{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6666666666666666, "eval_steps": 500, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.16593393683433533, "learning_rate": 6e-06, "loss": 1.2019, "step": 10 }, { "grad_norm": 0.2364676296710968, "learning_rate": 1.2666666666666668e-05, "loss": 1.2001, "step": 20 }, { "grad_norm": 0.3223605155944824, "learning_rate": 1.9333333333333333e-05, "loss": 1.1652, "step": 30 }, { "grad_norm": 0.288097620010376, "learning_rate": 2.6000000000000002e-05, "loss": 1.1203, "step": 40 }, { "grad_norm": 0.2957666516304016, "learning_rate": 3.266666666666667e-05, "loss": 1.1073, "step": 50 }, { "grad_norm": 0.3105441927909851, "learning_rate": 3.933333333333333e-05, "loss": 1.0981, "step": 60 }, { "grad_norm": 0.2903534770011902, "learning_rate": 4.600000000000001e-05, "loss": 1.0992, "step": 70 }, { "grad_norm": 0.2657025158405304, "learning_rate": 5.266666666666666e-05, "loss": 1.0949, "step": 80 }, { "grad_norm": 0.22208809852600098, "learning_rate": 5.9333333333333343e-05, "loss": 1.0883, "step": 90 }, { "grad_norm": 0.32055118680000305, "learning_rate": 6.6e-05, "loss": 1.0812, "step": 100 }, { "grad_norm": 0.2965445816516876, "learning_rate": 7.266666666666667e-05, "loss": 1.0723, "step": 110 }, { "grad_norm": 0.3756393790245056, "learning_rate": 7.933333333333334e-05, "loss": 1.0464, "step": 120 }, { "grad_norm": 0.5002507567405701, "learning_rate": 8.6e-05, "loss": 1.0023, "step": 130 }, { "grad_norm": 0.4813876152038574, "learning_rate": 9.266666666666666e-05, "loss": 0.9625, "step": 140 }, { "grad_norm": 0.513647198677063, "learning_rate": 9.933333333333334e-05, "loss": 0.9308, "step": 150 }, { "grad_norm": 0.8362369537353516, "learning_rate": 9.999753945398704e-05, "loss": 0.8938, "step": 160 }, { "grad_norm": 0.5613930821418762, "learning_rate": 9.998903417374228e-05, "loss": 0.862, "step": 170 }, { "grad_norm": 0.5763223767280579, "learning_rate": 9.997445481536973e-05, "loss": 0.8221, "step": 180 }, { "grad_norm": 0.5941640734672546, "learning_rate": 9.995380315038119e-05, "loss": 0.7922, "step": 190 }, { "grad_norm": 0.7477675080299377, "learning_rate": 9.99270816881235e-05, "loss": 0.7702, "step": 200 }, { "grad_norm": 0.5304926633834839, "learning_rate": 9.989429367547377e-05, "loss": 0.7553, "step": 210 }, { "grad_norm": 0.5700099468231201, "learning_rate": 9.985544309644475e-05, "loss": 0.7369, "step": 220 }, { "grad_norm": 0.7928358316421509, "learning_rate": 9.98105346717008e-05, "loss": 0.7222, "step": 230 }, { "grad_norm": 0.7667017579078674, "learning_rate": 9.97595738579843e-05, "loss": 0.7114, "step": 240 }, { "grad_norm": 0.5060045123100281, "learning_rate": 9.970256684745258e-05, "loss": 0.6861, "step": 250 }, { "grad_norm": 0.6732467412948608, "learning_rate": 9.963952056692549e-05, "loss": 0.675, "step": 260 }, { "grad_norm": 0.6021248698234558, "learning_rate": 9.957044267704384e-05, "loss": 0.6767, "step": 270 }, { "grad_norm": 0.6704761981964111, "learning_rate": 9.949534157133844e-05, "loss": 0.6512, "step": 280 }, { "grad_norm": 0.7285646200180054, "learning_rate": 9.941422637521035e-05, "loss": 0.6512, "step": 290 }, { "grad_norm": 0.7110922336578369, "learning_rate": 9.932710694482191e-05, "loss": 0.6196, "step": 300 }, { "grad_norm": 0.6850391626358032, "learning_rate": 9.923399386589933e-05, "loss": 0.6118, "step": 310 }, { "grad_norm": 0.6997315883636475, "learning_rate": 9.913489845244626e-05, "loss": 0.5941, "step": 320 }, { "grad_norm": 0.760143518447876, "learning_rate": 9.902983274536912e-05, "loss": 0.5891, "step": 330 }, { "grad_norm": 0.759657084941864, "learning_rate": 9.891880951101407e-05, "loss": 0.5812, "step": 340 }, { "grad_norm": 0.8276827931404114, "learning_rate": 9.880184223961573e-05, "loss": 0.5674, "step": 350 }, { "grad_norm": 0.5555763244628906, "learning_rate": 9.867894514365802e-05, "loss": 0.5544, "step": 360 }, { "grad_norm": 0.6141887307167053, "learning_rate": 9.855013315614725e-05, "loss": 0.5504, "step": 370 }, { "grad_norm": 0.7782378792762756, "learning_rate": 9.841542192879762e-05, "loss": 0.5447, "step": 380 }, { "grad_norm": 0.8065223693847656, "learning_rate": 9.82748278301294e-05, "loss": 0.5435, "step": 390 }, { "grad_norm": 0.7179561257362366, "learning_rate": 9.812836794348004e-05, "loss": 0.5344, "step": 400 }, { "grad_norm": 0.7837668657302856, "learning_rate": 9.797606006492841e-05, "loss": 0.5249, "step": 410 }, { "grad_norm": 0.7675067782402039, "learning_rate": 9.781792270113241e-05, "loss": 0.5199, "step": 420 }, { "grad_norm": 0.8279472589492798, "learning_rate": 9.765397506708023e-05, "loss": 0.5071, "step": 430 }, { "grad_norm": 0.8260670304298401, "learning_rate": 9.748423708375563e-05, "loss": 0.4926, "step": 440 }, { "grad_norm": 0.881014347076416, "learning_rate": 9.730872937571739e-05, "loss": 0.4907, "step": 450 }, { "grad_norm": 0.9371006488800049, "learning_rate": 9.712747326859315e-05, "loss": 0.4815, "step": 460 }, { "grad_norm": 0.8164715766906738, "learning_rate": 9.69404907864883e-05, "loss": 0.4755, "step": 470 }, { "grad_norm": 0.7007074952125549, "learning_rate": 9.674780464930979e-05, "loss": 0.4649, "step": 480 }, { "grad_norm": 0.9710970520973206, "learning_rate": 9.654943827000548e-05, "loss": 0.4578, "step": 490 }, { "grad_norm": 0.7813271284103394, "learning_rate": 9.634541575171929e-05, "loss": 0.4481, "step": 500 }, { "grad_norm": 0.8395043611526489, "learning_rate": 9.613576188486253e-05, "loss": 0.4437, "step": 510 }, { "grad_norm": 0.8199443221092224, "learning_rate": 9.59205021441015e-05, "loss": 0.4302, "step": 520 }, { "grad_norm": 0.7755788564682007, "learning_rate": 9.569966268526232e-05, "loss": 0.4159, "step": 530 }, { "grad_norm": 0.8591013550758362, "learning_rate": 9.54732703421526e-05, "loss": 0.3954, "step": 540 }, { "grad_norm": 0.8200550079345703, "learning_rate": 9.524135262330098e-05, "loss": 0.3959, "step": 550 }, { "grad_norm": 0.8531577587127686, "learning_rate": 9.50039377086147e-05, "loss": 0.3763, "step": 560 }, { "grad_norm": 0.7548572421073914, "learning_rate": 9.476105444595534e-05, "loss": 0.3647, "step": 570 }, { "grad_norm": 0.8562734723091125, "learning_rate": 9.451273234763371e-05, "loss": 0.3526, "step": 580 }, { "grad_norm": 1.0484408140182495, "learning_rate": 9.425900158682385e-05, "loss": 0.3541, "step": 590 }, { "grad_norm": 1.114036202430725, "learning_rate": 9.399989299389661e-05, "loss": 0.3495, "step": 600 }, { "grad_norm": 1.0067291259765625, "learning_rate": 9.373543805267368e-05, "loss": 0.3446, "step": 610 }, { "grad_norm": 1.0836315155029297, "learning_rate": 9.346566889660193e-05, "loss": 0.3169, "step": 620 }, { "grad_norm": 1.1337170600891113, "learning_rate": 9.319061830484898e-05, "loss": 0.3029, "step": 630 }, { "grad_norm": 1.0124084949493408, "learning_rate": 9.291031969832026e-05, "loss": 0.3035, "step": 640 }, { "grad_norm": 1.1703892946243286, "learning_rate": 9.262480713559808e-05, "loss": 0.3065, "step": 650 }, { "grad_norm": 0.8515862226486206, "learning_rate": 9.233411530880326e-05, "loss": 0.2916, "step": 660 }, { "grad_norm": 1.3676186800003052, "learning_rate": 9.20382795393797e-05, "loss": 0.2933, "step": 670 }, { "grad_norm": 0.8166829943656921, "learning_rate": 9.173733577380258e-05, "loss": 0.2724, "step": 680 }, { "grad_norm": 1.028581142425537, "learning_rate": 9.143132057921058e-05, "loss": 0.26, "step": 690 }, { "grad_norm": 1.035443663597107, "learning_rate": 9.112027113896262e-05, "loss": 0.2551, "step": 700 }, { "grad_norm": 1.22231924533844, "learning_rate": 9.080422524811982e-05, "loss": 0.2613, "step": 710 }, { "grad_norm": 1.0958259105682373, "learning_rate": 9.048322130885305e-05, "loss": 0.2384, "step": 720 }, { "grad_norm": 0.8523461222648621, "learning_rate": 9.015729832577681e-05, "loss": 0.2329, "step": 730 }, { "grad_norm": 0.964102029800415, "learning_rate": 8.982649590120982e-05, "loss": 0.2094, "step": 740 }, { "grad_norm": 0.98223876953125, "learning_rate": 8.949085423036296e-05, "loss": 0.2132, "step": 750 }, { "grad_norm": 1.0459887981414795, "learning_rate": 8.91504140964553e-05, "loss": 0.1981, "step": 760 }, { "grad_norm": 1.3035297393798828, "learning_rate": 8.880521686575857e-05, "loss": 0.1954, "step": 770 }, { "grad_norm": 0.9774158596992493, "learning_rate": 8.845530448257085e-05, "loss": 0.1891, "step": 780 }, { "grad_norm": 1.1295584440231323, "learning_rate": 8.810071946411989e-05, "loss": 0.1783, "step": 790 }, { "grad_norm": 0.8923147916793823, "learning_rate": 8.774150489539707e-05, "loss": 0.1765, "step": 800 }, { "grad_norm": 1.3010443449020386, "learning_rate": 8.737770442392212e-05, "loss": 0.1742, "step": 810 }, { "grad_norm": 0.9801855683326721, "learning_rate": 8.700936225443959e-05, "loss": 0.1587, "step": 820 }, { "grad_norm": 1.0634956359863281, "learning_rate": 8.663652314354765e-05, "loss": 0.1646, "step": 830 }, { "grad_norm": 0.8811051845550537, "learning_rate": 8.625923239425978e-05, "loss": 0.1485, "step": 840 }, { "grad_norm": 1.0639090538024902, "learning_rate": 8.587753585050004e-05, "loss": 0.1585, "step": 850 }, { "grad_norm": 0.9883255362510681, "learning_rate": 8.549147989153276e-05, "loss": 0.1443, "step": 860 }, { "grad_norm": 1.4074150323867798, "learning_rate": 8.510111142632698e-05, "loss": 0.1388, "step": 870 }, { "grad_norm": 1.2396678924560547, "learning_rate": 8.470647788785665e-05, "loss": 0.1283, "step": 880 }, { "grad_norm": 1.1257601976394653, "learning_rate": 8.430762722733714e-05, "loss": 0.1183, "step": 890 }, { "grad_norm": 0.95068359375, "learning_rate": 8.390460790839882e-05, "loss": 0.1372, "step": 900 }, { "grad_norm": 1.1750320196151733, "learning_rate": 8.349746890119826e-05, "loss": 0.1125, "step": 910 }, { "grad_norm": 1.0115416049957275, "learning_rate": 8.308625967646795e-05, "loss": 0.1298, "step": 920 }, { "grad_norm": 0.9820786118507385, "learning_rate": 8.267103019950529e-05, "loss": 0.1199, "step": 930 }, { "grad_norm": 0.9424199461936951, "learning_rate": 8.225183092410128e-05, "loss": 0.1097, "step": 940 }, { "grad_norm": 0.804672122001648, "learning_rate": 8.182871278641009e-05, "loss": 0.1264, "step": 950 }, { "grad_norm": 1.0112346410751343, "learning_rate": 8.140172719875979e-05, "loss": 0.107, "step": 960 }, { "grad_norm": 0.9198870062828064, "learning_rate": 8.097092604340542e-05, "loss": 0.1114, "step": 970 }, { "grad_norm": 0.8466675877571106, "learning_rate": 8.053636166622476e-05, "loss": 0.1178, "step": 980 }, { "grad_norm": 0.9316571354866028, "learning_rate": 8.009808687035798e-05, "loss": 0.1163, "step": 990 }, { "grad_norm": 1.0110657215118408, "learning_rate": 7.965615490979163e-05, "loss": 0.1149, "step": 1000 }, { "grad_norm": 0.7280577421188354, "learning_rate": 7.921061948288773e-05, "loss": 0.105, "step": 1010 }, { "grad_norm": 0.9328599572181702, "learning_rate": 7.87615347258591e-05, "loss": 0.1069, "step": 1020 }, { "grad_norm": 0.9885624051094055, "learning_rate": 7.830895520619128e-05, "loss": 0.1069, "step": 1030 }, { "grad_norm": 0.8272039890289307, "learning_rate": 7.785293591601217e-05, "loss": 0.1128, "step": 1040 }, { "grad_norm": 1.0371506214141846, "learning_rate": 7.739353226541009e-05, "loss": 0.1034, "step": 1050 }, { "grad_norm": 0.8586914539337158, "learning_rate": 7.693080007570084e-05, "loss": 0.1159, "step": 1060 }, { "grad_norm": 0.7747752070426941, "learning_rate": 7.646479557264513e-05, "loss": 0.1115, "step": 1070 }, { "grad_norm": 0.8298690915107727, "learning_rate": 7.599557537961663e-05, "loss": 0.1106, "step": 1080 }, { "grad_norm": 0.8888349533081055, "learning_rate": 7.552319651072164e-05, "loss": 0.0915, "step": 1090 }, { "grad_norm": 0.8841614127159119, "learning_rate": 7.504771636387163e-05, "loss": 0.1019, "step": 1100 }, { "grad_norm": 1.0035121440887451, "learning_rate": 7.456919271380875e-05, "loss": 0.1104, "step": 1110 }, { "grad_norm": 0.8126855492591858, "learning_rate": 7.408768370508576e-05, "loss": 0.1106, "step": 1120 }, { "grad_norm": 0.9205325245857239, "learning_rate": 7.36032478450011e-05, "loss": 0.1164, "step": 1130 }, { "grad_norm": 0.8639265298843384, "learning_rate": 7.311594399648957e-05, "loss": 0.1162, "step": 1140 }, { "grad_norm": 0.9142033457756042, "learning_rate": 7.262583137097018e-05, "loss": 0.1087, "step": 1150 }, { "grad_norm": 0.9747472405433655, "learning_rate": 7.213296952115144e-05, "loss": 0.1083, "step": 1160 }, { "grad_norm": 0.9333903789520264, "learning_rate": 7.16374183337951e-05, "loss": 0.107, "step": 1170 }, { "grad_norm": 0.9898130297660828, "learning_rate": 7.113923802243957e-05, "loss": 0.1063, "step": 1180 }, { "grad_norm": 0.9155701398849487, "learning_rate": 7.06384891200834e-05, "loss": 0.1067, "step": 1190 }, { "grad_norm": 0.8125437498092651, "learning_rate": 7.013523247183e-05, "loss": 0.109, "step": 1200 }, { "grad_norm": 1.1674683094024658, "learning_rate": 6.962952922749457e-05, "loss": 0.1063, "step": 1210 }, { "grad_norm": 0.7635449767112732, "learning_rate": 6.912144083417376e-05, "loss": 0.1014, "step": 1220 }, { "grad_norm": 0.9461978673934937, "learning_rate": 6.861102902877946e-05, "loss": 0.1098, "step": 1230 }, { "grad_norm": 0.9153079986572266, "learning_rate": 6.809835583053715e-05, "loss": 0.0984, "step": 1240 }, { "grad_norm": 0.9081976413726807, "learning_rate": 6.758348353345014e-05, "loss": 0.1082, "step": 1250 }, { "grad_norm": 0.8417534232139587, "learning_rate": 6.706647469873031e-05, "loss": 0.0989, "step": 1260 }, { "grad_norm": 0.7554053664207458, "learning_rate": 6.654739214719641e-05, "loss": 0.1092, "step": 1270 }, { "grad_norm": 0.8323029279708862, "learning_rate": 6.602629895164081e-05, "loss": 0.1084, "step": 1280 }, { "grad_norm": 0.8539698719978333, "learning_rate": 6.550325842916559e-05, "loss": 0.1096, "step": 1290 }, { "grad_norm": 0.8106352686882019, "learning_rate": 6.497833413348909e-05, "loss": 0.0989, "step": 1300 }, { "grad_norm": 0.7995074987411499, "learning_rate": 6.445158984722358e-05, "loss": 0.0983, "step": 1310 }, { "grad_norm": 0.7924811244010925, "learning_rate": 6.39230895741251e-05, "loss": 0.1061, "step": 1320 }, { "grad_norm": 0.8590250015258789, "learning_rate": 6.339289753131649e-05, "loss": 0.0937, "step": 1330 }, { "grad_norm": 0.8390591740608215, "learning_rate": 6.286107814148454e-05, "loss": 0.1028, "step": 1340 }, { "grad_norm": 0.9238244891166687, "learning_rate": 6.232769602505203e-05, "loss": 0.0928, "step": 1350 }, { "grad_norm": 0.9588078260421753, "learning_rate": 6.179281599232591e-05, "loss": 0.1027, "step": 1360 }, { "grad_norm": 0.9879070520401001, "learning_rate": 6.125650303562221e-05, "loss": 0.0845, "step": 1370 }, { "grad_norm": 0.8994210362434387, "learning_rate": 6.071882232136901e-05, "loss": 0.0996, "step": 1380 }, { "grad_norm": 0.7739790081977844, "learning_rate": 6.017983918218812e-05, "loss": 0.091, "step": 1390 }, { "grad_norm": 0.7537702322006226, "learning_rate": 5.963961910895676e-05, "loss": 0.0993, "step": 1400 }, { "grad_norm": 0.7310131788253784, "learning_rate": 5.909822774284971e-05, "loss": 0.0863, "step": 1410 }, { "grad_norm": 0.7676811814308167, "learning_rate": 5.85557308673635e-05, "loss": 0.0992, "step": 1420 }, { "grad_norm": 0.8523777723312378, "learning_rate": 5.8012194400323116e-05, "loss": 0.0932, "step": 1430 }, { "grad_norm": 0.7514760494232178, "learning_rate": 5.746768438587245e-05, "loss": 0.0952, "step": 1440 }, { "grad_norm": 0.818895697593689, "learning_rate": 5.692226698644938e-05, "loss": 0.1012, "step": 1450 }, { "grad_norm": 0.924919605255127, "learning_rate": 5.637600847474656e-05, "loss": 0.0953, "step": 1460 }, { "grad_norm": 0.8371473550796509, "learning_rate": 5.5828975225658666e-05, "loss": 0.093, "step": 1470 }, { "grad_norm": 0.7313709259033203, "learning_rate": 5.52812337082173e-05, "loss": 0.1001, "step": 1480 }, { "grad_norm": 0.8753471374511719, "learning_rate": 5.473285047751451e-05, "loss": 0.0922, "step": 1490 }, { "grad_norm": 0.7444728016853333, "learning_rate": 5.418389216661579e-05, "loss": 0.1086, "step": 1500 }, { "grad_norm": 0.9741567373275757, "learning_rate": 5.363442547846356e-05, "loss": 0.0941, "step": 1510 }, { "grad_norm": 0.7972890138626099, "learning_rate": 5.308451717777228e-05, "loss": 0.0982, "step": 1520 }, { "grad_norm": 0.6457614302635193, "learning_rate": 5.2534234082915976e-05, "loss": 0.0928, "step": 1530 }, { "grad_norm": 0.7331373691558838, "learning_rate": 5.198364305780922e-05, "loss": 0.1023, "step": 1540 }, { "grad_norm": 0.7099182605743408, "learning_rate": 5.143281100378261e-05, "loss": 0.0982, "step": 1550 }, { "grad_norm": 0.7106317281723022, "learning_rate": 5.088180485145378e-05, "loss": 0.092, "step": 1560 }, { "grad_norm": 0.7194976210594177, "learning_rate": 5.033069155259471e-05, "loss": 0.0849, "step": 1570 }, { "grad_norm": 0.59505295753479, "learning_rate": 4.97795380719966e-05, "loss": 0.0845, "step": 1580 }, { "grad_norm": 0.6942402124404907, "learning_rate": 4.9228411379333014e-05, "loss": 0.0911, "step": 1590 }, { "grad_norm": 0.7630301713943481, "learning_rate": 4.867737844102261e-05, "loss": 0.0938, "step": 1600 }, { "grad_norm": 0.8009137511253357, "learning_rate": 4.812650621209209e-05, "loss": 0.0825, "step": 1610 }, { "grad_norm": 0.8127044439315796, "learning_rate": 4.7575861628040635e-05, "loss": 0.08, "step": 1620 }, { "grad_norm": 0.8295362591743469, "learning_rate": 4.702551159670672e-05, "loss": 0.0839, "step": 1630 }, { "grad_norm": 0.7284867763519287, "learning_rate": 4.647552299013828e-05, "loss": 0.0876, "step": 1640 }, { "grad_norm": 0.7745282649993896, "learning_rate": 4.5925962636467126e-05, "loss": 0.0964, "step": 1650 }, { "grad_norm": 0.7279770374298096, "learning_rate": 4.537689731178883e-05, "loss": 0.0746, "step": 1660 }, { "grad_norm": 0.7272283434867859, "learning_rate": 4.482839373204891e-05, "loss": 0.086, "step": 1670 }, { "grad_norm": 0.8412664532661438, "learning_rate": 4.428051854493623e-05, "loss": 0.094, "step": 1680 }, { "grad_norm": 0.7905071973800659, "learning_rate": 4.373333832178478e-05, "loss": 0.0889, "step": 1690 }, { "grad_norm": 0.7173388004302979, "learning_rate": 4.3186919549484784e-05, "loss": 0.0873, "step": 1700 }, { "grad_norm": 0.7447967529296875, "learning_rate": 4.264132862240387e-05, "loss": 0.0852, "step": 1710 }, { "grad_norm": 0.6910264492034912, "learning_rate": 4.209663183431969e-05, "loss": 0.093, "step": 1720 }, { "grad_norm": 0.6057111620903015, "learning_rate": 4.155289537036466e-05, "loss": 0.0898, "step": 1730 }, { "grad_norm": 0.7384588718414307, "learning_rate": 4.1010185298983984e-05, "loss": 0.0844, "step": 1740 }, { "grad_norm": 0.658176064491272, "learning_rate": 4.046856756390767e-05, "loss": 0.087, "step": 1750 }, { "grad_norm": 0.6321943402290344, "learning_rate": 3.9928107976137906e-05, "loss": 0.0895, "step": 1760 }, { "grad_norm": 0.6504121422767639, "learning_rate": 3.9388872205952526e-05, "loss": 0.0845, "step": 1770 }, { "grad_norm": 0.564155101776123, "learning_rate": 3.8850925774925425e-05, "loss": 0.0838, "step": 1780 }, { "grad_norm": 0.5792074799537659, "learning_rate": 3.831433404796521e-05, "loss": 0.0902, "step": 1790 }, { "grad_norm": 0.7973079085350037, "learning_rate": 3.777916222537285e-05, "loss": 0.0866, "step": 1800 }, { "grad_norm": 0.7829543948173523, "learning_rate": 3.7245475334919246e-05, "loss": 0.087, "step": 1810 }, { "grad_norm": 0.7178475260734558, "learning_rate": 3.6713338223943867e-05, "loss": 0.0843, "step": 1820 }, { "grad_norm": 0.5880443453788757, "learning_rate": 3.618281555147522e-05, "loss": 0.0845, "step": 1830 }, { "grad_norm": 0.6559892296791077, "learning_rate": 3.5653971780374295e-05, "loss": 0.0745, "step": 1840 }, { "grad_norm": 0.6737433671951294, "learning_rate": 3.512687116950182e-05, "loss": 0.08, "step": 1850 }, { "grad_norm": 0.7049776911735535, "learning_rate": 3.460157776591018e-05, "loss": 0.0909, "step": 1860 }, { "grad_norm": 0.7571489214897156, "learning_rate": 3.407815539706124e-05, "loss": 0.0912, "step": 1870 }, { "grad_norm": 0.6241227388381958, "learning_rate": 3.355666766307084e-05, "loss": 0.076, "step": 1880 }, { "grad_norm": 0.6526145339012146, "learning_rate": 3.3037177928980735e-05, "loss": 0.074, "step": 1890 }, { "grad_norm": 0.6423455476760864, "learning_rate": 3.251974931705933e-05, "loss": 0.094, "step": 1900 }, { "grad_norm": 0.5167522430419922, "learning_rate": 3.2004444699131727e-05, "loss": 0.0877, "step": 1910 }, { "grad_norm": 0.6353688836097717, "learning_rate": 3.1491326688940345e-05, "loss": 0.0812, "step": 1920 }, { "grad_norm": 0.7268972396850586, "learning_rate": 3.098045763453678e-05, "loss": 0.0867, "step": 1930 }, { "grad_norm": 0.521489679813385, "learning_rate": 3.0471899610706038e-05, "loss": 0.0856, "step": 1940 }, { "grad_norm": 0.61240154504776, "learning_rate": 2.9965714411423972e-05, "loss": 0.0839, "step": 1950 }, { "grad_norm": 0.6277241110801697, "learning_rate": 2.9461963542348737e-05, "loss": 0.0751, "step": 1960 }, { "grad_norm": 0.6723438501358032, "learning_rate": 2.8960708213347366e-05, "loss": 0.0881, "step": 1970 }, { "grad_norm": 0.6539252996444702, "learning_rate": 2.846200933105829e-05, "loss": 0.0938, "step": 1980 }, { "grad_norm": 0.6006316542625427, "learning_rate": 2.7965927491490705e-05, "loss": 0.0797, "step": 1990 }, { "grad_norm": 0.9345439672470093, "learning_rate": 2.747252297266162e-05, "loss": 0.0779, "step": 2000 } ], "logging_steps": 10, "max_steps": 3000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }