| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998001998001999, | |
| "eval_steps": 500, | |
| "global_step": 1251, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 4.9992117348705316e-05, | |
| "loss": 2.3263, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.9968474365716575e-05, | |
| "loss": 1.9238, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.992908596058501e-05, | |
| "loss": 1.8386, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.9873976972115636e-05, | |
| "loss": 1.7599, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 4.980318215270361e-05, | |
| "loss": 1.7486, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.9716746146418905e-05, | |
| "loss": 1.7243, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 4.9614723460853294e-05, | |
| "loss": 1.684, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 4.949717843274711e-05, | |
| "loss": 1.7019, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 4.93641851874178e-05, | |
| "loss": 1.6845, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.921582759201557e-05, | |
| "loss": 1.6427, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.905219920263573e-05, | |
| "loss": 1.6465, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.887340320532111e-05, | |
| "loss": 1.6448, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.86795523509917e-05, | |
| "loss": 1.6385, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.847076888434251e-05, | |
| "loss": 1.6059, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.824718446675464e-05, | |
| "loss": 1.6419, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.800894009326801e-05, | |
| "loss": 1.6043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.775618600366812e-05, | |
| "loss": 1.5975, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.7489081587743125e-05, | |
| "loss": 1.5865, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.7207795284770605e-05, | |
| "loss": 1.5931, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.69125044772978e-05, | |
| "loss": 1.5763, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.6603395379281975e-05, | |
| "loss": 1.5786, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.628066291866166e-05, | |
| "loss": 1.5494, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.5944510614432734e-05, | |
| "loss": 1.5663, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.55951504483069e-05, | |
| "loss": 1.5687, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.523280273103342e-05, | |
| "loss": 1.5446, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.485769596346849e-05, | |
| "loss": 1.5486, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.447006669247991e-05, | |
| "loss": 1.5278, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.407015936177762e-05, | |
| "loss": 1.5244, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.3658226157764704e-05, | |
| "loss": 1.5223, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.3234526850505456e-05, | |
| "loss": 1.5536, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.2799328629911323e-05, | |
| "loss": 1.5282, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.2352905937247654e-05, | |
| "loss": 1.5201, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.189554029206776e-05, | |
| "loss": 1.495, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.142752011468326e-05, | |
| "loss": 1.5137, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.094914054428272e-05, | |
| "loss": 1.4861, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.0460703252813326e-05, | |
| "loss": 1.5195, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.996251625474293e-05, | |
| "loss": 1.4974, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.945489371282237e-05, | |
| "loss": 1.5008, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.89381557399706e-05, | |
| "loss": 1.5005, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.84126281974077e-05, | |
| "loss": 1.4808, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.787864248916276e-05, | |
| "loss": 1.4921, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.7336535353086544e-05, | |
| "loss": 1.4761, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.6786648648500495e-05, | |
| "loss": 1.4805, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 3.622932914061611e-05, | |
| "loss": 1.4614, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.566492828186063e-05, | |
| "loss": 1.4709, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.509380199024684e-05, | |
| "loss": 1.497, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.451631042492693e-05, | |
| "loss": 1.4642, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.393281775907167e-05, | |
| "loss": 1.4579, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.3343691950218514e-05, | |
| "loss": 1.4586, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.2749304508233056e-05, | |
| "loss": 1.4664, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.215003026103041e-05, | |
| "loss": 1.4452, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.1546247118204235e-05, | |
| "loss": 1.4393, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.093833583271233e-05, | |
| "loss": 1.4462, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.0326679760769226e-05, | |
| "loss": 1.4334, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 2.9711664620097107e-05, | |
| "loss": 1.4495, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 2.9093678246687574e-05, | |
| "loss": 1.4096, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.847311035022753e-05, | |
| "loss": 1.4371, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.7850352268343594e-05, | |
| "loss": 1.425, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2.7225796719819778e-05, | |
| "loss": 1.4343, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.6599837556944353e-05, | |
| "loss": 1.4293, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.597286951714176e-05, | |
| "loss": 1.4137, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.534528797404646e-05, | |
| "loss": 1.4065, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.4717488688175512e-05, | |
| "loss": 1.4444, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.408986755735719e-05, | |
| "loss": 1.4088, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.3462820367073057e-05, | |
| "loss": 1.3999, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.2836742540870818e-05, | |
| "loss": 1.4124, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.2212028891005457e-05, | |
| "loss": 1.3908, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.1589073369465844e-05, | |
| "loss": 1.4017, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.0968268819543848e-05, | |
| "loss": 1.3934, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.035000672810261e-05, | |
| "loss": 1.3971, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.9734676978700177e-05, | |
| "loss": 1.3923, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 1.91226676057242e-05, | |
| "loss": 1.3962, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 1.8514364549692794e-05, | |
| "loss": 1.37, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.7910151413875717e-05, | |
| "loss": 1.3808, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 1.731040922238956e-05, | |
| "loss": 1.3868, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.6715516179919322e-05, | |
| "loss": 1.3889, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.6125847433217883e-05, | |
| "loss": 1.378, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.5541774834534024e-05, | |
| "loss": 1.3709, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.4963666707117852e-05, | |
| "loss": 1.3563, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.4391887612951685e-05, | |
| "loss": 1.3687, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.3826798122852868e-05, | |
| "loss": 1.3729, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.3268754589093468e-05, | |
| "loss": 1.3498, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.2718108920680193e-05, | |
| "loss": 1.3399, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.2175208361436328e-05, | |
| "loss": 1.3618, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.1640395271025572e-05, | |
| "loss": 1.3448, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.1114006909055876e-05, | |
| "loss": 1.3443, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.059637522239949e-05, | |
| "loss": 1.3523, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.0087826635863198e-05, | |
| "loss": 1.3462, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 9.58868184634095e-06, | |
| "loss": 1.3398, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 9.099255620578451e-06, | |
| "loss": 1.3426, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.619856596677459e-06, | |
| "loss": 1.3455, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.150787089464893e-06, | |
| "loss": 1.3317, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 7.69234289984942e-06, | |
| "loss": 1.3206, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 7.244813128285835e-06, | |
| "loss": 1.3388, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 6.8084799924647724e-06, | |
| "loss": 1.349, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 6.383618649342893e-06, | |
| "loss": 1.3558, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5.970497021625507e-06, | |
| "loss": 1.3238, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5.569375628811296e-06, | |
| "loss": 1.3243, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5.180507422905584e-06, | |
| "loss": 1.3033, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.804137628905692e-06, | |
| "loss": 1.3483, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 4.440503590159145e-06, | |
| "loss": 1.3327, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.089834618692048e-06, | |
| "loss": 1.3165, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.7523518506021594e-06, | |
| "loss": 1.3148, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 3.428268106607782e-06, | |
| "loss": 1.3136, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.1177877578404485e-06, | |
| "loss": 1.3285, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 2.8211065969660167e-06, | |
| "loss": 1.3074, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.5384117147154384e-06, | |
| "loss": 1.3004, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.26988138190308e-06, | |
| "loss": 1.3128, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.015684937006981e-06, | |
| "loss": 1.31, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 1.775982679381985e-06, | |
| "loss": 1.3127, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.5509257681730032e-06, | |
| "loss": 1.3108, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.340656126992243e-06, | |
| "loss": 1.2913, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.145306354420439e-06, | |
| "loss": 1.3126, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 9.649996403886086e-07, | |
| "loss": 1.3001, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 7.99849688492979e-07, | |
| "loss": 1.3104, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 6.499606442921163e-07, | |
| "loss": 1.3257, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.154270296314878e-07, | |
| "loss": 1.296, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.9633368303683696e-07, | |
| "loss": 1.3004, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.9275570621398463e-07, | |
| "loss": 1.2962, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.047584166887717e-07, | |
| "loss": 1.311, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.3239730661706307e-07, | |
| "loss": 1.3142, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 7.571800779069271e-08, | |
| "loss": 1.3046, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.4756262861521116e-08, | |
| "loss": 1.2985, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.53790280169009e-09, | |
| "loss": 1.313, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.883061428071159e-11, | |
| "loss": 1.306, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1251, | |
| "total_flos": 5.24659992301142e+17, | |
| "train_loss": 1.4494417681873177, | |
| "train_runtime": 7393.5069, | |
| "train_samples_per_second": 10.831, | |
| "train_steps_per_second": 0.169 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1251, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "total_flos": 5.24659992301142e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |