{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 1206, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0008291873963515755, "grad_norm": 24.117380589781583, "learning_rate": 8.264462809917357e-08, "loss": 1.4263, "step": 1 }, { "epoch": 0.0041459369817578775, "grad_norm": 22.360999037367943, "learning_rate": 4.132231404958678e-07, "loss": 1.4076, "step": 5 }, { "epoch": 0.008291873963515755, "grad_norm": 8.103797569647355, "learning_rate": 8.264462809917356e-07, "loss": 1.3065, "step": 10 }, { "epoch": 0.012437810945273632, "grad_norm": 10.622518895271666, "learning_rate": 1.2396694214876035e-06, "loss": 1.161, "step": 15 }, { "epoch": 0.01658374792703151, "grad_norm": 2.9542061172740928, "learning_rate": 1.6528925619834712e-06, "loss": 0.9983, "step": 20 }, { "epoch": 0.020729684908789386, "grad_norm": 3.0228806635514696, "learning_rate": 2.066115702479339e-06, "loss": 0.9593, "step": 25 }, { "epoch": 0.024875621890547265, "grad_norm": 2.656334692534878, "learning_rate": 2.479338842975207e-06, "loss": 0.9235, "step": 30 }, { "epoch": 0.02902155887230514, "grad_norm": 2.1702370957074213, "learning_rate": 2.8925619834710743e-06, "loss": 0.9044, "step": 35 }, { "epoch": 0.03316749585406302, "grad_norm": 2.341195712446691, "learning_rate": 3.3057851239669424e-06, "loss": 0.8998, "step": 40 }, { "epoch": 0.03731343283582089, "grad_norm": 2.2826028497769526, "learning_rate": 3.71900826446281e-06, "loss": 0.8781, "step": 45 }, { "epoch": 0.04145936981757877, "grad_norm": 2.395866474689141, "learning_rate": 4.132231404958678e-06, "loss": 0.8757, "step": 50 }, { "epoch": 0.04560530679933665, "grad_norm": 2.329664924545399, "learning_rate": 4.5454545454545455e-06, "loss": 0.8535, "step": 55 }, { "epoch": 0.04975124378109453, "grad_norm": 2.319973421427611, "learning_rate": 4.958677685950414e-06, "loss": 0.8501, "step": 60 }, { "epoch": 0.0538971807628524, "grad_norm": 2.34618245693918, "learning_rate": 5.371900826446281e-06, "loss": 0.8459, "step": 65 }, { "epoch": 0.05804311774461028, "grad_norm": 2.4876099171019286, "learning_rate": 5.785123966942149e-06, "loss": 0.8359, "step": 70 }, { "epoch": 0.06218905472636816, "grad_norm": 2.4937280574508316, "learning_rate": 6.198347107438017e-06, "loss": 0.8296, "step": 75 }, { "epoch": 0.06633499170812604, "grad_norm": 2.5000631568636784, "learning_rate": 6.611570247933885e-06, "loss": 0.8298, "step": 80 }, { "epoch": 0.07048092868988391, "grad_norm": 2.3026309618556406, "learning_rate": 7.0247933884297525e-06, "loss": 0.8324, "step": 85 }, { "epoch": 0.07462686567164178, "grad_norm": 2.441626244973287, "learning_rate": 7.43801652892562e-06, "loss": 0.8237, "step": 90 }, { "epoch": 0.07877280265339967, "grad_norm": 2.6334045094131935, "learning_rate": 7.851239669421489e-06, "loss": 0.8173, "step": 95 }, { "epoch": 0.08291873963515754, "grad_norm": 2.8743818026826933, "learning_rate": 8.264462809917356e-06, "loss": 0.8179, "step": 100 }, { "epoch": 0.08706467661691543, "grad_norm": 2.4782540737477805, "learning_rate": 8.677685950413224e-06, "loss": 0.8138, "step": 105 }, { "epoch": 0.0912106135986733, "grad_norm": 2.516716969960893, "learning_rate": 9.090909090909091e-06, "loss": 0.8006, "step": 110 }, { "epoch": 0.09535655058043117, "grad_norm": 2.661516962429658, "learning_rate": 9.50413223140496e-06, "loss": 0.8088, "step": 115 }, { "epoch": 0.09950248756218906, "grad_norm": 2.169914953698175, "learning_rate": 9.917355371900828e-06, "loss": 0.795, "step": 120 }, { "epoch": 0.10364842454394693, "grad_norm": 2.6794554187970383, "learning_rate": 9.999664652243188e-06, "loss": 0.7991, "step": 125 }, { "epoch": 0.1077943615257048, "grad_norm": 2.812395361063442, "learning_rate": 9.99830237907608e-06, "loss": 0.7874, "step": 130 }, { "epoch": 0.11194029850746269, "grad_norm": 2.5526772224584495, "learning_rate": 9.995892506564461e-06, "loss": 0.7899, "step": 135 }, { "epoch": 0.11608623548922056, "grad_norm": 2.3409010885917465, "learning_rate": 9.992435539796e-06, "loss": 0.7802, "step": 140 }, { "epoch": 0.12023217247097844, "grad_norm": 2.517749639364559, "learning_rate": 9.987932203319917e-06, "loss": 0.7742, "step": 145 }, { "epoch": 0.12437810945273632, "grad_norm": 2.5710403396477086, "learning_rate": 9.982383440995146e-06, "loss": 0.7634, "step": 150 }, { "epoch": 0.1285240464344942, "grad_norm": 2.423809126521337, "learning_rate": 9.975790415792497e-06, "loss": 0.767, "step": 155 }, { "epoch": 0.13266998341625208, "grad_norm": 2.2531176128656347, "learning_rate": 9.968154509550914e-06, "loss": 0.7651, "step": 160 }, { "epoch": 0.13681592039800994, "grad_norm": 2.2909698896750497, "learning_rate": 9.959477322687852e-06, "loss": 0.7589, "step": 165 }, { "epoch": 0.14096185737976782, "grad_norm": 2.272355603940266, "learning_rate": 9.949760673863846e-06, "loss": 0.7499, "step": 170 }, { "epoch": 0.1451077943615257, "grad_norm": 2.3263130353605814, "learning_rate": 9.93900659960133e-06, "loss": 0.7523, "step": 175 }, { "epoch": 0.14925373134328357, "grad_norm": 2.3867379201836374, "learning_rate": 9.927217353857809e-06, "loss": 0.75, "step": 180 }, { "epoch": 0.15339966832504145, "grad_norm": 2.3095796755747413, "learning_rate": 9.914395407553444e-06, "loss": 0.7287, "step": 185 }, { "epoch": 0.15754560530679934, "grad_norm": 4.162860668791136, "learning_rate": 9.900543448053164e-06, "loss": 0.7347, "step": 190 }, { "epoch": 0.16169154228855723, "grad_norm": 2.4891248095680356, "learning_rate": 9.885664378603432e-06, "loss": 0.7325, "step": 195 }, { "epoch": 0.16583747927031509, "grad_norm": 2.4243428099512183, "learning_rate": 9.869761317723744e-06, "loss": 0.7334, "step": 200 }, { "epoch": 0.16998341625207297, "grad_norm": 2.281996216090256, "learning_rate": 9.85283759855301e-06, "loss": 0.7336, "step": 205 }, { "epoch": 0.17412935323383086, "grad_norm": 2.5493532152498295, "learning_rate": 9.834896768150963e-06, "loss": 0.7208, "step": 210 }, { "epoch": 0.17827529021558872, "grad_norm": 2.3851937555382015, "learning_rate": 9.81594258675473e-06, "loss": 0.7262, "step": 215 }, { "epoch": 0.1824212271973466, "grad_norm": 2.375711001816407, "learning_rate": 9.795979026990717e-06, "loss": 0.7027, "step": 220 }, { "epoch": 0.1865671641791045, "grad_norm": 2.4218230611759064, "learning_rate": 9.775010273041975e-06, "loss": 0.6911, "step": 225 }, { "epoch": 0.19071310116086235, "grad_norm": 2.3750995885676707, "learning_rate": 9.753040719771249e-06, "loss": 0.7037, "step": 230 }, { "epoch": 0.19485903814262023, "grad_norm": 2.4845749432051534, "learning_rate": 9.730074971799837e-06, "loss": 0.7067, "step": 235 }, { "epoch": 0.19900497512437812, "grad_norm": 2.352728065720678, "learning_rate": 9.706117842542517e-06, "loss": 0.6847, "step": 240 }, { "epoch": 0.20315091210613598, "grad_norm": 2.5880146309365957, "learning_rate": 9.681174353198687e-06, "loss": 0.6761, "step": 245 }, { "epoch": 0.20729684908789386, "grad_norm": 2.265463148401158, "learning_rate": 9.655249731699973e-06, "loss": 0.6792, "step": 250 }, { "epoch": 0.21144278606965175, "grad_norm": 2.3094585077391243, "learning_rate": 9.628349411614503e-06, "loss": 0.6889, "step": 255 }, { "epoch": 0.2155887230514096, "grad_norm": 2.5538802594012315, "learning_rate": 9.600479031008072e-06, "loss": 0.6829, "step": 260 }, { "epoch": 0.2197346600331675, "grad_norm": 2.411078499465308, "learning_rate": 9.571644431262463e-06, "loss": 0.6754, "step": 265 }, { "epoch": 0.22388059701492538, "grad_norm": 2.360420446008403, "learning_rate": 9.54185165585114e-06, "loss": 0.6626, "step": 270 }, { "epoch": 0.22802653399668324, "grad_norm": 2.459578001858168, "learning_rate": 9.511106949072588e-06, "loss": 0.669, "step": 275 }, { "epoch": 0.23217247097844113, "grad_norm": 2.4402362565595115, "learning_rate": 9.479416754741577e-06, "loss": 0.6615, "step": 280 }, { "epoch": 0.236318407960199, "grad_norm": 2.5822351337198906, "learning_rate": 9.446787714838579e-06, "loss": 0.6514, "step": 285 }, { "epoch": 0.24046434494195687, "grad_norm": 2.313401890971625, "learning_rate": 9.413226668117679e-06, "loss": 0.6526, "step": 290 }, { "epoch": 0.24461028192371476, "grad_norm": 2.2602171932852277, "learning_rate": 9.37874064867323e-06, "loss": 0.6407, "step": 295 }, { "epoch": 0.24875621890547264, "grad_norm": 2.199439032927717, "learning_rate": 9.343336884465577e-06, "loss": 0.6538, "step": 300 }, { "epoch": 0.25290215588723053, "grad_norm": 2.3885639020372613, "learning_rate": 9.307022795806125e-06, "loss": 0.6489, "step": 305 }, { "epoch": 0.2570480928689884, "grad_norm": 2.1637672184495114, "learning_rate": 9.26980599380213e-06, "loss": 0.6378, "step": 310 }, { "epoch": 0.26119402985074625, "grad_norm": 2.430992287073813, "learning_rate": 9.231694278761455e-06, "loss": 0.6277, "step": 315 }, { "epoch": 0.26533996683250416, "grad_norm": 2.241700824439163, "learning_rate": 9.192695638557723e-06, "loss": 0.6269, "step": 320 }, { "epoch": 0.269485903814262, "grad_norm": 2.3110446586037723, "learning_rate": 9.1528182469561e-06, "loss": 0.6188, "step": 325 }, { "epoch": 0.2736318407960199, "grad_norm": 2.3953066920058266, "learning_rate": 9.112070461900178e-06, "loss": 0.6165, "step": 330 }, { "epoch": 0.2777777777777778, "grad_norm": 2.373760618925254, "learning_rate": 9.070460823760197e-06, "loss": 0.6083, "step": 335 }, { "epoch": 0.28192371475953565, "grad_norm": 2.316668729079351, "learning_rate": 9.027998053543079e-06, "loss": 0.6108, "step": 340 }, { "epoch": 0.2860696517412935, "grad_norm": 2.2857072319973835, "learning_rate": 8.984691051064576e-06, "loss": 0.6232, "step": 345 }, { "epoch": 0.2902155887230514, "grad_norm": 2.098710949700429, "learning_rate": 8.94054889308395e-06, "loss": 0.6051, "step": 350 }, { "epoch": 0.2943615257048093, "grad_norm": 2.20767713179006, "learning_rate": 8.895580831401563e-06, "loss": 0.6128, "step": 355 }, { "epoch": 0.29850746268656714, "grad_norm": 2.2284584055532575, "learning_rate": 8.849796290919787e-06, "loss": 0.6021, "step": 360 }, { "epoch": 0.30265339966832505, "grad_norm": 2.1494719513708054, "learning_rate": 8.803204867667624e-06, "loss": 0.5985, "step": 365 }, { "epoch": 0.3067993366500829, "grad_norm": 2.197939057621055, "learning_rate": 8.755816326789469e-06, "loss": 0.5939, "step": 370 }, { "epoch": 0.31094527363184077, "grad_norm": 2.2386284592458914, "learning_rate": 8.70764060049842e-06, "loss": 0.593, "step": 375 }, { "epoch": 0.3150912106135987, "grad_norm": 2.401742728390247, "learning_rate": 8.658687785994579e-06, "loss": 0.5929, "step": 380 }, { "epoch": 0.31923714759535654, "grad_norm": 2.185601347943659, "learning_rate": 8.608968143348765e-06, "loss": 0.5785, "step": 385 }, { "epoch": 0.32338308457711445, "grad_norm": 2.382435915886427, "learning_rate": 8.558492093352098e-06, "loss": 0.5941, "step": 390 }, { "epoch": 0.3275290215588723, "grad_norm": 2.245237966412665, "learning_rate": 8.50727021533189e-06, "loss": 0.5791, "step": 395 }, { "epoch": 0.33167495854063017, "grad_norm": 2.210905541427763, "learning_rate": 8.455313244934324e-06, "loss": 0.5776, "step": 400 }, { "epoch": 0.3358208955223881, "grad_norm": 2.25124182944938, "learning_rate": 8.402632071874348e-06, "loss": 0.5761, "step": 405 }, { "epoch": 0.33996683250414594, "grad_norm": 2.318367828560386, "learning_rate": 8.349237737653288e-06, "loss": 0.5642, "step": 410 }, { "epoch": 0.3441127694859038, "grad_norm": 2.22226285780482, "learning_rate": 8.29514143324466e-06, "loss": 0.5717, "step": 415 }, { "epoch": 0.3482587064676617, "grad_norm": 2.3365625686335454, "learning_rate": 8.24035449674863e-06, "loss": 0.5742, "step": 420 }, { "epoch": 0.3524046434494196, "grad_norm": 2.397392257225569, "learning_rate": 8.184888411015655e-06, "loss": 0.5572, "step": 425 }, { "epoch": 0.35655058043117743, "grad_norm": 2.170211523773955, "learning_rate": 8.128754801239781e-06, "loss": 0.5643, "step": 430 }, { "epoch": 0.36069651741293535, "grad_norm": 2.2345632725077036, "learning_rate": 8.071965432522107e-06, "loss": 0.5511, "step": 435 }, { "epoch": 0.3648424543946932, "grad_norm": 2.5642658922512034, "learning_rate": 8.01453220740492e-06, "loss": 0.5549, "step": 440 }, { "epoch": 0.36898839137645106, "grad_norm": 2.2643156664457664, "learning_rate": 7.956467163377037e-06, "loss": 0.548, "step": 445 }, { "epoch": 0.373134328358209, "grad_norm": 2.2598271385003073, "learning_rate": 7.89778247035085e-06, "loss": 0.549, "step": 450 }, { "epoch": 0.37728026533996684, "grad_norm": 2.2096973686709043, "learning_rate": 7.838490428111625e-06, "loss": 0.5481, "step": 455 }, { "epoch": 0.3814262023217247, "grad_norm": 2.2158742606108013, "learning_rate": 7.77860346373957e-06, "loss": 0.5353, "step": 460 }, { "epoch": 0.3855721393034826, "grad_norm": 2.227206784578813, "learning_rate": 7.718134129005238e-06, "loss": 0.5458, "step": 465 }, { "epoch": 0.38971807628524047, "grad_norm": 2.250456554114917, "learning_rate": 7.657095097738793e-06, "loss": 0.5346, "step": 470 }, { "epoch": 0.3938640132669983, "grad_norm": 2.0913145200295173, "learning_rate": 7.59549916317367e-06, "loss": 0.5319, "step": 475 }, { "epoch": 0.39800995024875624, "grad_norm": 2.2021018890450277, "learning_rate": 7.533359235265248e-06, "loss": 0.5399, "step": 480 }, { "epoch": 0.4021558872305141, "grad_norm": 2.470025254943181, "learning_rate": 7.470688337985029e-06, "loss": 0.5286, "step": 485 }, { "epoch": 0.40630182421227196, "grad_norm": 2.118719596596625, "learning_rate": 7.407499606590934e-06, "loss": 0.5307, "step": 490 }, { "epoch": 0.41044776119402987, "grad_norm": 2.2191071823240054, "learning_rate": 7.343806284874268e-06, "loss": 0.5163, "step": 495 }, { "epoch": 0.41459369817578773, "grad_norm": 2.2680281384787158, "learning_rate": 7.279621722383939e-06, "loss": 0.5201, "step": 500 }, { "epoch": 0.4187396351575456, "grad_norm": 2.192485128822605, "learning_rate": 7.214959371628522e-06, "loss": 0.5147, "step": 505 }, { "epoch": 0.4228855721393035, "grad_norm": 2.0954471941939348, "learning_rate": 7.149832785256718e-06, "loss": 0.5222, "step": 510 }, { "epoch": 0.42703150912106136, "grad_norm": 2.2786753291313606, "learning_rate": 7.084255613216855e-06, "loss": 0.5131, "step": 515 }, { "epoch": 0.4311774461028192, "grad_norm": 2.255057502517376, "learning_rate": 7.018241599895974e-06, "loss": 0.5044, "step": 520 }, { "epoch": 0.43532338308457713, "grad_norm": 2.231896861869262, "learning_rate": 6.95180458123913e-06, "loss": 0.5021, "step": 525 }, { "epoch": 0.439469320066335, "grad_norm": 2.150100810777524, "learning_rate": 6.8849584818494984e-06, "loss": 0.5063, "step": 530 }, { "epoch": 0.44361525704809285, "grad_norm": 2.166139345995876, "learning_rate": 6.817717312069913e-06, "loss": 0.5013, "step": 535 }, { "epoch": 0.44776119402985076, "grad_norm": 2.1458812545141024, "learning_rate": 6.750095165046415e-06, "loss": 0.4915, "step": 540 }, { "epoch": 0.4519071310116086, "grad_norm": 2.1508468010441106, "learning_rate": 6.682106213774459e-06, "loss": 0.4909, "step": 545 }, { "epoch": 0.4560530679933665, "grad_norm": 2.1329384520928216, "learning_rate": 6.6137647081283776e-06, "loss": 0.4884, "step": 550 }, { "epoch": 0.4601990049751244, "grad_norm": 2.204604952138092, "learning_rate": 6.545084971874738e-06, "loss": 0.4898, "step": 555 }, { "epoch": 0.46434494195688225, "grad_norm": 2.151094585470471, "learning_rate": 6.476081399670212e-06, "loss": 0.5026, "step": 560 }, { "epoch": 0.4684908789386401, "grad_norm": 2.1220218077497583, "learning_rate": 6.406768454044581e-06, "loss": 0.483, "step": 565 }, { "epoch": 0.472636815920398, "grad_norm": 2.351276627704081, "learning_rate": 6.337160662369519e-06, "loss": 0.4898, "step": 570 }, { "epoch": 0.4767827529021559, "grad_norm": 2.097791645235577, "learning_rate": 6.267272613813789e-06, "loss": 0.4867, "step": 575 }, { "epoch": 0.48092868988391374, "grad_norm": 2.254147716445627, "learning_rate": 6.19711895628548e-06, "loss": 0.4838, "step": 580 }, { "epoch": 0.48507462686567165, "grad_norm": 2.132491206903525, "learning_rate": 6.126714393361939e-06, "loss": 0.4687, "step": 585 }, { "epoch": 0.4892205638474295, "grad_norm": 2.1755427175233675, "learning_rate": 6.056073681208038e-06, "loss": 0.4846, "step": 590 }, { "epoch": 0.49336650082918737, "grad_norm": 2.2166671419673936, "learning_rate": 5.985211625483403e-06, "loss": 0.4691, "step": 595 }, { "epoch": 0.4975124378109453, "grad_norm": 2.074430457553342, "learning_rate": 5.914143078239293e-06, "loss": 0.4668, "step": 600 }, { "epoch": 0.5016583747927031, "grad_norm": 2.1855010339066285, "learning_rate": 5.842882934805731e-06, "loss": 0.4747, "step": 605 }, { "epoch": 0.5058043117744611, "grad_norm": 2.080130024063031, "learning_rate": 5.771446130669589e-06, "loss": 0.4648, "step": 610 }, { "epoch": 0.5099502487562189, "grad_norm": 2.2706734365013586, "learning_rate": 5.6998476383442345e-06, "loss": 0.4672, "step": 615 }, { "epoch": 0.5140961857379768, "grad_norm": 2.2061487275873675, "learning_rate": 5.628102464231429e-06, "loss": 0.4572, "step": 620 }, { "epoch": 0.5182421227197347, "grad_norm": 2.106202128986693, "learning_rate": 5.556225645476119e-06, "loss": 0.4551, "step": 625 }, { "epoch": 0.5223880597014925, "grad_norm": 2.164397168766167, "learning_rate": 5.4842322468147926e-06, "loss": 0.4631, "step": 630 }, { "epoch": 0.5265339966832504, "grad_norm": 2.1067639245390164, "learning_rate": 5.412137357418037e-06, "loss": 0.4558, "step": 635 }, { "epoch": 0.5306799336650083, "grad_norm": 2.042765318334228, "learning_rate": 5.339956087727985e-06, "loss": 0.4517, "step": 640 }, { "epoch": 0.5348258706467661, "grad_norm": 2.2878931581220607, "learning_rate": 5.2677035662913116e-06, "loss": 0.4509, "step": 645 }, { "epoch": 0.538971807628524, "grad_norm": 2.257982283268261, "learning_rate": 5.195394936588409e-06, "loss": 0.4465, "step": 650 }, { "epoch": 0.543117744610282, "grad_norm": 2.0663328687466027, "learning_rate": 5.123045353859465e-06, "loss": 0.4543, "step": 655 }, { "epoch": 0.5472636815920398, "grad_norm": 2.1437144866342983, "learning_rate": 5.050669981928056e-06, "loss": 0.4525, "step": 660 }, { "epoch": 0.5514096185737977, "grad_norm": 2.0883208963142113, "learning_rate": 4.978283990022936e-06, "loss": 0.4438, "step": 665 }, { "epoch": 0.5555555555555556, "grad_norm": 2.111420551110085, "learning_rate": 4.905902549598719e-06, "loss": 0.445, "step": 670 }, { "epoch": 0.5597014925373134, "grad_norm": 2.068453334060506, "learning_rate": 4.833540831156062e-06, "loss": 0.4402, "step": 675 }, { "epoch": 0.5638474295190713, "grad_norm": 2.158910191803973, "learning_rate": 4.761214001062079e-06, "loss": 0.4426, "step": 680 }, { "epoch": 0.5679933665008292, "grad_norm": 2.1065328754001693, "learning_rate": 4.688937218371592e-06, "loss": 0.4454, "step": 685 }, { "epoch": 0.572139303482587, "grad_norm": 2.019371521833108, "learning_rate": 4.616725631649938e-06, "loss": 0.434, "step": 690 }, { "epoch": 0.5762852404643449, "grad_norm": 2.162342441008238, "learning_rate": 4.544594375797969e-06, "loss": 0.4359, "step": 695 }, { "epoch": 0.5804311774461028, "grad_norm": 2.1532621197943596, "learning_rate": 4.472558568879901e-06, "loss": 0.4264, "step": 700 }, { "epoch": 0.5845771144278606, "grad_norm": 2.045817805816866, "learning_rate": 4.400633308954713e-06, "loss": 0.4377, "step": 705 }, { "epoch": 0.5887230514096186, "grad_norm": 2.068014268281018, "learning_rate": 4.3288336709117246e-06, "loss": 0.4286, "step": 710 }, { "epoch": 0.5928689883913765, "grad_norm": 2.0015792155885372, "learning_rate": 4.257174703311032e-06, "loss": 0.4177, "step": 715 }, { "epoch": 0.5970149253731343, "grad_norm": 1.991509167040826, "learning_rate": 4.185671425229477e-06, "loss": 0.4229, "step": 720 }, { "epoch": 0.6011608623548922, "grad_norm": 2.1167861439632065, "learning_rate": 4.11433882311277e-06, "loss": 0.4168, "step": 725 }, { "epoch": 0.6053067993366501, "grad_norm": 2.157855739678933, "learning_rate": 4.043191847634469e-06, "loss": 0.4269, "step": 730 }, { "epoch": 0.6094527363184079, "grad_norm": 2.079976376382546, "learning_rate": 3.9722454105624545e-06, "loss": 0.4089, "step": 735 }, { "epoch": 0.6135986733001658, "grad_norm": 2.0902836848679125, "learning_rate": 3.901514381633555e-06, "loss": 0.4155, "step": 740 }, { "epoch": 0.6177446102819237, "grad_norm": 2.212253263846612, "learning_rate": 3.831013585436985e-06, "loss": 0.4058, "step": 745 }, { "epoch": 0.6218905472636815, "grad_norm": 2.031272866177863, "learning_rate": 3.7607577983072486e-06, "loss": 0.4215, "step": 750 }, { "epoch": 0.6260364842454395, "grad_norm": 2.0730340821988236, "learning_rate": 3.6907617452271394e-06, "loss": 0.4088, "step": 755 }, { "epoch": 0.6301824212271974, "grad_norm": 2.026454332990267, "learning_rate": 3.621040096741526e-06, "loss": 0.4093, "step": 760 }, { "epoch": 0.6343283582089553, "grad_norm": 2.0635438234787986, "learning_rate": 3.55160746588254e-06, "loss": 0.4111, "step": 765 }, { "epoch": 0.6384742951907131, "grad_norm": 2.2500499736874193, "learning_rate": 3.482478405106803e-06, "loss": 0.4059, "step": 770 }, { "epoch": 0.642620232172471, "grad_norm": 2.049125956105953, "learning_rate": 3.4136674032453787e-06, "loss": 0.4055, "step": 775 }, { "epoch": 0.6467661691542289, "grad_norm": 2.1798987402102656, "learning_rate": 3.34518888246703e-06, "loss": 0.4001, "step": 780 }, { "epoch": 0.6509121061359867, "grad_norm": 2.018383080122737, "learning_rate": 3.2770571952554674e-06, "loss": 0.4115, "step": 785 }, { "epoch": 0.6550580431177446, "grad_norm": 2.116608719766938, "learning_rate": 3.2092866214011984e-06, "loss": 0.3994, "step": 790 }, { "epoch": 0.6592039800995025, "grad_norm": 2.0325657099793966, "learning_rate": 3.141891365008609e-06, "loss": 0.3934, "step": 795 }, { "epoch": 0.6633499170812603, "grad_norm": 1.951948440418877, "learning_rate": 3.0748855515189104e-06, "loss": 0.3988, "step": 800 }, { "epoch": 0.6674958540630183, "grad_norm": 2.1780405134646665, "learning_rate": 3.00828322474958e-06, "loss": 0.3905, "step": 805 }, { "epoch": 0.6716417910447762, "grad_norm": 2.0809299186938452, "learning_rate": 2.942098343950891e-06, "loss": 0.3944, "step": 810 }, { "epoch": 0.675787728026534, "grad_norm": 1.9769415360760876, "learning_rate": 2.8763447808801914e-06, "loss": 0.3952, "step": 815 }, { "epoch": 0.6799336650082919, "grad_norm": 2.0165191387468577, "learning_rate": 2.8110363168944976e-06, "loss": 0.3943, "step": 820 }, { "epoch": 0.6840796019900498, "grad_norm": 2.067073539421952, "learning_rate": 2.7461866400620506e-06, "loss": 0.397, "step": 825 }, { "epoch": 0.6882255389718076, "grad_norm": 2.2430133872078235, "learning_rate": 2.6818093422934254e-06, "loss": 0.3873, "step": 830 }, { "epoch": 0.6923714759535655, "grad_norm": 2.0056674691322414, "learning_rate": 2.617917916492776e-06, "loss": 0.3862, "step": 835 }, { "epoch": 0.6965174129353234, "grad_norm": 2.2368944940213895, "learning_rate": 2.5545257537298497e-06, "loss": 0.3881, "step": 840 }, { "epoch": 0.7006633499170812, "grad_norm": 2.0486303817757694, "learning_rate": 2.491646140433346e-06, "loss": 0.3826, "step": 845 }, { "epoch": 0.7048092868988391, "grad_norm": 1.9653848895410453, "learning_rate": 2.4292922556061877e-06, "loss": 0.383, "step": 850 }, { "epoch": 0.7089552238805971, "grad_norm": 1.9970130911409754, "learning_rate": 2.367477168063326e-06, "loss": 0.3908, "step": 855 }, { "epoch": 0.7131011608623549, "grad_norm": 2.2828359832017986, "learning_rate": 2.3062138336926406e-06, "loss": 0.3785, "step": 860 }, { "epoch": 0.7172470978441128, "grad_norm": 1.9936084042443714, "learning_rate": 2.245515092739488e-06, "loss": 0.3789, "step": 865 }, { "epoch": 0.7213930348258707, "grad_norm": 2.0968907429438306, "learning_rate": 2.185393667115513e-06, "loss": 0.3763, "step": 870 }, { "epoch": 0.7255389718076285, "grad_norm": 2.094302381182494, "learning_rate": 2.125862157732245e-06, "loss": 0.3824, "step": 875 }, { "epoch": 0.7296849087893864, "grad_norm": 2.0254226506847335, "learning_rate": 2.066933041860059e-06, "loss": 0.3717, "step": 880 }, { "epoch": 0.7338308457711443, "grad_norm": 2.088557256342498, "learning_rate": 2.0086186705130545e-06, "loss": 0.3774, "step": 885 }, { "epoch": 0.7379767827529021, "grad_norm": 1.9729363418312684, "learning_rate": 1.9509312658603954e-06, "loss": 0.3617, "step": 890 }, { "epoch": 0.74212271973466, "grad_norm": 2.004972020498788, "learning_rate": 1.8938829186646484e-06, "loss": 0.3703, "step": 895 }, { "epoch": 0.746268656716418, "grad_norm": 2.068508600154368, "learning_rate": 1.8374855857476687e-06, "loss": 0.3742, "step": 900 }, { "epoch": 0.7504145936981758, "grad_norm": 1.9735024401776045, "learning_rate": 1.7817510874845585e-06, "loss": 0.3726, "step": 905 }, { "epoch": 0.7545605306799337, "grad_norm": 1.9048148982657027, "learning_rate": 1.7266911053262196e-06, "loss": 0.3573, "step": 910 }, { "epoch": 0.7587064676616916, "grad_norm": 2.008169389953658, "learning_rate": 1.6723171793510363e-06, "loss": 0.3648, "step": 915 }, { "epoch": 0.7628524046434494, "grad_norm": 2.0403043389904836, "learning_rate": 1.6186407058461622e-06, "loss": 0.3672, "step": 920 }, { "epoch": 0.7669983416252073, "grad_norm": 2.0206754317295568, "learning_rate": 1.5656729349189742e-06, "loss": 0.3584, "step": 925 }, { "epoch": 0.7711442786069652, "grad_norm": 2.21355887603079, "learning_rate": 1.5134249681391416e-06, "loss": 0.3629, "step": 930 }, { "epoch": 0.775290215588723, "grad_norm": 1.9815337955525554, "learning_rate": 1.4619077562118477e-06, "loss": 0.3602, "step": 935 }, { "epoch": 0.7794361525704809, "grad_norm": 2.095624811245843, "learning_rate": 1.411132096682606e-06, "loss": 0.3625, "step": 940 }, { "epoch": 0.7835820895522388, "grad_norm": 2.0717460611601846, "learning_rate": 1.3611086316742057e-06, "loss": 0.358, "step": 945 }, { "epoch": 0.7877280265339967, "grad_norm": 2.077072842707322, "learning_rate": 1.3118478456562073e-06, "loss": 0.361, "step": 950 }, { "epoch": 0.7918739635157546, "grad_norm": 2.092256674883774, "learning_rate": 1.2633600632474962e-06, "loss": 0.3597, "step": 955 }, { "epoch": 0.7960199004975125, "grad_norm": 2.0064897074384715, "learning_rate": 1.2156554470523364e-06, "loss": 0.3568, "step": 960 }, { "epoch": 0.8001658374792703, "grad_norm": 2.0443529466183583, "learning_rate": 1.1687439955303764e-06, "loss": 0.3609, "step": 965 }, { "epoch": 0.8043117744610282, "grad_norm": 2.1390446036117057, "learning_rate": 1.1226355409010686e-06, "loss": 0.3484, "step": 970 }, { "epoch": 0.8084577114427861, "grad_norm": 2.0533181729015073, "learning_rate": 1.0773397470829145e-06, "loss": 0.3563, "step": 975 }, { "epoch": 0.8126036484245439, "grad_norm": 1.9694189898925665, "learning_rate": 1.032866107667999e-06, "loss": 0.3555, "step": 980 }, { "epoch": 0.8167495854063018, "grad_norm": 2.06716828543327, "learning_rate": 9.892239439322243e-07, "loss": 0.3565, "step": 985 }, { "epoch": 0.8208955223880597, "grad_norm": 1.9932579734924687, "learning_rate": 9.464224028816427e-07, "loss": 0.3458, "step": 990 }, { "epoch": 0.8250414593698175, "grad_norm": 2.054408370219731, "learning_rate": 9.044704553353323e-07, "loss": 0.3552, "step": 995 }, { "epoch": 0.8291873963515755, "grad_norm": 1.973455889397997, "learning_rate": 8.633768940451981e-07, "loss": 0.3486, "step": 1000 }, { "epoch": 0.8333333333333334, "grad_norm": 1.8931502640580207, "learning_rate": 8.231503318530814e-07, "loss": 0.347, "step": 1005 }, { "epoch": 0.8374792703150912, "grad_norm": 2.001049515702371, "learning_rate": 7.837991998855899e-07, "loss": 0.3455, "step": 1010 }, { "epoch": 0.8416252072968491, "grad_norm": 1.9995964909333828, "learning_rate": 7.453317457870096e-07, "loss": 0.3484, "step": 1015 }, { "epoch": 0.845771144278607, "grad_norm": 2.013571417338608, "learning_rate": 7.077560319906696e-07, "loss": 0.3532, "step": 1020 }, { "epoch": 0.8499170812603648, "grad_norm": 1.936993522187304, "learning_rate": 6.710799340291341e-07, "loss": 0.3551, "step": 1025 }, { "epoch": 0.8540630182421227, "grad_norm": 2.0795836765220437, "learning_rate": 6.353111388835564e-07, "loss": 0.3508, "step": 1030 }, { "epoch": 0.8582089552238806, "grad_norm": 2.152120228931756, "learning_rate": 6.00457143372557e-07, "loss": 0.3456, "step": 1035 }, { "epoch": 0.8623548922056384, "grad_norm": 2.0979576731637204, "learning_rate": 5.665252525809583e-07, "loss": 0.3413, "step": 1040 }, { "epoch": 0.8665008291873963, "grad_norm": 2.053994823878588, "learning_rate": 5.335225783287051e-07, "loss": 0.3447, "step": 1045 }, { "epoch": 0.8706467661691543, "grad_norm": 2.0748523280459885, "learning_rate": 5.014560376802913e-07, "loss": 0.3522, "step": 1050 }, { "epoch": 0.8747927031509121, "grad_norm": 1.9949057030771349, "learning_rate": 4.703323514950042e-07, "loss": 0.3422, "step": 1055 }, { "epoch": 0.87893864013267, "grad_norm": 2.041787728436122, "learning_rate": 4.401580430182928e-07, "loss": 0.3468, "step": 1060 }, { "epoch": 0.8830845771144279, "grad_norm": 2.047401944679116, "learning_rate": 4.1093943651455305e-07, "loss": 0.3449, "step": 1065 }, { "epoch": 0.8872305140961857, "grad_norm": 2.026607234375352, "learning_rate": 3.826826559416219e-07, "loss": 0.3436, "step": 1070 }, { "epoch": 0.8913764510779436, "grad_norm": 1.9784143122665, "learning_rate": 3.5539362366724784e-07, "loss": 0.3414, "step": 1075 }, { "epoch": 0.8955223880597015, "grad_norm": 2.0742460744946887, "learning_rate": 3.290780592278148e-07, "loss": 0.3388, "step": 1080 }, { "epoch": 0.8996683250414593, "grad_norm": 2.0665904257359564, "learning_rate": 3.0374147812958387e-07, "loss": 0.3375, "step": 1085 }, { "epoch": 0.9038142620232172, "grad_norm": 2.0474865317583153, "learning_rate": 2.7938919069268654e-07, "loss": 0.3416, "step": 1090 }, { "epoch": 0.9079601990049752, "grad_norm": 2.01615513079154, "learning_rate": 2.5602630093813253e-07, "loss": 0.3459, "step": 1095 }, { "epoch": 0.912106135986733, "grad_norm": 2.0584349650509006, "learning_rate": 2.3365770551805223e-07, "loss": 0.3454, "step": 1100 }, { "epoch": 0.9162520729684909, "grad_norm": 1.9418495675468312, "learning_rate": 2.1228809268940164e-07, "loss": 0.343, "step": 1105 }, { "epoch": 0.9203980099502488, "grad_norm": 2.1058130546046856, "learning_rate": 1.919219413313478e-07, "loss": 0.334, "step": 1110 }, { "epoch": 0.9245439469320066, "grad_norm": 1.9826703763375784, "learning_rate": 1.725635200065323e-07, "loss": 0.3333, "step": 1115 }, { "epoch": 0.9286898839137645, "grad_norm": 2.128017210498838, "learning_rate": 1.5421688606642392e-07, "loss": 0.3405, "step": 1120 }, { "epoch": 0.9328358208955224, "grad_norm": 2.221737755918679, "learning_rate": 1.3688588480092913e-07, "loss": 0.3444, "step": 1125 }, { "epoch": 0.9369817578772802, "grad_norm": 1.973211526295206, "learning_rate": 1.205741486324552e-07, "loss": 0.3371, "step": 1130 }, { "epoch": 0.9411276948590381, "grad_norm": 2.1094021028143146, "learning_rate": 1.0528509635458873e-07, "loss": 0.3372, "step": 1135 }, { "epoch": 0.945273631840796, "grad_norm": 2.105030991712497, "learning_rate": 9.102193241554757e-08, "loss": 0.3431, "step": 1140 }, { "epoch": 0.9494195688225538, "grad_norm": 2.0343476094611295, "learning_rate": 7.778764624655433e-08, "loss": 0.3409, "step": 1145 }, { "epoch": 0.9535655058043118, "grad_norm": 2.2397713243211963, "learning_rate": 6.558501163527964e-08, "loss": 0.3391, "step": 1150 }, { "epoch": 0.9577114427860697, "grad_norm": 1.9822421247052011, "learning_rate": 5.44165861444812e-08, "loss": 0.3405, "step": 1155 }, { "epoch": 0.9618573797678275, "grad_norm": 2.1147360132390713, "learning_rate": 4.428471057596362e-08, "loss": 0.3402, "step": 1160 }, { "epoch": 0.9660033167495854, "grad_norm": 2.0560516008756924, "learning_rate": 3.519150847996422e-08, "loss": 0.3381, "step": 1165 }, { "epoch": 0.9701492537313433, "grad_norm": 2.010722248694153, "learning_rate": 2.713888571007739e-08, "loss": 0.3408, "step": 1170 }, { "epoch": 0.9742951907131011, "grad_norm": 2.0190484921041985, "learning_rate": 2.012853002380466e-08, "loss": 0.3318, "step": 1175 }, { "epoch": 0.978441127694859, "grad_norm": 2.061213730599656, "learning_rate": 1.4161910728816009e-08, "loss": 0.3356, "step": 1180 }, { "epoch": 0.9825870646766169, "grad_norm": 1.9826574739124956, "learning_rate": 9.240278374995637e-09, "loss": 0.3386, "step": 1185 }, { "epoch": 0.9867330016583747, "grad_norm": 1.9287703091271367, "learning_rate": 5.364664492337746e-09, "loss": 0.3378, "step": 1190 }, { "epoch": 0.9908789386401327, "grad_norm": 2.131184018330049, "learning_rate": 2.5358813747500266e-09, "loss": 0.3374, "step": 1195 }, { "epoch": 0.9950248756218906, "grad_norm": 2.053000413572238, "learning_rate": 7.545219097987444e-10, "loss": 0.3383, "step": 1200 }, { "epoch": 0.9991708126036484, "grad_norm": 2.1341756433072914, "learning_rate": 2.0959454449243076e-11, "loss": 0.3423, "step": 1205 }, { "epoch": 1.0, "eval_loss": 0.3398309648036957, "eval_runtime": 124.7272, "eval_samples_per_second": 3.127, "eval_steps_per_second": 0.786, "step": 1206 }, { "epoch": 1.0, "step": 1206, "total_flos": 252459519836160.0, "train_loss": 0.5258088858912437, "train_runtime": 27195.6189, "train_samples_per_second": 1.419, "train_steps_per_second": 0.044 } ], "logging_steps": 5, "max_steps": 1206, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 252459519836160.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }