| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.998269896193771, | |
| "eval_steps": 500, | |
| "global_step": 2889, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01730103806228374, | |
| "grad_norm": 9.043228149414062, | |
| "learning_rate": 1.3793103448275863e-05, | |
| "loss": 1.0259, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03460207612456748, | |
| "grad_norm": 4.801871299743652, | |
| "learning_rate": 2.7586206896551727e-05, | |
| "loss": 0.6145, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05190311418685121, | |
| "grad_norm": 2.854033946990967, | |
| "learning_rate": 4.1379310344827587e-05, | |
| "loss": 0.3438, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06920415224913495, | |
| "grad_norm": 1.3458539247512817, | |
| "learning_rate": 5.517241379310345e-05, | |
| "loss": 0.242, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08650519031141868, | |
| "grad_norm": 2.376776695251465, | |
| "learning_rate": 6.896551724137931e-05, | |
| "loss": 0.2051, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10380622837370242, | |
| "grad_norm": 1.1847227811813354, | |
| "learning_rate": 8.275862068965517e-05, | |
| "loss": 0.1798, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12110726643598616, | |
| "grad_norm": 1.5719020366668701, | |
| "learning_rate": 9.655172413793105e-05, | |
| "loss": 0.1753, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1384083044982699, | |
| "grad_norm": 1.4353362321853638, | |
| "learning_rate": 0.0001103448275862069, | |
| "loss": 0.168, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15570934256055363, | |
| "grad_norm": 1.1386443376541138, | |
| "learning_rate": 0.00012413793103448277, | |
| "loss": 0.1318, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17301038062283736, | |
| "grad_norm": 0.8579152822494507, | |
| "learning_rate": 0.00013793103448275863, | |
| "loss": 0.1337, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1903114186851211, | |
| "grad_norm": 1.252938151359558, | |
| "learning_rate": 0.00015172413793103449, | |
| "loss": 0.1174, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20761245674740483, | |
| "grad_norm": 0.9858207702636719, | |
| "learning_rate": 0.00016551724137931035, | |
| "loss": 0.141, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22491349480968859, | |
| "grad_norm": 1.5159997940063477, | |
| "learning_rate": 0.0001793103448275862, | |
| "loss": 0.1132, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2422145328719723, | |
| "grad_norm": 1.087937593460083, | |
| "learning_rate": 0.0001931034482758621, | |
| "loss": 0.1008, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25951557093425603, | |
| "grad_norm": 1.02494215965271, | |
| "learning_rate": 0.0001999983615229662, | |
| "loss": 0.124, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2768166089965398, | |
| "grad_norm": 1.0136982202529907, | |
| "learning_rate": 0.00019998525402884653, | |
| "loss": 0.0977, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 1.4797533750534058, | |
| "learning_rate": 0.0001999590407586994, | |
| "loss": 0.1054, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.31141868512110726, | |
| "grad_norm": 0.7462301254272461, | |
| "learning_rate": 0.000199919725148484, | |
| "loss": 0.0972, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.328719723183391, | |
| "grad_norm": 0.9645005464553833, | |
| "learning_rate": 0.00019986731235157592, | |
| "loss": 0.11, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3460207612456747, | |
| "grad_norm": 1.327414631843567, | |
| "learning_rate": 0.00019980180923809214, | |
| "loss": 0.111, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3633217993079585, | |
| "grad_norm": 0.8332083821296692, | |
| "learning_rate": 0.00019972322439399, | |
| "loss": 0.092, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3806228373702422, | |
| "grad_norm": 0.9236223101615906, | |
| "learning_rate": 0.00019963156811994215, | |
| "loss": 0.0983, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.39792387543252594, | |
| "grad_norm": 0.8200174570083618, | |
| "learning_rate": 0.0001995268524299861, | |
| "loss": 0.0888, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.41522491349480967, | |
| "grad_norm": 0.8660184741020203, | |
| "learning_rate": 0.00019940909104994973, | |
| "loss": 0.0761, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.43252595155709345, | |
| "grad_norm": 0.826708972454071, | |
| "learning_rate": 0.00019927829941565186, | |
| "loss": 0.0794, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.44982698961937717, | |
| "grad_norm": 0.5673882961273193, | |
| "learning_rate": 0.00019913449467087916, | |
| "loss": 0.0795, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4671280276816609, | |
| "grad_norm": 0.9092473983764648, | |
| "learning_rate": 0.00019897769566513897, | |
| "loss": 0.0779, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4844290657439446, | |
| "grad_norm": 0.8896488547325134, | |
| "learning_rate": 0.00019880792295118852, | |
| "loss": 0.0982, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5017301038062284, | |
| "grad_norm": 1.0210949182510376, | |
| "learning_rate": 0.00019862519878234084, | |
| "loss": 0.0653, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5190311418685121, | |
| "grad_norm": 0.8563092350959778, | |
| "learning_rate": 0.00019842954710954812, | |
| "loss": 0.0807, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5363321799307958, | |
| "grad_norm": 0.9336950182914734, | |
| "learning_rate": 0.000198220993578262, | |
| "loss": 0.073, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5536332179930796, | |
| "grad_norm": 0.6465988159179688, | |
| "learning_rate": 0.00019799956552507233, | |
| "loss": 0.0828, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5709342560553633, | |
| "grad_norm": 1.0124980211257935, | |
| "learning_rate": 0.00019776529197412362, | |
| "loss": 0.0819, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.7318516373634338, | |
| "learning_rate": 0.00019751820363331097, | |
| "loss": 0.0667, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6055363321799307, | |
| "grad_norm": 0.8519881367683411, | |
| "learning_rate": 0.00019725833289025476, | |
| "loss": 0.0654, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6228373702422145, | |
| "grad_norm": 0.5366441607475281, | |
| "learning_rate": 0.00019698571380805552, | |
| "loss": 0.0726, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6401384083044983, | |
| "grad_norm": 0.570626974105835, | |
| "learning_rate": 0.00019670038212082886, | |
| "loss": 0.0554, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.657439446366782, | |
| "grad_norm": 0.7261837720870972, | |
| "learning_rate": 0.00019640237522902174, | |
| "loss": 0.0674, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6747404844290658, | |
| "grad_norm": 0.49231183528900146, | |
| "learning_rate": 0.00019609173219450998, | |
| "loss": 0.0734, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6920415224913494, | |
| "grad_norm": 0.6281108260154724, | |
| "learning_rate": 0.0001957684937354782, | |
| "loss": 0.0582, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7093425605536332, | |
| "grad_norm": 0.6687746644020081, | |
| "learning_rate": 0.00019543270222108268, | |
| "loss": 0.0714, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.726643598615917, | |
| "grad_norm": 0.6667928099632263, | |
| "learning_rate": 0.00019508440166589753, | |
| "loss": 0.0792, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7439446366782007, | |
| "grad_norm": 0.7570515871047974, | |
| "learning_rate": 0.00019472363772414563, | |
| "loss": 0.0625, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7612456747404844, | |
| "grad_norm": 0.5553678274154663, | |
| "learning_rate": 0.00019435045768371415, | |
| "loss": 0.063, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7785467128027682, | |
| "grad_norm": 0.6174982190132141, | |
| "learning_rate": 0.00019396491045995648, | |
| "loss": 0.0832, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7958477508650519, | |
| "grad_norm": 0.6104321479797363, | |
| "learning_rate": 0.00019356704658928035, | |
| "loss": 0.0564, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8131487889273357, | |
| "grad_norm": 0.5369688868522644, | |
| "learning_rate": 0.00019315691822252362, | |
| "loss": 0.0647, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8304498269896193, | |
| "grad_norm": 0.6868175268173218, | |
| "learning_rate": 0.0001927345791181187, | |
| "loss": 0.0586, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8477508650519031, | |
| "grad_norm": 0.6508586406707764, | |
| "learning_rate": 0.00019230008463504595, | |
| "loss": 0.0569, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8650519031141869, | |
| "grad_norm": 0.5308678150177002, | |
| "learning_rate": 0.00019185349172557724, | |
| "loss": 0.05, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 0.41809916496276855, | |
| "learning_rate": 0.00019139485892781118, | |
| "loss": 0.0528, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8996539792387543, | |
| "grad_norm": 0.6402536630630493, | |
| "learning_rate": 0.00019092424635799962, | |
| "loss": 0.0448, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.916955017301038, | |
| "grad_norm": 0.8051866888999939, | |
| "learning_rate": 0.0001904417157026683, | |
| "loss": 0.0567, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9342560553633218, | |
| "grad_norm": 0.4536950886249542, | |
| "learning_rate": 0.00018994733021053076, | |
| "loss": 0.0711, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9515570934256056, | |
| "grad_norm": 0.4057430922985077, | |
| "learning_rate": 0.00018944115468419809, | |
| "loss": 0.0491, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9688581314878892, | |
| "grad_norm": 0.3847835063934326, | |
| "learning_rate": 0.00018892325547168473, | |
| "loss": 0.0563, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.986159169550173, | |
| "grad_norm": 0.5546027421951294, | |
| "learning_rate": 0.00018839370045771182, | |
| "loss": 0.0664, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0034602076124568, | |
| "grad_norm": 0.5202723145484924, | |
| "learning_rate": 0.00018785255905480897, | |
| "loss": 0.0581, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.0207612456747406, | |
| "grad_norm": 0.729364812374115, | |
| "learning_rate": 0.00018729990219421594, | |
| "loss": 0.0485, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.0380622837370241, | |
| "grad_norm": 0.5977576375007629, | |
| "learning_rate": 0.0001867358023165851, | |
| "loss": 0.0545, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.055363321799308, | |
| "grad_norm": 0.5301733016967773, | |
| "learning_rate": 0.00018616033336248632, | |
| "loss": 0.0582, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0726643598615917, | |
| "grad_norm": 0.5742247700691223, | |
| "learning_rate": 0.00018557357076271475, | |
| "loss": 0.0655, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0899653979238755, | |
| "grad_norm": 0.7400846481323242, | |
| "learning_rate": 0.0001849755914284039, | |
| "loss": 0.0467, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.1072664359861593, | |
| "grad_norm": 0.7476756572723389, | |
| "learning_rate": 0.00018436647374094406, | |
| "loss": 0.0471, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1245674740484428, | |
| "grad_norm": 0.5191195011138916, | |
| "learning_rate": 0.00018374629754170854, | |
| "loss": 0.0585, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.1418685121107266, | |
| "grad_norm": 0.711711585521698, | |
| "learning_rate": 0.00018311514412158806, | |
| "loss": 0.0526, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.1591695501730104, | |
| "grad_norm": 0.4129495620727539, | |
| "learning_rate": 0.0001824730962103356, | |
| "loss": 0.0436, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.6265865564346313, | |
| "learning_rate": 0.0001818202379657222, | |
| "loss": 0.0672, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.193771626297578, | |
| "grad_norm": 0.5319189429283142, | |
| "learning_rate": 0.0001811566549625061, | |
| "loss": 0.0476, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.2110726643598615, | |
| "grad_norm": 0.4006735682487488, | |
| "learning_rate": 0.00018048243418121551, | |
| "loss": 0.0501, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2283737024221453, | |
| "grad_norm": 0.7269253134727478, | |
| "learning_rate": 0.00017979766399674776, | |
| "loss": 0.0548, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.245674740484429, | |
| "grad_norm": 0.2751833200454712, | |
| "learning_rate": 0.00017910243416678512, | |
| "loss": 0.0419, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.2629757785467128, | |
| "grad_norm": 0.411044180393219, | |
| "learning_rate": 0.00017839683582002982, | |
| "loss": 0.0591, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.2802768166089966, | |
| "grad_norm": 0.3086739480495453, | |
| "learning_rate": 0.00017768096144425902, | |
| "loss": 0.041, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.2975778546712804, | |
| "grad_norm": 0.49662747979164124, | |
| "learning_rate": 0.00017695490487420194, | |
| "loss": 0.0476, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.314878892733564, | |
| "grad_norm": 0.3548310101032257, | |
| "learning_rate": 0.0001762187612792401, | |
| "loss": 0.0427, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.3321799307958477, | |
| "grad_norm": 0.5193726420402527, | |
| "learning_rate": 0.00017547262715093291, | |
| "loss": 0.0495, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.3494809688581315, | |
| "grad_norm": 0.5085720419883728, | |
| "learning_rate": 0.00017471660029036987, | |
| "loss": 0.0518, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.3667820069204153, | |
| "grad_norm": 1.0562312602996826, | |
| "learning_rate": 0.00017395077979535088, | |
| "loss": 0.0414, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.3840830449826989, | |
| "grad_norm": 0.3967028558254242, | |
| "learning_rate": 0.00017317526604739708, | |
| "loss": 0.0526, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4013840830449826, | |
| "grad_norm": 0.5379968285560608, | |
| "learning_rate": 0.0001723901606985929, | |
| "loss": 0.0434, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.4186851211072664, | |
| "grad_norm": 0.4309251606464386, | |
| "learning_rate": 0.00017159556665826195, | |
| "loss": 0.0653, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.4359861591695502, | |
| "grad_norm": 0.5331924557685852, | |
| "learning_rate": 0.0001707915880794778, | |
| "loss": 0.0477, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.453287197231834, | |
| "grad_norm": 0.34852275252342224, | |
| "learning_rate": 0.0001699783303454121, | |
| "loss": 0.0403, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.3794853389263153, | |
| "learning_rate": 0.00016915590005552118, | |
| "loss": 0.0428, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.4878892733564013, | |
| "grad_norm": 0.5624513030052185, | |
| "learning_rate": 0.00016832440501157313, | |
| "loss": 0.0466, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.505190311418685, | |
| "grad_norm": 0.2839530110359192, | |
| "learning_rate": 0.0001674839542035178, | |
| "loss": 0.0452, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.5224913494809689, | |
| "grad_norm": 0.624341607093811, | |
| "learning_rate": 0.0001666346577952004, | |
| "loss": 0.0481, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.5397923875432526, | |
| "grad_norm": 0.5866113305091858, | |
| "learning_rate": 0.00016577662710992174, | |
| "loss": 0.058, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.5570934256055362, | |
| "grad_norm": 0.5155909061431885, | |
| "learning_rate": 0.00016490997461584617, | |
| "loss": 0.0467, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.57439446366782, | |
| "grad_norm": 0.33836182951927185, | |
| "learning_rate": 0.00016403481391125973, | |
| "loss": 0.0434, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.5916955017301038, | |
| "grad_norm": 0.3541756570339203, | |
| "learning_rate": 0.00016315125970967978, | |
| "loss": 0.0568, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.6089965397923875, | |
| "grad_norm": 0.4802621603012085, | |
| "learning_rate": 0.000162259427824819, | |
| "loss": 0.0616, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.6262975778546713, | |
| "grad_norm": 0.3919571340084076, | |
| "learning_rate": 0.00016135943515540455, | |
| "loss": 0.0382, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.643598615916955, | |
| "grad_norm": 0.40250062942504883, | |
| "learning_rate": 0.0001604513996698556, | |
| "loss": 0.0446, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6608996539792389, | |
| "grad_norm": 0.3435959219932556, | |
| "learning_rate": 0.00015953544039082012, | |
| "loss": 0.0473, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.6782006920415224, | |
| "grad_norm": 0.4662492275238037, | |
| "learning_rate": 0.00015861167737957397, | |
| "loss": 0.0373, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.6955017301038062, | |
| "grad_norm": 0.460500031709671, | |
| "learning_rate": 0.00015768023172028342, | |
| "loss": 0.0457, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.71280276816609, | |
| "grad_norm": 0.5573172569274902, | |
| "learning_rate": 0.00015674122550413396, | |
| "loss": 0.0421, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.7301038062283736, | |
| "grad_norm": 0.32693353295326233, | |
| "learning_rate": 0.00015579478181332684, | |
| "loss": 0.0375, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7474048442906573, | |
| "grad_norm": 0.3940613865852356, | |
| "learning_rate": 0.00015484102470494576, | |
| "loss": 0.0561, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 0.3881168067455292, | |
| "learning_rate": 0.00015388007919469603, | |
| "loss": 0.0361, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.782006920415225, | |
| "grad_norm": 0.3204861581325531, | |
| "learning_rate": 0.0001529120712405177, | |
| "loss": 0.0375, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.7993079584775087, | |
| "grad_norm": 0.47057387232780457, | |
| "learning_rate": 0.00015193712772607537, | |
| "loss": 0.0382, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.8166089965397925, | |
| "grad_norm": 0.29308807849884033, | |
| "learning_rate": 0.0001509553764441267, | |
| "loss": 0.0331, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.8339100346020762, | |
| "grad_norm": 0.4945255219936371, | |
| "learning_rate": 0.00014996694607977176, | |
| "loss": 0.0477, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.85121107266436, | |
| "grad_norm": 0.3753719925880432, | |
| "learning_rate": 0.00014897196619358526, | |
| "loss": 0.0333, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.8685121107266436, | |
| "grad_norm": 0.4464806914329529, | |
| "learning_rate": 0.0001479705672046341, | |
| "loss": 0.04, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.8858131487889274, | |
| "grad_norm": 0.4522175192832947, | |
| "learning_rate": 0.00014696288037338256, | |
| "loss": 0.0426, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.903114186851211, | |
| "grad_norm": 0.37199148535728455, | |
| "learning_rate": 0.00014594903778448705, | |
| "loss": 0.0432, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.9204152249134947, | |
| "grad_norm": 0.3355126678943634, | |
| "learning_rate": 0.00014492917232948263, | |
| "loss": 0.0356, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.9377162629757785, | |
| "grad_norm": 0.37697604298591614, | |
| "learning_rate": 0.00014390341768936413, | |
| "loss": 0.0359, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.9550173010380623, | |
| "grad_norm": 0.4980825185775757, | |
| "learning_rate": 0.00014287190831706372, | |
| "loss": 0.0374, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.972318339100346, | |
| "grad_norm": 0.3119935691356659, | |
| "learning_rate": 0.00014183477941982704, | |
| "loss": 0.034, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.9896193771626298, | |
| "grad_norm": 0.3332310914993286, | |
| "learning_rate": 0.00014079216694149076, | |
| "loss": 0.0386, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.0069204152249136, | |
| "grad_norm": 0.5753083229064941, | |
| "learning_rate": 0.00013974420754466328, | |
| "loss": 0.0356, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.0242214532871974, | |
| "grad_norm": 0.6476043462753296, | |
| "learning_rate": 0.00013869103859281165, | |
| "loss": 0.0401, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.041522491349481, | |
| "grad_norm": 0.6274977922439575, | |
| "learning_rate": 0.0001376327981322561, | |
| "loss": 0.0405, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.0588235294117645, | |
| "grad_norm": 0.5121188163757324, | |
| "learning_rate": 0.0001365696248740756, | |
| "loss": 0.0409, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.0761245674740483, | |
| "grad_norm": 0.38302546739578247, | |
| "learning_rate": 0.0001355016581759257, | |
| "loss": 0.0407, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.093425605536332, | |
| "grad_norm": 0.365703821182251, | |
| "learning_rate": 0.00013442903802377226, | |
| "loss": 0.0439, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.110726643598616, | |
| "grad_norm": 0.43399107456207275, | |
| "learning_rate": 0.00013335190501354227, | |
| "loss": 0.036, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.1280276816608996, | |
| "grad_norm": 0.31331631541252136, | |
| "learning_rate": 0.000132270400332695, | |
| "loss": 0.0303, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.1453287197231834, | |
| "grad_norm": 0.5099217295646667, | |
| "learning_rate": 0.00013118466574171564, | |
| "loss": 0.0339, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.162629757785467, | |
| "grad_norm": 0.39915943145751953, | |
| "learning_rate": 0.00013009484355553364, | |
| "loss": 0.0371, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.179930795847751, | |
| "grad_norm": 0.42217525839805603, | |
| "learning_rate": 0.00012900107662486857, | |
| "loss": 0.0271, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.1972318339100347, | |
| "grad_norm": 0.35948893427848816, | |
| "learning_rate": 0.00012790350831750556, | |
| "loss": 0.0298, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.2145328719723185, | |
| "grad_norm": 0.4188036620616913, | |
| "learning_rate": 0.0001268022824995032, | |
| "loss": 0.0413, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.2318339100346023, | |
| "grad_norm": 0.3547791838645935, | |
| "learning_rate": 0.0001256975435163359, | |
| "loss": 0.0325, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.2491349480968856, | |
| "grad_norm": 0.4535089433193207, | |
| "learning_rate": 0.00012458943617397344, | |
| "loss": 0.0382, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.2664359861591694, | |
| "grad_norm": 0.39500531554222107, | |
| "learning_rate": 0.00012347810571990055, | |
| "loss": 0.0307, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.283737024221453, | |
| "grad_norm": 0.41853851079940796, | |
| "learning_rate": 0.00012236369782407783, | |
| "loss": 0.0313, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.301038062283737, | |
| "grad_norm": 0.3748478591442108, | |
| "learning_rate": 0.0001212463585598481, | |
| "loss": 0.0412, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.3183391003460208, | |
| "grad_norm": 0.41274523735046387, | |
| "learning_rate": 0.00012012623438478931, | |
| "loss": 0.0473, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.3356401384083045, | |
| "grad_norm": 0.24082864820957184, | |
| "learning_rate": 0.0001190034721215176, | |
| "loss": 0.0347, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.45620933175086975, | |
| "learning_rate": 0.00011787821893844189, | |
| "loss": 0.0437, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.370242214532872, | |
| "grad_norm": 0.3050684630870819, | |
| "learning_rate": 0.00011675062233047364, | |
| "loss": 0.0369, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.387543252595156, | |
| "grad_norm": 0.42224356532096863, | |
| "learning_rate": 0.00011562083009969366, | |
| "loss": 0.0322, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.404844290657439, | |
| "grad_norm": 0.3712410032749176, | |
| "learning_rate": 0.00011448899033597855, | |
| "loss": 0.0386, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.422145328719723, | |
| "grad_norm": 0.22947420179843903, | |
| "learning_rate": 0.00011335525139758962, | |
| "loss": 0.0375, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.4394463667820068, | |
| "grad_norm": 0.42010223865509033, | |
| "learning_rate": 0.00011221976189172644, | |
| "loss": 0.0404, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.4567474048442905, | |
| "grad_norm": 0.40255504846572876, | |
| "learning_rate": 0.0001110826706550479, | |
| "loss": 0.0356, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.4740484429065743, | |
| "grad_norm": 0.6300822496414185, | |
| "learning_rate": 0.00010994412673416303, | |
| "loss": 0.0403, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.491349480968858, | |
| "grad_norm": 0.4776591658592224, | |
| "learning_rate": 0.00010880427936609455, | |
| "loss": 0.0293, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.508650519031142, | |
| "grad_norm": 0.5055949091911316, | |
| "learning_rate": 0.0001076632779587172, | |
| "loss": 0.039, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.5259515570934257, | |
| "grad_norm": 0.43698614835739136, | |
| "learning_rate": 0.00010652127207117386, | |
| "loss": 0.033, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.5432525951557095, | |
| "grad_norm": 0.4662177264690399, | |
| "learning_rate": 0.00010537841139427178, | |
| "loss": 0.0214, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.5605536332179932, | |
| "grad_norm": 0.28925833106040955, | |
| "learning_rate": 0.00010423484573086138, | |
| "loss": 0.0317, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.577854671280277, | |
| "grad_norm": 0.5237163305282593, | |
| "learning_rate": 0.00010309072497620081, | |
| "loss": 0.0295, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.595155709342561, | |
| "grad_norm": 0.34184929728507996, | |
| "learning_rate": 0.00010194619909830787, | |
| "loss": 0.0389, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.612456747404844, | |
| "grad_norm": 0.3619227111339569, | |
| "learning_rate": 0.00010080141811830277, | |
| "loss": 0.0278, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.629757785467128, | |
| "grad_norm": 0.27774831652641296, | |
| "learning_rate": 9.965653209074378e-05, | |
| "loss": 0.0278, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.6470588235294117, | |
| "grad_norm": 0.2855290472507477, | |
| "learning_rate": 9.851169108395842e-05, | |
| "loss": 0.031, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.6643598615916955, | |
| "grad_norm": 0.29521238803863525, | |
| "learning_rate": 9.736704516037317e-05, | |
| "loss": 0.0282, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.6816608996539792, | |
| "grad_norm": 0.30670079588890076, | |
| "learning_rate": 9.622274435684334e-05, | |
| "loss": 0.0376, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.698961937716263, | |
| "grad_norm": 0.2929821014404297, | |
| "learning_rate": 9.507893866498714e-05, | |
| "loss": 0.0261, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.716262975778547, | |
| "grad_norm": 0.23183377087116241, | |
| "learning_rate": 9.393577801152486e-05, | |
| "loss": 0.0253, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.7335640138408306, | |
| "grad_norm": 0.20076750218868256, | |
| "learning_rate": 9.279341223862705e-05, | |
| "loss": 0.0317, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.750865051903114, | |
| "grad_norm": 0.4574607014656067, | |
| "learning_rate": 9.165199108427364e-05, | |
| "loss": 0.0391, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.7681660899653977, | |
| "grad_norm": 0.4383868873119354, | |
| "learning_rate": 9.051166416262673e-05, | |
| "loss": 0.0303, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.7854671280276815, | |
| "grad_norm": 0.4536924362182617, | |
| "learning_rate": 8.937258094441953e-05, | |
| "loss": 0.0309, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.8027681660899653, | |
| "grad_norm": 0.2295350879430771, | |
| "learning_rate": 8.823489073736429e-05, | |
| "loss": 0.0308, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.820069204152249, | |
| "grad_norm": 0.363835871219635, | |
| "learning_rate": 8.70987426665814e-05, | |
| "loss": 0.0353, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.837370242214533, | |
| "grad_norm": 0.15470369160175323, | |
| "learning_rate": 8.596428565505245e-05, | |
| "loss": 0.0261, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.8546712802768166, | |
| "grad_norm": 0.45268356800079346, | |
| "learning_rate": 8.483166840409995e-05, | |
| "loss": 0.0298, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.8719723183391004, | |
| "grad_norm": 0.2832735478878021, | |
| "learning_rate": 8.370103937389595e-05, | |
| "loss": 0.0343, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.889273356401384, | |
| "grad_norm": 0.3091322183609009, | |
| "learning_rate": 8.257254676400237e-05, | |
| "loss": 0.0276, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.906574394463668, | |
| "grad_norm": 0.30403849482536316, | |
| "learning_rate": 8.144633849394527e-05, | |
| "loss": 0.0259, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.9238754325259517, | |
| "grad_norm": 0.41098812222480774, | |
| "learning_rate": 8.032256218382618e-05, | |
| "loss": 0.0416, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 0.23849168419837952, | |
| "learning_rate": 7.920136513497232e-05, | |
| "loss": 0.0285, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.9584775086505193, | |
| "grad_norm": 0.3104817569255829, | |
| "learning_rate": 7.808289431062892e-05, | |
| "loss": 0.026, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.9757785467128026, | |
| "grad_norm": 0.4680814743041992, | |
| "learning_rate": 7.696729631669564e-05, | |
| "loss": 0.0336, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.9930795847750864, | |
| "grad_norm": 0.3355979323387146, | |
| "learning_rate": 7.585471738250984e-05, | |
| "loss": 0.0307, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.01038062283737, | |
| "grad_norm": 0.5686395168304443, | |
| "learning_rate": 7.474530334167935e-05, | |
| "loss": 0.032, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.027681660899654, | |
| "grad_norm": 0.2684016823768616, | |
| "learning_rate": 7.363919961296699e-05, | |
| "loss": 0.0244, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.0449826989619377, | |
| "grad_norm": 0.2455485612154007, | |
| "learning_rate": 7.253655118122948e-05, | |
| "loss": 0.0245, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.0622837370242215, | |
| "grad_norm": 0.35358694195747375, | |
| "learning_rate": 7.143750257841333e-05, | |
| "loss": 0.0267, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.0795847750865053, | |
| "grad_norm": 0.18395821750164032, | |
| "learning_rate": 7.034219786460987e-05, | |
| "loss": 0.023, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.096885813148789, | |
| "grad_norm": 0.3733113408088684, | |
| "learning_rate": 6.925078060917245e-05, | |
| "loss": 0.0276, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.114186851211073, | |
| "grad_norm": 0.30367976427078247, | |
| "learning_rate": 6.816339387189763e-05, | |
| "loss": 0.0298, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.131487889273356, | |
| "grad_norm": 0.3499404489994049, | |
| "learning_rate": 6.708018018427343e-05, | |
| "loss": 0.023, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.14878892733564, | |
| "grad_norm": 0.24544458091259003, | |
| "learning_rate": 6.600128153079661e-05, | |
| "loss": 0.0187, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.1660899653979238, | |
| "grad_norm": 0.2629719376564026, | |
| "learning_rate": 6.492683933036183e-05, | |
| "loss": 0.0244, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.1833910034602075, | |
| "grad_norm": 0.46131864190101624, | |
| "learning_rate": 6.38569944177249e-05, | |
| "loss": 0.0294, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.2006920415224913, | |
| "grad_norm": 0.2149653136730194, | |
| "learning_rate": 6.279188702504252e-05, | |
| "loss": 0.0221, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.217993079584775, | |
| "grad_norm": 0.1689160317182541, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 0.0218, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.235294117647059, | |
| "grad_norm": 0.22369259595870972, | |
| "learning_rate": 6.0676442604966654e-05, | |
| "loss": 0.0181, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.2525951557093427, | |
| "grad_norm": 0.16078820824623108, | |
| "learning_rate": 5.9626382863869414e-05, | |
| "loss": 0.0236, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.2698961937716264, | |
| "grad_norm": 0.36825454235076904, | |
| "learning_rate": 5.8581615178973274e-05, | |
| "loss": 0.0264, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.28719723183391, | |
| "grad_norm": 0.3015175759792328, | |
| "learning_rate": 5.754227649538497e-05, | |
| "loss": 0.0306, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.304498269896194, | |
| "grad_norm": 0.4220731556415558, | |
| "learning_rate": 5.6508503046593484e-05, | |
| "loss": 0.0237, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.3217993079584773, | |
| "grad_norm": 0.4073288142681122, | |
| "learning_rate": 5.548043033661297e-05, | |
| "loss": 0.0164, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.339100346020761, | |
| "grad_norm": 0.27142080664634705, | |
| "learning_rate": 5.44581931222214e-05, | |
| "loss": 0.0255, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.356401384083045, | |
| "grad_norm": 0.25887903571128845, | |
| "learning_rate": 5.3441925395297065e-05, | |
| "loss": 0.0176, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.3737024221453287, | |
| "grad_norm": 0.3199214041233063, | |
| "learning_rate": 5.243176036525499e-05, | |
| "loss": 0.0276, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.3910034602076125, | |
| "grad_norm": 0.24466337263584137, | |
| "learning_rate": 5.142783044158668e-05, | |
| "loss": 0.0232, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.4083044982698962, | |
| "grad_norm": 0.38157567381858826, | |
| "learning_rate": 5.043026721650388e-05, | |
| "loss": 0.0342, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.42560553633218, | |
| "grad_norm": 0.2125539630651474, | |
| "learning_rate": 4.943920144769013e-05, | |
| "loss": 0.0268, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.442906574394464, | |
| "grad_norm": 0.3309932351112366, | |
| "learning_rate": 4.845476304116132e-05, | |
| "loss": 0.0214, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.4602076124567476, | |
| "grad_norm": 0.2618106007575989, | |
| "learning_rate": 4.74770810342379e-05, | |
| "loss": 0.0249, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.477508650519031, | |
| "grad_norm": 0.20525750517845154, | |
| "learning_rate": 4.650628357863113e-05, | |
| "loss": 0.0228, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.4948096885813147, | |
| "grad_norm": 0.2250107228755951, | |
| "learning_rate": 4.5542497923645456e-05, | |
| "loss": 0.0252, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.5121107266435985, | |
| "grad_norm": 0.3091624975204468, | |
| "learning_rate": 4.458585039949874e-05, | |
| "loss": 0.0204, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.3369143605232239, | |
| "learning_rate": 4.363646640076355e-05, | |
| "loss": 0.0227, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.546712802768166, | |
| "grad_norm": 0.2998056709766388, | |
| "learning_rate": 4.2694470369930697e-05, | |
| "loss": 0.0191, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.56401384083045, | |
| "grad_norm": 0.1255960911512375, | |
| "learning_rate": 4.175998578109756e-05, | |
| "loss": 0.0225, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.5813148788927336, | |
| "grad_norm": 0.21881087124347687, | |
| "learning_rate": 4.0833135123783683e-05, | |
| "loss": 0.0248, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.5986159169550174, | |
| "grad_norm": 0.20889127254486084, | |
| "learning_rate": 3.991403988687499e-05, | |
| "loss": 0.0249, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.615916955017301, | |
| "grad_norm": 0.2735111713409424, | |
| "learning_rate": 3.900282054269954e-05, | |
| "loss": 0.0207, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.633217993079585, | |
| "grad_norm": 0.3216061294078827, | |
| "learning_rate": 3.8099596531236357e-05, | |
| "loss": 0.0191, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.6505190311418687, | |
| "grad_norm": 0.31375864148139954, | |
| "learning_rate": 3.7204486244459334e-05, | |
| "loss": 0.0261, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.6678200692041525, | |
| "grad_norm": 0.37862274050712585, | |
| "learning_rate": 3.631760701081913e-05, | |
| "loss": 0.028, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.685121107266436, | |
| "grad_norm": 0.34272995591163635, | |
| "learning_rate": 3.5439075079863913e-05, | |
| "loss": 0.0249, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.7024221453287196, | |
| "grad_norm": 0.35570189356803894, | |
| "learning_rate": 3.456900560700158e-05, | |
| "loss": 0.019, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.7197231833910034, | |
| "grad_norm": 0.2902262806892395, | |
| "learning_rate": 3.370751263840581e-05, | |
| "loss": 0.0202, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.737024221453287, | |
| "grad_norm": 0.1935553103685379, | |
| "learning_rate": 3.285470909606696e-05, | |
| "loss": 0.026, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.754325259515571, | |
| "grad_norm": 0.1905568242073059, | |
| "learning_rate": 3.2010706762990736e-05, | |
| "loss": 0.0172, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.7716262975778547, | |
| "grad_norm": 0.27930760383605957, | |
| "learning_rate": 3.117561626854601e-05, | |
| "loss": 0.0149, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.7889273356401385, | |
| "grad_norm": 0.38801103830337524, | |
| "learning_rate": 3.0349547073963693e-05, | |
| "loss": 0.0202, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.8062283737024223, | |
| "grad_norm": 0.171518012881279, | |
| "learning_rate": 2.953260745798898e-05, | |
| "loss": 0.0226, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.8235294117647056, | |
| "grad_norm": 0.22121591866016388, | |
| "learning_rate": 2.8724904502688566e-05, | |
| "loss": 0.022, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.8408304498269894, | |
| "grad_norm": 0.36183586716651917, | |
| "learning_rate": 2.792654407941444e-05, | |
| "loss": 0.0207, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.858131487889273, | |
| "grad_norm": 0.49139508605003357, | |
| "learning_rate": 2.7137630834926788e-05, | |
| "loss": 0.0237, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.875432525951557, | |
| "grad_norm": 0.29921260476112366, | |
| "learning_rate": 2.635826817767708e-05, | |
| "loss": 0.021, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.8927335640138407, | |
| "grad_norm": 0.2761327922344208, | |
| "learning_rate": 2.5588558264253547e-05, | |
| "loss": 0.0183, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.9100346020761245, | |
| "grad_norm": 0.21798908710479736, | |
| "learning_rate": 2.4828601985990983e-05, | |
| "loss": 0.0301, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.9273356401384083, | |
| "grad_norm": 0.20128929615020752, | |
| "learning_rate": 2.407849895574592e-05, | |
| "loss": 0.0263, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.944636678200692, | |
| "grad_norm": 0.3022063672542572, | |
| "learning_rate": 2.3338347494839997e-05, | |
| "loss": 0.0249, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.961937716262976, | |
| "grad_norm": 0.13671119511127472, | |
| "learning_rate": 2.260824462017195e-05, | |
| "loss": 0.0189, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.9792387543252596, | |
| "grad_norm": 0.217565655708313, | |
| "learning_rate": 2.1888286031501216e-05, | |
| "loss": 0.022, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.9965397923875434, | |
| "grad_norm": 0.335618793964386, | |
| "learning_rate": 2.1178566098903674e-05, | |
| "loss": 0.0188, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.013840830449827, | |
| "grad_norm": 0.22030693292617798, | |
| "learning_rate": 2.047917785040202e-05, | |
| "loss": 0.0179, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.031141868512111, | |
| "grad_norm": 0.23955483734607697, | |
| "learning_rate": 1.9790212959771815e-05, | |
| "loss": 0.0194, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.048442906574395, | |
| "grad_norm": 0.4007752239704132, | |
| "learning_rate": 1.911176173452529e-05, | |
| "loss": 0.0213, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 4.0657439446366785, | |
| "grad_norm": 0.41095274686813354, | |
| "learning_rate": 1.8443913104073983e-05, | |
| "loss": 0.0227, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 4.083044982698962, | |
| "grad_norm": 0.29689717292785645, | |
| "learning_rate": 1.7786754608072154e-05, | |
| "loss": 0.0182, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 4.100346020761246, | |
| "grad_norm": 0.4345749616622925, | |
| "learning_rate": 1.7140372384942427e-05, | |
| "loss": 0.0261, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 4.117647058823529, | |
| "grad_norm": 0.2638370990753174, | |
| "learning_rate": 1.6504851160584854e-05, | |
| "loss": 0.0141, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 4.134948096885813, | |
| "grad_norm": 0.271502822637558, | |
| "learning_rate": 1.5880274237271442e-05, | |
| "loss": 0.0238, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 4.1522491349480966, | |
| "grad_norm": 0.32822224497795105, | |
| "learning_rate": 1.5266723482727075e-05, | |
| "loss": 0.0169, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.16955017301038, | |
| "grad_norm": 0.37470826506614685, | |
| "learning_rate": 1.4664279319398566e-05, | |
| "loss": 0.0189, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 4.186851211072664, | |
| "grad_norm": 0.25935235619544983, | |
| "learning_rate": 1.4073020713912987e-05, | |
| "loss": 0.0232, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 4.204152249134948, | |
| "grad_norm": 0.26056811213493347, | |
| "learning_rate": 1.349302516672717e-05, | |
| "loss": 0.022, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 4.221453287197232, | |
| "grad_norm": 0.2688888609409332, | |
| "learning_rate": 1.2924368701968936e-05, | |
| "loss": 0.0165, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 4.2387543252595155, | |
| "grad_norm": 0.2515370547771454, | |
| "learning_rate": 1.2367125857472283e-05, | |
| "loss": 0.0262, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.256055363321799, | |
| "grad_norm": 0.22485879063606262, | |
| "learning_rate": 1.1821369675007076e-05, | |
| "loss": 0.0215, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.273356401384083, | |
| "grad_norm": 0.16641457378864288, | |
| "learning_rate": 1.1287171690704923e-05, | |
| "loss": 0.0168, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.290657439446367, | |
| "grad_norm": 0.20574532449245453, | |
| "learning_rate": 1.076460192568246e-05, | |
| "loss": 0.0191, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.307958477508651, | |
| "grad_norm": 0.22523482143878937, | |
| "learning_rate": 1.0253728876863255e-05, | |
| "loss": 0.0228, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.325259515570934, | |
| "grad_norm": 0.23874153196811676, | |
| "learning_rate": 9.754619507999286e-06, | |
| "loss": 0.0243, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.342560553633218, | |
| "grad_norm": 0.19232115149497986, | |
| "learning_rate": 9.26733924089369e-06, | |
| "loss": 0.0192, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.359861591695502, | |
| "grad_norm": 0.21156880259513855, | |
| "learning_rate": 8.791951946825305e-06, | |
| "loss": 0.0167, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.377162629757786, | |
| "grad_norm": 0.13566535711288452, | |
| "learning_rate": 8.328519938176737e-06, | |
| "loss": 0.0159, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.3944636678200695, | |
| "grad_norm": 0.23292118310928345, | |
| "learning_rate": 7.877103960266574e-06, | |
| "loss": 0.0233, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.411764705882353, | |
| "grad_norm": 0.31070488691329956, | |
| "learning_rate": 7.437763183387048e-06, | |
| "loss": 0.0186, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.429065743944637, | |
| "grad_norm": 0.47433632612228394, | |
| "learning_rate": 7.010555195048241e-06, | |
| "loss": 0.0257, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.446366782006921, | |
| "grad_norm": 0.1894678920507431, | |
| "learning_rate": 6.59553599242958e-06, | |
| "loss": 0.0189, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.463667820069205, | |
| "grad_norm": 0.24350449442863464, | |
| "learning_rate": 6.1927599750399634e-06, | |
| "loss": 0.0213, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.4809688581314875, | |
| "grad_norm": 0.198005810379982, | |
| "learning_rate": 5.802279937587218e-06, | |
| "loss": 0.015, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.498269896193771, | |
| "grad_norm": 0.1783621609210968, | |
| "learning_rate": 5.424147063057938e-06, | |
| "loss": 0.0225, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.515570934256055, | |
| "grad_norm": 0.1628614068031311, | |
| "learning_rate": 5.058410916008494e-06, | |
| "loss": 0.0163, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.532871972318339, | |
| "grad_norm": 0.28581690788269043, | |
| "learning_rate": 4.70511943606835e-06, | |
| "loss": 0.0161, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.550173010380623, | |
| "grad_norm": 0.12549340724945068, | |
| "learning_rate": 4.364318931656186e-06, | |
| "loss": 0.0187, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.567474048442906, | |
| "grad_norm": 0.23058100044727325, | |
| "learning_rate": 4.0360540739100335e-06, | |
| "loss": 0.0234, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.58477508650519, | |
| "grad_norm": 0.21442987024784088, | |
| "learning_rate": 3.7203678908318327e-06, | |
| "loss": 0.0215, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.602076124567474, | |
| "grad_norm": 0.15471522510051727, | |
| "learning_rate": 3.417301761647429e-06, | |
| "loss": 0.0145, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.619377162629758, | |
| "grad_norm": 0.3764342963695526, | |
| "learning_rate": 3.1268954113827798e-06, | |
| "loss": 0.0256, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.6366782006920415, | |
| "grad_norm": 0.32330504059791565, | |
| "learning_rate": 2.8491869056568643e-06, | |
| "loss": 0.0196, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.653979238754325, | |
| "grad_norm": 0.27203860878944397, | |
| "learning_rate": 2.5842126456921633e-06, | |
| "loss": 0.0157, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.671280276816609, | |
| "grad_norm": 0.1730179786682129, | |
| "learning_rate": 2.3320073635432984e-06, | |
| "loss": 0.0184, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.688581314878893, | |
| "grad_norm": 0.2550993263721466, | |
| "learning_rate": 2.092604117544461e-06, | |
| "loss": 0.0146, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.23976610600948334, | |
| "learning_rate": 1.8660342879761817e-06, | |
| "loss": 0.0198, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.72318339100346, | |
| "grad_norm": 0.1788170337677002, | |
| "learning_rate": 1.6523275729521615e-06, | |
| "loss": 0.012, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.740484429065744, | |
| "grad_norm": 0.16593609750270844, | |
| "learning_rate": 1.4515119845264658e-06, | |
| "loss": 0.0118, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.757785467128028, | |
| "grad_norm": 0.2375115007162094, | |
| "learning_rate": 1.2636138450218382e-06, | |
| "loss": 0.0186, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.775086505190312, | |
| "grad_norm": 0.163288876414299, | |
| "learning_rate": 1.0886577835793831e-06, | |
| "loss": 0.0156, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.7923875432525955, | |
| "grad_norm": 0.19594237208366394, | |
| "learning_rate": 9.26666732930348e-07, | |
| "loss": 0.014, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.809688581314878, | |
| "grad_norm": 0.09780792146921158, | |
| "learning_rate": 7.776619263900387e-07, | |
| "loss": 0.0152, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.826989619377162, | |
| "grad_norm": 0.17942482233047485, | |
| "learning_rate": 6.416628950747461e-07, | |
| "loss": 0.0183, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.844290657439446, | |
| "grad_norm": 0.34686434268951416, | |
| "learning_rate": 5.186874653415718e-07, | |
| "loss": 0.0188, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.86159169550173, | |
| "grad_norm": 0.1715351790189743, | |
| "learning_rate": 4.087517564518528e-07, | |
| "loss": 0.0139, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.8788927335640135, | |
| "grad_norm": 0.21105673909187317, | |
| "learning_rate": 3.1187017845827337e-07, | |
| "loss": 0.0162, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.896193771626297, | |
| "grad_norm": 0.2553834319114685, | |
| "learning_rate": 2.2805543031604314e-07, | |
| "loss": 0.0241, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.913494809688581, | |
| "grad_norm": 0.2684378921985626, | |
| "learning_rate": 1.5731849821833954e-07, | |
| "loss": 0.0203, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.930795847750865, | |
| "grad_norm": 0.2061266154050827, | |
| "learning_rate": 9.966865415631521e-08, | |
| "loss": 0.0195, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.948096885813149, | |
| "grad_norm": 0.19463296234607697, | |
| "learning_rate": 5.5113454703692445e-08, | |
| "loss": 0.0146, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.965397923875432, | |
| "grad_norm": 0.27894893288612366, | |
| "learning_rate": 2.3658740026311077e-08, | |
| "loss": 0.0225, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.982698961937716, | |
| "grad_norm": 0.26400962471961975, | |
| "learning_rate": 5.3086331166074535e-09, | |
| "loss": 0.0168, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.998269896193771, | |
| "step": 2889, | |
| "total_flos": 1.012715054916768e+17, | |
| "train_loss": 0.04862415902999991, | |
| "train_runtime": 1280.7015, | |
| "train_samples_per_second": 36.093, | |
| "train_steps_per_second": 2.256 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2889, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 1.012715054916768e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |