| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3520, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.028409090909090908, | |
| "grad_norm": 1.4697265625, | |
| "learning_rate": 0.00019999601726381413, | |
| "loss": 2.9105, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.056818181818181816, | |
| "grad_norm": 1.3994140625, | |
| "learning_rate": 0.00019998406937250034, | |
| "loss": 2.5482, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08522727272727272, | |
| "grad_norm": 1.5947265625, | |
| "learning_rate": 0.00019996415727776455, | |
| "loss": 2.5436, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 1.22265625, | |
| "learning_rate": 0.0001999362825656992, | |
| "loss": 2.5159, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14204545454545456, | |
| "grad_norm": 1.2119140625, | |
| "learning_rate": 0.00019990044745665672, | |
| "loss": 2.4464, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17045454545454544, | |
| "grad_norm": 1.23828125, | |
| "learning_rate": 0.0001998566548050729, | |
| "loss": 2.3599, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.19886363636363635, | |
| "grad_norm": 1.2744140625, | |
| "learning_rate": 0.00019980490809923926, | |
| "loss": 2.5164, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 1.1611328125, | |
| "learning_rate": 0.00019974521146102537, | |
| "loss": 2.4555, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2556818181818182, | |
| "grad_norm": 1.2080078125, | |
| "learning_rate": 0.00019967756964555045, | |
| "loss": 2.3519, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2840909090909091, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.0001996019880408046, | |
| "loss": 2.3518, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 1.4345703125, | |
| "learning_rate": 0.0001995184726672197, | |
| "loss": 2.3879, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 1.16015625, | |
| "learning_rate": 0.00019942703017718975, | |
| "loss": 2.3626, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3693181818181818, | |
| "grad_norm": 1.103515625, | |
| "learning_rate": 0.000199327667854541, | |
| "loss": 2.3314, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3977272727272727, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 0.00019922039361395185, | |
| "loss": 2.4043, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.42613636363636365, | |
| "grad_norm": 1.0341796875, | |
| "learning_rate": 0.00019910521600032227, | |
| "loss": 2.3263, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.986328125, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 2.3119, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.48295454545454547, | |
| "grad_norm": 1.0888671875, | |
| "learning_rate": 0.00019885118798051605, | |
| "loss": 2.2697, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5113636363636364, | |
| "grad_norm": 0.9833984375, | |
| "learning_rate": 0.00019871235780887113, | |
| "loss": 2.2761, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5397727272727273, | |
| "grad_norm": 1.068359375, | |
| "learning_rate": 0.00019856566473163746, | |
| "loss": 2.3054, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "grad_norm": 1.10546875, | |
| "learning_rate": 0.0001984111204336116, | |
| "loss": 2.2658, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5965909090909091, | |
| "grad_norm": 1.021484375, | |
| "learning_rate": 0.00019824873722497694, | |
| "loss": 2.2965, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.986328125, | |
| "learning_rate": 0.00019807852804032305, | |
| "loss": 2.303, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6534090909090909, | |
| "grad_norm": 0.9755859375, | |
| "learning_rate": 0.00019790050643761552, | |
| "loss": 2.2398, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.94677734375, | |
| "learning_rate": 0.00019771468659711595, | |
| "loss": 2.257, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7102272727272727, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.0001975210833202524, | |
| "loss": 2.288, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7386363636363636, | |
| "grad_norm": 1.1494140625, | |
| "learning_rate": 0.00019731971202844036, | |
| "loss": 2.264, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7670454545454546, | |
| "grad_norm": 0.98779296875, | |
| "learning_rate": 0.00019711058876185447, | |
| "loss": 2.2359, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7954545454545454, | |
| "grad_norm": 1.0556640625, | |
| "learning_rate": 0.00019689373017815073, | |
| "loss": 2.287, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8238636363636364, | |
| "grad_norm": 0.85009765625, | |
| "learning_rate": 0.00019666915355113975, | |
| "loss": 2.2565, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8522727272727273, | |
| "grad_norm": 1.0673828125, | |
| "learning_rate": 0.00019643687676941068, | |
| "loss": 2.2853, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8806818181818182, | |
| "grad_norm": 0.9814453125, | |
| "learning_rate": 0.00019619691833490643, | |
| "loss": 2.2209, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.87744140625, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 2.2515, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.9345703125, | |
| "learning_rate": 0.0001956940335732209, | |
| "loss": 2.193, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9659090909090909, | |
| "grad_norm": 0.87451171875, | |
| "learning_rate": 0.0001954311473031864, | |
| "loss": 2.3074, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9943181818181818, | |
| "grad_norm": 0.91650390625, | |
| "learning_rate": 0.00019516065949147943, | |
| "loss": 2.2416, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0227272727272727, | |
| "grad_norm": 0.9345703125, | |
| "learning_rate": 0.00019488259168373197, | |
| "loss": 2.2029, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0511363636363635, | |
| "grad_norm": 1.904296875, | |
| "learning_rate": 0.00019459696602935837, | |
| "loss": 2.0812, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0795454545454546, | |
| "grad_norm": 0.93603515625, | |
| "learning_rate": 0.00019430380527979123, | |
| "loss": 2.1138, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.1079545454545454, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00019400313278666902, | |
| "loss": 2.1646, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 1.1142578125, | |
| "learning_rate": 0.0001936949724999762, | |
| "loss": 2.0984, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.1647727272727273, | |
| "grad_norm": 1.2470703125, | |
| "learning_rate": 0.00019337934896613516, | |
| "loss": 2.0974, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.1931818181818181, | |
| "grad_norm": 0.96044921875, | |
| "learning_rate": 0.00019305628732605137, | |
| "loss": 2.095, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.2215909090909092, | |
| "grad_norm": 0.9833984375, | |
| "learning_rate": 0.00019272581331311047, | |
| "loss": 2.1586, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.068359375, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 2.1677, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.2784090909090908, | |
| "grad_norm": 1.0478515625, | |
| "learning_rate": 0.00019204273405225587, | |
| "loss": 2.1594, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.3068181818181819, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00019169018321483198, | |
| "loss": 2.1347, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.3352272727272727, | |
| "grad_norm": 1.091796875, | |
| "learning_rate": 0.00019133032882119653, | |
| "loss": 2.0907, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 2.1914, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.3920454545454546, | |
| "grad_norm": 1.025390625, | |
| "learning_rate": 0.00019058882460117974, | |
| "loss": 2.1875, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.4204545454545454, | |
| "grad_norm": 1.0380859375, | |
| "learning_rate": 0.00019020723383911215, | |
| "loss": 2.1611, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.4488636363636362, | |
| "grad_norm": 0.9951171875, | |
| "learning_rate": 0.00018981845764475582, | |
| "loss": 2.113, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.4772727272727273, | |
| "grad_norm": 1.208984375, | |
| "learning_rate": 0.00018942252698597113, | |
| "loss": 2.1395, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.5056818181818183, | |
| "grad_norm": 1.0732421875, | |
| "learning_rate": 0.00018901947340050528, | |
| "loss": 2.1425, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.5340909090909092, | |
| "grad_norm": 1.0830078125, | |
| "learning_rate": 0.00018860932899348028, | |
| "loss": 2.0784, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 1.23046875, | |
| "learning_rate": 0.0001881921264348355, | |
| "loss": 2.0838, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.5909090909090908, | |
| "grad_norm": 1.30859375, | |
| "learning_rate": 0.00018776789895672558, | |
| "loss": 2.1142, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.6193181818181817, | |
| "grad_norm": 1.4462890625, | |
| "learning_rate": 0.00018733668035087302, | |
| "loss": 2.1271, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.6477272727272727, | |
| "grad_norm": 1.123046875, | |
| "learning_rate": 0.00018689850496587674, | |
| "loss": 2.0622, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.6761363636363638, | |
| "grad_norm": 1.119140625, | |
| "learning_rate": 0.00018645340770447595, | |
| "loss": 2.1017, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.7045454545454546, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 0.00018600142402077006, | |
| "loss": 2.0523, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.7329545454545454, | |
| "grad_norm": 1.0556640625, | |
| "learning_rate": 0.00018554258991739452, | |
| "loss": 2.1738, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.7613636363636362, | |
| "grad_norm": 1.1376953125, | |
| "learning_rate": 0.0001850769419426531, | |
| "loss": 2.107, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.7897727272727273, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.0001846045171876065, | |
| "loss": 2.1752, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.0634765625, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 2.0615, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.8465909090909092, | |
| "grad_norm": 1.205078125, | |
| "learning_rate": 0.00018363948839685636, | |
| "loss": 2.101, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.05859375, | |
| "learning_rate": 0.00018314696123025454, | |
| "loss": 2.1308, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.9034090909090908, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 0.000182647811015428, | |
| "loss": 2.1138, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.9318181818181817, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00018214207751204918, | |
| "loss": 2.0551, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.9602272727272727, | |
| "grad_norm": 1.2529296875, | |
| "learning_rate": 0.0001816298010041806, | |
| "loss": 2.1243, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.9886363636363638, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 0.0001811110222970659, | |
| "loss": 2.0663, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.0170454545454546, | |
| "grad_norm": 1.345703125, | |
| "learning_rate": 0.0001805857827138798, | |
| "loss": 2.1003, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.0454545454545454, | |
| "grad_norm": 1.0517578125, | |
| "learning_rate": 0.00018005412409243606, | |
| "loss": 2.0274, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.0738636363636362, | |
| "grad_norm": 1.021484375, | |
| "learning_rate": 0.0001795160887818553, | |
| "loss": 1.9693, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.102272727272727, | |
| "grad_norm": 1.1298828125, | |
| "learning_rate": 0.0001789717196391916, | |
| "loss": 2.0038, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.1306818181818183, | |
| "grad_norm": 1.080078125, | |
| "learning_rate": 0.00017842106002601856, | |
| "loss": 2.0249, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.159090909090909, | |
| "grad_norm": 1.146484375, | |
| "learning_rate": 0.00017786415380497553, | |
| "loss": 1.965, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 0.0001773010453362737, | |
| "loss": 1.9959, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.215909090909091, | |
| "grad_norm": 1.1259765625, | |
| "learning_rate": 0.00017673177947416258, | |
| "loss": 1.9325, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.2443181818181817, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.00017615640156335712, | |
| "loss": 2.031, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 1.24609375, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 1.9761, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.3011363636363638, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 0.0001749874934051401, | |
| "loss": 2.0198, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.3295454545454546, | |
| "grad_norm": 1.25390625, | |
| "learning_rate": 0.00017439405626678496, | |
| "loss": 1.9408, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.3579545454545454, | |
| "grad_norm": 1.1611328125, | |
| "learning_rate": 0.00017379469329043167, | |
| "loss": 2.0076, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.3863636363636362, | |
| "grad_norm": 1.1669921875, | |
| "learning_rate": 0.00017318945221817255, | |
| "loss": 2.0166, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.4147727272727275, | |
| "grad_norm": 1.212890625, | |
| "learning_rate": 0.00017257838126031797, | |
| "loss": 2.0593, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.4431818181818183, | |
| "grad_norm": 1.2314453125, | |
| "learning_rate": 0.00017196152909155628, | |
| "loss": 1.9916, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.471590909090909, | |
| "grad_norm": 1.1845703125, | |
| "learning_rate": 0.00017133894484707655, | |
| "loss": 1.9788, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.087890625, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 1.9791, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.528409090909091, | |
| "grad_norm": 1.2236328125, | |
| "learning_rate": 0.00017007677895070357, | |
| "loss": 2.0564, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.5568181818181817, | |
| "grad_norm": 1.2998046875, | |
| "learning_rate": 0.00016943729783628608, | |
| "loss": 1.9301, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.5852272727272725, | |
| "grad_norm": 1.1318359375, | |
| "learning_rate": 0.00016879228571309377, | |
| "loss": 1.9204, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.6136363636363638, | |
| "grad_norm": 1.1044921875, | |
| "learning_rate": 0.00016814179395938913, | |
| "loss": 2.0031, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.6420454545454546, | |
| "grad_norm": 1.2138671875, | |
| "learning_rate": 0.00016748587438991303, | |
| "loss": 1.9994, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.6704545454545454, | |
| "grad_norm": 1.1650390625, | |
| "learning_rate": 0.00016682457925175763, | |
| "loss": 2.0172, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.6988636363636362, | |
| "grad_norm": 1.26953125, | |
| "learning_rate": 0.00016615796122020443, | |
| "loss": 1.9542, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.423828125, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 1.9322, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.7556818181818183, | |
| "grad_norm": 1.3388671875, | |
| "learning_rate": 0.00016480896929376907, | |
| "loss": 1.9912, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.784090909090909, | |
| "grad_norm": 1.2529296875, | |
| "learning_rate": 0.0001641267028524661, | |
| "loss": 1.9818, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 0.00016343932841636456, | |
| "loss": 2.0132, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.840909090909091, | |
| "grad_norm": 1.4873046875, | |
| "learning_rate": 0.0001627469007380852, | |
| "loss": 1.9735, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.8693181818181817, | |
| "grad_norm": 1.1220703125, | |
| "learning_rate": 0.00016204947497276345, | |
| "loss": 1.9642, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.8977272727272725, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00016134710667365596, | |
| "loss": 1.9564, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.9261363636363638, | |
| "grad_norm": 1.142578125, | |
| "learning_rate": 0.00016063985178771555, | |
| "loss": 1.9701, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.9545454545454546, | |
| "grad_norm": 1.17578125, | |
| "learning_rate": 0.0001599277666511347, | |
| "loss": 1.913, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.9829545454545454, | |
| "grad_norm": 1.1396484375, | |
| "learning_rate": 0.00015921090798485832, | |
| "loss": 2.0151, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.0113636363636362, | |
| "grad_norm": 1.2216796875, | |
| "learning_rate": 0.0001584893328900653, | |
| "loss": 1.9299, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.039772727272727, | |
| "grad_norm": 1.1787109375, | |
| "learning_rate": 0.0001577630988436206, | |
| "loss": 1.899, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.0681818181818183, | |
| "grad_norm": 1.31640625, | |
| "learning_rate": 0.0001570322636934964, | |
| "loss": 1.8602, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.096590909090909, | |
| "grad_norm": 1.1279296875, | |
| "learning_rate": 0.00015629688565416478, | |
| "loss": 1.8292, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 1.2705078125, | |
| "learning_rate": 0.00015555702330196023, | |
| "loss": 1.891, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.153409090909091, | |
| "grad_norm": 1.208984375, | |
| "learning_rate": 0.000154812735570414, | |
| "loss": 1.8526, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 1.4580078125, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 1.8514, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.210227272727273, | |
| "grad_norm": 1.330078125, | |
| "learning_rate": 0.00015331112146121104, | |
| "loss": 1.8956, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.2386363636363638, | |
| "grad_norm": 1.771484375, | |
| "learning_rate": 0.00015255391469421128, | |
| "loss": 1.8339, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.2670454545454546, | |
| "grad_norm": 1.7021484375, | |
| "learning_rate": 0.00015179252175965633, | |
| "loss": 1.8952, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.2954545454545454, | |
| "grad_norm": 1.5029296875, | |
| "learning_rate": 0.00015102700330609, | |
| "loss": 1.8727, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.3238636363636362, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.00015025742031067317, | |
| "loss": 1.8974, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.3522727272727275, | |
| "grad_norm": 1.6708984375, | |
| "learning_rate": 0.00014948383407432678, | |
| "loss": 1.8314, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.3806818181818183, | |
| "grad_norm": 1.6494140625, | |
| "learning_rate": 0.00014870630621684872, | |
| "loss": 1.8768, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.409090909090909, | |
| "grad_norm": 1.1337890625, | |
| "learning_rate": 0.0001479248986720057, | |
| "loss": 1.8845, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.4375, | |
| "grad_norm": 1.2841796875, | |
| "learning_rate": 0.0001471396736825998, | |
| "loss": 1.8475, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.465909090909091, | |
| "grad_norm": 1.537109375, | |
| "learning_rate": 0.00014635069379551055, | |
| "loss": 1.8741, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.4943181818181817, | |
| "grad_norm": 1.58203125, | |
| "learning_rate": 0.00014555802185671298, | |
| "loss": 1.8274, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.5227272727272725, | |
| "grad_norm": 1.306640625, | |
| "learning_rate": 0.00014476172100627127, | |
| "loss": 1.8671, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.5511363636363638, | |
| "grad_norm": 1.4541015625, | |
| "learning_rate": 0.00014396185467330973, | |
| "loss": 1.8998, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.5795454545454546, | |
| "grad_norm": 1.3642578125, | |
| "learning_rate": 0.00014315848657096004, | |
| "loss": 1.8484, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.6079545454545454, | |
| "grad_norm": 1.31640625, | |
| "learning_rate": 0.00014235168069128657, | |
| "loss": 1.9098, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 1.51171875, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 1.8746, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.6647727272727275, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.00014072801293228188, | |
| "loss": 1.9036, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.6931818181818183, | |
| "grad_norm": 1.205078125, | |
| "learning_rate": 0.00013991128038575741, | |
| "loss": 1.9155, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.721590909090909, | |
| "grad_norm": 1.373046875, | |
| "learning_rate": 0.00013909136871722067, | |
| "loss": 1.8956, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.2607421875, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 1.8862, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.778409090909091, | |
| "grad_norm": 1.5869140625, | |
| "learning_rate": 0.00013744226950148972, | |
| "loss": 1.8749, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.8068181818181817, | |
| "grad_norm": 1.3154296875, | |
| "learning_rate": 0.00013661321331283796, | |
| "loss": 1.8831, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.8352272727272725, | |
| "grad_norm": 1.25, | |
| "learning_rate": 0.00013578124070879534, | |
| "loss": 1.8459, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.8636363636363638, | |
| "grad_norm": 1.1337890625, | |
| "learning_rate": 0.00013494641795990986, | |
| "loss": 1.8644, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.8920454545454546, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 0.00013410881156375684, | |
| "loss": 1.8569, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.9204545454545454, | |
| "grad_norm": 1.76953125, | |
| "learning_rate": 0.00013326848823964243, | |
| "loss": 1.884, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.9488636363636362, | |
| "grad_norm": 1.3408203125, | |
| "learning_rate": 0.00013242551492328875, | |
| "loss": 1.8759, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.9772727272727275, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.0001315799587615025, | |
| "loss": 1.8104, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.005681818181818, | |
| "grad_norm": 1.169921875, | |
| "learning_rate": 0.0001307318871068261, | |
| "loss": 1.8617, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.034090909090909, | |
| "grad_norm": 1.2158203125, | |
| "learning_rate": 0.00012988136751217291, | |
| "loss": 1.7548, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.0625, | |
| "grad_norm": 1.30078125, | |
| "learning_rate": 0.00012902846772544624, | |
| "loss": 1.7547, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.090909090909091, | |
| "grad_norm": 1.2236328125, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 1.7734, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.119318181818182, | |
| "grad_norm": 1.259765625, | |
| "learning_rate": 0.000127315799509942, | |
| "loss": 1.7146, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.1477272727272725, | |
| "grad_norm": 1.4580078125, | |
| "learning_rate": 0.0001264561675032779, | |
| "loss": 1.788, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.176136363636363, | |
| "grad_norm": 1.5205078125, | |
| "learning_rate": 0.00012559442813790076, | |
| "loss": 1.8247, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.204545454545454, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 0.00012473065005542155, | |
| "loss": 1.7217, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.232954545454546, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 0.00012386490205984488, | |
| "loss": 1.772, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.261363636363637, | |
| "grad_norm": 1.2919921875, | |
| "learning_rate": 0.00012299725311208808, | |
| "loss": 1.7837, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.2897727272727275, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00012212777232448838, | |
| "loss": 1.7954, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.318181818181818, | |
| "grad_norm": 1.5068359375, | |
| "learning_rate": 0.00012125652895529766, | |
| "loss": 1.7202, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.346590909090909, | |
| "grad_norm": 1.56640625, | |
| "learning_rate": 0.00012038359240316589, | |
| "loss": 1.7618, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 1.388671875, | |
| "learning_rate": 0.00011950903220161285, | |
| "loss": 1.758, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.403409090909091, | |
| "grad_norm": 1.6416015625, | |
| "learning_rate": 0.0001186329180134898, | |
| "loss": 1.8011, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.431818181818182, | |
| "grad_norm": 1.4912109375, | |
| "learning_rate": 0.00011775531962543036, | |
| "loss": 1.7798, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.4602272727272725, | |
| "grad_norm": 1.65234375, | |
| "learning_rate": 0.0001168763069422916, | |
| "loss": 1.7664, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.488636363636363, | |
| "grad_norm": 1.37109375, | |
| "learning_rate": 0.00011599594998158602, | |
| "loss": 1.7301, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.517045454545455, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 0.00011511431886790407, | |
| "loss": 1.7749, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 1.3369140625, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 1.7646, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.573863636363637, | |
| "grad_norm": 1.2080078125, | |
| "learning_rate": 0.00011334751518184061, | |
| "loss": 1.7962, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.6022727272727275, | |
| "grad_norm": 1.3974609375, | |
| "learning_rate": 0.0001124624833437186, | |
| "loss": 1.805, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.630681818181818, | |
| "grad_norm": 1.392578125, | |
| "learning_rate": 0.00011157645880992902, | |
| "loss": 1.7486, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.659090909090909, | |
| "grad_norm": 1.404296875, | |
| "learning_rate": 0.00011068951215651132, | |
| "loss": 1.7962, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.6875, | |
| "grad_norm": 1.380859375, | |
| "learning_rate": 0.0001098017140329561, | |
| "loss": 1.7722, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.715909090909091, | |
| "grad_norm": 1.580078125, | |
| "learning_rate": 0.0001089131351565776, | |
| "loss": 1.7477, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.744318181818182, | |
| "grad_norm": 1.490234375, | |
| "learning_rate": 0.00010802384630688078, | |
| "loss": 1.7825, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.7727272727272725, | |
| "grad_norm": 1.330078125, | |
| "learning_rate": 0.00010713391831992323, | |
| "loss": 1.7713, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.801136363636363, | |
| "grad_norm": 1.4033203125, | |
| "learning_rate": 0.00010624342208267292, | |
| "loss": 1.74, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.829545454545455, | |
| "grad_norm": 1.6806640625, | |
| "learning_rate": 0.00010535242852736151, | |
| "loss": 1.8173, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.857954545454545, | |
| "grad_norm": 1.603515625, | |
| "learning_rate": 0.00010446100862583459, | |
| "loss": 1.7703, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.886363636363637, | |
| "grad_norm": 1.4755859375, | |
| "learning_rate": 0.00010356923338389806, | |
| "loss": 1.8032, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.9147727272727275, | |
| "grad_norm": 1.4833984375, | |
| "learning_rate": 0.00010267717383566246, | |
| "loss": 1.8321, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.943181818181818, | |
| "grad_norm": 1.4462890625, | |
| "learning_rate": 0.0001017849010378846, | |
| "loss": 1.792, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.971590909090909, | |
| "grad_norm": 1.333984375, | |
| "learning_rate": 0.00010089248606430775, | |
| "loss": 1.7765, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 0.0001, | |
| "loss": 1.8083, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.028409090909091, | |
| "grad_norm": 1.482421875, | |
| "learning_rate": 9.910751393569229e-05, | |
| "loss": 1.7054, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 5.056818181818182, | |
| "grad_norm": 1.30859375, | |
| "learning_rate": 9.821509896211539e-05, | |
| "loss": 1.6561, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.0852272727272725, | |
| "grad_norm": 1.3037109375, | |
| "learning_rate": 9.732282616433756e-05, | |
| "loss": 1.6723, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 5.113636363636363, | |
| "grad_norm": 1.283203125, | |
| "learning_rate": 9.643076661610196e-05, | |
| "loss": 1.6675, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.142045454545454, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 9.553899137416545e-05, | |
| "loss": 1.6865, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 5.170454545454546, | |
| "grad_norm": 1.6669921875, | |
| "learning_rate": 9.464757147263849e-05, | |
| "loss": 1.7227, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.198863636363637, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.37565779173271e-05, | |
| "loss": 1.7106, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 5.2272727272727275, | |
| "grad_norm": 1.4619140625, | |
| "learning_rate": 9.286608168007678e-05, | |
| "loss": 1.7399, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 5.255681818181818, | |
| "grad_norm": 1.4873046875, | |
| "learning_rate": 9.197615369311925e-05, | |
| "loss": 1.6774, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 5.284090909090909, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 9.108686484342241e-05, | |
| "loss": 1.675, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 5.3125, | |
| "grad_norm": 1.3916015625, | |
| "learning_rate": 9.019828596704394e-05, | |
| "loss": 1.6782, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 5.340909090909091, | |
| "grad_norm": 1.376953125, | |
| "learning_rate": 8.931048784348875e-05, | |
| "loss": 1.6725, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 5.369318181818182, | |
| "grad_norm": 1.50390625, | |
| "learning_rate": 8.842354119007099e-05, | |
| "loss": 1.6885, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 5.3977272727272725, | |
| "grad_norm": 1.3623046875, | |
| "learning_rate": 8.753751665628141e-05, | |
| "loss": 1.6701, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.426136363636363, | |
| "grad_norm": 2.06640625, | |
| "learning_rate": 8.665248481815941e-05, | |
| "loss": 1.7138, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 1.341796875, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 1.6776, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 5.482954545454546, | |
| "grad_norm": 1.62109375, | |
| "learning_rate": 8.488568113209593e-05, | |
| "loss": 1.6245, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 5.511363636363637, | |
| "grad_norm": 1.4208984375, | |
| "learning_rate": 8.400405001841399e-05, | |
| "loss": 1.6695, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 5.5397727272727275, | |
| "grad_norm": 1.439453125, | |
| "learning_rate": 8.312369305770843e-05, | |
| "loss": 1.7594, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 5.568181818181818, | |
| "grad_norm": 1.873046875, | |
| "learning_rate": 8.224468037456969e-05, | |
| "loss": 1.6707, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 5.596590909090909, | |
| "grad_norm": 1.6162109375, | |
| "learning_rate": 8.136708198651021e-05, | |
| "loss": 1.7541, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 5.625, | |
| "grad_norm": 1.51171875, | |
| "learning_rate": 8.049096779838719e-05, | |
| "loss": 1.6912, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 5.653409090909091, | |
| "grad_norm": 1.6669921875, | |
| "learning_rate": 7.961640759683416e-05, | |
| "loss": 1.6736, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 5.681818181818182, | |
| "grad_norm": 1.654296875, | |
| "learning_rate": 7.874347104470234e-05, | |
| "loss": 1.6927, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.7102272727272725, | |
| "grad_norm": 1.8779296875, | |
| "learning_rate": 7.787222767551164e-05, | |
| "loss": 1.6773, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.738636363636363, | |
| "grad_norm": 1.55078125, | |
| "learning_rate": 7.700274688791196e-05, | |
| "loss": 1.6698, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.767045454545455, | |
| "grad_norm": 1.400390625, | |
| "learning_rate": 7.613509794015517e-05, | |
| "loss": 1.6778, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 5.795454545454545, | |
| "grad_norm": 1.3837890625, | |
| "learning_rate": 7.526934994457844e-05, | |
| "loss": 1.7331, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 5.823863636363637, | |
| "grad_norm": 1.6669921875, | |
| "learning_rate": 7.440557186209926e-05, | |
| "loss": 1.6646, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.8522727272727275, | |
| "grad_norm": 1.4169921875, | |
| "learning_rate": 7.354383249672212e-05, | |
| "loss": 1.6603, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 5.880681818181818, | |
| "grad_norm": 2.63671875, | |
| "learning_rate": 7.268420049005807e-05, | |
| "loss": 1.6668, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.909090909090909, | |
| "grad_norm": 1.6923828125, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 1.6975, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.9375, | |
| "grad_norm": 1.7412109375, | |
| "learning_rate": 7.097153227455379e-05, | |
| "loss": 1.6874, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.965909090909091, | |
| "grad_norm": 1.4111328125, | |
| "learning_rate": 7.011863248782711e-05, | |
| "loss": 1.7143, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.994318181818182, | |
| "grad_norm": 1.3232421875, | |
| "learning_rate": 6.92681128931739e-05, | |
| "loss": 1.6843, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 6.0227272727272725, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 6.842004123849752e-05, | |
| "loss": 1.5913, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.051136363636363, | |
| "grad_norm": 1.6064453125, | |
| "learning_rate": 6.757448507671128e-05, | |
| "loss": 1.6041, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 6.079545454545454, | |
| "grad_norm": 1.4404296875, | |
| "learning_rate": 6.673151176035762e-05, | |
| "loss": 1.6045, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.107954545454546, | |
| "grad_norm": 1.39453125, | |
| "learning_rate": 6.589118843624315e-05, | |
| "loss": 1.5673, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 6.136363636363637, | |
| "grad_norm": 1.68359375, | |
| "learning_rate": 6.505358204009017e-05, | |
| "loss": 1.603, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 6.1647727272727275, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 6.421875929120469e-05, | |
| "loss": 1.6078, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 6.193181818181818, | |
| "grad_norm": 2.29296875, | |
| "learning_rate": 6.338678668716209e-05, | |
| "loss": 1.5896, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 6.221590909090909, | |
| "grad_norm": 1.5146484375, | |
| "learning_rate": 6.25577304985103e-05, | |
| "loss": 1.6124, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 1.6001, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.278409090909091, | |
| "grad_norm": 1.6513671875, | |
| "learning_rate": 6.090863128277938e-05, | |
| "loss": 1.6603, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 6.306818181818182, | |
| "grad_norm": 1.4775390625, | |
| "learning_rate": 6.008871961424258e-05, | |
| "loss": 1.6338, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 6.3352272727272725, | |
| "grad_norm": 1.5576171875, | |
| "learning_rate": 5.9271987067718125e-05, | |
| "loss": 1.6557, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 1.44921875, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 1.5855, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 6.392045454545454, | |
| "grad_norm": 1.35546875, | |
| "learning_rate": 5.7648319308713464e-05, | |
| "loss": 1.6695, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 6.420454545454546, | |
| "grad_norm": 1.3837890625, | |
| "learning_rate": 5.684151342903992e-05, | |
| "loss": 1.6349, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 6.448863636363637, | |
| "grad_norm": 1.669921875, | |
| "learning_rate": 5.603814532669032e-05, | |
| "loss": 1.6279, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 6.4772727272727275, | |
| "grad_norm": 1.5185546875, | |
| "learning_rate": 5.5238278993728756e-05, | |
| "loss": 1.5647, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 6.505681818181818, | |
| "grad_norm": 1.9384765625, | |
| "learning_rate": 5.4441978143287066e-05, | |
| "loss": 1.6135, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 6.534090909090909, | |
| "grad_norm": 1.4892578125, | |
| "learning_rate": 5.364930620448946e-05, | |
| "loss": 1.5748, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 6.5625, | |
| "grad_norm": 1.4716796875, | |
| "learning_rate": 5.286032631740023e-05, | |
| "loss": 1.644, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 6.590909090909091, | |
| "grad_norm": 1.513671875, | |
| "learning_rate": 5.207510132799436e-05, | |
| "loss": 1.6049, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 6.619318181818182, | |
| "grad_norm": 1.392578125, | |
| "learning_rate": 5.1293693783151275e-05, | |
| "loss": 1.6798, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 6.6477272727272725, | |
| "grad_norm": 1.748046875, | |
| "learning_rate": 5.051616592567323e-05, | |
| "loss": 1.6548, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 6.676136363636363, | |
| "grad_norm": 1.4150390625, | |
| "learning_rate": 4.9742579689326874e-05, | |
| "loss": 1.599, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 6.704545454545455, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.8972996693910054e-05, | |
| "loss": 1.6521, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 6.732954545454545, | |
| "grad_norm": 1.7119140625, | |
| "learning_rate": 4.82074782403437e-05, | |
| "loss": 1.6407, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 6.761363636363637, | |
| "grad_norm": 1.619140625, | |
| "learning_rate": 4.7446085305788725e-05, | |
| "loss": 1.6925, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 6.7897727272727275, | |
| "grad_norm": 1.900390625, | |
| "learning_rate": 4.668887853878896e-05, | |
| "loss": 1.6109, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 6.818181818181818, | |
| "grad_norm": 1.6572265625, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 1.6055, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.846590909090909, | |
| "grad_norm": 1.3701171875, | |
| "learning_rate": 4.518726442958599e-05, | |
| "loss": 1.5787, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 6.875, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.444297669803981e-05, | |
| "loss": 1.6368, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 6.903409090909091, | |
| "grad_norm": 1.49609375, | |
| "learning_rate": 4.370311434583525e-05, | |
| "loss": 1.6239, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 6.931818181818182, | |
| "grad_norm": 1.7333984375, | |
| "learning_rate": 4.296773630650358e-05, | |
| "loss": 1.5953, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 6.9602272727272725, | |
| "grad_norm": 1.302734375, | |
| "learning_rate": 4.223690115637944e-05, | |
| "loss": 1.5964, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 6.988636363636363, | |
| "grad_norm": 1.4521484375, | |
| "learning_rate": 4.15106671099347e-05, | |
| "loss": 1.6531, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.017045454545454, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 4.0789092015141714e-05, | |
| "loss": 1.5872, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 7.045454545454546, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 4.007223334886531e-05, | |
| "loss": 1.5554, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 7.073863636363637, | |
| "grad_norm": 1.517578125, | |
| "learning_rate": 3.9360148212284475e-05, | |
| "loss": 1.5189, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 7.1022727272727275, | |
| "grad_norm": 1.4599609375, | |
| "learning_rate": 3.865289332634407e-05, | |
| "loss": 1.5523, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.130681818181818, | |
| "grad_norm": 1.6494140625, | |
| "learning_rate": 3.7950525027236585e-05, | |
| "loss": 1.5453, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 7.159090909090909, | |
| "grad_norm": 1.4560546875, | |
| "learning_rate": 3.725309926191479e-05, | |
| "loss": 1.5612, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 7.1875, | |
| "grad_norm": 1.630859375, | |
| "learning_rate": 3.6560671583635467e-05, | |
| "loss": 1.6003, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 7.215909090909091, | |
| "grad_norm": 1.3662109375, | |
| "learning_rate": 3.5873297147533915e-05, | |
| "loss": 1.5927, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 7.244318181818182, | |
| "grad_norm": 1.43359375, | |
| "learning_rate": 3.519103070623096e-05, | |
| "loss": 1.545, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 1.5654296875, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 1.5439, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 7.301136363636363, | |
| "grad_norm": 1.845703125, | |
| "learning_rate": 3.384203877979559e-05, | |
| "loss": 1.5953, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 7.329545454545454, | |
| "grad_norm": 1.80859375, | |
| "learning_rate": 3.3175420748242406e-05, | |
| "loss": 1.5458, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 7.357954545454546, | |
| "grad_norm": 1.4873046875, | |
| "learning_rate": 3.2514125610086955e-05, | |
| "loss": 1.5612, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 7.386363636363637, | |
| "grad_norm": 1.5927734375, | |
| "learning_rate": 3.185820604061088e-05, | |
| "loss": 1.5528, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 7.4147727272727275, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 3.1207714286906256e-05, | |
| "loss": 1.5178, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 7.443181818181818, | |
| "grad_norm": 1.455078125, | |
| "learning_rate": 3.056270216371395e-05, | |
| "loss": 1.5612, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 7.471590909090909, | |
| "grad_norm": 1.580078125, | |
| "learning_rate": 2.9923221049296446e-05, | |
| "loss": 1.5789, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.5400390625, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 1.5545, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 7.528409090909091, | |
| "grad_norm": 1.4794921875, | |
| "learning_rate": 2.8661055152923456e-05, | |
| "loss": 1.5833, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 7.556818181818182, | |
| "grad_norm": 1.5126953125, | |
| "learning_rate": 2.8038470908443714e-05, | |
| "loss": 1.5997, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 7.5852272727272725, | |
| "grad_norm": 1.5869140625, | |
| "learning_rate": 2.742161873968202e-05, | |
| "loss": 1.5628, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 7.613636363636363, | |
| "grad_norm": 1.4697265625, | |
| "learning_rate": 2.681054778182748e-05, | |
| "loss": 1.5693, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 7.642045454545455, | |
| "grad_norm": 1.572265625, | |
| "learning_rate": 2.6205306709568354e-05, | |
| "loss": 1.6006, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 7.670454545454545, | |
| "grad_norm": 1.6005859375, | |
| "learning_rate": 2.5605943733215042e-05, | |
| "loss": 1.5808, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 7.698863636363637, | |
| "grad_norm": 1.3994140625, | |
| "learning_rate": 2.501250659485992e-05, | |
| "loss": 1.5769, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 7.7272727272727275, | |
| "grad_norm": 1.6845703125, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 1.6105, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 7.755681818181818, | |
| "grad_norm": 1.5673828125, | |
| "learning_rate": 2.38435984366429e-05, | |
| "loss": 1.5196, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 7.784090909090909, | |
| "grad_norm": 2.056640625, | |
| "learning_rate": 2.3268220525837437e-05, | |
| "loss": 1.5696, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 7.8125, | |
| "grad_norm": 1.521484375, | |
| "learning_rate": 2.26989546637263e-05, | |
| "loss": 1.5564, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 7.840909090909091, | |
| "grad_norm": 1.744140625, | |
| "learning_rate": 2.2135846195024513e-05, | |
| "loss": 1.5608, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 7.869318181818182, | |
| "grad_norm": 1.6572265625, | |
| "learning_rate": 2.157893997398146e-05, | |
| "loss": 1.5823, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 7.8977272727272725, | |
| "grad_norm": 1.552734375, | |
| "learning_rate": 2.1028280360808407e-05, | |
| "loss": 1.5475, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 7.926136363636363, | |
| "grad_norm": 1.919921875, | |
| "learning_rate": 2.0483911218144715e-05, | |
| "loss": 1.5616, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 7.954545454545455, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.994587590756397e-05, | |
| "loss": 1.5512, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 7.982954545454545, | |
| "grad_norm": 1.666015625, | |
| "learning_rate": 1.941421728612023e-05, | |
| "loss": 1.6264, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 8.011363636363637, | |
| "grad_norm": 1.4365234375, | |
| "learning_rate": 1.8888977702934085e-05, | |
| "loss": 1.5251, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 8.039772727272727, | |
| "grad_norm": 1.52734375, | |
| "learning_rate": 1.837019899581943e-05, | |
| "loss": 1.5552, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 8.068181818181818, | |
| "grad_norm": 1.443359375, | |
| "learning_rate": 1.7857922487950874e-05, | |
| "loss": 1.5443, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 8.096590909090908, | |
| "grad_norm": 1.4052734375, | |
| "learning_rate": 1.7352188984572026e-05, | |
| "loss": 1.5339, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 8.125, | |
| "grad_norm": 1.7041015625, | |
| "learning_rate": 1.6853038769745467e-05, | |
| "loss": 1.5153, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 8.153409090909092, | |
| "grad_norm": 1.615234375, | |
| "learning_rate": 1.6360511603143648e-05, | |
| "loss": 1.5135, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 1.568359375, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 1.5008, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 8.210227272727273, | |
| "grad_norm": 1.654296875, | |
| "learning_rate": 1.5395482812393514e-05, | |
| "loss": 1.5273, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 8.238636363636363, | |
| "grad_norm": 2.16015625, | |
| "learning_rate": 1.4923058057346929e-05, | |
| "loss": 1.548, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 8.267045454545455, | |
| "grad_norm": 1.5595703125, | |
| "learning_rate": 1.4457410082605483e-05, | |
| "loss": 1.516, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 8.295454545454545, | |
| "grad_norm": 1.5810546875, | |
| "learning_rate": 1.3998575979229944e-05, | |
| "loss": 1.5658, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 8.323863636363637, | |
| "grad_norm": 1.5517578125, | |
| "learning_rate": 1.3546592295524074e-05, | |
| "loss": 1.5368, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 8.352272727272727, | |
| "grad_norm": 1.4638671875, | |
| "learning_rate": 1.3101495034123313e-05, | |
| "loss": 1.5138, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 8.380681818181818, | |
| "grad_norm": 1.9521484375, | |
| "learning_rate": 1.2663319649127026e-05, | |
| "loss": 1.5193, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 8.409090909090908, | |
| "grad_norm": 1.3642578125, | |
| "learning_rate": 1.2232101043274436e-05, | |
| "loss": 1.5136, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 8.4375, | |
| "grad_norm": 1.376953125, | |
| "learning_rate": 1.1807873565164506e-05, | |
| "loss": 1.487, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 8.465909090909092, | |
| "grad_norm": 1.869140625, | |
| "learning_rate": 1.139067100651976e-05, | |
| "loss": 1.6046, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 8.494318181818182, | |
| "grad_norm": 1.9716796875, | |
| "learning_rate": 1.0980526599494733e-05, | |
| "loss": 1.5021, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 8.522727272727273, | |
| "grad_norm": 1.427734375, | |
| "learning_rate": 1.057747301402887e-05, | |
| "loss": 1.5226, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.551136363636363, | |
| "grad_norm": 2.681640625, | |
| "learning_rate": 1.0181542355244167e-05, | |
| "loss": 1.5702, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 8.579545454545455, | |
| "grad_norm": 1.568359375, | |
| "learning_rate": 9.792766160887868e-06, | |
| "loss": 1.5281, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 8.607954545454545, | |
| "grad_norm": 1.7783203125, | |
| "learning_rate": 9.411175398820271e-06, | |
| "loss": 1.5334, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 8.636363636363637, | |
| "grad_norm": 1.3994140625, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 1.5025, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 8.664772727272727, | |
| "grad_norm": 1.4384765625, | |
| "learning_rate": 8.669671178803485e-06, | |
| "loss": 1.6133, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 8.693181818181818, | |
| "grad_norm": 1.6689453125, | |
| "learning_rate": 8.309816785168034e-06, | |
| "loss": 1.4961, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 8.721590909090908, | |
| "grad_norm": 1.4951171875, | |
| "learning_rate": 7.95726594774413e-06, | |
| "loss": 1.5019, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 1.66796875, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 1.4871, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 8.778409090909092, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 7.2741866868895395e-06, | |
| "loss": 1.5037, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 8.806818181818182, | |
| "grad_norm": 1.6630859375, | |
| "learning_rate": 6.943712673948644e-06, | |
| "loss": 1.5642, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 8.835227272727273, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 6.620651033864844e-06, | |
| "loss": 1.5448, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 8.863636363636363, | |
| "grad_norm": 1.607421875, | |
| "learning_rate": 6.3050275000238414e-06, | |
| "loss": 1.5506, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 8.892045454545455, | |
| "grad_norm": 1.5, | |
| "learning_rate": 5.996867213330992e-06, | |
| "loss": 1.5726, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 8.920454545454545, | |
| "grad_norm": 1.669921875, | |
| "learning_rate": 5.696194720208792e-06, | |
| "loss": 1.5486, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 8.948863636363637, | |
| "grad_norm": 1.9248046875, | |
| "learning_rate": 5.403033970641647e-06, | |
| "loss": 1.5235, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 8.977272727272727, | |
| "grad_norm": 1.5751953125, | |
| "learning_rate": 5.1174083162680465e-06, | |
| "loss": 1.5525, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 9.005681818181818, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 4.839340508520562e-06, | |
| "loss": 1.5572, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 9.034090909090908, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.568852696813619e-06, | |
| "loss": 1.5289, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 9.0625, | |
| "grad_norm": 1.5341796875, | |
| "learning_rate": 4.305966426779118e-06, | |
| "loss": 1.5175, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 1.6708984375, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 1.5103, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 9.119318181818182, | |
| "grad_norm": 1.697265625, | |
| "learning_rate": 3.8030816650935776e-06, | |
| "loss": 1.5117, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 9.147727272727273, | |
| "grad_norm": 1.548828125, | |
| "learning_rate": 3.5631232305893046e-06, | |
| "loss": 1.4877, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 9.176136363636363, | |
| "grad_norm": 1.6552734375, | |
| "learning_rate": 3.330846448860259e-06, | |
| "loss": 1.5132, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 9.204545454545455, | |
| "grad_norm": 1.5751953125, | |
| "learning_rate": 3.1062698218492724e-06, | |
| "loss": 1.488, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 9.232954545454545, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 2.889411238145545e-06, | |
| "loss": 1.5202, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 9.261363636363637, | |
| "grad_norm": 1.51953125, | |
| "learning_rate": 2.6802879715596585e-06, | |
| "loss": 1.5516, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 9.289772727272727, | |
| "grad_norm": 1.626953125, | |
| "learning_rate": 2.4789166797476228e-06, | |
| "loss": 1.4951, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 9.318181818181818, | |
| "grad_norm": 1.8134765625, | |
| "learning_rate": 2.2853134028840594e-06, | |
| "loss": 1.5397, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 9.346590909090908, | |
| "grad_norm": 1.62109375, | |
| "learning_rate": 2.0994935623844692e-06, | |
| "loss": 1.5236, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 9.375, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.921471959676957e-06, | |
| "loss": 1.4827, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 9.403409090909092, | |
| "grad_norm": 1.5693359375, | |
| "learning_rate": 1.751262775023077e-06, | |
| "loss": 1.5021, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 9.431818181818182, | |
| "grad_norm": 1.46484375, | |
| "learning_rate": 1.5888795663883904e-06, | |
| "loss": 1.4966, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 9.460227272727273, | |
| "grad_norm": 1.6455078125, | |
| "learning_rate": 1.4343352683625411e-06, | |
| "loss": 1.5455, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 9.488636363636363, | |
| "grad_norm": 1.666015625, | |
| "learning_rate": 1.2876421911288905e-06, | |
| "loss": 1.4876, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 9.517045454545455, | |
| "grad_norm": 1.6396484375, | |
| "learning_rate": 1.148812019483958e-06, | |
| "loss": 1.5912, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 9.545454545454545, | |
| "grad_norm": 2.1328125, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 1.5355, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 9.573863636363637, | |
| "grad_norm": 2.03515625, | |
| "learning_rate": 8.947839996777285e-07, | |
| "loss": 1.5147, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 9.602272727272727, | |
| "grad_norm": 1.806640625, | |
| "learning_rate": 7.796063860481595e-07, | |
| "loss": 1.53, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 9.630681818181818, | |
| "grad_norm": 1.5048828125, | |
| "learning_rate": 6.723321454590092e-07, | |
| "loss": 1.5051, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 9.659090909090908, | |
| "grad_norm": 1.701171875, | |
| "learning_rate": 5.729698228102653e-07, | |
| "loss": 1.4818, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 9.6875, | |
| "grad_norm": 1.6064453125, | |
| "learning_rate": 4.815273327803182e-07, | |
| "loss": 1.5438, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 9.715909090909092, | |
| "grad_norm": 2.04296875, | |
| "learning_rate": 3.9801195919541014e-07, | |
| "loss": 1.551, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 9.744318181818182, | |
| "grad_norm": 1.8115234375, | |
| "learning_rate": 3.224303544495766e-07, | |
| "loss": 1.4888, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 9.772727272727273, | |
| "grad_norm": 1.923828125, | |
| "learning_rate": 2.547885389746485e-07, | |
| "loss": 1.5491, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 9.801136363636363, | |
| "grad_norm": 2.728515625, | |
| "learning_rate": 1.9509190076074657e-07, | |
| "loss": 1.5374, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 9.829545454545455, | |
| "grad_norm": 1.5556640625, | |
| "learning_rate": 1.4334519492711362e-07, | |
| "loss": 1.5385, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 9.857954545454545, | |
| "grad_norm": 1.5185546875, | |
| "learning_rate": 9.955254334328423e-08, | |
| "loss": 1.4884, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 9.886363636363637, | |
| "grad_norm": 1.48046875, | |
| "learning_rate": 6.37174343008251e-08, | |
| "loss": 1.5733, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 9.914772727272727, | |
| "grad_norm": 1.8115234375, | |
| "learning_rate": 3.584272223546847e-08, | |
| "loss": 1.5452, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 9.943181818181818, | |
| "grad_norm": 1.5380859375, | |
| "learning_rate": 1.593062749967178e-08, | |
| "loss": 1.5122, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 9.971590909090908, | |
| "grad_norm": 1.63671875, | |
| "learning_rate": 3.982736185859093e-09, | |
| "loss": 1.5273, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.5048828125, | |
| "learning_rate": 0.0, | |
| "loss": 1.5284, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 3520, | |
| "total_flos": 2.31165810180096e+17, | |
| "train_loss": 1.80265548852357, | |
| "train_runtime": 3529.4078, | |
| "train_samples_per_second": 3.989, | |
| "train_steps_per_second": 0.997 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3520, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2.31165810180096e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |