| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4023, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007457121551081283, | |
| "grad_norm": 1.4923322722027186, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6314, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014914243102162566, | |
| "grad_norm": 1.4277498493155112, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5957, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02237136465324385, | |
| "grad_norm": 1.3160638879347244, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5858, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02982848620432513, | |
| "grad_norm": 1.3286844636967108, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6083, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.037285607755406416, | |
| "grad_norm": 1.3457945081617781, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5833, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0447427293064877, | |
| "grad_norm": 1.1775640475079898, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5615, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05219985085756898, | |
| "grad_norm": 0.7851058774790824, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5599, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05965697240865026, | |
| "grad_norm": 0.6748980341595908, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5738, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06711409395973154, | |
| "grad_norm": 0.6098300182337207, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5513, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07457121551081283, | |
| "grad_norm": 0.6353294403272493, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5723, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08202833706189411, | |
| "grad_norm": 0.5674389209790128, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5666, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0894854586129754, | |
| "grad_norm": 0.6279447283125179, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5475, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09694258016405667, | |
| "grad_norm": 0.576962331333649, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5532, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10439970171513796, | |
| "grad_norm": 0.5745640803688579, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5561, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11185682326621924, | |
| "grad_norm": 0.568035600435591, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5542, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11931394481730052, | |
| "grad_norm": 0.5903955926152153, | |
| "learning_rate": 5e-06, | |
| "loss": 0.55, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1267710663683818, | |
| "grad_norm": 0.6272192755787677, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5577, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1342281879194631, | |
| "grad_norm": 0.621732319302288, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5579, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14168530947054436, | |
| "grad_norm": 0.582261312726868, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5691, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.14914243102162567, | |
| "grad_norm": 0.6038562541450465, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5441, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15659955257270694, | |
| "grad_norm": 0.6342330659293374, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5438, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16405667412378822, | |
| "grad_norm": 0.5605346796698036, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5552, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1715137956748695, | |
| "grad_norm": 0.5998445612806728, | |
| "learning_rate": 5e-06, | |
| "loss": 0.57, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1789709172259508, | |
| "grad_norm": 0.5714415956910529, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5588, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.18642803877703207, | |
| "grad_norm": 0.6368975283895788, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5435, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19388516032811334, | |
| "grad_norm": 0.5912645424796484, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5371, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.20134228187919462, | |
| "grad_norm": 0.5671791863380606, | |
| "learning_rate": 5e-06, | |
| "loss": 0.554, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.20879940343027592, | |
| "grad_norm": 0.5825115142497552, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5648, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2162565249813572, | |
| "grad_norm": 0.6148559353437262, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5626, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.22371364653243847, | |
| "grad_norm": 0.5992171256664753, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5495, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23117076808351977, | |
| "grad_norm": 0.5756923092743548, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5492, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.23862788963460105, | |
| "grad_norm": 0.5571239833558982, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5382, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.24608501118568232, | |
| "grad_norm": 0.6142758618521919, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5626, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2535421327367636, | |
| "grad_norm": 0.6071098554908272, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5537, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2609992542878449, | |
| "grad_norm": 0.5503873105742949, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5577, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2684563758389262, | |
| "grad_norm": 0.5837181817071831, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5514, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2759134973900075, | |
| "grad_norm": 0.6207096163948669, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5415, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2833706189410887, | |
| "grad_norm": 0.5430609166419872, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5641, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.29082774049217003, | |
| "grad_norm": 0.6540545927611947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5481, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.29828486204325133, | |
| "grad_norm": 0.5461956945004067, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5513, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3057419835943326, | |
| "grad_norm": 0.6695239493079291, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5638, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3131991051454139, | |
| "grad_norm": 0.586611622356162, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5492, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.32065622669649513, | |
| "grad_norm": 0.610109513110288, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5607, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.32811334824757643, | |
| "grad_norm": 0.6127769350008854, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5645, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.33557046979865773, | |
| "grad_norm": 0.6680654140600404, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5549, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.343027591349739, | |
| "grad_norm": 0.5826310904520773, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5548, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3504847129008203, | |
| "grad_norm": 0.6239330754324169, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5583, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3579418344519016, | |
| "grad_norm": 0.5877553356456545, | |
| "learning_rate": 5e-06, | |
| "loss": 0.545, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.36539895600298283, | |
| "grad_norm": 0.5829722537190941, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5395, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.37285607755406414, | |
| "grad_norm": 0.6621755183448562, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5456, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.38031319910514544, | |
| "grad_norm": 0.5591184552346828, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5367, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.3877703206562267, | |
| "grad_norm": 0.565634611705387, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5716, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.395227442207308, | |
| "grad_norm": 0.5782734602622104, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5545, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.40268456375838924, | |
| "grad_norm": 0.5695093375447697, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5646, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.41014168530947054, | |
| "grad_norm": 0.6396895709952651, | |
| "learning_rate": 5e-06, | |
| "loss": 0.559, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.41759880686055184, | |
| "grad_norm": 0.5626775345590961, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5398, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4250559284116331, | |
| "grad_norm": 0.6561792759757238, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5506, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4325130499627144, | |
| "grad_norm": 0.6503627380598381, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5624, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4399701715137957, | |
| "grad_norm": 0.5712768208534026, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5493, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.44742729306487694, | |
| "grad_norm": 0.632567133823459, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5495, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45488441461595824, | |
| "grad_norm": 0.5891856850275913, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5422, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.46234153616703955, | |
| "grad_norm": 0.5745761726721252, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5575, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4697986577181208, | |
| "grad_norm": 0.6304716115534362, | |
| "learning_rate": 5e-06, | |
| "loss": 0.552, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4772557792692021, | |
| "grad_norm": 0.5804599760466073, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5574, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.48471290082028334, | |
| "grad_norm": 0.5920559180390651, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5515, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.49217002237136465, | |
| "grad_norm": 0.5791705205755153, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5574, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.49962714392244595, | |
| "grad_norm": 0.6201405707302831, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5585, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5070842654735273, | |
| "grad_norm": 0.5753619878991011, | |
| "learning_rate": 5e-06, | |
| "loss": 0.556, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5145413870246085, | |
| "grad_norm": 0.5594987960581536, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5437, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5219985085756897, | |
| "grad_norm": 0.5703994092235947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5568, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5294556301267711, | |
| "grad_norm": 0.5834034437186236, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5431, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5369127516778524, | |
| "grad_norm": 0.560106669559214, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5552, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5443698732289336, | |
| "grad_norm": 0.5679786278378056, | |
| "learning_rate": 5e-06, | |
| "loss": 0.554, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.551826994780015, | |
| "grad_norm": 0.6429988192429542, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5495, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5592841163310962, | |
| "grad_norm": 0.6267938520430484, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5561, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5667412378821775, | |
| "grad_norm": 0.6138480606756376, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5532, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5741983594332588, | |
| "grad_norm": 0.6398337341244397, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5567, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5816554809843401, | |
| "grad_norm": 0.6201612138500541, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5411, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5891126025354213, | |
| "grad_norm": 0.5777153613729241, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5598, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5965697240865027, | |
| "grad_norm": 0.5831925884746647, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5505, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6040268456375839, | |
| "grad_norm": 0.5811154005536924, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5301, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6114839671886652, | |
| "grad_norm": 0.5578916945797876, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5616, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6189410887397464, | |
| "grad_norm": 0.6251996052710133, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5444, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6263982102908278, | |
| "grad_norm": 0.5440172219846166, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5596, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.633855331841909, | |
| "grad_norm": 0.5904453872308907, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5469, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6413124533929903, | |
| "grad_norm": 0.5659960741097526, | |
| "learning_rate": 5e-06, | |
| "loss": 0.562, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6487695749440716, | |
| "grad_norm": 0.5836414884442678, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5388, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6562266964951529, | |
| "grad_norm": 0.6185085505610105, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5406, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6636838180462341, | |
| "grad_norm": 0.5986837831143031, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5549, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "grad_norm": 0.5921990630549645, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5291, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6785980611483967, | |
| "grad_norm": 0.5802544521725337, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5523, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.686055182699478, | |
| "grad_norm": 0.5851196561131752, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5448, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6935123042505593, | |
| "grad_norm": 0.583877657890954, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5415, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7009694258016406, | |
| "grad_norm": 0.5598625156605288, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5712, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7084265473527218, | |
| "grad_norm": 0.6934347616736215, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5519, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7158836689038032, | |
| "grad_norm": 0.5977390121445122, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5551, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7233407904548844, | |
| "grad_norm": 0.5908530451006745, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5561, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7307979120059657, | |
| "grad_norm": 0.5657057055912034, | |
| "learning_rate": 5e-06, | |
| "loss": 0.553, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.738255033557047, | |
| "grad_norm": 0.5866240008585512, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5403, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7457121551081283, | |
| "grad_norm": 0.5925600343817397, | |
| "learning_rate": 5e-06, | |
| "loss": 0.548, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7531692766592095, | |
| "grad_norm": 0.5905026803607015, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5466, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7606263982102909, | |
| "grad_norm": 0.5939700140149752, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5519, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7680835197613721, | |
| "grad_norm": 0.6358723434503033, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5465, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7755406413124534, | |
| "grad_norm": 0.5893886156719156, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5478, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7829977628635347, | |
| "grad_norm": 0.6124289839560092, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5578, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.790454884414616, | |
| "grad_norm": 0.5876929465940476, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5486, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.7979120059656972, | |
| "grad_norm": 0.5765792583371309, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5453, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8053691275167785, | |
| "grad_norm": 0.5763345735177611, | |
| "learning_rate": 5e-06, | |
| "loss": 0.54, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8128262490678598, | |
| "grad_norm": 0.6392240553326028, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5536, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8202833706189411, | |
| "grad_norm": 0.5783160074020695, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5645, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8277404921700223, | |
| "grad_norm": 0.580827750660942, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5618, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8351976137211037, | |
| "grad_norm": 0.5805219027688908, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5378, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8426547352721849, | |
| "grad_norm": 0.5631938645784198, | |
| "learning_rate": 5e-06, | |
| "loss": 0.552, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.8501118568232662, | |
| "grad_norm": 0.5460444743065537, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5474, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.8575689783743475, | |
| "grad_norm": 0.619229281669363, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5568, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.8650260999254288, | |
| "grad_norm": 0.5892594363626477, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5343, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.87248322147651, | |
| "grad_norm": 0.6139047400145734, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5428, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8799403430275914, | |
| "grad_norm": 0.5966906346273294, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5356, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.8873974645786726, | |
| "grad_norm": 0.6535776740165109, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5499, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.8948545861297539, | |
| "grad_norm": 0.5662037475274219, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5288, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9023117076808352, | |
| "grad_norm": 0.5995000278562762, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5563, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9097688292319165, | |
| "grad_norm": 0.6346956605590084, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5392, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.9172259507829977, | |
| "grad_norm": 0.5657793485267905, | |
| "learning_rate": 5e-06, | |
| "loss": 0.551, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.9246830723340791, | |
| "grad_norm": 0.5681455949615055, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5603, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9321401938851603, | |
| "grad_norm": 0.6009029524168309, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5468, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9395973154362416, | |
| "grad_norm": 0.6145844854677124, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5609, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.947054436987323, | |
| "grad_norm": 0.5696932141639989, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5592, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.9545115585384042, | |
| "grad_norm": 0.5890062683374891, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5554, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.9619686800894854, | |
| "grad_norm": 0.5938117712349276, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5646, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.9694258016405667, | |
| "grad_norm": 0.6257865475961188, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5559, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.976882923191648, | |
| "grad_norm": 0.5860984667723387, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5486, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.9843400447427293, | |
| "grad_norm": 0.5609191548184437, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5341, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.9917971662938105, | |
| "grad_norm": 0.6025440403637561, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5446, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.9992542878448919, | |
| "grad_norm": 0.5572472436768862, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5404, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5463822484016418, | |
| "eval_runtime": 339.2131, | |
| "eval_samples_per_second": 26.626, | |
| "eval_steps_per_second": 0.419, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 1.0067114093959733, | |
| "grad_norm": 0.6160017220578684, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4824, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.0141685309470545, | |
| "grad_norm": 0.5591512386853066, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4797, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.0216256524981358, | |
| "grad_norm": 0.6118294509759679, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4686, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.029082774049217, | |
| "grad_norm": 0.5583913619404066, | |
| "learning_rate": 5e-06, | |
| "loss": 0.469, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.0365398956002982, | |
| "grad_norm": 0.5818799856636863, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4903, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.0439970171513795, | |
| "grad_norm": 0.6171364008097712, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4956, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0514541387024607, | |
| "grad_norm": 0.6061235511507647, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4933, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.0589112602535422, | |
| "grad_norm": 0.6107761071918465, | |
| "learning_rate": 5e-06, | |
| "loss": 0.486, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.0663683818046235, | |
| "grad_norm": 0.5547904203912964, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4968, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.0738255033557047, | |
| "grad_norm": 0.5709552455416576, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4902, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.081282624906786, | |
| "grad_norm": 0.6263774656448823, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4703, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.0887397464578672, | |
| "grad_norm": 0.5856371980460074, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4816, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.0961968680089484, | |
| "grad_norm": 0.5914165561619208, | |
| "learning_rate": 5e-06, | |
| "loss": 0.479, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.10365398956003, | |
| "grad_norm": 0.6094767293533612, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4834, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.5628759979218878, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4807, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.1185682326621924, | |
| "grad_norm": 0.5604069640766214, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4945, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1260253542132737, | |
| "grad_norm": 0.6056154435588564, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4756, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.133482475764355, | |
| "grad_norm": 0.5813579925633866, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4753, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.1409395973154361, | |
| "grad_norm": 0.5387240892913252, | |
| "learning_rate": 5e-06, | |
| "loss": 0.474, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.1483967188665176, | |
| "grad_norm": 0.5334644195163835, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4786, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.1558538404175989, | |
| "grad_norm": 0.6302845458131542, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4658, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.1633109619686801, | |
| "grad_norm": 0.5458667914397936, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4884, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.1707680835197614, | |
| "grad_norm": 0.5839445403317228, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4907, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.1782252050708426, | |
| "grad_norm": 0.5603070510925319, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4872, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.1856823266219239, | |
| "grad_norm": 0.6298096241603139, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4816, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.1931394481730053, | |
| "grad_norm": 0.6031228261187916, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4676, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2005965697240866, | |
| "grad_norm": 0.5890973152880289, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4789, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.2080536912751678, | |
| "grad_norm": 0.5915397514486846, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4816, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.215510812826249, | |
| "grad_norm": 0.5786222706794956, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4852, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.2229679343773303, | |
| "grad_norm": 0.5699021496499598, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4986, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.2304250559284116, | |
| "grad_norm": 0.5949853159150295, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4811, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.2378821774794928, | |
| "grad_norm": 0.5673027327267307, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4886, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.2453392990305743, | |
| "grad_norm": 0.560192582433626, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4656, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.2527964205816555, | |
| "grad_norm": 0.6108254414303492, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4715, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.2602535421327368, | |
| "grad_norm": 0.5590093744521378, | |
| "learning_rate": 5e-06, | |
| "loss": 0.471, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.267710663683818, | |
| "grad_norm": 0.5914565694717694, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4736, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.2751677852348993, | |
| "grad_norm": 0.5674996558281405, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4798, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.2826249067859807, | |
| "grad_norm": 0.6498966266280284, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4792, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.290082028337062, | |
| "grad_norm": 0.5750798444633526, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4843, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.2975391498881432, | |
| "grad_norm": 0.5825772193243091, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4869, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.3049962714392245, | |
| "grad_norm": 0.613520592437996, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4838, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.3124533929903057, | |
| "grad_norm": 0.6147541770677961, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4746, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.319910514541387, | |
| "grad_norm": 0.6241611435750868, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4931, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.3273676360924682, | |
| "grad_norm": 0.5926826684472789, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4871, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.3348247576435495, | |
| "grad_norm": 0.6066670454162487, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4841, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.342281879194631, | |
| "grad_norm": 0.6069960632718007, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4862, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.3497390007457122, | |
| "grad_norm": 0.5479395919590173, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4766, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.3571961222967934, | |
| "grad_norm": 0.5639904817681118, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4812, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.3646532438478747, | |
| "grad_norm": 0.5858843238376426, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4686, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.372110365398956, | |
| "grad_norm": 0.5734924915807309, | |
| "learning_rate": 5e-06, | |
| "loss": 0.477, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.3795674869500374, | |
| "grad_norm": 0.5902217700915133, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4889, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.3870246085011186, | |
| "grad_norm": 0.5854924372795176, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4792, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.3944817300521999, | |
| "grad_norm": 0.6350410086083018, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4897, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.4019388516032811, | |
| "grad_norm": 0.5606930586639369, | |
| "learning_rate": 5e-06, | |
| "loss": 0.496, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.4093959731543624, | |
| "grad_norm": 0.5586778910160224, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4671, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.4168530947054436, | |
| "grad_norm": 0.5732737315369484, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4786, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4243102162565249, | |
| "grad_norm": 4.027507437517355, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4886, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.4317673378076063, | |
| "grad_norm": 0.5991590420531947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.497, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.4392244593586876, | |
| "grad_norm": 0.5828725709335938, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4899, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.4466815809097688, | |
| "grad_norm": 0.5473354356011987, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4775, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.45413870246085, | |
| "grad_norm": 0.6210979610287483, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4907, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.4615958240119313, | |
| "grad_norm": 0.5638329844959816, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4759, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.4690529455630128, | |
| "grad_norm": 0.5723878428151293, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4897, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.476510067114094, | |
| "grad_norm": 0.5521756950405313, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4746, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.4839671886651753, | |
| "grad_norm": 0.6691296670250785, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4885, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.4914243102162565, | |
| "grad_norm": 0.6141910929035425, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4884, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.4988814317673378, | |
| "grad_norm": 0.5400373664496432, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4799, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.506338553318419, | |
| "grad_norm": 0.5645402137800272, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4832, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.5137956748695003, | |
| "grad_norm": 0.59297682178836, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4765, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.5212527964205815, | |
| "grad_norm": 0.6099720079302091, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4876, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.5287099179716628, | |
| "grad_norm": 0.5797412589969116, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4867, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.5361670395227442, | |
| "grad_norm": 0.6160597774931952, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4952, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.5436241610738255, | |
| "grad_norm": 0.5775983688665581, | |
| "learning_rate": 5e-06, | |
| "loss": 0.476, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.5510812826249067, | |
| "grad_norm": 0.5846473827271578, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4937, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.5585384041759882, | |
| "grad_norm": 0.5702728266559334, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4974, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.5659955257270695, | |
| "grad_norm": 0.5870580247571568, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4877, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.5734526472781507, | |
| "grad_norm": 0.5817243215012845, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4873, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.580909768829232, | |
| "grad_norm": 0.5315491273649968, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4798, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.5883668903803132, | |
| "grad_norm": 0.6200889922410328, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4699, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.5958240119313944, | |
| "grad_norm": 0.6029121739064351, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4995, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.6032811334824757, | |
| "grad_norm": 0.6028582417864272, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4959, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.610738255033557, | |
| "grad_norm": 0.5893432209254877, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4783, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.6181953765846382, | |
| "grad_norm": 0.5884082032907385, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4823, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.6256524981357197, | |
| "grad_norm": 0.5921720057393691, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4761, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.633109619686801, | |
| "grad_norm": 0.5776696223531537, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4823, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.6405667412378822, | |
| "grad_norm": 0.5645109392805745, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4827, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6480238627889636, | |
| "grad_norm": 0.5532331562154945, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4878, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.6554809843400449, | |
| "grad_norm": 0.5877417623919632, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4854, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.6629381058911261, | |
| "grad_norm": 0.5528378617144938, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4876, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.6703952274422074, | |
| "grad_norm": 0.6104249956677287, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4976, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.6778523489932886, | |
| "grad_norm": 0.5676314169193764, | |
| "learning_rate": 5e-06, | |
| "loss": 0.473, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.6853094705443699, | |
| "grad_norm": 0.5696949870525951, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4866, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.692766592095451, | |
| "grad_norm": 0.5702619592671349, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4909, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.7002237136465324, | |
| "grad_norm": 0.5973494243465649, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4914, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.7076808351976136, | |
| "grad_norm": 0.5545197259083948, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4804, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.7151379567486948, | |
| "grad_norm": 0.5905506685708721, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4925, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.7225950782997763, | |
| "grad_norm": 0.6320966702911732, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4913, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.7300521998508576, | |
| "grad_norm": 0.5417972662105043, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4769, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.7375093214019388, | |
| "grad_norm": 0.597965564481012, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4898, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.7449664429530203, | |
| "grad_norm": 0.6144634291593705, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4826, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.7524235645041015, | |
| "grad_norm": 0.5987390856089427, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4777, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.7598806860551828, | |
| "grad_norm": 0.5693070234792923, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4934, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.767337807606264, | |
| "grad_norm": 0.5879521460157956, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5053, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.7747949291573453, | |
| "grad_norm": 0.5680219111185565, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4817, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.7822520507084265, | |
| "grad_norm": 0.6501364801989891, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4853, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.7897091722595078, | |
| "grad_norm": 0.615030215815147, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4791, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.797166293810589, | |
| "grad_norm": 0.5785652647623134, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4809, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.8046234153616703, | |
| "grad_norm": 0.6044924431994139, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4929, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.8120805369127517, | |
| "grad_norm": 0.5426508657118521, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4847, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.819537658463833, | |
| "grad_norm": 0.6009223844637949, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4906, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.8269947800149142, | |
| "grad_norm": 0.5743711413707431, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4975, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.8344519015659957, | |
| "grad_norm": 0.5983412684564712, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4814, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.841909023117077, | |
| "grad_norm": 0.5560773314183736, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4811, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.8493661446681582, | |
| "grad_norm": 0.5992209963884128, | |
| "learning_rate": 5e-06, | |
| "loss": 0.496, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.8568232662192394, | |
| "grad_norm": 0.5387716311327034, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4804, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.8642803877703207, | |
| "grad_norm": 0.5990433932479565, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4863, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.871737509321402, | |
| "grad_norm": 0.5341574484686165, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4909, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.8791946308724832, | |
| "grad_norm": 0.5857022850771958, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4831, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.8866517524235644, | |
| "grad_norm": 0.596806200047546, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5042, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.8941088739746457, | |
| "grad_norm": 0.5831993654340557, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5004, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.901565995525727, | |
| "grad_norm": 0.5256203832071005, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4802, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.9090231170768084, | |
| "grad_norm": 0.6256000165719325, | |
| "learning_rate": 5e-06, | |
| "loss": 0.49, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.9164802386278896, | |
| "grad_norm": 0.6025652662604613, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4961, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.9239373601789709, | |
| "grad_norm": 0.5851336978699863, | |
| "learning_rate": 5e-06, | |
| "loss": 0.493, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.9313944817300523, | |
| "grad_norm": 0.5841404426808496, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4788, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.9388516032811336, | |
| "grad_norm": 0.5553137000045634, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4944, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.9463087248322148, | |
| "grad_norm": 0.5954485746159562, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4924, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.953765846383296, | |
| "grad_norm": 0.573701644653239, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4909, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.9612229679343773, | |
| "grad_norm": 0.6460568307562583, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4981, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.9686800894854586, | |
| "grad_norm": 0.5881894639448034, | |
| "learning_rate": 5e-06, | |
| "loss": 0.498, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.9761372110365398, | |
| "grad_norm": 0.5442840215629223, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4793, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.983594332587621, | |
| "grad_norm": 0.5772375310283397, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4877, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.9910514541387023, | |
| "grad_norm": 0.5624955248751579, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4852, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.9985085756897838, | |
| "grad_norm": 0.5618877188539807, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4804, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.5512435436248779, | |
| "eval_runtime": 342.0704, | |
| "eval_samples_per_second": 26.404, | |
| "eval_steps_per_second": 0.415, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 2.005965697240865, | |
| "grad_norm": 0.6558184793140344, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4215, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.0134228187919465, | |
| "grad_norm": 0.6034125785194365, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4033, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.0208799403430278, | |
| "grad_norm": 0.5610165111292696, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3907, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.028337061894109, | |
| "grad_norm": 0.580517844768164, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3977, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.0357941834451903, | |
| "grad_norm": 0.5514277294221982, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4112, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.0432513049962715, | |
| "grad_norm": 0.60029875754166, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4176, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.0507084265473527, | |
| "grad_norm": 0.6089301446363884, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4026, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.058165548098434, | |
| "grad_norm": 0.5849276057096263, | |
| "learning_rate": 5e-06, | |
| "loss": 0.409, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.0656226696495152, | |
| "grad_norm": 0.5950355197672823, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4024, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.0730797912005965, | |
| "grad_norm": 0.5889941012752544, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4175, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.0805369127516777, | |
| "grad_norm": 0.5611907053787162, | |
| "learning_rate": 5e-06, | |
| "loss": 0.403, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.087994034302759, | |
| "grad_norm": 0.5934324272584153, | |
| "learning_rate": 5e-06, | |
| "loss": 0.413, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.0954511558538402, | |
| "grad_norm": 0.5774168779978719, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4159, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.1029082774049215, | |
| "grad_norm": 0.564340240629499, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4107, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.110365398956003, | |
| "grad_norm": 0.6144699399372608, | |
| "learning_rate": 5e-06, | |
| "loss": 0.415, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.1178225205070844, | |
| "grad_norm": 0.623817632576955, | |
| "learning_rate": 5e-06, | |
| "loss": 0.415, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.1252796420581657, | |
| "grad_norm": 0.5835224903490361, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4096, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.132736763609247, | |
| "grad_norm": 0.6186331596636564, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4015, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.140193885160328, | |
| "grad_norm": 0.5790608361059295, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4068, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.1476510067114094, | |
| "grad_norm": 0.5869215595357605, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4215, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.1551081282624907, | |
| "grad_norm": 0.5936702509963696, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4141, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.162565249813572, | |
| "grad_norm": 0.6381922955234081, | |
| "learning_rate": 5e-06, | |
| "loss": 0.407, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.170022371364653, | |
| "grad_norm": 0.5862737710378301, | |
| "learning_rate": 5e-06, | |
| "loss": 0.417, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.1774794929157344, | |
| "grad_norm": 0.5822255490097745, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4225, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.1849366144668156, | |
| "grad_norm": 0.5764505027268744, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4138, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 2.192393736017897, | |
| "grad_norm": 0.6073760733719726, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4112, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.1998508575689786, | |
| "grad_norm": 0.6091808835643772, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4208, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.20730797912006, | |
| "grad_norm": 0.6239375451432535, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4167, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.214765100671141, | |
| "grad_norm": 0.6130674669978408, | |
| "learning_rate": 5e-06, | |
| "loss": 0.413, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.5668324865537339, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4126, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.2296793437733036, | |
| "grad_norm": 0.5765769707939168, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4189, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 2.237136465324385, | |
| "grad_norm": 0.6216905758163017, | |
| "learning_rate": 5e-06, | |
| "loss": 0.416, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.244593586875466, | |
| "grad_norm": 0.642406145303932, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4182, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 2.2520507084265473, | |
| "grad_norm": 0.6153726361933832, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4401, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.2595078299776286, | |
| "grad_norm": 0.6162592369179899, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4093, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.26696495152871, | |
| "grad_norm": 0.6062140595912686, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4201, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.274422073079791, | |
| "grad_norm": 0.6489911104793248, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4216, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.2818791946308723, | |
| "grad_norm": 0.5973147232668122, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4225, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.289336316181954, | |
| "grad_norm": 0.6029950937606031, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4187, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.2967934377330352, | |
| "grad_norm": 0.6231584364415401, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4238, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.3042505592841165, | |
| "grad_norm": 0.5561184622741537, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4219, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.3117076808351977, | |
| "grad_norm": 0.5941537102506256, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4246, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.319164802386279, | |
| "grad_norm": 0.6410414670886565, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4071, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 2.3266219239373602, | |
| "grad_norm": 0.5732228808060483, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4114, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.3340790454884415, | |
| "grad_norm": 0.5736385479610295, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4154, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 2.3415361670395227, | |
| "grad_norm": 0.5563198902731237, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4342, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.348993288590604, | |
| "grad_norm": 0.6010188015263107, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4115, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.356450410141685, | |
| "grad_norm": 0.6377162414933872, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4129, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.3639075316927665, | |
| "grad_norm": 0.5983638229139512, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4116, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 2.3713646532438477, | |
| "grad_norm": 0.6406800867031649, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4257, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.378821774794929, | |
| "grad_norm": 0.6165353698446668, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4332, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.3862788963460106, | |
| "grad_norm": 0.5950518706856058, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4166, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.393736017897092, | |
| "grad_norm": 0.5601525861175598, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4214, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.401193139448173, | |
| "grad_norm": 0.6338595567107324, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4175, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.4086502609992544, | |
| "grad_norm": 0.613546741805029, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4205, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 2.4161073825503356, | |
| "grad_norm": 0.6203881329193865, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4106, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.423564504101417, | |
| "grad_norm": 0.587446398589364, | |
| "learning_rate": 5e-06, | |
| "loss": 0.435, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.431021625652498, | |
| "grad_norm": 0.6180935141585874, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4148, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 2.4384787472035794, | |
| "grad_norm": 0.5966462669608655, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4071, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 2.4459358687546606, | |
| "grad_norm": 0.6318483897236675, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4222, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 2.453392990305742, | |
| "grad_norm": 0.6445808903137813, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4152, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 2.460850111856823, | |
| "grad_norm": 0.6047687718392507, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4051, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.4683072334079044, | |
| "grad_norm": 0.6113185157541442, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4221, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 2.4757643549589856, | |
| "grad_norm": 0.5912448445164308, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4116, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 2.4832214765100673, | |
| "grad_norm": 0.6710834935010201, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4211, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 2.4906785980611486, | |
| "grad_norm": 0.5855123869668963, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4176, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 2.49813571961223, | |
| "grad_norm": 0.5672278638197414, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4283, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.505592841163311, | |
| "grad_norm": 0.6623637796001947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4207, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 2.5130499627143923, | |
| "grad_norm": 0.5429235298023167, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4239, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 2.5205070842654735, | |
| "grad_norm": 0.5553083911368453, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4266, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 2.527964205816555, | |
| "grad_norm": 0.5876357465241916, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4262, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 2.535421327367636, | |
| "grad_norm": 0.6222644485385007, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4082, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.5428784489187173, | |
| "grad_norm": 0.6327192739316518, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4123, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 2.5503355704697985, | |
| "grad_norm": 0.5833750402335252, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4177, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 2.5577926920208798, | |
| "grad_norm": 0.642842699957225, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4238, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 2.5652498135719615, | |
| "grad_norm": 0.6050955007783061, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4333, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 2.5727069351230423, | |
| "grad_norm": 0.616598274296138, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4253, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.580164056674124, | |
| "grad_norm": 0.6524589708431916, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4249, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 2.587621178225205, | |
| "grad_norm": 0.5751310148688968, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4177, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 2.5950782997762865, | |
| "grad_norm": 0.627879388176911, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4161, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 2.6025354213273677, | |
| "grad_norm": 0.6201429646996841, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4163, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 2.609992542878449, | |
| "grad_norm": 0.5820543874841265, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4186, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.61744966442953, | |
| "grad_norm": 0.6017708715046648, | |
| "learning_rate": 5e-06, | |
| "loss": 0.428, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 2.6249067859806114, | |
| "grad_norm": 0.6272846852335328, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4171, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 2.6323639075316927, | |
| "grad_norm": 0.6171944158126793, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4208, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 2.639821029082774, | |
| "grad_norm": 0.5757983007676734, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4232, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 2.647278150633855, | |
| "grad_norm": 0.615459114965357, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4145, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 2.6547352721849364, | |
| "grad_norm": 0.642264448040409, | |
| "learning_rate": 5e-06, | |
| "loss": 0.428, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 2.662192393736018, | |
| "grad_norm": 0.5989193246950945, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4271, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 2.669649515287099, | |
| "grad_norm": 0.5516635613380466, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4153, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 2.6771066368381806, | |
| "grad_norm": 0.6619116578052296, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4157, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 2.684563758389262, | |
| "grad_norm": 0.6371720904571588, | |
| "learning_rate": 5e-06, | |
| "loss": 0.427, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.692020879940343, | |
| "grad_norm": 0.5757772064087978, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4141, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 2.6994780014914244, | |
| "grad_norm": 0.6037614796783571, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4169, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 2.7069351230425056, | |
| "grad_norm": 0.6349534794022433, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4363, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 2.714392244593587, | |
| "grad_norm": 0.5873004454383765, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4247, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 2.721849366144668, | |
| "grad_norm": 0.5840849381722206, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4275, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 2.7293064876957494, | |
| "grad_norm": 0.5723518390413971, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4294, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 2.7367636092468306, | |
| "grad_norm": 0.5792593078543581, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4274, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 2.744220730797912, | |
| "grad_norm": 0.5770536517441345, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4184, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 2.751677852348993, | |
| "grad_norm": 0.5803564642723817, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4172, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 2.759134973900075, | |
| "grad_norm": 0.610053209093683, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4189, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.7665920954511556, | |
| "grad_norm": 0.6101684327110989, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4219, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 2.7740492170022373, | |
| "grad_norm": 0.5904638045362951, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4199, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 2.7815063385533185, | |
| "grad_norm": 0.6123335251469906, | |
| "learning_rate": 5e-06, | |
| "loss": 0.417, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 2.7889634601043998, | |
| "grad_norm": 0.6433911863159033, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4247, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 2.796420581655481, | |
| "grad_norm": 0.6031149493474106, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4185, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.8038777032065623, | |
| "grad_norm": 0.5895736114737712, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4289, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 2.8113348247576435, | |
| "grad_norm": 0.5882027658002406, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4295, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 2.8187919463087248, | |
| "grad_norm": 0.6697395749462909, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4278, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 2.826249067859806, | |
| "grad_norm": 0.61486633354577, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4194, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 2.8337061894108873, | |
| "grad_norm": 0.6062660800570471, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4191, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.841163310961969, | |
| "grad_norm": 0.5823531490175686, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4249, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 2.8486204325130497, | |
| "grad_norm": 0.6594942240638506, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4141, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 2.8560775540641314, | |
| "grad_norm": 0.5894447911469776, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4226, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 2.8635346756152127, | |
| "grad_norm": 0.6257621624046176, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4237, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 2.870991797166294, | |
| "grad_norm": 0.5507526853748216, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4247, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 2.878448918717375, | |
| "grad_norm": 0.555588132655637, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4259, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 2.8859060402684564, | |
| "grad_norm": 0.617068246539021, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4307, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 2.8933631618195377, | |
| "grad_norm": 0.5957184006665759, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4302, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 2.900820283370619, | |
| "grad_norm": 0.5821306913096179, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4197, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 2.9082774049217, | |
| "grad_norm": 0.5948347831059887, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4196, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.9157345264727814, | |
| "grad_norm": 0.5748015527924519, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4234, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 2.9231916480238627, | |
| "grad_norm": 0.573330338580251, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4155, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 2.930648769574944, | |
| "grad_norm": 0.5959038101452754, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4245, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 2.9381058911260256, | |
| "grad_norm": 0.6327154198497399, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4141, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 2.9455630126771064, | |
| "grad_norm": 0.5879319738064069, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4111, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 2.953020134228188, | |
| "grad_norm": 0.5850847757389654, | |
| "learning_rate": 5e-06, | |
| "loss": 0.413, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 2.9604772557792693, | |
| "grad_norm": 0.6440418621171623, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4254, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 2.9679343773303506, | |
| "grad_norm": 0.6104743322838082, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4116, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 2.975391498881432, | |
| "grad_norm": 0.5820609145361088, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4188, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 2.982848620432513, | |
| "grad_norm": 0.6266879443777426, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4393, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.9903057419835943, | |
| "grad_norm": 0.5751703131935023, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4298, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 2.9977628635346756, | |
| "grad_norm": 0.6279401240117848, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4292, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5737335681915283, | |
| "eval_runtime": 341.8895, | |
| "eval_samples_per_second": 26.418, | |
| "eval_steps_per_second": 0.415, | |
| "step": 4023 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 4023, | |
| "total_flos": 2108813507297280.0, | |
| "train_loss": 0.4856802497430432, | |
| "train_runtime": 54595.2546, | |
| "train_samples_per_second": 9.429, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4023, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2108813507297280.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |