{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.9580936729663105, "eval_steps": 500, "global_step": 3600, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.008216926869350863, "grad_norm": 19.974408737204264, "learning_rate": 9.240246406570842e-07, "loss": 8.1444, "step": 10 }, { "epoch": 0.016433853738701727, "grad_norm": 17.651529960979847, "learning_rate": 1.9507186858316226e-06, "loss": 7.9807, "step": 20 }, { "epoch": 0.024650780608052588, "grad_norm": 11.386632405383008, "learning_rate": 2.9774127310061602e-06, "loss": 7.2669, "step": 30 }, { "epoch": 0.03286770747740345, "grad_norm": 7.046114654796961, "learning_rate": 4.004106776180698e-06, "loss": 6.2786, "step": 40 }, { "epoch": 0.041084634346754315, "grad_norm": 6.151945803126112, "learning_rate": 5.030800821355237e-06, "loss": 5.4899, "step": 50 }, { "epoch": 0.049301561216105176, "grad_norm": 8.193676876026643, "learning_rate": 6.057494866529775e-06, "loss": 4.6739, "step": 60 }, { "epoch": 0.05751848808545604, "grad_norm": 6.679224346201021, "learning_rate": 7.084188911704312e-06, "loss": 3.4236, "step": 70 }, { "epoch": 0.0657354149548069, "grad_norm": 2.6366005239013384, "learning_rate": 8.11088295687885e-06, "loss": 2.3599, "step": 80 }, { "epoch": 0.07395234182415776, "grad_norm": 1.375969201844041, "learning_rate": 9.137577002053388e-06, "loss": 1.9964, "step": 90 }, { "epoch": 0.08216926869350863, "grad_norm": 1.7699452853204731, "learning_rate": 1.0164271047227925e-05, "loss": 1.8573, "step": 100 }, { "epoch": 0.09038619556285948, "grad_norm": 8.30917131347657, "learning_rate": 1.1190965092402464e-05, "loss": 1.7505, "step": 110 }, { "epoch": 0.09860312243221035, "grad_norm": 6.75443517080473, "learning_rate": 1.2217659137577003e-05, "loss": 1.6748, "step": 120 }, { "epoch": 0.10682004930156122, "grad_norm": 6.3489758184708975, "learning_rate": 1.324435318275154e-05, "loss": 1.6308, "step": 130 }, { "epoch": 0.11503697617091208, "grad_norm": 7.53791236524009, "learning_rate": 1.427104722792608e-05, "loss": 1.5869, "step": 140 }, { "epoch": 0.12325390304026294, "grad_norm": 8.054957976158356, "learning_rate": 1.5297741273100617e-05, "loss": 1.5436, "step": 150 }, { "epoch": 0.1314708299096138, "grad_norm": 7.410251002629914, "learning_rate": 1.6324435318275156e-05, "loss": 1.5198, "step": 160 }, { "epoch": 0.13968775677896467, "grad_norm": 7.667346984161794, "learning_rate": 1.735112936344969e-05, "loss": 1.4864, "step": 170 }, { "epoch": 0.14790468364831552, "grad_norm": 3.427899808429963, "learning_rate": 1.837782340862423e-05, "loss": 1.4329, "step": 180 }, { "epoch": 0.1561216105176664, "grad_norm": 7.389097005777736, "learning_rate": 1.940451745379877e-05, "loss": 1.3795, "step": 190 }, { "epoch": 0.16433853738701726, "grad_norm": 3.8578784994944915, "learning_rate": 2.0431211498973308e-05, "loss": 1.3356, "step": 200 }, { "epoch": 0.1725554642563681, "grad_norm": 6.847593619711181, "learning_rate": 2.1457905544147843e-05, "loss": 1.298, "step": 210 }, { "epoch": 0.18077239112571897, "grad_norm": 3.9848894805058768, "learning_rate": 2.2484599589322382e-05, "loss": 1.2761, "step": 220 }, { "epoch": 0.18898931799506985, "grad_norm": 3.42148486562202, "learning_rate": 2.351129363449692e-05, "loss": 1.2514, "step": 230 }, { "epoch": 0.1972062448644207, "grad_norm": 4.183793781638511, "learning_rate": 2.453798767967146e-05, "loss": 1.212, "step": 240 }, { "epoch": 0.20542317173377156, "grad_norm": 4.417796453724179, "learning_rate": 2.5564681724845996e-05, "loss": 1.1858, "step": 250 }, { "epoch": 0.21364009860312244, "grad_norm": 3.1317362646162863, "learning_rate": 2.6591375770020538e-05, "loss": 1.1511, "step": 260 }, { "epoch": 0.2218570254724733, "grad_norm": 3.6674629817218523, "learning_rate": 2.7618069815195074e-05, "loss": 1.1253, "step": 270 }, { "epoch": 0.23007395234182415, "grad_norm": 2.4646163322898187, "learning_rate": 2.8644763860369613e-05, "loss": 1.0806, "step": 280 }, { "epoch": 0.23829087921117503, "grad_norm": 3.79598763836086, "learning_rate": 2.9671457905544148e-05, "loss": 1.0437, "step": 290 }, { "epoch": 0.2465078060805259, "grad_norm": 2.495574663825656, "learning_rate": 3.069815195071869e-05, "loss": 0.9992, "step": 300 }, { "epoch": 0.25472473294987674, "grad_norm": 2.708197176131291, "learning_rate": 3.172484599589323e-05, "loss": 0.9809, "step": 310 }, { "epoch": 0.2629416598192276, "grad_norm": 3.2301936096498336, "learning_rate": 3.2751540041067765e-05, "loss": 0.9391, "step": 320 }, { "epoch": 0.27115858668857845, "grad_norm": 3.3281900964809314, "learning_rate": 3.37782340862423e-05, "loss": 0.9122, "step": 330 }, { "epoch": 0.27937551355792933, "grad_norm": 3.7547304430550508, "learning_rate": 3.480492813141684e-05, "loss": 0.8853, "step": 340 }, { "epoch": 0.2875924404272802, "grad_norm": 2.960454105842976, "learning_rate": 3.583162217659138e-05, "loss": 0.8593, "step": 350 }, { "epoch": 0.29580936729663104, "grad_norm": 3.4887397124991772, "learning_rate": 3.6858316221765914e-05, "loss": 0.8366, "step": 360 }, { "epoch": 0.3040262941659819, "grad_norm": 3.759040242435043, "learning_rate": 3.788501026694045e-05, "loss": 0.808, "step": 370 }, { "epoch": 0.3122432210353328, "grad_norm": 3.5316624341034393, "learning_rate": 3.891170431211499e-05, "loss": 0.7784, "step": 380 }, { "epoch": 0.32046014790468363, "grad_norm": 2.536712563451594, "learning_rate": 3.993839835728953e-05, "loss": 0.7638, "step": 390 }, { "epoch": 0.3286770747740345, "grad_norm": 3.1494878741303958, "learning_rate": 4.096509240246407e-05, "loss": 0.7436, "step": 400 }, { "epoch": 0.3368940016433854, "grad_norm": 3.429889342571355, "learning_rate": 4.1991786447638605e-05, "loss": 0.7202, "step": 410 }, { "epoch": 0.3451109285127362, "grad_norm": 3.6597537437672423, "learning_rate": 4.301848049281315e-05, "loss": 0.7059, "step": 420 }, { "epoch": 0.3533278553820871, "grad_norm": 3.3239802253491155, "learning_rate": 4.404517453798768e-05, "loss": 0.6915, "step": 430 }, { "epoch": 0.36154478225143794, "grad_norm": 3.087880747049263, "learning_rate": 4.507186858316222e-05, "loss": 0.668, "step": 440 }, { "epoch": 0.3697617091207888, "grad_norm": 3.083327651120632, "learning_rate": 4.6098562628336754e-05, "loss": 0.6437, "step": 450 }, { "epoch": 0.3779786359901397, "grad_norm": 3.062369129556137, "learning_rate": 4.7125256673511296e-05, "loss": 0.6347, "step": 460 }, { "epoch": 0.3861955628594905, "grad_norm": 3.9335033286871117, "learning_rate": 4.815195071868583e-05, "loss": 0.6111, "step": 470 }, { "epoch": 0.3944124897288414, "grad_norm": 2.845179486951232, "learning_rate": 4.9178644763860374e-05, "loss": 0.5961, "step": 480 }, { "epoch": 0.4026294165981923, "grad_norm": 2.979224322509434, "learning_rate": 5.020533880903491e-05, "loss": 0.5813, "step": 490 }, { "epoch": 0.4108463434675431, "grad_norm": 3.6005231646388736, "learning_rate": 5.1232032854209445e-05, "loss": 0.559, "step": 500 }, { "epoch": 0.419063270336894, "grad_norm": 3.781201309516029, "learning_rate": 5.225872689938398e-05, "loss": 0.5505, "step": 510 }, { "epoch": 0.4272801972062449, "grad_norm": 2.820314146132875, "learning_rate": 5.328542094455853e-05, "loss": 0.5206, "step": 520 }, { "epoch": 0.4354971240755957, "grad_norm": 3.3247760436321783, "learning_rate": 5.4312114989733065e-05, "loss": 0.5191, "step": 530 }, { "epoch": 0.4437140509449466, "grad_norm": 3.185990273496067, "learning_rate": 5.53388090349076e-05, "loss": 0.4937, "step": 540 }, { "epoch": 0.4519309778142975, "grad_norm": 3.0364259829392717, "learning_rate": 5.6365503080082136e-05, "loss": 0.4812, "step": 550 }, { "epoch": 0.4601479046836483, "grad_norm": 2.981635296387684, "learning_rate": 5.739219712525667e-05, "loss": 0.4634, "step": 560 }, { "epoch": 0.4683648315529992, "grad_norm": 3.9637927429253934, "learning_rate": 5.8418891170431214e-05, "loss": 0.4496, "step": 570 }, { "epoch": 0.47658175842235007, "grad_norm": 3.422720069544419, "learning_rate": 5.944558521560575e-05, "loss": 0.4345, "step": 580 }, { "epoch": 0.4847986852917009, "grad_norm": 2.618638333473933, "learning_rate": 6.04722792607803e-05, "loss": 0.4169, "step": 590 }, { "epoch": 0.4930156121610518, "grad_norm": 3.4310336539551147, "learning_rate": 6.149897330595483e-05, "loss": 0.4069, "step": 600 }, { "epoch": 0.5012325390304027, "grad_norm": 2.695323794945264, "learning_rate": 6.252566735112937e-05, "loss": 0.3941, "step": 610 }, { "epoch": 0.5094494658997535, "grad_norm": 3.2005546191921357, "learning_rate": 6.35523613963039e-05, "loss": 0.3844, "step": 620 }, { "epoch": 0.5176663927691043, "grad_norm": 2.698949275451768, "learning_rate": 6.457905544147844e-05, "loss": 0.3685, "step": 630 }, { "epoch": 0.5258833196384552, "grad_norm": 3.0068086367783846, "learning_rate": 6.560574948665298e-05, "loss": 0.3654, "step": 640 }, { "epoch": 0.5341002465078061, "grad_norm": 2.699112912753321, "learning_rate": 6.663244353182751e-05, "loss": 0.3543, "step": 650 }, { "epoch": 0.5423171733771569, "grad_norm": 2.2278954979820833, "learning_rate": 6.765913757700205e-05, "loss": 0.3413, "step": 660 }, { "epoch": 0.5505341002465078, "grad_norm": 2.245661320124028, "learning_rate": 6.86858316221766e-05, "loss": 0.3332, "step": 670 }, { "epoch": 0.5587510271158587, "grad_norm": 2.0476933338187524, "learning_rate": 6.971252566735113e-05, "loss": 0.3261, "step": 680 }, { "epoch": 0.5669679539852095, "grad_norm": 2.389837965434361, "learning_rate": 7.073921971252567e-05, "loss": 0.3164, "step": 690 }, { "epoch": 0.5751848808545604, "grad_norm": 2.1365833016872644, "learning_rate": 7.17659137577002e-05, "loss": 0.3038, "step": 700 }, { "epoch": 0.5834018077239113, "grad_norm": 2.7541557475997234, "learning_rate": 7.279260780287475e-05, "loss": 0.2988, "step": 710 }, { "epoch": 0.5916187345932621, "grad_norm": 2.0589479996604774, "learning_rate": 7.381930184804929e-05, "loss": 0.2872, "step": 720 }, { "epoch": 0.599835661462613, "grad_norm": 2.991548851558623, "learning_rate": 7.484599589322382e-05, "loss": 0.2872, "step": 730 }, { "epoch": 0.6080525883319639, "grad_norm": 1.6552071864163582, "learning_rate": 7.587268993839836e-05, "loss": 0.2762, "step": 740 }, { "epoch": 0.6162695152013147, "grad_norm": 1.789564293031145, "learning_rate": 7.689938398357291e-05, "loss": 0.2661, "step": 750 }, { "epoch": 0.6244864420706656, "grad_norm": 1.9221080215457937, "learning_rate": 7.792607802874744e-05, "loss": 0.2633, "step": 760 }, { "epoch": 0.6327033689400164, "grad_norm": 1.9268969992622877, "learning_rate": 7.895277207392198e-05, "loss": 0.2562, "step": 770 }, { "epoch": 0.6409202958093673, "grad_norm": 1.854381700776246, "learning_rate": 7.997946611909651e-05, "loss": 0.2486, "step": 780 }, { "epoch": 0.6491372226787182, "grad_norm": 2.3180166811886704, "learning_rate": 8.100616016427105e-05, "loss": 0.2409, "step": 790 }, { "epoch": 0.657354149548069, "grad_norm": 1.7752686819968808, "learning_rate": 8.203285420944559e-05, "loss": 0.2323, "step": 800 }, { "epoch": 0.6655710764174199, "grad_norm": 1.4459478240777937, "learning_rate": 8.305954825462012e-05, "loss": 0.2258, "step": 810 }, { "epoch": 0.6737880032867708, "grad_norm": 1.5953025726188175, "learning_rate": 8.408624229979466e-05, "loss": 0.2255, "step": 820 }, { "epoch": 0.6820049301561216, "grad_norm": 1.6045621144575808, "learning_rate": 8.51129363449692e-05, "loss": 0.2169, "step": 830 }, { "epoch": 0.6902218570254725, "grad_norm": 1.7224271707297807, "learning_rate": 8.613963039014374e-05, "loss": 0.2136, "step": 840 }, { "epoch": 0.6984387838948234, "grad_norm": 1.310776192683364, "learning_rate": 8.716632443531828e-05, "loss": 0.2056, "step": 850 }, { "epoch": 0.7066557107641742, "grad_norm": 1.4088847490761656, "learning_rate": 8.819301848049281e-05, "loss": 0.1985, "step": 860 }, { "epoch": 0.714872637633525, "grad_norm": 1.60851910808792, "learning_rate": 8.921971252566735e-05, "loss": 0.1943, "step": 870 }, { "epoch": 0.7230895645028759, "grad_norm": 1.5206100365618338, "learning_rate": 9.02464065708419e-05, "loss": 0.1922, "step": 880 }, { "epoch": 0.7313064913722268, "grad_norm": 1.3408978875102782, "learning_rate": 9.127310061601643e-05, "loss": 0.1851, "step": 890 }, { "epoch": 0.7395234182415776, "grad_norm": 1.3916859583046408, "learning_rate": 9.229979466119097e-05, "loss": 0.1805, "step": 900 }, { "epoch": 0.7477403451109285, "grad_norm": 1.0284073963140599, "learning_rate": 9.332648870636552e-05, "loss": 0.1765, "step": 910 }, { "epoch": 0.7559572719802794, "grad_norm": 1.312106418490207, "learning_rate": 9.435318275154005e-05, "loss": 0.1729, "step": 920 }, { "epoch": 0.7641741988496302, "grad_norm": 1.2873051569354634, "learning_rate": 9.537987679671459e-05, "loss": 0.1697, "step": 930 }, { "epoch": 0.772391125718981, "grad_norm": 1.3089626745248004, "learning_rate": 9.640657084188912e-05, "loss": 0.1644, "step": 940 }, { "epoch": 0.780608052588332, "grad_norm": 1.1995762867540671, "learning_rate": 9.743326488706366e-05, "loss": 0.1605, "step": 950 }, { "epoch": 0.7888249794576828, "grad_norm": 1.2716058027661918, "learning_rate": 9.84599589322382e-05, "loss": 0.1568, "step": 960 }, { "epoch": 0.7970419063270336, "grad_norm": 1.0777715246378874, "learning_rate": 9.948665297741273e-05, "loss": 0.1524, "step": 970 }, { "epoch": 0.8052588331963846, "grad_norm": 1.2259646342544421, "learning_rate": 9.999991965235563e-05, "loss": 0.1501, "step": 980 }, { "epoch": 0.8134757600657354, "grad_norm": 1.2693664502136377, "learning_rate": 9.999927687274992e-05, "loss": 0.1462, "step": 990 }, { "epoch": 0.8216926869350862, "grad_norm": 0.9595367386056208, "learning_rate": 9.999799132180184e-05, "loss": 0.1427, "step": 1000 }, { "epoch": 0.8299096138044372, "grad_norm": 0.9879321961046859, "learning_rate": 9.999606301603793e-05, "loss": 0.1389, "step": 1010 }, { "epoch": 0.838126540673788, "grad_norm": 1.0545188211346712, "learning_rate": 9.999349198024775e-05, "loss": 0.1374, "step": 1020 }, { "epoch": 0.8463434675431388, "grad_norm": 0.9381854736129261, "learning_rate": 9.999027824748352e-05, "loss": 0.134, "step": 1030 }, { "epoch": 0.8545603944124898, "grad_norm": 1.018456405955215, "learning_rate": 9.998642185905977e-05, "loss": 0.1294, "step": 1040 }, { "epoch": 0.8627773212818406, "grad_norm": 0.817274757757614, "learning_rate": 9.998192286455273e-05, "loss": 0.1277, "step": 1050 }, { "epoch": 0.8709942481511914, "grad_norm": 0.877627283927687, "learning_rate": 9.997678132179969e-05, "loss": 0.1259, "step": 1060 }, { "epoch": 0.8792111750205424, "grad_norm": 0.9719564910761921, "learning_rate": 9.997099729689839e-05, "loss": 0.1211, "step": 1070 }, { "epoch": 0.8874281018898932, "grad_norm": 1.0155987976096954, "learning_rate": 9.996457086420598e-05, "loss": 0.1172, "step": 1080 }, { "epoch": 0.895645028759244, "grad_norm": 0.905332195577956, "learning_rate": 9.995750210633822e-05, "loss": 0.1151, "step": 1090 }, { "epoch": 0.903861955628595, "grad_norm": 0.875108702397198, "learning_rate": 9.994979111416829e-05, "loss": 0.1114, "step": 1100 }, { "epoch": 0.9120788824979458, "grad_norm": 0.8315209452920128, "learning_rate": 9.994143798682573e-05, "loss": 0.1108, "step": 1110 }, { "epoch": 0.9202958093672966, "grad_norm": 0.8419546904865063, "learning_rate": 9.993244283169513e-05, "loss": 0.1091, "step": 1120 }, { "epoch": 0.9285127362366475, "grad_norm": 0.8298844114966769, "learning_rate": 9.99228057644147e-05, "loss": 0.1037, "step": 1130 }, { "epoch": 0.9367296631059984, "grad_norm": 0.8012814539297765, "learning_rate": 9.991252690887485e-05, "loss": 0.1023, "step": 1140 }, { "epoch": 0.9449465899753492, "grad_norm": 0.8677793896939191, "learning_rate": 9.990160639721657e-05, "loss": 0.1008, "step": 1150 }, { "epoch": 0.9531635168447001, "grad_norm": 0.7447389790617776, "learning_rate": 9.989004436982974e-05, "loss": 0.0992, "step": 1160 }, { "epoch": 0.961380443714051, "grad_norm": 0.6685887225271923, "learning_rate": 9.987784097535127e-05, "loss": 0.0961, "step": 1170 }, { "epoch": 0.9695973705834018, "grad_norm": 0.7320800211798315, "learning_rate": 9.986499637066331e-05, "loss": 0.0915, "step": 1180 }, { "epoch": 0.9778142974527527, "grad_norm": 0.7347748013970976, "learning_rate": 9.985151072089112e-05, "loss": 0.0922, "step": 1190 }, { "epoch": 0.9860312243221035, "grad_norm": 0.7658898129650706, "learning_rate": 9.983738419940096e-05, "loss": 0.0887, "step": 1200 }, { "epoch": 0.9942481511914544, "grad_norm": 0.6371211246975768, "learning_rate": 9.982261698779795e-05, "loss": 0.0879, "step": 1210 }, { "epoch": 1.0024650780608053, "grad_norm": 0.6728408512142455, "learning_rate": 9.980720927592366e-05, "loss": 0.0862, "step": 1220 }, { "epoch": 1.0106820049301561, "grad_norm": 0.7086330539017783, "learning_rate": 9.979116126185364e-05, "loss": 0.0784, "step": 1230 }, { "epoch": 1.018898931799507, "grad_norm": 0.6735114050298749, "learning_rate": 9.977447315189493e-05, "loss": 0.0754, "step": 1240 }, { "epoch": 1.0271158586688578, "grad_norm": 0.610617286707555, "learning_rate": 9.975714516058344e-05, "loss": 0.0744, "step": 1250 }, { "epoch": 1.0353327855382086, "grad_norm": 0.6567215002986171, "learning_rate": 9.973917751068111e-05, "loss": 0.0724, "step": 1260 }, { "epoch": 1.0435497124075597, "grad_norm": 0.6711754808875117, "learning_rate": 9.972057043317308e-05, "loss": 0.0721, "step": 1270 }, { "epoch": 1.0517666392769105, "grad_norm": 0.6518939760978527, "learning_rate": 9.970132416726474e-05, "loss": 0.0703, "step": 1280 }, { "epoch": 1.0599835661462613, "grad_norm": 0.6612508351869697, "learning_rate": 9.968143896037864e-05, "loss": 0.0702, "step": 1290 }, { "epoch": 1.0682004930156122, "grad_norm": 0.6539738578193367, "learning_rate": 9.966091506815128e-05, "loss": 0.068, "step": 1300 }, { "epoch": 1.076417419884963, "grad_norm": 0.5743874788903558, "learning_rate": 9.963975275442989e-05, "loss": 0.0706, "step": 1310 }, { "epoch": 1.0846343467543138, "grad_norm": 0.6303386908851022, "learning_rate": 9.961795229126896e-05, "loss": 0.0666, "step": 1320 }, { "epoch": 1.0928512736236646, "grad_norm": 0.589250923788632, "learning_rate": 9.959551395892681e-05, "loss": 0.0662, "step": 1330 }, { "epoch": 1.1010682004930157, "grad_norm": 0.576469666142936, "learning_rate": 9.957243804586197e-05, "loss": 0.0648, "step": 1340 }, { "epoch": 1.1092851273623665, "grad_norm": 0.6424675739164545, "learning_rate": 9.954872484872944e-05, "loss": 0.0635, "step": 1350 }, { "epoch": 1.1175020542317173, "grad_norm": 0.6304193275617966, "learning_rate": 9.952437467237687e-05, "loss": 0.0632, "step": 1360 }, { "epoch": 1.1257189811010682, "grad_norm": 0.60085541104229, "learning_rate": 9.949938782984072e-05, "loss": 0.0627, "step": 1370 }, { "epoch": 1.133935907970419, "grad_norm": 0.6387032797411912, "learning_rate": 9.947376464234215e-05, "loss": 0.0619, "step": 1380 }, { "epoch": 1.14215283483977, "grad_norm": 0.590359230196097, "learning_rate": 9.944750543928297e-05, "loss": 0.0601, "step": 1390 }, { "epoch": 1.1503697617091209, "grad_norm": 0.6016993560357249, "learning_rate": 9.942061055824127e-05, "loss": 0.0585, "step": 1400 }, { "epoch": 1.1585866885784717, "grad_norm": 0.6227776527484776, "learning_rate": 9.939308034496728e-05, "loss": 0.0585, "step": 1410 }, { "epoch": 1.1668036154478225, "grad_norm": 0.5840424383749039, "learning_rate": 9.936491515337874e-05, "loss": 0.059, "step": 1420 }, { "epoch": 1.1750205423171733, "grad_norm": 0.602439104016672, "learning_rate": 9.933611534555644e-05, "loss": 0.056, "step": 1430 }, { "epoch": 1.1832374691865242, "grad_norm": 0.5131524987589036, "learning_rate": 9.930668129173956e-05, "loss": 0.0565, "step": 1440 }, { "epoch": 1.191454396055875, "grad_norm": 0.5607165065501415, "learning_rate": 9.927661337032093e-05, "loss": 0.0554, "step": 1450 }, { "epoch": 1.199671322925226, "grad_norm": 0.5938542954233547, "learning_rate": 9.924591196784205e-05, "loss": 0.0547, "step": 1460 }, { "epoch": 1.2078882497945769, "grad_norm": 0.5689585616491131, "learning_rate": 9.92145774789883e-05, "loss": 0.0544, "step": 1470 }, { "epoch": 1.2161051766639277, "grad_norm": 0.5849055658080314, "learning_rate": 9.918261030658371e-05, "loss": 0.0519, "step": 1480 }, { "epoch": 1.2243221035332785, "grad_norm": 0.5248794740274818, "learning_rate": 9.915001086158588e-05, "loss": 0.0514, "step": 1490 }, { "epoch": 1.2325390304026294, "grad_norm": 0.5972426102710096, "learning_rate": 9.911677956308068e-05, "loss": 0.0525, "step": 1500 }, { "epoch": 1.2407559572719804, "grad_norm": 0.5374698363190101, "learning_rate": 9.908291683827678e-05, "loss": 0.0505, "step": 1510 }, { "epoch": 1.2489728841413312, "grad_norm": 0.5061181252518886, "learning_rate": 9.904842312250024e-05, "loss": 0.0503, "step": 1520 }, { "epoch": 1.257189811010682, "grad_norm": 0.5022441363216565, "learning_rate": 9.901329885918898e-05, "loss": 0.0492, "step": 1530 }, { "epoch": 1.2654067378800329, "grad_norm": 0.5451769166028749, "learning_rate": 9.897754449988687e-05, "loss": 0.0508, "step": 1540 }, { "epoch": 1.2736236647493837, "grad_norm": 0.584898627104412, "learning_rate": 9.894116050423812e-05, "loss": 0.0491, "step": 1550 }, { "epoch": 1.2818405916187345, "grad_norm": 0.5371972267407287, "learning_rate": 9.890414733998131e-05, "loss": 0.0473, "step": 1560 }, { "epoch": 1.2900575184880854, "grad_norm": 0.6017034266346077, "learning_rate": 9.886650548294333e-05, "loss": 0.0484, "step": 1570 }, { "epoch": 1.2982744453574364, "grad_norm": 0.5010695241137435, "learning_rate": 9.882823541703332e-05, "loss": 0.0464, "step": 1580 }, { "epoch": 1.3064913722267872, "grad_norm": 0.5676853943666109, "learning_rate": 9.878933763423644e-05, "loss": 0.0483, "step": 1590 }, { "epoch": 1.314708299096138, "grad_norm": 0.5418403716225952, "learning_rate": 9.874981263460749e-05, "loss": 0.0461, "step": 1600 }, { "epoch": 1.322925225965489, "grad_norm": 0.5584739393482486, "learning_rate": 9.87096609262646e-05, "loss": 0.0442, "step": 1610 }, { "epoch": 1.3311421528348397, "grad_norm": 0.4890574045825851, "learning_rate": 9.866888302538255e-05, "loss": 0.0456, "step": 1620 }, { "epoch": 1.3393590797041908, "grad_norm": 0.5311077678345358, "learning_rate": 9.862747945618629e-05, "loss": 0.046, "step": 1630 }, { "epoch": 1.3475760065735414, "grad_norm": 0.545353294627644, "learning_rate": 9.858545075094402e-05, "loss": 0.0443, "step": 1640 }, { "epoch": 1.3557929334428924, "grad_norm": 0.5284222343562959, "learning_rate": 9.854279744996054e-05, "loss": 0.0435, "step": 1650 }, { "epoch": 1.3640098603122432, "grad_norm": 0.46725321717942186, "learning_rate": 9.849952010157014e-05, "loss": 0.0438, "step": 1660 }, { "epoch": 1.372226787181594, "grad_norm": 0.47587942407600886, "learning_rate": 9.845561926212967e-05, "loss": 0.0431, "step": 1670 }, { "epoch": 1.380443714050945, "grad_norm": 0.42647419036174755, "learning_rate": 9.841109549601133e-05, "loss": 0.0423, "step": 1680 }, { "epoch": 1.3886606409202957, "grad_norm": 0.47348266699383407, "learning_rate": 9.83659493755954e-05, "loss": 0.0417, "step": 1690 }, { "epoch": 1.3968775677896468, "grad_norm": 0.5734279662260182, "learning_rate": 9.832018148126291e-05, "loss": 0.0418, "step": 1700 }, { "epoch": 1.4050944946589976, "grad_norm": 0.5079390469387594, "learning_rate": 9.827379240138824e-05, "loss": 0.0407, "step": 1710 }, { "epoch": 1.4133114215283484, "grad_norm": 0.48071880121247074, "learning_rate": 9.822678273233139e-05, "loss": 0.0407, "step": 1720 }, { "epoch": 1.4215283483976993, "grad_norm": 0.4533213147750345, "learning_rate": 9.817915307843046e-05, "loss": 0.0392, "step": 1730 }, { "epoch": 1.42974527526705, "grad_norm": 0.45354884644459964, "learning_rate": 9.813090405199388e-05, "loss": 0.0386, "step": 1740 }, { "epoch": 1.437962202136401, "grad_norm": 0.4060068627853837, "learning_rate": 9.80820362732924e-05, "loss": 0.0382, "step": 1750 }, { "epoch": 1.4461791290057517, "grad_norm": 0.4507780656354325, "learning_rate": 9.803255037055128e-05, "loss": 0.0382, "step": 1760 }, { "epoch": 1.4543960558751028, "grad_norm": 0.5120472100440739, "learning_rate": 9.798244697994214e-05, "loss": 0.0376, "step": 1770 }, { "epoch": 1.4626129827444536, "grad_norm": 0.4669544207241051, "learning_rate": 9.793172674557475e-05, "loss": 0.0368, "step": 1780 }, { "epoch": 1.4708299096138044, "grad_norm": 0.41686952393568316, "learning_rate": 9.78803903194888e-05, "loss": 0.036, "step": 1790 }, { "epoch": 1.4790468364831553, "grad_norm": 0.47436080111491863, "learning_rate": 9.782843836164553e-05, "loss": 0.0354, "step": 1800 }, { "epoch": 1.487263763352506, "grad_norm": 0.47129658653308115, "learning_rate": 9.777587153991916e-05, "loss": 0.0352, "step": 1810 }, { "epoch": 1.4954806902218571, "grad_norm": 0.4459800537917433, "learning_rate": 9.77226905300884e-05, "loss": 0.0361, "step": 1820 }, { "epoch": 1.5036976170912077, "grad_norm": 0.43894409691624636, "learning_rate": 9.766889601582774e-05, "loss": 0.0341, "step": 1830 }, { "epoch": 1.5119145439605588, "grad_norm": 0.4681621495969991, "learning_rate": 9.761448868869863e-05, "loss": 0.0365, "step": 1840 }, { "epoch": 1.5201314708299096, "grad_norm": 0.44125169083676535, "learning_rate": 9.755946924814055e-05, "loss": 0.0349, "step": 1850 }, { "epoch": 1.5283483976992605, "grad_norm": 0.42454765007479894, "learning_rate": 9.750383840146218e-05, "loss": 0.034, "step": 1860 }, { "epoch": 1.5365653245686115, "grad_norm": 0.4312790936546121, "learning_rate": 9.744759686383211e-05, "loss": 0.0342, "step": 1870 }, { "epoch": 1.544782251437962, "grad_norm": 0.46580009646158177, "learning_rate": 9.739074535826978e-05, "loss": 0.0335, "step": 1880 }, { "epoch": 1.5529991783073132, "grad_norm": 0.4275717475424306, "learning_rate": 9.733328461563612e-05, "loss": 0.0344, "step": 1890 }, { "epoch": 1.561216105176664, "grad_norm": 0.4756865714632423, "learning_rate": 9.72752153746242e-05, "loss": 0.034, "step": 1900 }, { "epoch": 1.5694330320460148, "grad_norm": 0.41615409042671186, "learning_rate": 9.721653838174972e-05, "loss": 0.0322, "step": 1910 }, { "epoch": 1.5776499589153656, "grad_norm": 0.4261218829495612, "learning_rate": 9.715725439134132e-05, "loss": 0.0323, "step": 1920 }, { "epoch": 1.5858668857847165, "grad_norm": 0.4103285392171169, "learning_rate": 9.709736416553106e-05, "loss": 0.0321, "step": 1930 }, { "epoch": 1.5940838126540675, "grad_norm": 0.42730940931321443, "learning_rate": 9.703686847424449e-05, "loss": 0.0323, "step": 1940 }, { "epoch": 1.6023007395234181, "grad_norm": 0.37284494211308117, "learning_rate": 9.697576809519079e-05, "loss": 0.0306, "step": 1950 }, { "epoch": 1.6105176663927692, "grad_norm": 0.4347235376642582, "learning_rate": 9.691406381385278e-05, "loss": 0.0309, "step": 1960 }, { "epoch": 1.61873459326212, "grad_norm": 0.41859821688218857, "learning_rate": 9.685175642347679e-05, "loss": 0.0306, "step": 1970 }, { "epoch": 1.6269515201314708, "grad_norm": 0.458193281780139, "learning_rate": 9.678884672506252e-05, "loss": 0.0304, "step": 1980 }, { "epoch": 1.6351684470008216, "grad_norm": 0.4467297131068921, "learning_rate": 9.672533552735268e-05, "loss": 0.0307, "step": 1990 }, { "epoch": 1.6433853738701725, "grad_norm": 0.4246649874690625, "learning_rate": 9.666122364682263e-05, "loss": 0.029, "step": 2000 }, { "epoch": 1.6516023007395235, "grad_norm": 0.44552544340004335, "learning_rate": 9.659651190766992e-05, "loss": 0.0295, "step": 2010 }, { "epoch": 1.6598192276088741, "grad_norm": 0.4164644781602266, "learning_rate": 9.653120114180356e-05, "loss": 0.0291, "step": 2020 }, { "epoch": 1.6680361544782252, "grad_norm": 0.4579243177302215, "learning_rate": 9.646529218883349e-05, "loss": 0.0302, "step": 2030 }, { "epoch": 1.676253081347576, "grad_norm": 0.48543566357037554, "learning_rate": 9.639878589605969e-05, "loss": 0.0291, "step": 2040 }, { "epoch": 1.6844700082169268, "grad_norm": 0.46469944987261236, "learning_rate": 9.63316831184613e-05, "loss": 0.0292, "step": 2050 }, { "epoch": 1.6926869350862779, "grad_norm": 0.45357303638368995, "learning_rate": 9.626398471868564e-05, "loss": 0.0287, "step": 2060 }, { "epoch": 1.7009038619556285, "grad_norm": 0.42210490946569273, "learning_rate": 9.619569156703714e-05, "loss": 0.0282, "step": 2070 }, { "epoch": 1.7091207888249795, "grad_norm": 0.41082451259817443, "learning_rate": 9.612680454146609e-05, "loss": 0.0276, "step": 2080 }, { "epoch": 1.7173377156943304, "grad_norm": 0.41363141928585667, "learning_rate": 9.605732452755742e-05, "loss": 0.0273, "step": 2090 }, { "epoch": 1.7255546425636812, "grad_norm": 0.4206911941733154, "learning_rate": 9.598725241851932e-05, "loss": 0.0287, "step": 2100 }, { "epoch": 1.733771569433032, "grad_norm": 0.3951369498197376, "learning_rate": 9.591658911517162e-05, "loss": 0.0275, "step": 2110 }, { "epoch": 1.7419884963023828, "grad_norm": 0.4071839799145019, "learning_rate": 9.584533552593447e-05, "loss": 0.0263, "step": 2120 }, { "epoch": 1.7502054231717339, "grad_norm": 0.35587647953674884, "learning_rate": 9.577349256681636e-05, "loss": 0.0274, "step": 2130 }, { "epoch": 1.7584223500410845, "grad_norm": 0.43711475000779765, "learning_rate": 9.570106116140256e-05, "loss": 0.0268, "step": 2140 }, { "epoch": 1.7666392769104355, "grad_norm": 0.3991641744179894, "learning_rate": 9.56280422408432e-05, "loss": 0.0268, "step": 2150 }, { "epoch": 1.7748562037797864, "grad_norm": 0.4076460125133768, "learning_rate": 9.555443674384122e-05, "loss": 0.0255, "step": 2160 }, { "epoch": 1.7830731306491372, "grad_norm": 0.4348100956752934, "learning_rate": 9.548024561664039e-05, "loss": 0.0277, "step": 2170 }, { "epoch": 1.7912900575184882, "grad_norm": 0.4034844759662755, "learning_rate": 9.540546981301313e-05, "loss": 0.0254, "step": 2180 }, { "epoch": 1.7995069843878388, "grad_norm": 0.4132706716223311, "learning_rate": 9.53301102942482e-05, "loss": 0.0253, "step": 2190 }, { "epoch": 1.80772391125719, "grad_norm": 0.37701202488605456, "learning_rate": 9.525416802913841e-05, "loss": 0.0246, "step": 2200 }, { "epoch": 1.8159408381265407, "grad_norm": 0.39051860707046865, "learning_rate": 9.51776439939681e-05, "loss": 0.0257, "step": 2210 }, { "epoch": 1.8241577649958916, "grad_norm": 0.4106016394020398, "learning_rate": 9.510053917250061e-05, "loss": 0.0257, "step": 2220 }, { "epoch": 1.8323746918652424, "grad_norm": 0.3866184281474688, "learning_rate": 9.502285455596572e-05, "loss": 0.0245, "step": 2230 }, { "epoch": 1.8405916187345932, "grad_norm": 0.4021017932429677, "learning_rate": 9.494459114304672e-05, "loss": 0.0244, "step": 2240 }, { "epoch": 1.8488085456039443, "grad_norm": 0.3920008094917924, "learning_rate": 9.48657499398678e-05, "loss": 0.0248, "step": 2250 }, { "epoch": 1.8570254724732949, "grad_norm": 0.4004129162133985, "learning_rate": 9.47863319599809e-05, "loss": 0.0249, "step": 2260 }, { "epoch": 1.865242399342646, "grad_norm": 0.3782452193154733, "learning_rate": 9.470633822435282e-05, "loss": 0.0246, "step": 2270 }, { "epoch": 1.8734593262119967, "grad_norm": 0.36635346240750793, "learning_rate": 9.462576976135207e-05, "loss": 0.0228, "step": 2280 }, { "epoch": 1.8816762530813476, "grad_norm": 0.38839169002774604, "learning_rate": 9.454462760673559e-05, "loss": 0.0242, "step": 2290 }, { "epoch": 1.8898931799506984, "grad_norm": 0.37421172265077624, "learning_rate": 9.446291280363551e-05, "loss": 0.0239, "step": 2300 }, { "epoch": 1.8981101068200492, "grad_norm": 0.3832912238881792, "learning_rate": 9.43806264025457e-05, "loss": 0.023, "step": 2310 }, { "epoch": 1.9063270336894003, "grad_norm": 0.3814194428999421, "learning_rate": 9.429776946130826e-05, "loss": 0.0241, "step": 2320 }, { "epoch": 1.9145439605587509, "grad_norm": 0.39219173513731687, "learning_rate": 9.421434304509992e-05, "loss": 0.0228, "step": 2330 }, { "epoch": 1.922760887428102, "grad_norm": 0.3543509894192511, "learning_rate": 9.413034822641844e-05, "loss": 0.0232, "step": 2340 }, { "epoch": 1.9309778142974527, "grad_norm": 0.35733995786931505, "learning_rate": 9.404578608506865e-05, "loss": 0.0228, "step": 2350 }, { "epoch": 1.9391947411668036, "grad_norm": 0.386740275684357, "learning_rate": 9.396065770814867e-05, "loss": 0.0232, "step": 2360 }, { "epoch": 1.9474116680361546, "grad_norm": 0.44141491887351664, "learning_rate": 9.387496419003599e-05, "loss": 0.0233, "step": 2370 }, { "epoch": 1.9556285949055052, "grad_norm": 0.3802346697580586, "learning_rate": 9.378870663237328e-05, "loss": 0.022, "step": 2380 }, { "epoch": 1.9638455217748563, "grad_norm": 0.3692117355721275, "learning_rate": 9.37018861440543e-05, "loss": 0.0223, "step": 2390 }, { "epoch": 1.972062448644207, "grad_norm": 0.39846659821583263, "learning_rate": 9.361450384120963e-05, "loss": 0.0215, "step": 2400 }, { "epoch": 1.980279375513558, "grad_norm": 0.3262153453252359, "learning_rate": 9.352656084719231e-05, "loss": 0.0212, "step": 2410 }, { "epoch": 1.9884963023829088, "grad_norm": 0.34632528206609053, "learning_rate": 9.343805829256343e-05, "loss": 0.0214, "step": 2420 }, { "epoch": 1.9967132292522596, "grad_norm": 0.3654934223559427, "learning_rate": 9.334899731507756e-05, "loss": 0.021, "step": 2430 }, { "epoch": 2.0049301561216106, "grad_norm": 0.26848148846319253, "learning_rate": 9.325937905966813e-05, "loss": 0.0164, "step": 2440 }, { "epoch": 2.0131470829909612, "grad_norm": 0.29112160593983694, "learning_rate": 9.316920467843273e-05, "loss": 0.0126, "step": 2450 }, { "epoch": 2.0213640098603123, "grad_norm": 0.25419379869277514, "learning_rate": 9.307847533061831e-05, "loss": 0.0126, "step": 2460 }, { "epoch": 2.029580936729663, "grad_norm": 0.284347026269926, "learning_rate": 9.29871921826062e-05, "loss": 0.0115, "step": 2470 }, { "epoch": 2.037797863599014, "grad_norm": 0.2804187708379741, "learning_rate": 9.289535640789721e-05, "loss": 0.0109, "step": 2480 }, { "epoch": 2.046014790468365, "grad_norm": 0.21561701614022483, "learning_rate": 9.280296918709653e-05, "loss": 0.0112, "step": 2490 }, { "epoch": 2.0542317173377156, "grad_norm": 0.2770066140071947, "learning_rate": 9.271003170789846e-05, "loss": 0.0106, "step": 2500 }, { "epoch": 2.0624486442070666, "grad_norm": 0.3110912363010226, "learning_rate": 9.261654516507126e-05, "loss": 0.0108, "step": 2510 }, { "epoch": 2.0706655710764172, "grad_norm": 0.27296541462941787, "learning_rate": 9.252251076044172e-05, "loss": 0.0109, "step": 2520 }, { "epoch": 2.0788824979457683, "grad_norm": 0.24002967576989534, "learning_rate": 9.242792970287972e-05, "loss": 0.0108, "step": 2530 }, { "epoch": 2.0870994248151193, "grad_norm": 0.29764076691462776, "learning_rate": 9.233280320828274e-05, "loss": 0.0117, "step": 2540 }, { "epoch": 2.09531635168447, "grad_norm": 0.2913901487844858, "learning_rate": 9.223713249956016e-05, "loss": 0.0108, "step": 2550 }, { "epoch": 2.103533278553821, "grad_norm": 0.24293966139484074, "learning_rate": 9.214091880661752e-05, "loss": 0.0104, "step": 2560 }, { "epoch": 2.1117502054231716, "grad_norm": 0.24074539605557427, "learning_rate": 9.204416336634084e-05, "loss": 0.0099, "step": 2570 }, { "epoch": 2.1199671322925226, "grad_norm": 0.2999981940123419, "learning_rate": 9.194686742258059e-05, "loss": 0.0104, "step": 2580 }, { "epoch": 2.1281840591618733, "grad_norm": 0.25277770668402233, "learning_rate": 9.184903222613574e-05, "loss": 0.0102, "step": 2590 }, { "epoch": 2.1364009860312243, "grad_norm": 0.26178176275955506, "learning_rate": 9.175065903473768e-05, "loss": 0.0103, "step": 2600 }, { "epoch": 2.1446179129005754, "grad_norm": 0.26730537093378065, "learning_rate": 9.165174911303409e-05, "loss": 0.0103, "step": 2610 }, { "epoch": 2.152834839769926, "grad_norm": 0.2931413736706241, "learning_rate": 9.15523037325726e-05, "loss": 0.0108, "step": 2620 }, { "epoch": 2.161051766639277, "grad_norm": 0.2870887376128566, "learning_rate": 9.145232417178453e-05, "loss": 0.0102, "step": 2630 }, { "epoch": 2.1692686935086276, "grad_norm": 0.24704281922605292, "learning_rate": 9.135181171596838e-05, "loss": 0.0103, "step": 2640 }, { "epoch": 2.1774856203779787, "grad_norm": 0.26790802194061564, "learning_rate": 9.125076765727338e-05, "loss": 0.011, "step": 2650 }, { "epoch": 2.1857025472473293, "grad_norm": 0.3449249477967815, "learning_rate": 9.114919329468282e-05, "loss": 0.0116, "step": 2660 }, { "epoch": 2.1939194741166803, "grad_norm": 0.30244563420360604, "learning_rate": 9.104708993399738e-05, "loss": 0.011, "step": 2670 }, { "epoch": 2.2021364009860314, "grad_norm": 0.30736920260370454, "learning_rate": 9.094445888781829e-05, "loss": 0.0107, "step": 2680 }, { "epoch": 2.210353327855382, "grad_norm": 0.30219679969169827, "learning_rate": 9.084130147553057e-05, "loss": 0.0116, "step": 2690 }, { "epoch": 2.218570254724733, "grad_norm": 0.3286996402717758, "learning_rate": 9.073761902328597e-05, "loss": 0.0102, "step": 2700 }, { "epoch": 2.2267871815940836, "grad_norm": 0.30798545297089197, "learning_rate": 9.063341286398592e-05, "loss": 0.0106, "step": 2710 }, { "epoch": 2.2350041084634347, "grad_norm": 0.33718978166952857, "learning_rate": 9.052868433726448e-05, "loss": 0.0105, "step": 2720 }, { "epoch": 2.2432210353327857, "grad_norm": 0.2954643606610759, "learning_rate": 9.042343478947103e-05, "loss": 0.011, "step": 2730 }, { "epoch": 2.2514379622021363, "grad_norm": 0.30272236743742853, "learning_rate": 9.0317665573653e-05, "loss": 0.011, "step": 2740 }, { "epoch": 2.2596548890714874, "grad_norm": 0.3302443829248607, "learning_rate": 9.021137804953846e-05, "loss": 0.0112, "step": 2750 }, { "epoch": 2.267871815940838, "grad_norm": 0.29949297685046744, "learning_rate": 9.010457358351867e-05, "loss": 0.0109, "step": 2760 }, { "epoch": 2.276088742810189, "grad_norm": 0.2724599152940893, "learning_rate": 8.99972535486305e-05, "loss": 0.0114, "step": 2770 }, { "epoch": 2.28430566967954, "grad_norm": 0.31846220224170235, "learning_rate": 8.988941932453872e-05, "loss": 0.0118, "step": 2780 }, { "epoch": 2.2925225965488907, "grad_norm": 0.29238177724291164, "learning_rate": 8.978107229751842e-05, "loss": 0.0108, "step": 2790 }, { "epoch": 2.3007395234182417, "grad_norm": 0.27814751669130666, "learning_rate": 8.967221386043699e-05, "loss": 0.0112, "step": 2800 }, { "epoch": 2.3089564502875923, "grad_norm": 0.29768803452504616, "learning_rate": 8.956284541273632e-05, "loss": 0.0108, "step": 2810 }, { "epoch": 2.3171733771569434, "grad_norm": 0.30479438136210424, "learning_rate": 8.945296836041487e-05, "loss": 0.011, "step": 2820 }, { "epoch": 2.325390304026294, "grad_norm": 0.30400201170563734, "learning_rate": 8.934258411600945e-05, "loss": 0.0111, "step": 2830 }, { "epoch": 2.333607230895645, "grad_norm": 0.279262780416122, "learning_rate": 8.923169409857718e-05, "loss": 0.011, "step": 2840 }, { "epoch": 2.341824157764996, "grad_norm": 0.3105378625590925, "learning_rate": 8.912029973367717e-05, "loss": 0.0107, "step": 2850 }, { "epoch": 2.3500410846343467, "grad_norm": 0.29400831967333496, "learning_rate": 8.900840245335225e-05, "loss": 0.0113, "step": 2860 }, { "epoch": 2.3582580115036977, "grad_norm": 0.2893100065249251, "learning_rate": 8.889600369611051e-05, "loss": 0.0114, "step": 2870 }, { "epoch": 2.3664749383730483, "grad_norm": 0.2815383019080243, "learning_rate": 8.878310490690686e-05, "loss": 0.011, "step": 2880 }, { "epoch": 2.3746918652423994, "grad_norm": 0.3427726387197821, "learning_rate": 8.866970753712442e-05, "loss": 0.0113, "step": 2890 }, { "epoch": 2.38290879211175, "grad_norm": 0.278157815073135, "learning_rate": 8.855581304455584e-05, "loss": 0.0108, "step": 2900 }, { "epoch": 2.391125718981101, "grad_norm": 0.29741130920284337, "learning_rate": 8.844142289338466e-05, "loss": 0.0104, "step": 2910 }, { "epoch": 2.399342645850452, "grad_norm": 0.2729299367333149, "learning_rate": 8.832653855416634e-05, "loss": 0.0104, "step": 2920 }, { "epoch": 2.4075595727198027, "grad_norm": 0.2865288059261725, "learning_rate": 8.821116150380946e-05, "loss": 0.011, "step": 2930 }, { "epoch": 2.4157764995891537, "grad_norm": 0.2791177339025258, "learning_rate": 8.80952932255567e-05, "loss": 0.0107, "step": 2940 }, { "epoch": 2.4239934264585044, "grad_norm": 0.2980724885727162, "learning_rate": 8.797893520896579e-05, "loss": 0.011, "step": 2950 }, { "epoch": 2.4322103533278554, "grad_norm": 0.31076617373261356, "learning_rate": 8.786208894989032e-05, "loss": 0.0108, "step": 2960 }, { "epoch": 2.4404272801972064, "grad_norm": 0.30152315696501025, "learning_rate": 8.774475595046055e-05, "loss": 0.0111, "step": 2970 }, { "epoch": 2.448644207066557, "grad_norm": 0.2609839163777236, "learning_rate": 8.762693771906409e-05, "loss": 0.0109, "step": 2980 }, { "epoch": 2.456861133935908, "grad_norm": 0.30973305475716906, "learning_rate": 8.750863577032652e-05, "loss": 0.011, "step": 2990 }, { "epoch": 2.4650780608052587, "grad_norm": 0.3393765462966183, "learning_rate": 8.738985162509186e-05, "loss": 0.0112, "step": 3000 }, { "epoch": 2.4732949876746098, "grad_norm": 0.3048152571960306, "learning_rate": 8.727058681040308e-05, "loss": 0.011, "step": 3010 }, { "epoch": 2.481511914543961, "grad_norm": 0.25716354456976137, "learning_rate": 8.715084285948248e-05, "loss": 0.0107, "step": 3020 }, { "epoch": 2.4897288414133114, "grad_norm": 0.3076118802762377, "learning_rate": 8.703062131171192e-05, "loss": 0.0109, "step": 3030 }, { "epoch": 2.4979457682826625, "grad_norm": 0.2897555835996962, "learning_rate": 8.690992371261302e-05, "loss": 0.0105, "step": 3040 }, { "epoch": 2.506162695152013, "grad_norm": 0.2990726423763531, "learning_rate": 8.678875161382745e-05, "loss": 0.0114, "step": 3050 }, { "epoch": 2.514379622021364, "grad_norm": 0.3308837352116956, "learning_rate": 8.666710657309672e-05, "loss": 0.011, "step": 3060 }, { "epoch": 2.5225965488907147, "grad_norm": 0.2783330323448899, "learning_rate": 8.654499015424243e-05, "loss": 0.0106, "step": 3070 }, { "epoch": 2.5308134757600658, "grad_norm": 0.2834321980541149, "learning_rate": 8.642240392714593e-05, "loss": 0.0108, "step": 3080 }, { "epoch": 2.5390304026294164, "grad_norm": 0.21923295269567925, "learning_rate": 8.629934946772831e-05, "loss": 0.0105, "step": 3090 }, { "epoch": 2.5472473294987674, "grad_norm": 0.3113319375295575, "learning_rate": 8.617582835793003e-05, "loss": 0.011, "step": 3100 }, { "epoch": 2.5554642563681185, "grad_norm": 0.22522445036106728, "learning_rate": 8.605184218569067e-05, "loss": 0.0108, "step": 3110 }, { "epoch": 2.563681183237469, "grad_norm": 0.29129457748014026, "learning_rate": 8.592739254492844e-05, "loss": 0.0114, "step": 3120 }, { "epoch": 2.57189811010682, "grad_norm": 0.298278233144733, "learning_rate": 8.580248103551973e-05, "loss": 0.0104, "step": 3130 }, { "epoch": 2.5801150369761707, "grad_norm": 0.28438151731996136, "learning_rate": 8.567710926327855e-05, "loss": 0.0109, "step": 3140 }, { "epoch": 2.5883319638455218, "grad_norm": 0.29615207997528636, "learning_rate": 8.555127883993582e-05, "loss": 0.0111, "step": 3150 }, { "epoch": 2.596548890714873, "grad_norm": 0.2668249857546702, "learning_rate": 8.542499138311877e-05, "loss": 0.0104, "step": 3160 }, { "epoch": 2.6047658175842234, "grad_norm": 0.29885476401362265, "learning_rate": 8.529824851633004e-05, "loss": 0.0109, "step": 3170 }, { "epoch": 2.6129827444535745, "grad_norm": 0.2964635585154554, "learning_rate": 8.517105186892683e-05, "loss": 0.0103, "step": 3180 }, { "epoch": 2.621199671322925, "grad_norm": 0.26419372229458143, "learning_rate": 8.50434030761e-05, "loss": 0.0102, "step": 3190 }, { "epoch": 2.629416598192276, "grad_norm": 0.26957526996934655, "learning_rate": 8.4915303778853e-05, "loss": 0.0105, "step": 3200 }, { "epoch": 2.637633525061627, "grad_norm": 0.30995887012931245, "learning_rate": 8.478675562398077e-05, "loss": 0.0105, "step": 3210 }, { "epoch": 2.645850451930978, "grad_norm": 0.2885157857565514, "learning_rate": 8.465776026404866e-05, "loss": 0.0112, "step": 3220 }, { "epoch": 2.654067378800329, "grad_norm": 0.33415184613805593, "learning_rate": 8.452831935737102e-05, "loss": 0.0106, "step": 3230 }, { "epoch": 2.6622843056696794, "grad_norm": 0.2978809771443008, "learning_rate": 8.439843456799005e-05, "loss": 0.0101, "step": 3240 }, { "epoch": 2.6705012325390305, "grad_norm": 0.25393760859554587, "learning_rate": 8.426810756565429e-05, "loss": 0.0107, "step": 3250 }, { "epoch": 2.6787181594083815, "grad_norm": 0.2839448149328719, "learning_rate": 8.413734002579723e-05, "loss": 0.0111, "step": 3260 }, { "epoch": 2.686935086277732, "grad_norm": 0.2721771832897388, "learning_rate": 8.40061336295157e-05, "loss": 0.0111, "step": 3270 }, { "epoch": 2.6951520131470827, "grad_norm": 0.2872702777954717, "learning_rate": 8.387449006354837e-05, "loss": 0.0103, "step": 3280 }, { "epoch": 2.703368940016434, "grad_norm": 0.29596031948807433, "learning_rate": 8.374241102025391e-05, "loss": 0.0104, "step": 3290 }, { "epoch": 2.711585866885785, "grad_norm": 0.27424169300037154, "learning_rate": 8.360989819758935e-05, "loss": 0.0103, "step": 3300 }, { "epoch": 2.7198027937551355, "grad_norm": 0.2699351448831838, "learning_rate": 8.347695329908827e-05, "loss": 0.0103, "step": 3310 }, { "epoch": 2.7280197206244865, "grad_norm": 0.26678359974892135, "learning_rate": 8.334357803383875e-05, "loss": 0.0102, "step": 3320 }, { "epoch": 2.736236647493837, "grad_norm": 0.2948773860549256, "learning_rate": 8.32097741164616e-05, "loss": 0.0096, "step": 3330 }, { "epoch": 2.744453574363188, "grad_norm": 0.3090800658579484, "learning_rate": 8.307554326708815e-05, "loss": 0.0102, "step": 3340 }, { "epoch": 2.752670501232539, "grad_norm": 0.2735640172287781, "learning_rate": 8.294088721133821e-05, "loss": 0.0097, "step": 3350 }, { "epoch": 2.76088742810189, "grad_norm": 0.26265127181426384, "learning_rate": 8.28058076802979e-05, "loss": 0.0103, "step": 3360 }, { "epoch": 2.769104354971241, "grad_norm": 0.3267515866056434, "learning_rate": 8.267030641049738e-05, "loss": 0.0105, "step": 3370 }, { "epoch": 2.7773212818405915, "grad_norm": 0.2807694986872119, "learning_rate": 8.25343851438885e-05, "loss": 0.0101, "step": 3380 }, { "epoch": 2.7855382087099425, "grad_norm": 0.2918400047001943, "learning_rate": 8.239804562782241e-05, "loss": 0.0103, "step": 3390 }, { "epoch": 2.7937551355792936, "grad_norm": 0.3368518827581203, "learning_rate": 8.226128961502715e-05, "loss": 0.0099, "step": 3400 }, { "epoch": 2.801972062448644, "grad_norm": 0.30338648344652236, "learning_rate": 8.212411886358505e-05, "loss": 0.0105, "step": 3410 }, { "epoch": 2.810188989317995, "grad_norm": 0.26222245141291084, "learning_rate": 8.198653513691019e-05, "loss": 0.0096, "step": 3420 }, { "epoch": 2.818405916187346, "grad_norm": 0.31220931761376275, "learning_rate": 8.184854020372567e-05, "loss": 0.0107, "step": 3430 }, { "epoch": 2.826622843056697, "grad_norm": 0.26292349810527665, "learning_rate": 8.171013583804095e-05, "loss": 0.0099, "step": 3440 }, { "epoch": 2.834839769926048, "grad_norm": 0.28178421575879453, "learning_rate": 8.157132381912894e-05, "loss": 0.01, "step": 3450 }, { "epoch": 2.8430566967953985, "grad_norm": 0.2794905489623295, "learning_rate": 8.143210593150317e-05, "loss": 0.0109, "step": 3460 }, { "epoch": 2.851273623664749, "grad_norm": 0.30819250761179967, "learning_rate": 8.129248396489493e-05, "loss": 0.0105, "step": 3470 }, { "epoch": 2.8594905505341, "grad_norm": 0.30996445249233995, "learning_rate": 8.115245971423016e-05, "loss": 0.0103, "step": 3480 }, { "epoch": 2.867707477403451, "grad_norm": 0.2998585643222057, "learning_rate": 8.101203497960639e-05, "loss": 0.0103, "step": 3490 }, { "epoch": 2.875924404272802, "grad_norm": 0.2886522658424423, "learning_rate": 8.087121156626965e-05, "loss": 0.0102, "step": 3500 }, { "epoch": 2.884141331142153, "grad_norm": 0.2896678741036796, "learning_rate": 8.072999128459119e-05, "loss": 0.0101, "step": 3510 }, { "epoch": 2.8923582580115035, "grad_norm": 0.2668371079506528, "learning_rate": 8.05883759500443e-05, "loss": 0.0101, "step": 3520 }, { "epoch": 2.9005751848808545, "grad_norm": 0.262945737927121, "learning_rate": 8.044636738318082e-05, "loss": 0.01, "step": 3530 }, { "epoch": 2.9087921117502056, "grad_norm": 0.3051556790714007, "learning_rate": 8.030396740960795e-05, "loss": 0.0106, "step": 3540 }, { "epoch": 2.917009038619556, "grad_norm": 0.2681398854835005, "learning_rate": 8.01611778599646e-05, "loss": 0.0095, "step": 3550 }, { "epoch": 2.9252259654889072, "grad_norm": 0.2812885203155075, "learning_rate": 8.001800056989792e-05, "loss": 0.0096, "step": 3560 }, { "epoch": 2.933442892358258, "grad_norm": 0.28431854280176366, "learning_rate": 7.987443738003973e-05, "loss": 0.009, "step": 3570 }, { "epoch": 2.941659819227609, "grad_norm": 0.25500549263286926, "learning_rate": 7.973049013598277e-05, "loss": 0.0092, "step": 3580 }, { "epoch": 2.94987674609696, "grad_norm": 0.2804777447797684, "learning_rate": 7.958616068825709e-05, "loss": 0.0099, "step": 3590 }, { "epoch": 2.9580936729663105, "grad_norm": 0.2837446971120676, "learning_rate": 7.944145089230619e-05, "loss": 0.0098, "step": 3600 } ], "logging_steps": 10, "max_steps": 9736, "num_input_tokens_seen": 0, "num_train_epochs": 8, "save_steps": 1800, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 212655470542848.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }