06, |
|
"loss": 0.7961, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 33.43790817260742, |
|
"learning_rate": 6.650000000000001e-06, |
|
"loss": 0.7626, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 1.2334959506988525, |
|
"learning_rate": 7.15e-06, |
|
"loss": 0.7732, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 1.7489686012268066, |
|
"learning_rate": 7.65e-06, |
|
"loss": 0.7715, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 3.358823776245117, |
|
"learning_rate": 8.15e-06, |
|
"loss": 0.7778, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 4.348245620727539, |
|
"learning_rate": 8.65e-06, |
|
"loss": 0.7619, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.9023125767707825, |
|
"learning_rate": 9.15e-06, |
|
"loss": 0.7746, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.9158114790916443, |
|
"learning_rate": 9.65e-06, |
|
"loss": 0.7592, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.6256226301193237, |
|
"learning_rate": 1.0150000000000001e-05, |
|
"loss": 0.7787, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.3805778920650482, |
|
"learning_rate": 1.065e-05, |
|
"loss": 0.7811, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.6968041062355042, |
|
"learning_rate": 1.115e-05, |
|
"loss": 0.73, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.9035410284996033, |
|
"learning_rate": 1.1650000000000002e-05, |
|
"loss": 0.7518, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 1.6727488040924072, |
|
"learning_rate": 1.215e-05, |
|
"loss": 0.7805, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.4153461158275604, |
|
"learning_rate": 1.2650000000000001e-05, |
|
"loss": 0.7706, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 1.4846651554107666, |
|
"learning_rate": 1.3150000000000001e-05, |
|
"loss": 0.7492, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 2.6631388664245605, |
|
"learning_rate": 1.3650000000000001e-05, |
|
"loss": 0.7681, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.6325013041496277, |
|
"learning_rate": 1.415e-05, |
|
"loss": 0.7753, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.877907395362854, |
|
"learning_rate": 1.465e-05, |
|
"loss": 0.7188, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.37142279744148254, |
|
"learning_rate": 1.515e-05, |
|
"loss": 0.7204, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.7337246537208557, |
|
"learning_rate": 1.565e-05, |
|
"loss": 0.778, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.5847220420837402, |
|
"learning_rate": 1.6150000000000003e-05, |
|
"loss": 0.7288, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 3.745180606842041, |
|
"learning_rate": 1.665e-05, |
|
"loss": 0.7531, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.357301265001297, |
|
"learning_rate": 1.7150000000000004e-05, |
|
"loss": 0.7448, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.9032486081123352, |
|
"learning_rate": 1.765e-05, |
|
"loss": 0.7651, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.2864232361316681, |
|
"learning_rate": 1.815e-05, |
|
"loss": 0.7193, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 1.8560261726379395, |
|
"learning_rate": 1.865e-05, |
|
"loss": 0.7421, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.2971792221069336, |
|
"learning_rate": 1.915e-05, |
|
"loss": 0.7171, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.40850627422332764, |
|
"learning_rate": 1.9650000000000003e-05, |
|
"loss": 0.7459, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.3934139311313629, |
|
"learning_rate": 2.0150000000000002e-05, |
|
"loss": 0.7205, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 1.8674131631851196, |
|
"learning_rate": 2.065e-05, |
|
"loss": 0.752, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.36707818508148193, |
|
"learning_rate": 2.115e-05, |
|
"loss": 0.7494, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 0.5044310092926025, |
|
"learning_rate": 2.165e-05, |
|
"loss": 0.7595, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 1.9025150537490845, |
|
"learning_rate": 2.215e-05, |
|
"loss": 0.7379, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 0.5004140138626099, |
|
"learning_rate": 2.265e-05, |
|
"loss": 0.7886, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 3.544482707977295, |
|
"learning_rate": 2.3150000000000004e-05, |
|
"loss": 0.7259, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.33023321628570557, |
|
"learning_rate": 2.365e-05, |
|
"loss": 0.7333, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.3548080325126648, |
|
"learning_rate": 2.415e-05, |
|
"loss": 0.7527, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.42450150847435, |
|
"learning_rate": 2.465e-05, |
|
"loss": 0.7443, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"eval_loss": 0.7428915500640869, |
|
"eval_runtime": 68.3143, |
|
"eval_samples_per_second": 29.276, |
|
"eval_steps_per_second": 0.922, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.3299656808376312, |
|
"learning_rate": 2.515e-05, |
|
"loss": 0.7447, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.27258285880088806, |
|
"learning_rate": 2.5650000000000003e-05, |
|
"loss": 0.7278, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.23753009736537933, |
|
"learning_rate": 2.6150000000000002e-05, |
|
"loss": 0.7721, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 1.1074901819229126, |
|
"learning_rate": 2.6650000000000004e-05, |
|
"loss": 0.7383, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.428067147731781, |
|
"learning_rate": 2.7150000000000003e-05, |
|
"loss": 0.7508, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.35956478118896484, |
|
"learning_rate": 2.7650000000000005e-05, |
|
"loss": 0.7461, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.23017314076423645, |
|
"learning_rate": 2.815e-05, |
|
"loss": 0.7463, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.20370900630950928, |
|
"learning_rate": 2.865e-05, |
|
"loss": 0.7405, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.24586541950702667, |
|
"learning_rate": 2.915e-05, |
|
"loss": 0.6964, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.27857884764671326, |
|
"learning_rate": 2.965e-05, |
|
"loss": 0.7376, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.2655833065509796, |
|
"learning_rate": 3.015e-05, |
|
"loss": 0.7496, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.17174670100212097, |
|
"learning_rate": 3.065e-05, |
|
"loss": 0.7219, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.22892743349075317, |
|
"learning_rate": 3.115e-05, |
|
"loss": 0.7111, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.5913789868354797, |
|
"learning_rate": 3.1650000000000004e-05, |
|
"loss": 0.7097, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.25620004534721375, |
|
"learning_rate": 3.215e-05, |
|
"loss": 0.7306, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.3311476409435272, |
|
"learning_rate": 3.265e-05, |
|
"loss": 0.7406, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.4708893597126007, |
|
"learning_rate": 3.3150000000000006e-05, |
|
"loss": 0.7071, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.22469288110733032, |
|
"learning_rate": 3.3650000000000005e-05, |
|
"loss": 0.7116, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.2909330725669861, |
|
"learning_rate": 3.415e-05, |
|
"loss": 0.7592, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.21274766325950623, |
|
"learning_rate": 3.465e-05, |
|
"loss": 0.7144, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.23929230868816376, |
|
"learning_rate": 3.515e-05, |
|
"loss": 0.7382, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.3021218180656433, |
|
"learning_rate": 3.565e-05, |
|
"loss": 0.7287, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.23557321727275848, |
|
"learning_rate": 3.615e-05, |
|
"loss": 0.7475, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.22648084163665771, |
|
"learning_rate": 3.665e-05, |
|
"loss": 0.7457, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.2707761824131012, |
|
"learning_rate": 3.715e-05, |
|
"loss": 0.7322, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.2799661457538605, |
|
"learning_rate": 3.765e-05, |
|
"loss": 0.7682, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.326861172914505, |
|
"learning_rate": 3.8150000000000006e-05, |
|
"loss": 0.73, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.31985723972320557, |
|
"learning_rate": 3.8650000000000004e-05, |
|
"loss": 0.744, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.21637533605098724, |
|
"learning_rate": 3.915e-05, |
|
"loss": 0.7161, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.26102888584136963, |
|
"learning_rate": 3.965e-05, |
|
"loss": 0.7553, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.2469174712896347, |
|
"learning_rate": 4.015000000000001e-05, |
|
"loss": 0.7444, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.2022310495376587, |
|
"learning_rate": 4.065e-05, |
|
"loss": 0.7089, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.1729898750782013, |
|
"learning_rate": 4.115e-05, |
|
"loss": 0.7122, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.22450707852840424, |
|
"learning_rate": 4.165e-05, |
|
"loss": 0.7099, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.22389698028564453, |
|
"learning_rate": 4.215e-05, |
|
"loss": 0.7445, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.24962696433067322, |
|
"learning_rate": 4.265e-05, |
|
"loss": 0.7052, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.23780404031276703, |
|
"learning_rate": 4.315e-05, |
|
"loss": 0.7423, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.2434270679950714, |
|
"learning_rate": 4.3650000000000004e-05, |
|
"loss": 0.7291, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.3332350552082062, |
|
"learning_rate": 4.415e-05, |
|
"loss": 0.7306, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.5084495544433594, |
|
"learning_rate": 4.465e-05, |
|
"loss": 0.735, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.27552103996276855, |
|
"learning_rate": 4.5150000000000006e-05, |
|
"loss": 0.7082, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.2184416800737381, |
|
"learning_rate": 4.5650000000000005e-05, |
|
"loss": 0.7066, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.21738290786743164, |
|
"learning_rate": 4.6150000000000004e-05, |
|
"loss": 0.6977, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.21888123452663422, |
|
"learning_rate": 4.665e-05, |
|
"loss": 0.7083, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.2516074776649475, |
|
"learning_rate": 4.715e-05, |
|
"loss": 0.7451, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.39361557364463806, |
|
"learning_rate": 4.765e-05, |
|
"loss": 0.7013, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.2461744099855423, |
|
"learning_rate": 4.815e-05, |
|
"loss": 0.7325, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.2554089426994324, |
|
"learning_rate": 4.8650000000000003e-05, |
|
"loss": 0.7384, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.2717374861240387, |
|
"learning_rate": 4.915e-05, |
|
"loss": 0.7172, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.35919347405433655, |
|
"learning_rate": 4.965e-05, |
|
"loss": 0.6851, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"eval_loss": 0.7169972658157349, |
|
"eval_runtime": 67.4497, |
|
"eval_samples_per_second": 29.652, |
|
"eval_steps_per_second": 0.934, |
|
"step": 1000 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 20690, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 5, |
|
"save_steps": 500, |
|
"total_flos": 2.1633806834671288e+18, |
|
"train_batch_size": 4, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|