{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 8.0, "eval_steps": 500, "global_step": 9736, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.008216926869350863, "grad_norm": 19.83480665335734, "learning_rate": 9.240246406570842e-07, "loss": 8.2641, "step": 10 }, { "epoch": 0.016433853738701727, "grad_norm": 18.704025706899706, "learning_rate": 1.9507186858316226e-06, "loss": 8.1054, "step": 20 }, { "epoch": 0.024650780608052588, "grad_norm": 9.254313873894626, "learning_rate": 2.9774127310061602e-06, "loss": 7.4171, "step": 30 }, { "epoch": 0.03286770747740345, "grad_norm": 5.336886587772746, "learning_rate": 4.004106776180698e-06, "loss": 6.652, "step": 40 }, { "epoch": 0.041084634346754315, "grad_norm": 3.356383707656696, "learning_rate": 5.030800821355237e-06, "loss": 6.1747, "step": 50 }, { "epoch": 0.049301561216105176, "grad_norm": 3.3258958472365467, "learning_rate": 6.057494866529775e-06, "loss": 5.8018, "step": 60 }, { "epoch": 0.05751848808545604, "grad_norm": 4.0537058731983, "learning_rate": 7.084188911704312e-06, "loss": 5.3316, "step": 70 }, { "epoch": 0.0657354149548069, "grad_norm": 4.086247611679589, "learning_rate": 8.11088295687885e-06, "loss": 4.6534, "step": 80 }, { "epoch": 0.07395234182415776, "grad_norm": 6.521478113854127, "learning_rate": 9.137577002053388e-06, "loss": 3.924, "step": 90 }, { "epoch": 0.08216926869350863, "grad_norm": 5.719373874300257, "learning_rate": 1.0164271047227925e-05, "loss": 3.3464, "step": 100 }, { "epoch": 0.09038619556285948, "grad_norm": 3.578443884259996, "learning_rate": 1.1190965092402464e-05, "loss": 2.9684, "step": 110 }, { "epoch": 0.09860312243221035, "grad_norm": 5.144310812867023, "learning_rate": 1.2217659137577003e-05, "loss": 2.734, "step": 120 }, { "epoch": 0.10682004930156122, "grad_norm": 5.024178909720029, "learning_rate": 1.324435318275154e-05, "loss": 2.5647, "step": 130 }, { "epoch": 0.11503697617091208, "grad_norm": 7.639677795548909, "learning_rate": 1.427104722792608e-05, "loss": 2.4388, "step": 140 }, { "epoch": 0.12325390304026294, "grad_norm": 7.4036645978760225, "learning_rate": 1.5297741273100617e-05, "loss": 2.3493, "step": 150 }, { "epoch": 0.1314708299096138, "grad_norm": 7.12561810579708, "learning_rate": 1.6324435318275156e-05, "loss": 2.2697, "step": 160 }, { "epoch": 0.13968775677896467, "grad_norm": 9.222649571442956, "learning_rate": 1.735112936344969e-05, "loss": 2.1975, "step": 170 }, { "epoch": 0.14790468364831552, "grad_norm": 8.868924318680982, "learning_rate": 1.837782340862423e-05, "loss": 2.1464, "step": 180 }, { "epoch": 0.1561216105176664, "grad_norm": 11.22877634105566, "learning_rate": 1.940451745379877e-05, "loss": 2.0921, "step": 190 }, { "epoch": 0.16433853738701726, "grad_norm": 14.840583315298584, "learning_rate": 2.0431211498973308e-05, "loss": 2.0518, "step": 200 }, { "epoch": 0.1725554642563681, "grad_norm": 12.147153984880905, "learning_rate": 2.1457905544147843e-05, "loss": 1.9688, "step": 210 }, { "epoch": 0.18077239112571897, "grad_norm": 17.004164152742906, "learning_rate": 2.2484599589322382e-05, "loss": 1.9184, "step": 220 }, { "epoch": 0.18898931799506985, "grad_norm": 54.36301855761989, "learning_rate": 2.351129363449692e-05, "loss": 1.8491, "step": 230 }, { "epoch": 0.1972062448644207, "grad_norm": 13.223675766771267, "learning_rate": 2.453798767967146e-05, "loss": 1.8062, "step": 240 }, { "epoch": 0.20542317173377156, "grad_norm": 23.739675599865905, "learning_rate": 2.5564681724845996e-05, "loss": 1.7007, "step": 250 }, { "epoch": 0.21364009860312244, "grad_norm": 13.136946015186366, "learning_rate": 2.6591375770020538e-05, "loss": 1.6032, "step": 260 }, { "epoch": 0.2218570254724733, "grad_norm": 11.684546950836488, "learning_rate": 2.7618069815195074e-05, "loss": 1.4973, "step": 270 }, { "epoch": 0.23007395234182415, "grad_norm": 7.667957804275019, "learning_rate": 2.8644763860369613e-05, "loss": 1.3794, "step": 280 }, { "epoch": 0.23829087921117503, "grad_norm": 12.12659183227395, "learning_rate": 2.9671457905544148e-05, "loss": 1.3322, "step": 290 }, { "epoch": 0.2465078060805259, "grad_norm": 8.972817044811347, "learning_rate": 3.069815195071869e-05, "loss": 1.2359, "step": 300 }, { "epoch": 0.25472473294987674, "grad_norm": 4.3028554301503465, "learning_rate": 3.172484599589323e-05, "loss": 1.1852, "step": 310 }, { "epoch": 0.2629416598192276, "grad_norm": 5.3453439233449185, "learning_rate": 3.2751540041067765e-05, "loss": 1.1183, "step": 320 }, { "epoch": 0.27115858668857845, "grad_norm": 6.559451348524855, "learning_rate": 3.37782340862423e-05, "loss": 1.0608, "step": 330 }, { "epoch": 0.27937551355792933, "grad_norm": 6.374692391110214, "learning_rate": 3.480492813141684e-05, "loss": 1.0151, "step": 340 }, { "epoch": 0.2875924404272802, "grad_norm": 7.303560110424278, "learning_rate": 3.583162217659138e-05, "loss": 0.9712, "step": 350 }, { "epoch": 0.29580936729663104, "grad_norm": 7.0001296670132485, "learning_rate": 3.6858316221765914e-05, "loss": 0.926, "step": 360 }, { "epoch": 0.3040262941659819, "grad_norm": 6.545632008564383, "learning_rate": 3.788501026694045e-05, "loss": 0.8996, "step": 370 }, { "epoch": 0.3122432210353328, "grad_norm": 4.842087687844629, "learning_rate": 3.891170431211499e-05, "loss": 0.8519, "step": 380 }, { "epoch": 0.32046014790468363, "grad_norm": 5.689801303683759, "learning_rate": 3.993839835728953e-05, "loss": 0.8223, "step": 390 }, { "epoch": 0.3286770747740345, "grad_norm": 6.971495473441899, "learning_rate": 4.096509240246407e-05, "loss": 0.7828, "step": 400 }, { "epoch": 0.3368940016433854, "grad_norm": 6.36358312425706, "learning_rate": 4.1991786447638605e-05, "loss": 0.7489, "step": 410 }, { "epoch": 0.3451109285127362, "grad_norm": 8.676168539975894, "learning_rate": 4.301848049281315e-05, "loss": 0.731, "step": 420 }, { "epoch": 0.3533278553820871, "grad_norm": 7.918859246440221, "learning_rate": 4.404517453798768e-05, "loss": 0.7064, "step": 430 }, { "epoch": 0.36154478225143794, "grad_norm": 8.806726830982097, "learning_rate": 4.507186858316222e-05, "loss": 0.6826, "step": 440 }, { "epoch": 0.3697617091207888, "grad_norm": 6.737818041811399, "learning_rate": 4.6098562628336754e-05, "loss": 0.6548, "step": 450 }, { "epoch": 0.3779786359901397, "grad_norm": 6.523855090754944, "learning_rate": 4.7125256673511296e-05, "loss": 0.6279, "step": 460 }, { "epoch": 0.3861955628594905, "grad_norm": 8.120644576403212, "learning_rate": 4.815195071868583e-05, "loss": 0.6037, "step": 470 }, { "epoch": 0.3944124897288414, "grad_norm": 5.5098094876695285, "learning_rate": 4.9178644763860374e-05, "loss": 0.5764, "step": 480 }, { "epoch": 0.4026294165981923, "grad_norm": 8.171893404017268, "learning_rate": 5.020533880903491e-05, "loss": 0.5662, "step": 490 }, { "epoch": 0.4108463434675431, "grad_norm": 5.545430210957229, "learning_rate": 5.1232032854209445e-05, "loss": 0.5416, "step": 500 }, { "epoch": 0.419063270336894, "grad_norm": 6.687063853266749, "learning_rate": 5.225872689938398e-05, "loss": 0.5087, "step": 510 }, { "epoch": 0.4272801972062449, "grad_norm": 4.47366213437882, "learning_rate": 5.328542094455853e-05, "loss": 0.4881, "step": 520 }, { "epoch": 0.4354971240755957, "grad_norm": 4.172813447018652, "learning_rate": 5.4312114989733065e-05, "loss": 0.4621, "step": 530 }, { "epoch": 0.4437140509449466, "grad_norm": 4.220516576800102, "learning_rate": 5.53388090349076e-05, "loss": 0.4487, "step": 540 }, { "epoch": 0.4519309778142975, "grad_norm": 4.709217957842771, "learning_rate": 5.6365503080082136e-05, "loss": 0.4246, "step": 550 }, { "epoch": 0.4601479046836483, "grad_norm": 3.6943640807724925, "learning_rate": 5.739219712525667e-05, "loss": 0.4002, "step": 560 }, { "epoch": 0.4683648315529992, "grad_norm": 4.324443086003805, "learning_rate": 5.8418891170431214e-05, "loss": 0.3769, "step": 570 }, { "epoch": 0.47658175842235007, "grad_norm": 4.712485210987473, "learning_rate": 5.944558521560575e-05, "loss": 0.3605, "step": 580 }, { "epoch": 0.4847986852917009, "grad_norm": 2.5269576298422347, "learning_rate": 6.04722792607803e-05, "loss": 0.3417, "step": 590 }, { "epoch": 0.4930156121610518, "grad_norm": 3.54537374699048, "learning_rate": 6.149897330595483e-05, "loss": 0.3316, "step": 600 }, { "epoch": 0.5012325390304027, "grad_norm": 3.530633503104792, "learning_rate": 6.252566735112937e-05, "loss": 0.313, "step": 610 }, { "epoch": 0.5094494658997535, "grad_norm": 4.839997750327561, "learning_rate": 6.35523613963039e-05, "loss": 0.2974, "step": 620 }, { "epoch": 0.5176663927691043, "grad_norm": 4.197641178153233, "learning_rate": 6.457905544147844e-05, "loss": 0.2745, "step": 630 }, { "epoch": 0.5258833196384552, "grad_norm": 2.705058855982426, "learning_rate": 6.560574948665298e-05, "loss": 0.2615, "step": 640 }, { "epoch": 0.5341002465078061, "grad_norm": 2.8099029424585313, "learning_rate": 6.663244353182751e-05, "loss": 0.2503, "step": 650 }, { "epoch": 0.5423171733771569, "grad_norm": 2.5734835696398712, "learning_rate": 6.765913757700205e-05, "loss": 0.2415, "step": 660 }, { "epoch": 0.5505341002465078, "grad_norm": 2.7836726382586425, "learning_rate": 6.86858316221766e-05, "loss": 0.2344, "step": 670 }, { "epoch": 0.5587510271158587, "grad_norm": 2.603051010989796, "learning_rate": 6.971252566735113e-05, "loss": 0.2154, "step": 680 }, { "epoch": 0.5669679539852095, "grad_norm": 2.1107040737971587, "learning_rate": 7.073921971252567e-05, "loss": 0.1976, "step": 690 }, { "epoch": 0.5751848808545604, "grad_norm": 2.803095444807186, "learning_rate": 7.17659137577002e-05, "loss": 0.1864, "step": 700 }, { "epoch": 0.5834018077239113, "grad_norm": 2.1670666899861075, "learning_rate": 7.279260780287475e-05, "loss": 0.178, "step": 710 }, { "epoch": 0.5916187345932621, "grad_norm": 1.7921531158206865, "learning_rate": 7.381930184804929e-05, "loss": 0.1652, "step": 720 }, { "epoch": 0.599835661462613, "grad_norm": 1.479382404686151, "learning_rate": 7.484599589322382e-05, "loss": 0.1576, "step": 730 }, { "epoch": 0.6080525883319639, "grad_norm": 1.743395608685881, "learning_rate": 7.587268993839836e-05, "loss": 0.1498, "step": 740 }, { "epoch": 0.6162695152013147, "grad_norm": 1.8303806232104347, "learning_rate": 7.689938398357291e-05, "loss": 0.136, "step": 750 }, { "epoch": 0.6244864420706656, "grad_norm": 2.312871831212618, "learning_rate": 7.792607802874744e-05, "loss": 0.1274, "step": 760 }, { "epoch": 0.6327033689400164, "grad_norm": 1.6901573163360035, "learning_rate": 7.895277207392198e-05, "loss": 0.1191, "step": 770 }, { "epoch": 0.6409202958093673, "grad_norm": 1.6314268202265299, "learning_rate": 7.997946611909651e-05, "loss": 0.1114, "step": 780 }, { "epoch": 0.6491372226787182, "grad_norm": 1.4508574236430893, "learning_rate": 8.100616016427105e-05, "loss": 0.101, "step": 790 }, { "epoch": 0.657354149548069, "grad_norm": 1.5649749543554101, "learning_rate": 8.203285420944559e-05, "loss": 0.0985, "step": 800 }, { "epoch": 0.6655710764174199, "grad_norm": 2.6845550965354428, "learning_rate": 8.305954825462012e-05, "loss": 0.1027, "step": 810 }, { "epoch": 0.6737880032867708, "grad_norm": 2.7764588978262426, "learning_rate": 8.408624229979466e-05, "loss": 0.1035, "step": 820 }, { "epoch": 0.6820049301561216, "grad_norm": 4.830497904460127, "learning_rate": 8.51129363449692e-05, "loss": 0.0976, "step": 830 }, { "epoch": 0.6902218570254725, "grad_norm": 1.0331763322679437, "learning_rate": 8.613963039014374e-05, "loss": 0.0879, "step": 840 }, { "epoch": 0.6984387838948234, "grad_norm": 0.596374501379962, "learning_rate": 8.716632443531828e-05, "loss": 0.074, "step": 850 }, { "epoch": 0.7066557107641742, "grad_norm": 0.8196280038018171, "learning_rate": 8.819301848049281e-05, "loss": 0.0692, "step": 860 }, { "epoch": 0.714872637633525, "grad_norm": 0.9801217376138596, "learning_rate": 8.921971252566735e-05, "loss": 0.065, "step": 870 }, { "epoch": 0.7230895645028759, "grad_norm": 0.5051290053812035, "learning_rate": 9.02464065708419e-05, "loss": 0.0596, "step": 880 }, { "epoch": 0.7313064913722268, "grad_norm": 0.9667645246911198, "learning_rate": 9.127310061601643e-05, "loss": 0.0556, "step": 890 }, { "epoch": 0.7395234182415776, "grad_norm": 0.46438651540167103, "learning_rate": 9.229979466119097e-05, "loss": 0.0545, "step": 900 }, { "epoch": 0.7477403451109285, "grad_norm": 0.40547338931365595, "learning_rate": 9.332648870636552e-05, "loss": 0.0494, "step": 910 }, { "epoch": 0.7559572719802794, "grad_norm": 0.4441111869944399, "learning_rate": 9.435318275154005e-05, "loss": 0.0466, "step": 920 }, { "epoch": 0.7641741988496302, "grad_norm": 0.8183415391457746, "learning_rate": 9.537987679671459e-05, "loss": 0.0453, "step": 930 }, { "epoch": 0.772391125718981, "grad_norm": 0.5138559384321786, "learning_rate": 9.640657084188912e-05, "loss": 0.0436, "step": 940 }, { "epoch": 0.780608052588332, "grad_norm": 0.474231944626131, "learning_rate": 9.743326488706366e-05, "loss": 0.0415, "step": 950 }, { "epoch": 0.7888249794576828, "grad_norm": 0.4407732436458076, "learning_rate": 9.84599589322382e-05, "loss": 0.04, "step": 960 }, { "epoch": 0.7970419063270336, "grad_norm": 1.888951859317562, "learning_rate": 9.948665297741273e-05, "loss": 0.0401, "step": 970 }, { "epoch": 0.8052588331963846, "grad_norm": 0.5937342056949868, "learning_rate": 9.999991965235563e-05, "loss": 0.0408, "step": 980 }, { "epoch": 0.8134757600657354, "grad_norm": 0.6643097529303563, "learning_rate": 9.999927687274992e-05, "loss": 0.0458, "step": 990 }, { "epoch": 0.8216926869350862, "grad_norm": 0.38763729674702113, "learning_rate": 9.999799132180184e-05, "loss": 0.0364, "step": 1000 }, { "epoch": 0.8299096138044372, "grad_norm": 0.3739679848963624, "learning_rate": 9.999606301603793e-05, "loss": 0.0332, "step": 1010 }, { "epoch": 0.838126540673788, "grad_norm": 0.3960365233672132, "learning_rate": 9.999349198024775e-05, "loss": 0.0327, "step": 1020 }, { "epoch": 0.8463434675431388, "grad_norm": 0.35789593639188133, "learning_rate": 9.999027824748352e-05, "loss": 0.0313, "step": 1030 }, { "epoch": 0.8545603944124898, "grad_norm": 0.3328319306799121, "learning_rate": 9.998642185905977e-05, "loss": 0.03, "step": 1040 }, { "epoch": 0.8627773212818406, "grad_norm": 0.397125770040445, "learning_rate": 9.998192286455273e-05, "loss": 0.0285, "step": 1050 }, { "epoch": 0.8709942481511914, "grad_norm": 0.35895726006131834, "learning_rate": 9.997678132179969e-05, "loss": 0.0283, "step": 1060 }, { "epoch": 0.8792111750205424, "grad_norm": 0.359528719136988, "learning_rate": 9.997099729689839e-05, "loss": 0.0272, "step": 1070 }, { "epoch": 0.8874281018898932, "grad_norm": 0.31783947017752745, "learning_rate": 9.996457086420598e-05, "loss": 0.026, "step": 1080 }, { "epoch": 0.895645028759244, "grad_norm": 0.24437306397975714, "learning_rate": 9.995750210633822e-05, "loss": 0.0252, "step": 1090 }, { "epoch": 0.903861955628595, "grad_norm": 0.2869935821010678, "learning_rate": 9.994979111416829e-05, "loss": 0.0243, "step": 1100 }, { "epoch": 0.9120788824979458, "grad_norm": 0.28967525917886927, "learning_rate": 9.994143798682573e-05, "loss": 0.0236, "step": 1110 }, { "epoch": 0.9202958093672966, "grad_norm": 0.4867480221470474, "learning_rate": 9.993244283169513e-05, "loss": 0.0232, "step": 1120 }, { "epoch": 0.9285127362366475, "grad_norm": 0.27522967583273783, "learning_rate": 9.99228057644147e-05, "loss": 0.0229, "step": 1130 }, { "epoch": 0.9367296631059984, "grad_norm": 0.2705586478283337, "learning_rate": 9.991252690887485e-05, "loss": 0.0217, "step": 1140 }, { "epoch": 0.9449465899753492, "grad_norm": 0.3133390383629242, "learning_rate": 9.990160639721657e-05, "loss": 0.0209, "step": 1150 }, { "epoch": 0.9531635168447001, "grad_norm": 0.3003241130339291, "learning_rate": 9.989004436982974e-05, "loss": 0.0204, "step": 1160 }, { "epoch": 0.961380443714051, "grad_norm": 0.27998609610081276, "learning_rate": 9.987784097535127e-05, "loss": 0.02, "step": 1170 }, { "epoch": 0.9695973705834018, "grad_norm": 0.3094626205699417, "learning_rate": 9.986499637066331e-05, "loss": 0.0198, "step": 1180 }, { "epoch": 0.9778142974527527, "grad_norm": 0.311495504428665, "learning_rate": 9.985151072089112e-05, "loss": 0.0193, "step": 1190 }, { "epoch": 0.9860312243221035, "grad_norm": 0.2810349187040472, "learning_rate": 9.983738419940096e-05, "loss": 0.0193, "step": 1200 }, { "epoch": 0.9942481511914544, "grad_norm": 0.3107871046473083, "learning_rate": 9.982261698779795e-05, "loss": 0.0185, "step": 1210 }, { "epoch": 1.0024650780608053, "grad_norm": 0.26519682492524527, "learning_rate": 9.980720927592366e-05, "loss": 0.0182, "step": 1220 }, { "epoch": 1.0106820049301561, "grad_norm": 0.2391845678539309, "learning_rate": 9.979116126185364e-05, "loss": 0.0173, "step": 1230 }, { "epoch": 1.018898931799507, "grad_norm": 0.26242451803570593, "learning_rate": 9.977447315189493e-05, "loss": 0.0166, "step": 1240 }, { "epoch": 1.0271158586688578, "grad_norm": 0.22409653311325814, "learning_rate": 9.975714516058344e-05, "loss": 0.0162, "step": 1250 }, { "epoch": 1.0353327855382086, "grad_norm": 0.2343121137078575, "learning_rate": 9.973917751068111e-05, "loss": 0.0155, "step": 1260 }, { "epoch": 1.0435497124075597, "grad_norm": 0.21473180531882663, "learning_rate": 9.972057043317308e-05, "loss": 0.0152, "step": 1270 }, { "epoch": 1.0517666392769105, "grad_norm": 0.1994724694575453, "learning_rate": 9.970132416726474e-05, "loss": 0.0151, "step": 1280 }, { "epoch": 1.0599835661462613, "grad_norm": 0.1979357989976122, "learning_rate": 9.968143896037864e-05, "loss": 0.0147, "step": 1290 }, { "epoch": 1.0682004930156122, "grad_norm": 0.23691205439082766, "learning_rate": 9.966091506815128e-05, "loss": 0.0144, "step": 1300 }, { "epoch": 1.076417419884963, "grad_norm": 2.057726124603243, "learning_rate": 9.963975275442989e-05, "loss": 0.0175, "step": 1310 }, { "epoch": 1.0846343467543138, "grad_norm": 0.3234261374936929, "learning_rate": 9.961795229126896e-05, "loss": 0.0165, "step": 1320 }, { "epoch": 1.0928512736236646, "grad_norm": 0.2509885643136706, "learning_rate": 9.959551395892681e-05, "loss": 0.0149, "step": 1330 }, { "epoch": 1.1010682004930157, "grad_norm": 0.19772103993153273, "learning_rate": 9.957243804586197e-05, "loss": 0.0141, "step": 1340 }, { "epoch": 1.1092851273623665, "grad_norm": 0.23068483781938592, "learning_rate": 9.954872484872944e-05, "loss": 0.0134, "step": 1350 }, { "epoch": 1.1175020542317173, "grad_norm": 0.21085093284095438, "learning_rate": 9.952437467237687e-05, "loss": 0.0133, "step": 1360 }, { "epoch": 1.1257189811010682, "grad_norm": 0.21244911066205022, "learning_rate": 9.949938782984072e-05, "loss": 0.0131, "step": 1370 }, { "epoch": 1.133935907970419, "grad_norm": 0.1841268164629968, "learning_rate": 9.947376464234215e-05, "loss": 0.0126, "step": 1380 }, { "epoch": 1.14215283483977, "grad_norm": 0.18885789130639805, "learning_rate": 9.944750543928297e-05, "loss": 0.0123, "step": 1390 }, { "epoch": 1.1503697617091209, "grad_norm": 0.18962637141794314, "learning_rate": 9.942061055824127e-05, "loss": 0.0121, "step": 1400 }, { "epoch": 1.1585866885784717, "grad_norm": 0.20049383873744792, "learning_rate": 9.939308034496728e-05, "loss": 0.0121, "step": 1410 }, { "epoch": 1.1668036154478225, "grad_norm": 0.19041767566091838, "learning_rate": 9.936491515337874e-05, "loss": 0.0119, "step": 1420 }, { "epoch": 1.1750205423171733, "grad_norm": 0.1995551393056727, "learning_rate": 9.933611534555644e-05, "loss": 0.0117, "step": 1430 }, { "epoch": 1.1832374691865242, "grad_norm": 0.16768458374812184, "learning_rate": 9.930668129173956e-05, "loss": 0.0116, "step": 1440 }, { "epoch": 1.191454396055875, "grad_norm": 0.18748344143729523, "learning_rate": 9.927661337032093e-05, "loss": 0.0115, "step": 1450 }, { "epoch": 1.199671322925226, "grad_norm": 0.17618964617531263, "learning_rate": 9.924591196784205e-05, "loss": 0.0113, "step": 1460 }, { "epoch": 1.2078882497945769, "grad_norm": 0.1915822510753366, "learning_rate": 9.92145774789883e-05, "loss": 0.011, "step": 1470 }, { "epoch": 1.2161051766639277, "grad_norm": 0.22348016077011987, "learning_rate": 9.918261030658371e-05, "loss": 0.011, "step": 1480 }, { "epoch": 1.2243221035332785, "grad_norm": 0.17910179979882987, "learning_rate": 9.915001086158588e-05, "loss": 0.0109, "step": 1490 }, { "epoch": 1.2325390304026294, "grad_norm": 0.20954080994397253, "learning_rate": 9.911677956308068e-05, "loss": 0.0106, "step": 1500 }, { "epoch": 1.2407559572719804, "grad_norm": 0.18491793792738792, "learning_rate": 9.908291683827678e-05, "loss": 0.0103, "step": 1510 }, { "epoch": 1.2489728841413312, "grad_norm": 0.1913002285241464, "learning_rate": 9.904842312250024e-05, "loss": 0.01, "step": 1520 }, { "epoch": 1.257189811010682, "grad_norm": 0.18665545298861494, "learning_rate": 9.901329885918898e-05, "loss": 0.0102, "step": 1530 }, { "epoch": 1.2654067378800329, "grad_norm": 0.20441793337405792, "learning_rate": 9.897754449988687e-05, "loss": 0.0099, "step": 1540 }, { "epoch": 1.2736236647493837, "grad_norm": 0.17232110916564725, "learning_rate": 9.894116050423812e-05, "loss": 0.01, "step": 1550 }, { "epoch": 1.2818405916187345, "grad_norm": 0.17117350565381512, "learning_rate": 9.890414733998131e-05, "loss": 0.0097, "step": 1560 }, { "epoch": 1.2900575184880854, "grad_norm": 0.16106675543143417, "learning_rate": 9.886650548294333e-05, "loss": 0.0098, "step": 1570 }, { "epoch": 1.2982744453574364, "grad_norm": 0.15673244249109636, "learning_rate": 9.882823541703332e-05, "loss": 0.0096, "step": 1580 }, { "epoch": 1.3064913722267872, "grad_norm": 0.18359994942806246, "learning_rate": 9.878933763423644e-05, "loss": 0.0095, "step": 1590 }, { "epoch": 1.314708299096138, "grad_norm": 0.16928325787860654, "learning_rate": 9.874981263460749e-05, "loss": 0.0092, "step": 1600 }, { "epoch": 1.322925225965489, "grad_norm": 0.17321373689712377, "learning_rate": 9.87096609262646e-05, "loss": 0.0091, "step": 1610 }, { "epoch": 1.3311421528348397, "grad_norm": 0.18291522891441084, "learning_rate": 9.866888302538255e-05, "loss": 0.009, "step": 1620 }, { "epoch": 1.3393590797041908, "grad_norm": 0.19719101776097683, "learning_rate": 9.862747945618629e-05, "loss": 0.0091, "step": 1630 }, { "epoch": 1.3475760065735414, "grad_norm": 0.1507248639899357, "learning_rate": 9.858545075094402e-05, "loss": 0.0089, "step": 1640 }, { "epoch": 1.3557929334428924, "grad_norm": 0.15900627243941012, "learning_rate": 9.854279744996054e-05, "loss": 0.0085, "step": 1650 }, { "epoch": 1.3640098603122432, "grad_norm": 0.1614739960911069, "learning_rate": 9.849952010157014e-05, "loss": 0.0084, "step": 1660 }, { "epoch": 1.372226787181594, "grad_norm": 0.18518862249859025, "learning_rate": 9.845561926212967e-05, "loss": 0.0085, "step": 1670 }, { "epoch": 1.380443714050945, "grad_norm": 0.16465865540140373, "learning_rate": 9.841109549601133e-05, "loss": 0.0087, "step": 1680 }, { "epoch": 1.3886606409202957, "grad_norm": 0.1778246547911174, "learning_rate": 9.83659493755954e-05, "loss": 0.0085, "step": 1690 }, { "epoch": 1.3968775677896468, "grad_norm": 0.17535152748688587, "learning_rate": 9.832018148126291e-05, "loss": 0.0082, "step": 1700 }, { "epoch": 1.4050944946589976, "grad_norm": 0.1616895940395759, "learning_rate": 9.827379240138824e-05, "loss": 0.008, "step": 1710 }, { "epoch": 1.4133114215283484, "grad_norm": 0.17228485317380646, "learning_rate": 9.822678273233139e-05, "loss": 0.0081, "step": 1720 }, { "epoch": 1.4215283483976993, "grad_norm": 0.14635147682270658, "learning_rate": 9.817915307843046e-05, "loss": 0.008, "step": 1730 }, { "epoch": 1.42974527526705, "grad_norm": 0.1577912852087264, "learning_rate": 9.813090405199388e-05, "loss": 0.0078, "step": 1740 }, { "epoch": 1.437962202136401, "grad_norm": 0.13215393793980348, "learning_rate": 9.80820362732924e-05, "loss": 0.0079, "step": 1750 }, { "epoch": 1.4461791290057517, "grad_norm": 0.14886527670507846, "learning_rate": 9.803255037055128e-05, "loss": 0.0077, "step": 1760 }, { "epoch": 1.4543960558751028, "grad_norm": 0.1663787864042773, "learning_rate": 9.798244697994214e-05, "loss": 0.0074, "step": 1770 }, { "epoch": 1.4626129827444536, "grad_norm": 0.14313267064422866, "learning_rate": 9.793172674557475e-05, "loss": 0.0072, "step": 1780 }, { "epoch": 1.4708299096138044, "grad_norm": 0.14881896112838372, "learning_rate": 9.78803903194888e-05, "loss": 0.0075, "step": 1790 }, { "epoch": 1.4790468364831553, "grad_norm": 0.1378603541071152, "learning_rate": 9.782843836164553e-05, "loss": 0.0071, "step": 1800 }, { "epoch": 1.487263763352506, "grad_norm": 0.13488096210212905, "learning_rate": 9.777587153991916e-05, "loss": 0.0069, "step": 1810 }, { "epoch": 1.4954806902218571, "grad_norm": 0.149195935264567, "learning_rate": 9.77226905300884e-05, "loss": 0.007, "step": 1820 }, { "epoch": 1.5036976170912077, "grad_norm": 0.14389230928854732, "learning_rate": 9.766889601582774e-05, "loss": 0.0071, "step": 1830 }, { "epoch": 1.5119145439605588, "grad_norm": 0.14038858287355682, "learning_rate": 9.761448868869863e-05, "loss": 0.0068, "step": 1840 }, { "epoch": 1.5201314708299096, "grad_norm": 0.14432103109386488, "learning_rate": 9.755946924814055e-05, "loss": 0.0069, "step": 1850 }, { "epoch": 1.5283483976992605, "grad_norm": 0.14793500060029685, "learning_rate": 9.750383840146218e-05, "loss": 0.0069, "step": 1860 }, { "epoch": 1.5365653245686115, "grad_norm": 0.1347785521160864, "learning_rate": 9.744759686383211e-05, "loss": 0.0069, "step": 1870 }, { "epoch": 1.544782251437962, "grad_norm": 0.13316136302347778, "learning_rate": 9.739074535826978e-05, "loss": 0.0065, "step": 1880 }, { "epoch": 1.5529991783073132, "grad_norm": 0.14574183234800075, "learning_rate": 9.733328461563612e-05, "loss": 0.0065, "step": 1890 }, { "epoch": 1.561216105176664, "grad_norm": 0.13326787078082328, "learning_rate": 9.72752153746242e-05, "loss": 0.0066, "step": 1900 }, { "epoch": 1.5694330320460148, "grad_norm": 0.14016833932250203, "learning_rate": 9.721653838174972e-05, "loss": 0.0064, "step": 1910 }, { "epoch": 1.5776499589153656, "grad_norm": 0.1220618263565421, "learning_rate": 9.715725439134132e-05, "loss": 0.0063, "step": 1920 }, { "epoch": 1.5858668857847165, "grad_norm": 0.14734607229286312, "learning_rate": 9.709736416553106e-05, "loss": 0.0064, "step": 1930 }, { "epoch": 1.5940838126540675, "grad_norm": 0.1513977921448163, "learning_rate": 9.703686847424449e-05, "loss": 0.0063, "step": 1940 }, { "epoch": 1.6023007395234181, "grad_norm": 0.15247244786036027, "learning_rate": 9.697576809519079e-05, "loss": 0.0063, "step": 1950 }, { "epoch": 1.6105176663927692, "grad_norm": 0.12449770873160382, "learning_rate": 9.691406381385278e-05, "loss": 0.0061, "step": 1960 }, { "epoch": 1.61873459326212, "grad_norm": 0.14780820552617843, "learning_rate": 9.685175642347679e-05, "loss": 0.0062, "step": 1970 }, { "epoch": 1.6269515201314708, "grad_norm": 0.12285378569676762, "learning_rate": 9.678884672506252e-05, "loss": 0.006, "step": 1980 }, { "epoch": 1.6351684470008216, "grad_norm": 0.1374507844474314, "learning_rate": 9.672533552735268e-05, "loss": 0.0059, "step": 1990 }, { "epoch": 1.6433853738701725, "grad_norm": 0.1340445045771385, "learning_rate": 9.666122364682263e-05, "loss": 0.0058, "step": 2000 }, { "epoch": 1.6516023007395235, "grad_norm": 1.5494672490847667, "learning_rate": 9.659651190766992e-05, "loss": 0.0473, "step": 2010 }, { "epoch": 1.6598192276088741, "grad_norm": 1.055588957101282, "learning_rate": 9.653120114180356e-05, "loss": 0.0213, "step": 2020 }, { "epoch": 1.6680361544782252, "grad_norm": 2.158511305689324, "learning_rate": 9.646529218883349e-05, "loss": 0.068, "step": 2030 }, { "epoch": 1.676253081347576, "grad_norm": 0.34278686660556884, "learning_rate": 9.639878589605969e-05, "loss": 0.025, "step": 2040 }, { "epoch": 1.6844700082169268, "grad_norm": 0.18106436706773044, "learning_rate": 9.63316831184613e-05, "loss": 0.0097, "step": 2050 }, { "epoch": 1.6926869350862779, "grad_norm": 0.14187259081013245, "learning_rate": 9.626398471868564e-05, "loss": 0.0079, "step": 2060 }, { "epoch": 1.7009038619556285, "grad_norm": 0.13032445672489365, "learning_rate": 9.619569156703714e-05, "loss": 0.007, "step": 2070 }, { "epoch": 1.7091207888249795, "grad_norm": 0.13900486225724332, "learning_rate": 9.612680454146609e-05, "loss": 0.0068, "step": 2080 }, { "epoch": 1.7173377156943304, "grad_norm": 0.1371829976046093, "learning_rate": 9.605732452755742e-05, "loss": 0.0062, "step": 2090 }, { "epoch": 1.7255546425636812, "grad_norm": 0.11435904716007088, "learning_rate": 9.598725241851932e-05, "loss": 0.0062, "step": 2100 }, { "epoch": 1.733771569433032, "grad_norm": 0.12003022629994674, "learning_rate": 9.591658911517162e-05, "loss": 0.006, "step": 2110 }, { "epoch": 1.7419884963023828, "grad_norm": 0.12472657638284222, "learning_rate": 9.584533552593447e-05, "loss": 0.0056, "step": 2120 }, { "epoch": 1.7502054231717339, "grad_norm": 0.10956857048247456, "learning_rate": 9.577349256681636e-05, "loss": 0.0057, "step": 2130 }, { "epoch": 1.7584223500410845, "grad_norm": 0.10206011618948153, "learning_rate": 9.570106116140256e-05, "loss": 0.0054, "step": 2140 }, { "epoch": 1.7666392769104355, "grad_norm": 0.12595281361553662, "learning_rate": 9.56280422408432e-05, "loss": 0.0054, "step": 2150 }, { "epoch": 1.7748562037797864, "grad_norm": 0.11843203344142937, "learning_rate": 9.555443674384122e-05, "loss": 0.0052, "step": 2160 }, { "epoch": 1.7830731306491372, "grad_norm": 0.10294032952850225, "learning_rate": 9.548024561664039e-05, "loss": 0.0051, "step": 2170 }, { "epoch": 1.7912900575184882, "grad_norm": 0.11977450499323275, "learning_rate": 9.540546981301313e-05, "loss": 0.0052, "step": 2180 }, { "epoch": 1.7995069843878388, "grad_norm": 0.10306048483524843, "learning_rate": 9.53301102942482e-05, "loss": 0.0051, "step": 2190 }, { "epoch": 1.80772391125719, "grad_norm": 0.11032309861470083, "learning_rate": 9.525416802913841e-05, "loss": 0.005, "step": 2200 }, { "epoch": 1.8159408381265407, "grad_norm": 0.12349361681027836, "learning_rate": 9.51776439939681e-05, "loss": 0.0051, "step": 2210 }, { "epoch": 1.8241577649958916, "grad_norm": 0.11522340458787386, "learning_rate": 9.510053917250061e-05, "loss": 0.0048, "step": 2220 }, { "epoch": 1.8323746918652424, "grad_norm": 0.10888918083024797, "learning_rate": 9.502285455596572e-05, "loss": 0.0048, "step": 2230 }, { "epoch": 1.8405916187345932, "grad_norm": 0.10904702080255187, "learning_rate": 9.494459114304672e-05, "loss": 0.0047, "step": 2240 }, { "epoch": 1.8488085456039443, "grad_norm": 0.39937298924235193, "learning_rate": 9.48657499398678e-05, "loss": 0.0059, "step": 2250 }, { "epoch": 1.8570254724732949, "grad_norm": 0.24868775681299268, "learning_rate": 9.47863319599809e-05, "loss": 0.0075, "step": 2260 }, { "epoch": 1.865242399342646, "grad_norm": 0.1633251832735737, "learning_rate": 9.470633822435282e-05, "loss": 0.0068, "step": 2270 }, { "epoch": 1.8734593262119967, "grad_norm": 0.10576097653379558, "learning_rate": 9.462576976135207e-05, "loss": 0.0054, "step": 2280 }, { "epoch": 1.8816762530813476, "grad_norm": 0.12656212811763418, "learning_rate": 9.454462760673559e-05, "loss": 0.005, "step": 2290 }, { "epoch": 1.8898931799506984, "grad_norm": 0.1000511957852473, "learning_rate": 9.446291280363551e-05, "loss": 0.0048, "step": 2300 }, { "epoch": 1.8981101068200492, "grad_norm": 0.1001106728471161, "learning_rate": 9.43806264025457e-05, "loss": 0.0046, "step": 2310 }, { "epoch": 1.9063270336894003, "grad_norm": 0.10653485492005862, "learning_rate": 9.429776946130826e-05, "loss": 0.0046, "step": 2320 }, { "epoch": 1.9145439605587509, "grad_norm": 0.09554139188139198, "learning_rate": 9.421434304509992e-05, "loss": 0.0044, "step": 2330 }, { "epoch": 1.922760887428102, "grad_norm": 0.11235081047459526, "learning_rate": 9.413034822641844e-05, "loss": 0.0043, "step": 2340 }, { "epoch": 1.9309778142974527, "grad_norm": 0.09925731232811477, "learning_rate": 9.404578608506865e-05, "loss": 0.0043, "step": 2350 }, { "epoch": 1.9391947411668036, "grad_norm": 0.10170180896026369, "learning_rate": 9.396065770814867e-05, "loss": 0.0044, "step": 2360 }, { "epoch": 1.9474116680361546, "grad_norm": 0.1094941960116376, "learning_rate": 9.387496419003599e-05, "loss": 0.0042, "step": 2370 }, { "epoch": 1.9556285949055052, "grad_norm": 0.09325155091207862, "learning_rate": 9.378870663237328e-05, "loss": 0.004, "step": 2380 }, { "epoch": 1.9638455217748563, "grad_norm": 0.09015117689020187, "learning_rate": 9.37018861440543e-05, "loss": 0.004, "step": 2390 }, { "epoch": 1.972062448644207, "grad_norm": 0.10243104002211424, "learning_rate": 9.361450384120963e-05, "loss": 0.0041, "step": 2400 }, { "epoch": 1.980279375513558, "grad_norm": 0.11356430772327424, "learning_rate": 9.352656084719231e-05, "loss": 0.004, "step": 2410 }, { "epoch": 1.9884963023829088, "grad_norm": 0.1133566447197813, "learning_rate": 9.343805829256343e-05, "loss": 0.0042, "step": 2420 }, { "epoch": 1.9967132292522596, "grad_norm": 0.09025115427016078, "learning_rate": 9.334899731507756e-05, "loss": 0.0039, "step": 2430 }, { "epoch": 2.0049301561216106, "grad_norm": 0.10161038188503023, "learning_rate": 9.325937905966813e-05, "loss": 0.0035, "step": 2440 }, { "epoch": 2.0131470829909612, "grad_norm": 0.08227738890594856, "learning_rate": 9.316920467843273e-05, "loss": 0.003, "step": 2450 }, { "epoch": 2.0213640098603123, "grad_norm": 0.08570367324292932, "learning_rate": 9.307847533061831e-05, "loss": 0.003, "step": 2460 }, { "epoch": 2.029580936729663, "grad_norm": 0.07659113085608305, "learning_rate": 9.29871921826062e-05, "loss": 0.0029, "step": 2470 }, { "epoch": 2.037797863599014, "grad_norm": 0.08735185109119072, "learning_rate": 9.289535640789721e-05, "loss": 0.0028, "step": 2480 }, { "epoch": 2.046014790468365, "grad_norm": 0.08112181526353192, "learning_rate": 9.280296918709653e-05, "loss": 0.0028, "step": 2490 }, { "epoch": 2.0542317173377156, "grad_norm": 0.09071733200524149, "learning_rate": 9.271003170789846e-05, "loss": 0.0028, "step": 2500 }, { "epoch": 2.0624486442070666, "grad_norm": 0.07888705433105933, "learning_rate": 9.261654516507126e-05, "loss": 0.0028, "step": 2510 }, { "epoch": 2.0706655710764172, "grad_norm": 0.08256418767761199, "learning_rate": 9.252251076044172e-05, "loss": 0.0027, "step": 2520 }, { "epoch": 2.0788824979457683, "grad_norm": 0.08816475273210793, "learning_rate": 9.242792970287972e-05, "loss": 0.0028, "step": 2530 }, { "epoch": 2.0870994248151193, "grad_norm": 0.11440139774626451, "learning_rate": 9.233280320828274e-05, "loss": 0.0028, "step": 2540 }, { "epoch": 2.09531635168447, "grad_norm": 0.0812607297477746, "learning_rate": 9.223713249956016e-05, "loss": 0.0028, "step": 2550 }, { "epoch": 2.103533278553821, "grad_norm": 0.09406446044734776, "learning_rate": 9.214091880661752e-05, "loss": 0.0028, "step": 2560 }, { "epoch": 2.1117502054231716, "grad_norm": 0.08675348503864556, "learning_rate": 9.204416336634084e-05, "loss": 0.0029, "step": 2570 }, { "epoch": 2.1199671322925226, "grad_norm": 0.08798127219029518, "learning_rate": 9.194686742258059e-05, "loss": 0.0029, "step": 2580 }, { "epoch": 2.1281840591618733, "grad_norm": 0.1042447423026133, "learning_rate": 9.184903222613574e-05, "loss": 0.0028, "step": 2590 }, { "epoch": 2.1364009860312243, "grad_norm": 0.07226523480692097, "learning_rate": 9.175065903473768e-05, "loss": 0.0028, "step": 2600 }, { "epoch": 2.1446179129005754, "grad_norm": 0.09060351493323598, "learning_rate": 9.165174911303409e-05, "loss": 0.0029, "step": 2610 }, { "epoch": 2.152834839769926, "grad_norm": 0.09147549430600695, "learning_rate": 9.15523037325726e-05, "loss": 0.0028, "step": 2620 }, { "epoch": 2.161051766639277, "grad_norm": 0.08581674950289743, "learning_rate": 9.145232417178453e-05, "loss": 0.0027, "step": 2630 }, { "epoch": 2.1692686935086276, "grad_norm": 0.0806685544659502, "learning_rate": 9.135181171596838e-05, "loss": 0.0027, "step": 2640 }, { "epoch": 2.1774856203779787, "grad_norm": 0.0867069998728415, "learning_rate": 9.125076765727338e-05, "loss": 0.0027, "step": 2650 }, { "epoch": 2.1857025472473293, "grad_norm": 0.09650448122353704, "learning_rate": 9.114919329468282e-05, "loss": 0.0026, "step": 2660 }, { "epoch": 2.1939194741166803, "grad_norm": 0.08346292571616859, "learning_rate": 9.104708993399738e-05, "loss": 0.0025, "step": 2670 }, { "epoch": 2.2021364009860314, "grad_norm": 0.0922034193015322, "learning_rate": 9.094445888781829e-05, "loss": 0.0027, "step": 2680 }, { "epoch": 2.210353327855382, "grad_norm": 0.08045238236970594, "learning_rate": 9.084130147553057e-05, "loss": 0.0026, "step": 2690 }, { "epoch": 2.218570254724733, "grad_norm": 0.0823407241176973, "learning_rate": 9.073761902328597e-05, "loss": 0.0026, "step": 2700 }, { "epoch": 2.2267871815940836, "grad_norm": 0.0918388311221415, "learning_rate": 9.063341286398592e-05, "loss": 0.0026, "step": 2710 }, { "epoch": 2.2350041084634347, "grad_norm": 0.08811224741688072, "learning_rate": 9.052868433726448e-05, "loss": 0.0027, "step": 2720 }, { "epoch": 2.2432210353327857, "grad_norm": 0.08713466229884091, "learning_rate": 9.042343478947103e-05, "loss": 0.0026, "step": 2730 }, { "epoch": 2.2514379622021363, "grad_norm": 0.08164204161119902, "learning_rate": 9.0317665573653e-05, "loss": 0.0027, "step": 2740 }, { "epoch": 2.2596548890714874, "grad_norm": 0.08621684454149689, "learning_rate": 9.021137804953846e-05, "loss": 0.0026, "step": 2750 }, { "epoch": 2.267871815940838, "grad_norm": 0.08581890628695187, "learning_rate": 9.010457358351867e-05, "loss": 0.0025, "step": 2760 }, { "epoch": 2.276088742810189, "grad_norm": 0.0880327255343587, "learning_rate": 8.99972535486305e-05, "loss": 0.0025, "step": 2770 }, { "epoch": 2.28430566967954, "grad_norm": 0.10198724504699996, "learning_rate": 8.988941932453872e-05, "loss": 0.0025, "step": 2780 }, { "epoch": 2.2925225965488907, "grad_norm": 0.08363470598937481, "learning_rate": 8.978107229751842e-05, "loss": 0.0026, "step": 2790 }, { "epoch": 2.3007395234182417, "grad_norm": 0.08701906271338153, "learning_rate": 8.967221386043699e-05, "loss": 0.0026, "step": 2800 }, { "epoch": 2.3089564502875923, "grad_norm": 0.08120255658359293, "learning_rate": 8.956284541273632e-05, "loss": 0.0026, "step": 2810 }, { "epoch": 2.3171733771569434, "grad_norm": 0.09844930162834767, "learning_rate": 8.945296836041487e-05, "loss": 0.0026, "step": 2820 }, { "epoch": 2.325390304026294, "grad_norm": 0.08546167195111991, "learning_rate": 8.934258411600945e-05, "loss": 0.0026, "step": 2830 }, { "epoch": 2.333607230895645, "grad_norm": 0.07598724139236995, "learning_rate": 8.923169409857718e-05, "loss": 0.0026, "step": 2840 }, { "epoch": 2.341824157764996, "grad_norm": 0.07855480397018198, "learning_rate": 8.912029973367717e-05, "loss": 0.0026, "step": 2850 }, { "epoch": 2.3500410846343467, "grad_norm": 0.08252273680081287, "learning_rate": 8.900840245335225e-05, "loss": 0.0026, "step": 2860 }, { "epoch": 2.3582580115036977, "grad_norm": 0.09490170806356914, "learning_rate": 8.889600369611051e-05, "loss": 0.0025, "step": 2870 }, { "epoch": 2.3664749383730483, "grad_norm": 0.0873514246728157, "learning_rate": 8.878310490690686e-05, "loss": 0.0026, "step": 2880 }, { "epoch": 2.3746918652423994, "grad_norm": 0.08433770615709385, "learning_rate": 8.866970753712442e-05, "loss": 0.0024, "step": 2890 }, { "epoch": 2.38290879211175, "grad_norm": 0.08543362259129167, "learning_rate": 8.855581304455584e-05, "loss": 0.0025, "step": 2900 }, { "epoch": 2.391125718981101, "grad_norm": 0.08675602957949538, "learning_rate": 8.844142289338466e-05, "loss": 0.0024, "step": 2910 }, { "epoch": 2.399342645850452, "grad_norm": 0.08568773220601615, "learning_rate": 8.832653855416634e-05, "loss": 0.0027, "step": 2920 }, { "epoch": 2.4075595727198027, "grad_norm": 0.07456118767226848, "learning_rate": 8.821116150380946e-05, "loss": 0.0025, "step": 2930 }, { "epoch": 2.4157764995891537, "grad_norm": 0.0704198355651112, "learning_rate": 8.80952932255567e-05, "loss": 0.0023, "step": 2940 }, { "epoch": 2.4239934264585044, "grad_norm": 0.08268043669392515, "learning_rate": 8.797893520896579e-05, "loss": 0.0023, "step": 2950 }, { "epoch": 2.4322103533278554, "grad_norm": 0.08227547263082345, "learning_rate": 8.786208894989032e-05, "loss": 0.0023, "step": 2960 }, { "epoch": 2.4404272801972064, "grad_norm": 0.08698253645225705, "learning_rate": 8.774475595046055e-05, "loss": 0.0025, "step": 2970 }, { "epoch": 2.448644207066557, "grad_norm": 0.08849442489572161, "learning_rate": 8.762693771906409e-05, "loss": 0.0023, "step": 2980 }, { "epoch": 2.456861133935908, "grad_norm": 0.07467727254896249, "learning_rate": 8.750863577032652e-05, "loss": 0.0023, "step": 2990 }, { "epoch": 2.4650780608052587, "grad_norm": 0.09244113926463907, "learning_rate": 8.738985162509186e-05, "loss": 0.0024, "step": 3000 }, { "epoch": 2.4732949876746098, "grad_norm": 0.08833101573406452, "learning_rate": 8.727058681040308e-05, "loss": 0.0024, "step": 3010 }, { "epoch": 2.481511914543961, "grad_norm": 0.07226683243114444, "learning_rate": 8.715084285948248e-05, "loss": 0.0024, "step": 3020 }, { "epoch": 2.4897288414133114, "grad_norm": 0.07268521459200765, "learning_rate": 8.703062131171192e-05, "loss": 0.0024, "step": 3030 }, { "epoch": 2.4979457682826625, "grad_norm": 0.09089659924385088, "learning_rate": 8.690992371261302e-05, "loss": 0.0025, "step": 3040 }, { "epoch": 2.506162695152013, "grad_norm": 0.07562339403468425, "learning_rate": 8.678875161382745e-05, "loss": 0.0024, "step": 3050 }, { "epoch": 2.514379622021364, "grad_norm": 0.08031934504490368, "learning_rate": 8.666710657309672e-05, "loss": 0.0023, "step": 3060 }, { "epoch": 2.5225965488907147, "grad_norm": 0.08457583970716735, "learning_rate": 8.654499015424243e-05, "loss": 0.0023, "step": 3070 }, { "epoch": 2.5308134757600658, "grad_norm": 0.0755537914601127, "learning_rate": 8.642240392714593e-05, "loss": 0.0024, "step": 3080 }, { "epoch": 2.5390304026294164, "grad_norm": 0.07873822493348767, "learning_rate": 8.629934946772831e-05, "loss": 0.0022, "step": 3090 }, { "epoch": 2.5472473294987674, "grad_norm": 0.08908105432532254, "learning_rate": 8.617582835793003e-05, "loss": 0.0022, "step": 3100 }, { "epoch": 2.5554642563681185, "grad_norm": 0.06999317221832577, "learning_rate": 8.605184218569067e-05, "loss": 0.0024, "step": 3110 }, { "epoch": 2.563681183237469, "grad_norm": 0.07464568302833668, "learning_rate": 8.592739254492844e-05, "loss": 0.0022, "step": 3120 }, { "epoch": 2.57189811010682, "grad_norm": 0.09956594996203859, "learning_rate": 8.580248103551973e-05, "loss": 0.0022, "step": 3130 }, { "epoch": 2.5801150369761707, "grad_norm": 0.09774738216358816, "learning_rate": 8.567710926327855e-05, "loss": 0.0024, "step": 3140 }, { "epoch": 2.5883319638455218, "grad_norm": 0.07882192410268168, "learning_rate": 8.555127883993582e-05, "loss": 0.0023, "step": 3150 }, { "epoch": 2.596548890714873, "grad_norm": 0.0762996815625555, "learning_rate": 8.542499138311877e-05, "loss": 0.0022, "step": 3160 }, { "epoch": 2.6047658175842234, "grad_norm": 0.07967271368798, "learning_rate": 8.529824851633004e-05, "loss": 0.0023, "step": 3170 }, { "epoch": 2.6129827444535745, "grad_norm": 0.09037714716671512, "learning_rate": 8.517105186892683e-05, "loss": 0.0021, "step": 3180 }, { "epoch": 2.621199671322925, "grad_norm": 0.06631171216954122, "learning_rate": 8.50434030761e-05, "loss": 0.0021, "step": 3190 }, { "epoch": 2.629416598192276, "grad_norm": 0.06516524243367872, "learning_rate": 8.4915303778853e-05, "loss": 0.002, "step": 3200 }, { "epoch": 2.637633525061627, "grad_norm": 0.07554930250673897, "learning_rate": 8.478675562398077e-05, "loss": 0.0021, "step": 3210 }, { "epoch": 2.645850451930978, "grad_norm": 0.08936807507539121, "learning_rate": 8.465776026404866e-05, "loss": 0.0021, "step": 3220 }, { "epoch": 2.654067378800329, "grad_norm": 0.10445169947837134, "learning_rate": 8.452831935737102e-05, "loss": 0.0022, "step": 3230 }, { "epoch": 2.6622843056696794, "grad_norm": 0.0747838347850065, "learning_rate": 8.439843456799005e-05, "loss": 0.0022, "step": 3240 }, { "epoch": 2.6705012325390305, "grad_norm": 0.07552693679165537, "learning_rate": 8.426810756565429e-05, "loss": 0.0022, "step": 3250 }, { "epoch": 2.6787181594083815, "grad_norm": 0.07514300993270172, "learning_rate": 8.413734002579723e-05, "loss": 0.0022, "step": 3260 }, { "epoch": 2.686935086277732, "grad_norm": 0.07037597595165376, "learning_rate": 8.40061336295157e-05, "loss": 0.0021, "step": 3270 }, { "epoch": 2.6951520131470827, "grad_norm": 0.07303202846814086, "learning_rate": 8.387449006354837e-05, "loss": 0.002, "step": 3280 }, { "epoch": 2.703368940016434, "grad_norm": 0.08480661725235046, "learning_rate": 8.374241102025391e-05, "loss": 0.002, "step": 3290 }, { "epoch": 2.711585866885785, "grad_norm": 0.07058395092941114, "learning_rate": 8.360989819758935e-05, "loss": 0.0021, "step": 3300 }, { "epoch": 2.7198027937551355, "grad_norm": 0.07837331908662558, "learning_rate": 8.347695329908827e-05, "loss": 0.0021, "step": 3310 }, { "epoch": 2.7280197206244865, "grad_norm": 0.07564552331177772, "learning_rate": 8.334357803383875e-05, "loss": 0.002, "step": 3320 }, { "epoch": 2.736236647493837, "grad_norm": 0.07717015273999385, "learning_rate": 8.32097741164616e-05, "loss": 0.0021, "step": 3330 }, { "epoch": 2.744453574363188, "grad_norm": 0.07721669499074643, "learning_rate": 8.307554326708815e-05, "loss": 0.002, "step": 3340 }, { "epoch": 2.752670501232539, "grad_norm": 0.07578771932626932, "learning_rate": 8.294088721133821e-05, "loss": 0.002, "step": 3350 }, { "epoch": 2.76088742810189, "grad_norm": 0.06375027550713391, "learning_rate": 8.28058076802979e-05, "loss": 0.002, "step": 3360 }, { "epoch": 2.769104354971241, "grad_norm": 0.06947732837366209, "learning_rate": 8.267030641049738e-05, "loss": 0.0019, "step": 3370 }, { "epoch": 2.7773212818405915, "grad_norm": 0.07035332876274755, "learning_rate": 8.25343851438885e-05, "loss": 0.0018, "step": 3380 }, { "epoch": 2.7855382087099425, "grad_norm": 0.06994042828521449, "learning_rate": 8.239804562782241e-05, "loss": 0.0019, "step": 3390 }, { "epoch": 2.7937551355792936, "grad_norm": 0.0739022246469536, "learning_rate": 8.226128961502715e-05, "loss": 0.0019, "step": 3400 }, { "epoch": 2.801972062448644, "grad_norm": 0.08090723440170883, "learning_rate": 8.212411886358505e-05, "loss": 0.0018, "step": 3410 }, { "epoch": 2.810188989317995, "grad_norm": 0.06668617007404189, "learning_rate": 8.198653513691019e-05, "loss": 0.0018, "step": 3420 }, { "epoch": 2.818405916187346, "grad_norm": 0.08112145655425776, "learning_rate": 8.184854020372567e-05, "loss": 0.0019, "step": 3430 }, { "epoch": 2.826622843056697, "grad_norm": 0.06838412780008499, "learning_rate": 8.171013583804095e-05, "loss": 0.0018, "step": 3440 }, { "epoch": 2.834839769926048, "grad_norm": 0.06801532642214224, "learning_rate": 8.157132381912894e-05, "loss": 0.0018, "step": 3450 }, { "epoch": 2.8430566967953985, "grad_norm": 0.07450069669807256, "learning_rate": 8.143210593150317e-05, "loss": 0.0018, "step": 3460 }, { "epoch": 2.851273623664749, "grad_norm": 0.07144863614082946, "learning_rate": 8.129248396489493e-05, "loss": 0.0018, "step": 3470 }, { "epoch": 2.8594905505341, "grad_norm": 0.06885416663157976, "learning_rate": 8.115245971423016e-05, "loss": 0.0019, "step": 3480 }, { "epoch": 2.867707477403451, "grad_norm": 0.07359978466021078, "learning_rate": 8.101203497960639e-05, "loss": 0.0019, "step": 3490 }, { "epoch": 2.875924404272802, "grad_norm": 0.06305662725881052, "learning_rate": 8.087121156626965e-05, "loss": 0.0017, "step": 3500 }, { "epoch": 2.884141331142153, "grad_norm": 0.06846870722129733, "learning_rate": 8.072999128459119e-05, "loss": 0.0019, "step": 3510 }, { "epoch": 2.8923582580115035, "grad_norm": 0.07241840468071385, "learning_rate": 8.05883759500443e-05, "loss": 0.0017, "step": 3520 }, { "epoch": 2.9005751848808545, "grad_norm": 0.06286985529378947, "learning_rate": 8.044636738318082e-05, "loss": 0.0016, "step": 3530 }, { "epoch": 2.9087921117502056, "grad_norm": 0.07986920999372275, "learning_rate": 8.030396740960795e-05, "loss": 0.0017, "step": 3540 }, { "epoch": 2.917009038619556, "grad_norm": 0.06958100326685528, "learning_rate": 8.01611778599646e-05, "loss": 0.0017, "step": 3550 }, { "epoch": 2.9252259654889072, "grad_norm": 0.06983385948648055, "learning_rate": 8.001800056989792e-05, "loss": 0.0017, "step": 3560 }, { "epoch": 2.933442892358258, "grad_norm": 0.06621268583147509, "learning_rate": 7.987443738003973e-05, "loss": 0.0018, "step": 3570 }, { "epoch": 2.941659819227609, "grad_norm": 0.06811521015654298, "learning_rate": 7.973049013598277e-05, "loss": 0.0017, "step": 3580 }, { "epoch": 2.94987674609696, "grad_norm": 0.06577394956829613, "learning_rate": 7.958616068825709e-05, "loss": 0.0018, "step": 3590 }, { "epoch": 2.9580936729663105, "grad_norm": 0.08652052889965613, "learning_rate": 7.944145089230619e-05, "loss": 0.0019, "step": 3600 }, { "epoch": 2.9663105998356616, "grad_norm": 0.07907359283151173, "learning_rate": 7.929636260846316e-05, "loss": 0.0017, "step": 3610 }, { "epoch": 2.974527526705012, "grad_norm": 0.07872677969953618, "learning_rate": 7.915089770192679e-05, "loss": 0.0016, "step": 3620 }, { "epoch": 2.9827444535743632, "grad_norm": 0.07503927412326125, "learning_rate": 7.900505804273759e-05, "loss": 0.0017, "step": 3630 }, { "epoch": 2.9909613804437143, "grad_norm": 0.07254751773517193, "learning_rate": 7.885884550575376e-05, "loss": 0.0017, "step": 3640 }, { "epoch": 2.999178307313065, "grad_norm": 0.0737068765701154, "learning_rate": 7.871226197062705e-05, "loss": 0.0016, "step": 3650 }, { "epoch": 3.007395234182416, "grad_norm": 0.047490842379613256, "learning_rate": 7.856530932177862e-05, "loss": 0.0012, "step": 3660 }, { "epoch": 3.0156121610517665, "grad_norm": 0.04477331642729452, "learning_rate": 7.841798944837483e-05, "loss": 0.001, "step": 3670 }, { "epoch": 3.0238290879211176, "grad_norm": 0.05277869090793632, "learning_rate": 7.827030424430292e-05, "loss": 0.0009, "step": 3680 }, { "epoch": 3.032046014790468, "grad_norm": 0.05192630198863309, "learning_rate": 7.812225560814669e-05, "loss": 0.0009, "step": 3690 }, { "epoch": 3.0402629416598193, "grad_norm": 0.050081385496367994, "learning_rate": 7.797384544316207e-05, "loss": 0.0009, "step": 3700 }, { "epoch": 3.0484798685291703, "grad_norm": 0.06481407384433083, "learning_rate": 7.782507565725267e-05, "loss": 0.0009, "step": 3710 }, { "epoch": 3.056696795398521, "grad_norm": 0.04721607919018233, "learning_rate": 7.767594816294528e-05, "loss": 0.001, "step": 3720 }, { "epoch": 3.064913722267872, "grad_norm": 0.05775547139759615, "learning_rate": 7.752646487736519e-05, "loss": 0.0009, "step": 3730 }, { "epoch": 3.0731306491372226, "grad_norm": 0.046089344040974235, "learning_rate": 7.737662772221165e-05, "loss": 0.0009, "step": 3740 }, { "epoch": 3.0813475760065736, "grad_norm": 0.056155203567558286, "learning_rate": 7.722643862373309e-05, "loss": 0.0008, "step": 3750 }, { "epoch": 3.089564502875924, "grad_norm": 0.046228211822813975, "learning_rate": 7.707589951270242e-05, "loss": 0.0009, "step": 3760 }, { "epoch": 3.0977814297452753, "grad_norm": 0.047344255947247574, "learning_rate": 7.692501232439214e-05, "loss": 0.0009, "step": 3770 }, { "epoch": 3.1059983566146263, "grad_norm": 0.053339070367864526, "learning_rate": 7.677377899854954e-05, "loss": 0.0009, "step": 3780 }, { "epoch": 3.114215283483977, "grad_norm": 0.03786126261309618, "learning_rate": 7.662220147937167e-05, "loss": 0.0009, "step": 3790 }, { "epoch": 3.122432210353328, "grad_norm": 0.05134081824746223, "learning_rate": 7.647028171548043e-05, "loss": 0.0009, "step": 3800 }, { "epoch": 3.1306491372226786, "grad_norm": 0.052376320834621, "learning_rate": 7.631802165989749e-05, "loss": 0.0008, "step": 3810 }, { "epoch": 3.1388660640920296, "grad_norm": 0.04894849288766951, "learning_rate": 7.616542327001915e-05, "loss": 0.0008, "step": 3820 }, { "epoch": 3.1470829909613807, "grad_norm": 0.05038320690015898, "learning_rate": 7.601248850759122e-05, "loss": 0.0008, "step": 3830 }, { "epoch": 3.1552999178307313, "grad_norm": 0.0657261484269386, "learning_rate": 7.585921933868379e-05, "loss": 0.0009, "step": 3840 }, { "epoch": 3.1635168447000823, "grad_norm": 3.168359972403995, "learning_rate": 7.570561773366596e-05, "loss": 0.0336, "step": 3850 }, { "epoch": 3.171733771569433, "grad_norm": 2.5751418151669765, "learning_rate": 7.555168566718046e-05, "loss": 0.0704, "step": 3860 }, { "epoch": 3.179950698438784, "grad_norm": 1.2137053590738942, "learning_rate": 7.539742511811833e-05, "loss": 0.0262, "step": 3870 }, { "epoch": 3.1881676253081346, "grad_norm": 0.5619696914357382, "learning_rate": 7.52428380695935e-05, "loss": 0.0383, "step": 3880 }, { "epoch": 3.1963845521774856, "grad_norm": 0.15515360670720688, "learning_rate": 7.508792650891713e-05, "loss": 0.0094, "step": 3890 }, { "epoch": 3.2046014790468367, "grad_norm": 0.07854377076706702, "learning_rate": 7.493269242757233e-05, "loss": 0.0041, "step": 3900 }, { "epoch": 3.2128184059161873, "grad_norm": 0.06398460012451318, "learning_rate": 7.477713782118829e-05, "loss": 0.0027, "step": 3910 }, { "epoch": 3.2210353327855383, "grad_norm": 0.05959020487106516, "learning_rate": 7.462126468951483e-05, "loss": 0.002, "step": 3920 }, { "epoch": 3.229252259654889, "grad_norm": 0.07895981054885065, "learning_rate": 7.446507503639654e-05, "loss": 0.0018, "step": 3930 }, { "epoch": 3.23746918652424, "grad_norm": 0.06113952062109854, "learning_rate": 7.430857086974714e-05, "loss": 0.0016, "step": 3940 }, { "epoch": 3.2456861133935906, "grad_norm": 0.05754540843556571, "learning_rate": 7.415175420152358e-05, "loss": 0.0014, "step": 3950 }, { "epoch": 3.2539030402629416, "grad_norm": 0.04961336104323965, "learning_rate": 7.399462704770023e-05, "loss": 0.0014, "step": 3960 }, { "epoch": 3.2621199671322927, "grad_norm": 0.05151553382652407, "learning_rate": 7.383719142824296e-05, "loss": 0.0013, "step": 3970 }, { "epoch": 3.2703368940016433, "grad_norm": 0.051454661455976, "learning_rate": 7.367944936708309e-05, "loss": 0.0012, "step": 3980 }, { "epoch": 3.2785538208709943, "grad_norm": 0.04530954103893932, "learning_rate": 7.352140289209149e-05, "loss": 0.0012, "step": 3990 }, { "epoch": 3.286770747740345, "grad_norm": 0.05885986312885061, "learning_rate": 7.336305403505246e-05, "loss": 0.0012, "step": 4000 }, { "epoch": 3.294987674609696, "grad_norm": 0.0460453852602048, "learning_rate": 7.320440483163757e-05, "loss": 0.0011, "step": 4010 }, { "epoch": 3.303204601479047, "grad_norm": 0.04443230603480501, "learning_rate": 7.304545732137955e-05, "loss": 0.0011, "step": 4020 }, { "epoch": 3.3114215283483976, "grad_norm": 0.04789780221442194, "learning_rate": 7.288621354764604e-05, "loss": 0.001, "step": 4030 }, { "epoch": 3.3196384552177487, "grad_norm": 0.042748414485179825, "learning_rate": 7.272667555761332e-05, "loss": 0.0011, "step": 4040 }, { "epoch": 3.3278553820870993, "grad_norm": 0.04642188378147712, "learning_rate": 7.256684540224006e-05, "loss": 0.0011, "step": 4050 }, { "epoch": 3.3360723089564503, "grad_norm": 0.05335217218286504, "learning_rate": 7.24067251362408e-05, "loss": 0.001, "step": 4060 }, { "epoch": 3.3442892358258014, "grad_norm": 0.050575978754545205, "learning_rate": 7.224631681805969e-05, "loss": 0.001, "step": 4070 }, { "epoch": 3.352506162695152, "grad_norm": 0.047797377267243354, "learning_rate": 7.208562250984395e-05, "loss": 0.001, "step": 4080 }, { "epoch": 3.360723089564503, "grad_norm": 0.04155249576757446, "learning_rate": 7.19246442774174e-05, "loss": 0.001, "step": 4090 }, { "epoch": 3.3689400164338537, "grad_norm": 0.04874116961943212, "learning_rate": 7.176338419025384e-05, "loss": 0.001, "step": 4100 }, { "epoch": 3.3771569433032047, "grad_norm": 0.04993451162248523, "learning_rate": 7.160184432145049e-05, "loss": 0.0009, "step": 4110 }, { "epoch": 3.3853738701725553, "grad_norm": 0.058322267886013586, "learning_rate": 7.144002674770142e-05, "loss": 0.001, "step": 4120 }, { "epoch": 3.3935907970419064, "grad_norm": 0.06024103937242288, "learning_rate": 7.127793354927062e-05, "loss": 0.0009, "step": 4130 }, { "epoch": 3.401807723911257, "grad_norm": 0.04365055097921593, "learning_rate": 7.111556680996551e-05, "loss": 0.001, "step": 4140 }, { "epoch": 3.410024650780608, "grad_norm": 0.04349525514676888, "learning_rate": 7.095292861711005e-05, "loss": 0.0009, "step": 4150 }, { "epoch": 3.418241577649959, "grad_norm": 0.04210826597697148, "learning_rate": 7.079002106151782e-05, "loss": 0.0009, "step": 4160 }, { "epoch": 3.4264585045193097, "grad_norm": 0.05144291876443696, "learning_rate": 7.06268462374653e-05, "loss": 0.0009, "step": 4170 }, { "epoch": 3.4346754313886607, "grad_norm": 0.0485868162793419, "learning_rate": 7.046340624266483e-05, "loss": 0.0009, "step": 4180 }, { "epoch": 3.4428923582580113, "grad_norm": 0.046691862164365515, "learning_rate": 7.029970317823771e-05, "loss": 0.0009, "step": 4190 }, { "epoch": 3.4511092851273624, "grad_norm": 0.04578967328008899, "learning_rate": 7.013573914868711e-05, "loss": 0.0009, "step": 4200 }, { "epoch": 3.4593262119967134, "grad_norm": 0.04393816584475013, "learning_rate": 6.997151626187116e-05, "loss": 0.0008, "step": 4210 }, { "epoch": 3.467543138866064, "grad_norm": 0.04634896559224033, "learning_rate": 6.980703662897565e-05, "loss": 0.0009, "step": 4220 }, { "epoch": 3.475760065735415, "grad_norm": 0.0482944694819107, "learning_rate": 6.96423023644871e-05, "loss": 0.0008, "step": 4230 }, { "epoch": 3.4839769926047657, "grad_norm": 0.04203105550281779, "learning_rate": 6.947731558616535e-05, "loss": 0.0008, "step": 4240 }, { "epoch": 3.4921939194741167, "grad_norm": 0.0465749345449833, "learning_rate": 6.93120784150166e-05, "loss": 0.0008, "step": 4250 }, { "epoch": 3.5004108463434678, "grad_norm": 0.045140987957123097, "learning_rate": 6.914659297526593e-05, "loss": 0.0008, "step": 4260 }, { "epoch": 3.5086277732128184, "grad_norm": 0.042297592126977576, "learning_rate": 6.898086139433003e-05, "loss": 0.0008, "step": 4270 }, { "epoch": 3.5168447000821694, "grad_norm": 0.04830929173540514, "learning_rate": 6.881488580278999e-05, "loss": 0.0008, "step": 4280 }, { "epoch": 3.52506162695152, "grad_norm": 0.04420957846578299, "learning_rate": 6.864866833436368e-05, "loss": 0.0008, "step": 4290 }, { "epoch": 3.533278553820871, "grad_norm": 0.07244059997824107, "learning_rate": 6.848221112587857e-05, "loss": 0.0008, "step": 4300 }, { "epoch": 3.541495480690222, "grad_norm": 0.047667856990312174, "learning_rate": 6.831551631724402e-05, "loss": 0.0008, "step": 4310 }, { "epoch": 3.5497124075595727, "grad_norm": 0.050174664576941114, "learning_rate": 6.8148586051424e-05, "loss": 0.0008, "step": 4320 }, { "epoch": 3.5579293344289233, "grad_norm": 0.045041164796578714, "learning_rate": 6.798142247440932e-05, "loss": 0.0008, "step": 4330 }, { "epoch": 3.5661462612982744, "grad_norm": 0.04967539131164685, "learning_rate": 6.781402773519021e-05, "loss": 0.0008, "step": 4340 }, { "epoch": 3.5743631881676254, "grad_norm": 0.05341541584045626, "learning_rate": 6.764640398572864e-05, "loss": 0.0008, "step": 4350 }, { "epoch": 3.582580115036976, "grad_norm": 0.043982046768756534, "learning_rate": 6.747855338093058e-05, "loss": 0.0008, "step": 4360 }, { "epoch": 3.590797041906327, "grad_norm": 0.047423357609113315, "learning_rate": 6.731047807861846e-05, "loss": 0.0008, "step": 4370 }, { "epoch": 3.5990139687756777, "grad_norm": 0.04608125535772885, "learning_rate": 6.714218023950323e-05, "loss": 0.0007, "step": 4380 }, { "epoch": 3.6072308956450287, "grad_norm": 0.03832215344417614, "learning_rate": 6.697366202715677e-05, "loss": 0.0007, "step": 4390 }, { "epoch": 3.61544782251438, "grad_norm": 0.04652649445328663, "learning_rate": 6.680492560798396e-05, "loss": 0.0007, "step": 4400 }, { "epoch": 3.6236647493837304, "grad_norm": 0.05305197226082353, "learning_rate": 6.663597315119487e-05, "loss": 0.0007, "step": 4410 }, { "epoch": 3.6318816762530814, "grad_norm": 0.05057609750718562, "learning_rate": 6.646680682877687e-05, "loss": 0.0008, "step": 4420 }, { "epoch": 3.640098603122432, "grad_norm": 0.04033944891093342, "learning_rate": 6.629742881546666e-05, "loss": 0.0007, "step": 4430 }, { "epoch": 3.648315529991783, "grad_norm": 0.04948102191122691, "learning_rate": 6.612784128872243e-05, "loss": 0.0008, "step": 4440 }, { "epoch": 3.656532456861134, "grad_norm": 0.047259959582598036, "learning_rate": 6.595804642869572e-05, "loss": 0.0007, "step": 4450 }, { "epoch": 3.6647493837304848, "grad_norm": 0.045113327814988856, "learning_rate": 6.578804641820352e-05, "loss": 0.0007, "step": 4460 }, { "epoch": 3.672966310599836, "grad_norm": 0.04670046591039699, "learning_rate": 6.561784344270014e-05, "loss": 0.0007, "step": 4470 }, { "epoch": 3.6811832374691864, "grad_norm": 0.046278939150937184, "learning_rate": 6.544743969024912e-05, "loss": 0.0007, "step": 4480 }, { "epoch": 3.6894001643385375, "grad_norm": 0.05534144137585495, "learning_rate": 6.527683735149512e-05, "loss": 0.0008, "step": 4490 }, { "epoch": 3.6976170912078885, "grad_norm": 0.05002450146642488, "learning_rate": 6.510603861963575e-05, "loss": 0.0007, "step": 4500 }, { "epoch": 3.705834018077239, "grad_norm": 0.043348809995981, "learning_rate": 6.493504569039336e-05, "loss": 0.0007, "step": 4510 }, { "epoch": 3.7140509449465897, "grad_norm": 0.05384704883851096, "learning_rate": 6.476386076198682e-05, "loss": 0.0008, "step": 4520 }, { "epoch": 3.7222678718159408, "grad_norm": 0.04234911358976624, "learning_rate": 6.459248603510331e-05, "loss": 0.0007, "step": 4530 }, { "epoch": 3.730484798685292, "grad_norm": 0.049820373871424475, "learning_rate": 6.442092371286996e-05, "loss": 0.0007, "step": 4540 }, { "epoch": 3.7387017255546424, "grad_norm": 0.03967221454712835, "learning_rate": 6.424917600082551e-05, "loss": 0.0006, "step": 4550 }, { "epoch": 3.7469186524239935, "grad_norm": 0.040322207216051634, "learning_rate": 6.407724510689208e-05, "loss": 0.0007, "step": 4560 }, { "epoch": 3.755135579293344, "grad_norm": 0.04706651146485133, "learning_rate": 6.390513324134666e-05, "loss": 0.0007, "step": 4570 }, { "epoch": 3.763352506162695, "grad_norm": 0.045706719059608596, "learning_rate": 6.373284261679274e-05, "loss": 0.0007, "step": 4580 }, { "epoch": 3.771569433032046, "grad_norm": 0.044021060463565076, "learning_rate": 6.356037544813186e-05, "loss": 0.0007, "step": 4590 }, { "epoch": 3.7797863599013968, "grad_norm": 0.03989425500166346, "learning_rate": 6.338773395253518e-05, "loss": 0.0006, "step": 4600 }, { "epoch": 3.788003286770748, "grad_norm": 0.040615858297838967, "learning_rate": 6.32149203494149e-05, "loss": 0.0007, "step": 4610 }, { "epoch": 3.7962202136400984, "grad_norm": 0.041885517563488475, "learning_rate": 6.30419368603958e-05, "loss": 0.0006, "step": 4620 }, { "epoch": 3.8044371405094495, "grad_norm": 0.036534009161008255, "learning_rate": 6.286878570928661e-05, "loss": 0.0007, "step": 4630 }, { "epoch": 3.8126540673788005, "grad_norm": 0.04338387181621446, "learning_rate": 6.26954691220515e-05, "loss": 0.0007, "step": 4640 }, { "epoch": 3.820870994248151, "grad_norm": 0.046759900017018734, "learning_rate": 6.252198932678139e-05, "loss": 0.0007, "step": 4650 }, { "epoch": 3.829087921117502, "grad_norm": 0.05082785353366562, "learning_rate": 6.234834855366536e-05, "loss": 0.0007, "step": 4660 }, { "epoch": 3.837304847986853, "grad_norm": 0.052059100017410494, "learning_rate": 6.217454903496194e-05, "loss": 0.0007, "step": 4670 }, { "epoch": 3.845521774856204, "grad_norm": 0.042815478981559, "learning_rate": 6.200059300497045e-05, "loss": 0.0007, "step": 4680 }, { "epoch": 3.853738701725555, "grad_norm": 0.050710592684603775, "learning_rate": 6.182648270000226e-05, "loss": 0.0007, "step": 4690 }, { "epoch": 3.8619556285949055, "grad_norm": 0.0425296848360912, "learning_rate": 6.1652220358352e-05, "loss": 0.0007, "step": 4700 }, { "epoch": 3.870172555464256, "grad_norm": 0.05028111538463843, "learning_rate": 6.147780822026888e-05, "loss": 0.0007, "step": 4710 }, { "epoch": 3.878389482333607, "grad_norm": 0.05054916636835842, "learning_rate": 6.13032485279278e-05, "loss": 0.0007, "step": 4720 }, { "epoch": 3.886606409202958, "grad_norm": 0.04559127850433833, "learning_rate": 6.112854352540058e-05, "loss": 0.0007, "step": 4730 }, { "epoch": 3.894823336072309, "grad_norm": 0.04285740509309393, "learning_rate": 6.095369545862708e-05, "loss": 0.0006, "step": 4740 }, { "epoch": 3.90304026294166, "grad_norm": 0.031659359638648216, "learning_rate": 6.0778706575386335e-05, "loss": 0.0007, "step": 4750 }, { "epoch": 3.9112571898110104, "grad_norm": 0.047331703200472405, "learning_rate": 6.060357912526766e-05, "loss": 0.0006, "step": 4760 }, { "epoch": 3.9194741166803615, "grad_norm": 0.04437034098908951, "learning_rate": 6.042831535964177e-05, "loss": 0.0007, "step": 4770 }, { "epoch": 3.9276910435497125, "grad_norm": 0.041788078569932614, "learning_rate": 6.025291753163174e-05, "loss": 0.0006, "step": 4780 }, { "epoch": 3.935907970419063, "grad_norm": 0.04064511384332702, "learning_rate": 6.007738789608414e-05, "loss": 0.0006, "step": 4790 }, { "epoch": 3.944124897288414, "grad_norm": 0.04516380903934509, "learning_rate": 5.990172870953997e-05, "loss": 0.0006, "step": 4800 }, { "epoch": 3.952341824157765, "grad_norm": 0.049161850675267314, "learning_rate": 5.972594223020576e-05, "loss": 0.0006, "step": 4810 }, { "epoch": 3.960558751027116, "grad_norm": 0.04236527013634945, "learning_rate": 5.955003071792438e-05, "loss": 0.0006, "step": 4820 }, { "epoch": 3.968775677896467, "grad_norm": 0.04471094991450066, "learning_rate": 5.937399643414611e-05, "loss": 0.0006, "step": 4830 }, { "epoch": 3.9769926047658175, "grad_norm": 0.03309602903316914, "learning_rate": 5.9197841641899575e-05, "loss": 0.0006, "step": 4840 }, { "epoch": 3.9852095316351686, "grad_norm": 0.037352264298995394, "learning_rate": 5.9021568605762534e-05, "loss": 0.0007, "step": 4850 }, { "epoch": 3.993426458504519, "grad_norm": 0.04838730112539713, "learning_rate": 5.884517959183291e-05, "loss": 0.0007, "step": 4860 }, { "epoch": 4.00164338537387, "grad_norm": 0.03309423486511311, "learning_rate": 5.8668676867699545e-05, "loss": 0.0006, "step": 4870 }, { "epoch": 4.009860312243221, "grad_norm": 0.03742795440763105, "learning_rate": 5.849206270241313e-05, "loss": 0.0004, "step": 4880 }, { "epoch": 4.018077239112572, "grad_norm": 0.025982402783236713, "learning_rate": 5.831533936645697e-05, "loss": 0.0004, "step": 4890 }, { "epoch": 4.0262941659819225, "grad_norm": 0.021173652969151354, "learning_rate": 5.813850913171782e-05, "loss": 0.0003, "step": 4900 }, { "epoch": 4.0345110928512735, "grad_norm": 0.03884524153412678, "learning_rate": 5.7961574271456766e-05, "loss": 0.0003, "step": 4910 }, { "epoch": 4.042728019720625, "grad_norm": 0.03237193231555927, "learning_rate": 5.778453706027979e-05, "loss": 0.0003, "step": 4920 }, { "epoch": 4.050944946589976, "grad_norm": 0.027744828422386958, "learning_rate": 5.7607399774108775e-05, "loss": 0.0003, "step": 4930 }, { "epoch": 4.059161873459326, "grad_norm": 0.02209600752156913, "learning_rate": 5.743016469015204e-05, "loss": 0.0003, "step": 4940 }, { "epoch": 4.067378800328677, "grad_norm": 0.0246809631719301, "learning_rate": 5.725283408687524e-05, "loss": 0.0003, "step": 4950 }, { "epoch": 4.075595727198028, "grad_norm": 0.03253094321739853, "learning_rate": 5.7075410243971896e-05, "loss": 0.0003, "step": 4960 }, { "epoch": 4.083812654067379, "grad_norm": 0.03526592397808173, "learning_rate": 5.689789544233425e-05, "loss": 0.0003, "step": 4970 }, { "epoch": 4.09202958093673, "grad_norm": 0.03315662111797274, "learning_rate": 5.6720291964023866e-05, "loss": 0.0003, "step": 4980 }, { "epoch": 4.10024650780608, "grad_norm": 0.029908518972376952, "learning_rate": 5.654260209224225e-05, "loss": 0.0003, "step": 4990 }, { "epoch": 4.108463434675431, "grad_norm": 0.017966906894539586, "learning_rate": 5.6364828111301636e-05, "loss": 0.0003, "step": 5000 }, { "epoch": 4.116680361544782, "grad_norm": 0.031733965598610296, "learning_rate": 5.618697230659545e-05, "loss": 0.0003, "step": 5010 }, { "epoch": 4.124897288414133, "grad_norm": 0.02796437906615926, "learning_rate": 5.600903696456907e-05, "loss": 0.0003, "step": 5020 }, { "epoch": 4.133114215283484, "grad_norm": 0.028540105491421153, "learning_rate": 5.583102437269032e-05, "loss": 0.0003, "step": 5030 }, { "epoch": 4.1413311421528345, "grad_norm": 0.019745798345455003, "learning_rate": 5.565293681942018e-05, "loss": 0.0003, "step": 5040 }, { "epoch": 4.1495480690221855, "grad_norm": 0.020293421667175984, "learning_rate": 5.5474776594183266e-05, "loss": 0.0002, "step": 5050 }, { "epoch": 4.157764995891537, "grad_norm": 0.01983241518804618, "learning_rate": 5.5296545987338423e-05, "loss": 0.0003, "step": 5060 }, { "epoch": 4.165981922760888, "grad_norm": 0.02538247416002366, "learning_rate": 5.511824729014936e-05, "loss": 0.0003, "step": 5070 }, { "epoch": 4.174198849630239, "grad_norm": 0.02400961391590281, "learning_rate": 5.493988279475504e-05, "loss": 0.0003, "step": 5080 }, { "epoch": 4.182415776499589, "grad_norm": 0.0241979191640819, "learning_rate": 5.476145479414039e-05, "loss": 0.0003, "step": 5090 }, { "epoch": 4.19063270336894, "grad_norm": 0.030998754525850693, "learning_rate": 5.458296558210665e-05, "loss": 0.0003, "step": 5100 }, { "epoch": 4.198849630238291, "grad_norm": 0.023175684797061263, "learning_rate": 5.440441745324205e-05, "loss": 0.0002, "step": 5110 }, { "epoch": 4.207066557107642, "grad_norm": 0.032065988331051926, "learning_rate": 5.422581270289219e-05, "loss": 0.0003, "step": 5120 }, { "epoch": 4.215283483976993, "grad_norm": 0.021129273846021066, "learning_rate": 5.4047153627130554e-05, "loss": 0.0003, "step": 5130 }, { "epoch": 4.223500410846343, "grad_norm": 0.027912163959137346, "learning_rate": 5.386844252272909e-05, "loss": 0.0003, "step": 5140 }, { "epoch": 4.231717337715694, "grad_norm": 0.022534689172164276, "learning_rate": 5.368968168712849e-05, "loss": 0.0003, "step": 5150 }, { "epoch": 4.239934264585045, "grad_norm": 0.030348531604707404, "learning_rate": 5.351087341840888e-05, "loss": 0.0003, "step": 5160 }, { "epoch": 4.248151191454396, "grad_norm": 0.028593431069463687, "learning_rate": 5.3332020015260096e-05, "loss": 0.0003, "step": 5170 }, { "epoch": 4.2563681183237465, "grad_norm": 0.027769940229289972, "learning_rate": 5.315312377695226e-05, "loss": 0.0003, "step": 5180 }, { "epoch": 4.264585045193098, "grad_norm": 0.03885996205827739, "learning_rate": 5.297418700330612e-05, "loss": 0.0003, "step": 5190 }, { "epoch": 4.272801972062449, "grad_norm": 0.028923939507343847, "learning_rate": 5.279521199466356e-05, "loss": 0.0003, "step": 5200 }, { "epoch": 4.2810188989318, "grad_norm": 0.027560956736764744, "learning_rate": 5.2616201051857975e-05, "loss": 0.0003, "step": 5210 }, { "epoch": 4.289235825801151, "grad_norm": 0.027556840470268764, "learning_rate": 5.243715647618473e-05, "loss": 0.0003, "step": 5220 }, { "epoch": 4.297452752670501, "grad_norm": 0.03381762969765711, "learning_rate": 5.225808056937156e-05, "loss": 0.0003, "step": 5230 }, { "epoch": 4.305669679539852, "grad_norm": 0.02539113935234555, "learning_rate": 5.2078975633548985e-05, "loss": 0.0003, "step": 5240 }, { "epoch": 4.313886606409203, "grad_norm": 0.026921250605595987, "learning_rate": 5.189984397122071e-05, "loss": 0.0003, "step": 5250 }, { "epoch": 4.322103533278554, "grad_norm": 0.025535227987355257, "learning_rate": 5.1720687885233996e-05, "loss": 0.0003, "step": 5260 }, { "epoch": 4.330320460147905, "grad_norm": 0.026774168094902517, "learning_rate": 5.1541509678750146e-05, "loss": 0.0003, "step": 5270 }, { "epoch": 4.338537387017255, "grad_norm": 0.028123957146302097, "learning_rate": 5.1362311655214765e-05, "loss": 0.0003, "step": 5280 }, { "epoch": 4.346754313886606, "grad_norm": 0.03153316274601846, "learning_rate": 5.118309611832828e-05, "loss": 0.0003, "step": 5290 }, { "epoch": 4.354971240755957, "grad_norm": 0.02906558723351918, "learning_rate": 5.100386537201621e-05, "loss": 0.0003, "step": 5300 }, { "epoch": 4.363188167625308, "grad_norm": 0.039144025680844775, "learning_rate": 5.082462172039966e-05, "loss": 0.0003, "step": 5310 }, { "epoch": 4.3714050944946585, "grad_norm": 0.030066461511613133, "learning_rate": 5.064536746776558e-05, "loss": 0.0003, "step": 5320 }, { "epoch": 4.37962202136401, "grad_norm": 0.027128262836743043, "learning_rate": 5.046610491853724e-05, "loss": 0.0003, "step": 5330 }, { "epoch": 4.387838948233361, "grad_norm": 0.03063359358812782, "learning_rate": 5.028683637724455e-05, "loss": 0.0003, "step": 5340 }, { "epoch": 4.396055875102712, "grad_norm": 0.028069030154914642, "learning_rate": 5.010756414849448e-05, "loss": 0.0003, "step": 5350 }, { "epoch": 4.404272801972063, "grad_norm": 0.032991717867976986, "learning_rate": 4.9928290536941355e-05, "loss": 0.0003, "step": 5360 }, { "epoch": 4.412489728841413, "grad_norm": 0.03563727981699905, "learning_rate": 4.9749017847257324e-05, "loss": 0.0003, "step": 5370 }, { "epoch": 4.420706655710764, "grad_norm": 0.026321403574470093, "learning_rate": 4.956974838410265e-05, "loss": 0.0003, "step": 5380 }, { "epoch": 4.428923582580115, "grad_norm": 0.029257997020529145, "learning_rate": 4.939048445209617e-05, "loss": 0.0003, "step": 5390 }, { "epoch": 4.437140509449466, "grad_norm": 0.025951152011347074, "learning_rate": 4.921122835578554e-05, "loss": 0.0002, "step": 5400 }, { "epoch": 4.445357436318817, "grad_norm": 0.022790767182579535, "learning_rate": 4.9031982399617715e-05, "loss": 0.0003, "step": 5410 }, { "epoch": 4.453574363188167, "grad_norm": 0.027285148692887937, "learning_rate": 4.885274888790936e-05, "loss": 0.0003, "step": 5420 }, { "epoch": 4.461791290057518, "grad_norm": 0.02796363692133071, "learning_rate": 4.8673530124817016e-05, "loss": 0.0003, "step": 5430 }, { "epoch": 4.470008216926869, "grad_norm": 0.01983742893775883, "learning_rate": 4.849432841430775e-05, "loss": 0.0003, "step": 5440 }, { "epoch": 4.47822514379622, "grad_norm": 0.017903281569223436, "learning_rate": 4.831514606012935e-05, "loss": 0.0003, "step": 5450 }, { "epoch": 4.486442070665571, "grad_norm": 0.035804302669490556, "learning_rate": 4.81359853657808e-05, "loss": 0.0003, "step": 5460 }, { "epoch": 4.494658997534922, "grad_norm": 0.020900607814891756, "learning_rate": 4.795684863448258e-05, "loss": 0.0003, "step": 5470 }, { "epoch": 4.502875924404273, "grad_norm": 0.02734038794819125, "learning_rate": 4.777773816914713e-05, "loss": 0.0003, "step": 5480 }, { "epoch": 4.511092851273624, "grad_norm": 0.029929425774254854, "learning_rate": 4.75986562723493e-05, "loss": 0.0003, "step": 5490 }, { "epoch": 4.519309778142975, "grad_norm": 0.02959560982639449, "learning_rate": 4.741960524629652e-05, "loss": 0.0003, "step": 5500 }, { "epoch": 4.527526705012326, "grad_norm": 0.03533501309766263, "learning_rate": 4.7240587392799516e-05, "loss": 0.0003, "step": 5510 }, { "epoch": 4.535743631881676, "grad_norm": 0.034782134547894486, "learning_rate": 4.7061605013242456e-05, "loss": 0.0003, "step": 5520 }, { "epoch": 4.543960558751027, "grad_norm": 0.024237927881660685, "learning_rate": 4.6882660408553536e-05, "loss": 0.0003, "step": 5530 }, { "epoch": 4.552177485620378, "grad_norm": 0.026262804022003455, "learning_rate": 4.6703755879175275e-05, "loss": 0.0003, "step": 5540 }, { "epoch": 4.560394412489729, "grad_norm": 0.022867261679024517, "learning_rate": 4.652489372503502e-05, "loss": 0.0003, "step": 5550 }, { "epoch": 4.56861133935908, "grad_norm": 0.032071903571258976, "learning_rate": 4.6346076245515394e-05, "loss": 0.0003, "step": 5560 }, { "epoch": 4.57682826622843, "grad_norm": 0.03709091220841175, "learning_rate": 4.616730573942467e-05, "loss": 0.0003, "step": 5570 }, { "epoch": 4.585045193097781, "grad_norm": 0.03497683584374609, "learning_rate": 4.598858450496722e-05, "loss": 0.0003, "step": 5580 }, { "epoch": 4.593262119967132, "grad_norm": 0.02911729572375434, "learning_rate": 4.580991483971403e-05, "loss": 0.0003, "step": 5590 }, { "epoch": 4.6014790468364835, "grad_norm": 0.028431830767967015, "learning_rate": 4.563129904057317e-05, "loss": 0.0003, "step": 5600 }, { "epoch": 4.6096959737058345, "grad_norm": 0.020467190529052527, "learning_rate": 4.5452739403760155e-05, "loss": 0.0003, "step": 5610 }, { "epoch": 4.617912900575185, "grad_norm": 0.032931557060531975, "learning_rate": 4.527423822476852e-05, "loss": 0.0003, "step": 5620 }, { "epoch": 4.626129827444536, "grad_norm": 0.024062112966267334, "learning_rate": 4.509579779834034e-05, "loss": 0.0003, "step": 5630 }, { "epoch": 4.634346754313887, "grad_norm": 0.03455239635353224, "learning_rate": 4.4917420418436627e-05, "loss": 0.0003, "step": 5640 }, { "epoch": 4.642563681183238, "grad_norm": 0.024975479976678344, "learning_rate": 4.47391083782079e-05, "loss": 0.0003, "step": 5650 }, { "epoch": 4.650780608052588, "grad_norm": 0.026189212834599903, "learning_rate": 4.456086396996467e-05, "loss": 0.0003, "step": 5660 }, { "epoch": 4.658997534921939, "grad_norm": 0.025344352338761436, "learning_rate": 4.4382689485148096e-05, "loss": 0.0003, "step": 5670 }, { "epoch": 4.66721446179129, "grad_norm": 0.026077309985734515, "learning_rate": 4.420458721430031e-05, "loss": 0.0003, "step": 5680 }, { "epoch": 4.675431388660641, "grad_norm": 0.027805581781903038, "learning_rate": 4.402655944703514e-05, "loss": 0.0003, "step": 5690 }, { "epoch": 4.683648315529992, "grad_norm": 0.03597635044226168, "learning_rate": 4.384860847200864e-05, "loss": 0.0003, "step": 5700 }, { "epoch": 4.691865242399342, "grad_norm": 0.022843678703533255, "learning_rate": 4.367073657688966e-05, "loss": 0.0003, "step": 5710 }, { "epoch": 4.700082169268693, "grad_norm": 0.039677810254041504, "learning_rate": 4.349294604833037e-05, "loss": 0.0003, "step": 5720 }, { "epoch": 4.708299096138044, "grad_norm": 0.032277815769057784, "learning_rate": 4.3315239171936956e-05, "loss": 0.0003, "step": 5730 }, { "epoch": 4.7165160230073955, "grad_norm": 0.02985009639251006, "learning_rate": 4.313761823224023e-05, "loss": 0.0003, "step": 5740 }, { "epoch": 4.724732949876746, "grad_norm": 0.026711206009289175, "learning_rate": 4.2960085512666225e-05, "loss": 0.0003, "step": 5750 }, { "epoch": 4.732949876746097, "grad_norm": 0.03745476456598732, "learning_rate": 4.2782643295506794e-05, "loss": 0.0003, "step": 5760 }, { "epoch": 4.741166803615448, "grad_norm": 0.031888981953617224, "learning_rate": 4.260529386189041e-05, "loss": 0.0003, "step": 5770 }, { "epoch": 4.749383730484799, "grad_norm": 0.03325098063833695, "learning_rate": 4.242803949175271e-05, "loss": 0.0003, "step": 5780 }, { "epoch": 4.75760065735415, "grad_norm": 0.02526571952387307, "learning_rate": 4.225088246380723e-05, "loss": 0.0003, "step": 5790 }, { "epoch": 4.7658175842235, "grad_norm": 0.028678235118990182, "learning_rate": 4.20738250555161e-05, "loss": 0.0003, "step": 5800 }, { "epoch": 4.774034511092851, "grad_norm": 0.02505231680442755, "learning_rate": 4.1896869543060846e-05, "loss": 0.0003, "step": 5810 }, { "epoch": 4.782251437962202, "grad_norm": 0.03602293451312027, "learning_rate": 4.1720018201312995e-05, "loss": 0.0003, "step": 5820 }, { "epoch": 4.790468364831553, "grad_norm": 0.025784019344654737, "learning_rate": 4.1543273303804886e-05, "loss": 0.0003, "step": 5830 }, { "epoch": 4.798685291700904, "grad_norm": 0.02836466583928212, "learning_rate": 4.136663712270052e-05, "loss": 0.0003, "step": 5840 }, { "epoch": 4.806902218570254, "grad_norm": 0.029594133987557485, "learning_rate": 4.119011192876624e-05, "loss": 0.0003, "step": 5850 }, { "epoch": 4.815119145439605, "grad_norm": 0.05062326191619077, "learning_rate": 4.101369999134156e-05, "loss": 0.0003, "step": 5860 }, { "epoch": 4.8233360723089564, "grad_norm": 0.024132562649026783, "learning_rate": 4.083740357831003e-05, "loss": 0.0003, "step": 5870 }, { "epoch": 4.8315529991783075, "grad_norm": 0.04509656173413742, "learning_rate": 4.0661224956070105e-05, "loss": 0.0004, "step": 5880 }, { "epoch": 4.8397699260476585, "grad_norm": 0.02815893869070155, "learning_rate": 4.048516638950592e-05, "loss": 0.0003, "step": 5890 }, { "epoch": 4.847986852917009, "grad_norm": 0.0336949207238929, "learning_rate": 4.030923014195821e-05, "loss": 0.0004, "step": 5900 }, { "epoch": 4.85620377978636, "grad_norm": 0.058924540187972, "learning_rate": 4.013341847519527e-05, "loss": 0.0003, "step": 5910 }, { "epoch": 4.864420706655711, "grad_norm": 0.02642943565735785, "learning_rate": 3.995773364938378e-05, "loss": 0.0003, "step": 5920 }, { "epoch": 4.872637633525062, "grad_norm": 0.024788158649896117, "learning_rate": 3.978217792305989e-05, "loss": 0.0003, "step": 5930 }, { "epoch": 4.880854560394413, "grad_norm": 0.03413158489002041, "learning_rate": 3.960675355309997e-05, "loss": 0.0003, "step": 5940 }, { "epoch": 4.889071487263763, "grad_norm": 0.029517898034993655, "learning_rate": 3.9431462794691827e-05, "loss": 0.0003, "step": 5950 }, { "epoch": 4.897288414133114, "grad_norm": 0.03679038258970327, "learning_rate": 3.925630790130558e-05, "loss": 0.0003, "step": 5960 }, { "epoch": 4.905505341002465, "grad_norm": 0.0338284105145708, "learning_rate": 3.908129112466467e-05, "loss": 0.0003, "step": 5970 }, { "epoch": 4.913722267871816, "grad_norm": 0.02331916062018235, "learning_rate": 3.890641471471705e-05, "loss": 0.0003, "step": 5980 }, { "epoch": 4.921939194741167, "grad_norm": 0.030219349998636718, "learning_rate": 3.8731680919606106e-05, "loss": 0.0003, "step": 5990 }, { "epoch": 4.930156121610517, "grad_norm": 0.036087832739331166, "learning_rate": 3.855709198564189e-05, "loss": 0.0003, "step": 6000 }, { "epoch": 4.9383730484798685, "grad_norm": 0.039606549405370325, "learning_rate": 3.838265015727206e-05, "loss": 0.0003, "step": 6010 }, { "epoch": 4.9465899753492195, "grad_norm": 0.04432870723931057, "learning_rate": 3.820835767705326e-05, "loss": 0.0003, "step": 6020 }, { "epoch": 4.954806902218571, "grad_norm": 0.02085758915110865, "learning_rate": 3.803421678562213e-05, "loss": 0.0003, "step": 6030 }, { "epoch": 4.963023829087922, "grad_norm": 0.02945409632034805, "learning_rate": 3.7860229721666494e-05, "loss": 0.0003, "step": 6040 }, { "epoch": 4.971240755957272, "grad_norm": 0.022809635896518556, "learning_rate": 3.768639872189672e-05, "loss": 0.0003, "step": 6050 }, { "epoch": 4.979457682826623, "grad_norm": 0.025470458484121158, "learning_rate": 3.7512726021016794e-05, "loss": 0.0003, "step": 6060 }, { "epoch": 4.987674609695974, "grad_norm": 0.026826519727473037, "learning_rate": 3.733921385169576e-05, "loss": 0.0003, "step": 6070 }, { "epoch": 4.995891536565325, "grad_norm": 0.031086590973493294, "learning_rate": 3.7165864444538806e-05, "loss": 0.0003, "step": 6080 }, { "epoch": 5.004108463434675, "grad_norm": 0.02171367098466048, "learning_rate": 3.6992680028058835e-05, "loss": 0.0002, "step": 6090 }, { "epoch": 5.012325390304026, "grad_norm": 0.0199100186189244, "learning_rate": 3.681966282864762e-05, "loss": 0.0002, "step": 6100 }, { "epoch": 5.020542317173377, "grad_norm": 0.010549471164456866, "learning_rate": 3.664681507054731e-05, "loss": 0.0002, "step": 6110 }, { "epoch": 5.028759244042728, "grad_norm": 0.025671768949555553, "learning_rate": 3.6474138975821726e-05, "loss": 0.0002, "step": 6120 }, { "epoch": 5.036976170912079, "grad_norm": 0.021566137252671754, "learning_rate": 3.630163676432786e-05, "loss": 0.0002, "step": 6130 }, { "epoch": 5.045193097781429, "grad_norm": 0.020977756350381693, "learning_rate": 3.612931065368741e-05, "loss": 0.0001, "step": 6140 }, { "epoch": 5.0534100246507805, "grad_norm": 0.012086958356686271, "learning_rate": 3.595716285925804e-05, "loss": 0.0002, "step": 6150 }, { "epoch": 5.0616269515201315, "grad_norm": 0.018026330095533824, "learning_rate": 3.57851955941052e-05, "loss": 0.0001, "step": 6160 }, { "epoch": 5.069843878389483, "grad_norm": 0.02459035481386552, "learning_rate": 3.5613411068973416e-05, "loss": 0.0002, "step": 6170 }, { "epoch": 5.078060805258834, "grad_norm": 0.024590141539745643, "learning_rate": 3.544181149225807e-05, "loss": 0.0002, "step": 6180 }, { "epoch": 5.086277732128184, "grad_norm": 0.010664808178114226, "learning_rate": 3.527039906997684e-05, "loss": 0.0002, "step": 6190 }, { "epoch": 5.094494658997535, "grad_norm": 0.012795118285109614, "learning_rate": 3.5099176005741445e-05, "loss": 0.0002, "step": 6200 }, { "epoch": 5.102711585866886, "grad_norm": 0.018494555520494727, "learning_rate": 3.492814450072934e-05, "loss": 0.0001, "step": 6210 }, { "epoch": 5.110928512736237, "grad_norm": 0.016335021691850356, "learning_rate": 3.4757306753655266e-05, "loss": 0.0001, "step": 6220 }, { "epoch": 5.119145439605587, "grad_norm": 0.007710298681706114, "learning_rate": 3.4586664960743184e-05, "loss": 0.0001, "step": 6230 }, { "epoch": 5.127362366474938, "grad_norm": 0.007400362534347697, "learning_rate": 3.441622131569789e-05, "loss": 0.0001, "step": 6240 }, { "epoch": 5.135579293344289, "grad_norm": 0.015092907858013422, "learning_rate": 3.424597800967691e-05, "loss": 0.0001, "step": 6250 }, { "epoch": 5.14379622021364, "grad_norm": 0.021106434532348384, "learning_rate": 3.407593723126223e-05, "loss": 0.0001, "step": 6260 }, { "epoch": 5.152013147082991, "grad_norm": 0.026164696919490152, "learning_rate": 3.390610116643227e-05, "loss": 0.0001, "step": 6270 }, { "epoch": 5.1602300739523415, "grad_norm": 0.01644369157557952, "learning_rate": 3.373647199853371e-05, "loss": 0.0001, "step": 6280 }, { "epoch": 5.1684470008216925, "grad_norm": 0.01612265741922029, "learning_rate": 3.3567051908253465e-05, "loss": 0.0001, "step": 6290 }, { "epoch": 5.1766639276910436, "grad_norm": 0.01174062218540938, "learning_rate": 3.3397843073590593e-05, "loss": 0.0001, "step": 6300 }, { "epoch": 5.184880854560395, "grad_norm": 0.02457439723718392, "learning_rate": 3.322884766982833e-05, "loss": 0.0001, "step": 6310 }, { "epoch": 5.193097781429746, "grad_norm": 0.015143624713557357, "learning_rate": 3.3060067869506204e-05, "loss": 0.0001, "step": 6320 }, { "epoch": 5.201314708299096, "grad_norm": 0.026444307119181755, "learning_rate": 3.289150584239192e-05, "loss": 0.0001, "step": 6330 }, { "epoch": 5.209531635168447, "grad_norm": 0.016226173213306588, "learning_rate": 3.272316375545363e-05, "loss": 0.0001, "step": 6340 }, { "epoch": 5.217748562037798, "grad_norm": 0.01427924067871283, "learning_rate": 3.255504377283205e-05, "loss": 0.0001, "step": 6350 }, { "epoch": 5.225965488907149, "grad_norm": 0.010231911350809125, "learning_rate": 3.2387148055812564e-05, "loss": 0.0001, "step": 6360 }, { "epoch": 5.2341824157765, "grad_norm": 0.01124171387137202, "learning_rate": 3.221947876279749e-05, "loss": 0.0001, "step": 6370 }, { "epoch": 5.24239934264585, "grad_norm": 0.00925846446332683, "learning_rate": 3.205203804927831e-05, "loss": 0.0001, "step": 6380 }, { "epoch": 5.250616269515201, "grad_norm": 0.021547131349480917, "learning_rate": 3.188482806780805e-05, "loss": 0.0001, "step": 6390 }, { "epoch": 5.258833196384552, "grad_norm": 0.01642639962260317, "learning_rate": 3.1717850967973436e-05, "loss": 0.0001, "step": 6400 }, { "epoch": 5.267050123253903, "grad_norm": 0.010649014704981072, "learning_rate": 3.155110889636743e-05, "loss": 0.0001, "step": 6410 }, { "epoch": 5.275267050123254, "grad_norm": 0.02444831687936009, "learning_rate": 3.1384603996561536e-05, "loss": 0.0001, "step": 6420 }, { "epoch": 5.2834839769926045, "grad_norm": 0.013939335373189869, "learning_rate": 3.121833840907827e-05, "loss": 0.0001, "step": 6430 }, { "epoch": 5.291700903861956, "grad_norm": 0.006439883255854533, "learning_rate": 3.105231427136364e-05, "loss": 0.0001, "step": 6440 }, { "epoch": 5.299917830731307, "grad_norm": 0.01819304641838225, "learning_rate": 3.0886533717759646e-05, "loss": 0.0001, "step": 6450 }, { "epoch": 5.308134757600658, "grad_norm": 0.019797807971433386, "learning_rate": 3.072099887947694e-05, "loss": 0.0001, "step": 6460 }, { "epoch": 5.316351684470008, "grad_norm": 0.022819589732322836, "learning_rate": 3.0555711884567274e-05, "loss": 0.0001, "step": 6470 }, { "epoch": 5.324568611339359, "grad_norm": 0.007610969515867313, "learning_rate": 3.039067485789624e-05, "loss": 0.0001, "step": 6480 }, { "epoch": 5.33278553820871, "grad_norm": 0.008902363401491229, "learning_rate": 3.0225889921115968e-05, "loss": 0.0001, "step": 6490 }, { "epoch": 5.341002465078061, "grad_norm": 0.011504943855497104, "learning_rate": 3.0061359192637807e-05, "loss": 0.0001, "step": 6500 }, { "epoch": 5.349219391947412, "grad_norm": 0.013216583630589683, "learning_rate": 2.9897084787605056e-05, "loss": 0.0001, "step": 6510 }, { "epoch": 5.357436318816762, "grad_norm": 0.016425876279932708, "learning_rate": 2.9733068817865867e-05, "loss": 0.0001, "step": 6520 }, { "epoch": 5.365653245686113, "grad_norm": 0.02110279530395085, "learning_rate": 2.9569313391946045e-05, "loss": 0.0001, "step": 6530 }, { "epoch": 5.373870172555464, "grad_norm": 0.013917973430989765, "learning_rate": 2.9405820615021946e-05, "loss": 0.0001, "step": 6540 }, { "epoch": 5.382087099424815, "grad_norm": 0.013245214872098964, "learning_rate": 2.924259258889336e-05, "loss": 0.0001, "step": 6550 }, { "epoch": 5.390304026294166, "grad_norm": 0.014348036307614406, "learning_rate": 2.9079631411956622e-05, "loss": 0.0001, "step": 6560 }, { "epoch": 5.3985209531635165, "grad_norm": 0.0121919105192544, "learning_rate": 2.8916939179177525e-05, "loss": 0.0001, "step": 6570 }, { "epoch": 5.406737880032868, "grad_norm": 0.018793908388556848, "learning_rate": 2.8754517982064365e-05, "loss": 0.0001, "step": 6580 }, { "epoch": 5.414954806902219, "grad_norm": 0.014823108811301932, "learning_rate": 2.8592369908641204e-05, "loss": 0.0001, "step": 6590 }, { "epoch": 5.42317173377157, "grad_norm": 0.022861883032126724, "learning_rate": 2.8430497043420874e-05, "loss": 0.0001, "step": 6600 }, { "epoch": 5.43138866064092, "grad_norm": 0.012867674691088932, "learning_rate": 2.8268901467378244e-05, "loss": 0.0001, "step": 6610 }, { "epoch": 5.439605587510271, "grad_norm": 0.017690714893954716, "learning_rate": 2.8107585257923468e-05, "loss": 0.0001, "step": 6620 }, { "epoch": 5.447822514379622, "grad_norm": 0.01544144610793372, "learning_rate": 2.794655048887527e-05, "loss": 0.0001, "step": 6630 }, { "epoch": 5.456039441248973, "grad_norm": 0.022426486680758053, "learning_rate": 2.7785799230434285e-05, "loss": 0.0001, "step": 6640 }, { "epoch": 5.464256368118324, "grad_norm": 0.007384029816730038, "learning_rate": 2.762533354915644e-05, "loss": 0.0001, "step": 6650 }, { "epoch": 5.472473294987674, "grad_norm": 0.014851393638684198, "learning_rate": 2.74651555079264e-05, "loss": 0.0001, "step": 6660 }, { "epoch": 5.480690221857025, "grad_norm": 0.024573288804553103, "learning_rate": 2.730526716593103e-05, "loss": 0.0001, "step": 6670 }, { "epoch": 5.488907148726376, "grad_norm": 0.014476804382629892, "learning_rate": 2.7145670578632954e-05, "loss": 0.0001, "step": 6680 }, { "epoch": 5.497124075595727, "grad_norm": 0.008276073361119152, "learning_rate": 2.698636779774409e-05, "loss": 0.0001, "step": 6690 }, { "epoch": 5.505341002465078, "grad_norm": 0.02374055114997714, "learning_rate": 2.6827360871199303e-05, "loss": 0.0001, "step": 6700 }, { "epoch": 5.513557929334429, "grad_norm": 0.02943848202691088, "learning_rate": 2.666865184313007e-05, "loss": 0.0001, "step": 6710 }, { "epoch": 5.52177485620378, "grad_norm": 0.017560351023291697, "learning_rate": 2.65102427538382e-05, "loss": 0.0001, "step": 6720 }, { "epoch": 5.529991783073131, "grad_norm": 0.008399203272705309, "learning_rate": 2.6352135639769603e-05, "loss": 0.0001, "step": 6730 }, { "epoch": 5.538208709942482, "grad_norm": 0.022044056870370733, "learning_rate": 2.6194332533488118e-05, "loss": 0.0001, "step": 6740 }, { "epoch": 5.546425636811833, "grad_norm": 0.011673273123667663, "learning_rate": 2.6036835463649373e-05, "loss": 0.0001, "step": 6750 }, { "epoch": 5.554642563681183, "grad_norm": 0.010214166324477926, "learning_rate": 2.5879646454974703e-05, "loss": 0.0001, "step": 6760 }, { "epoch": 5.562859490550534, "grad_norm": 0.0213837906667468, "learning_rate": 2.5722767528225145e-05, "loss": 0.0001, "step": 6770 }, { "epoch": 5.571076417419885, "grad_norm": 0.007372533438218981, "learning_rate": 2.5566200700175424e-05, "loss": 0.0001, "step": 6780 }, { "epoch": 5.579293344289236, "grad_norm": 0.011313494087360994, "learning_rate": 2.5409947983588066e-05, "loss": 0.0001, "step": 6790 }, { "epoch": 5.587510271158587, "grad_norm": 0.023040456741824047, "learning_rate": 2.525401138718748e-05, "loss": 0.0001, "step": 6800 }, { "epoch": 5.595727198027937, "grad_norm": 0.016739522386311052, "learning_rate": 2.509839291563417e-05, "loss": 0.0001, "step": 6810 }, { "epoch": 5.603944124897288, "grad_norm": 0.013564068847468084, "learning_rate": 2.494309456949895e-05, "loss": 0.0001, "step": 6820 }, { "epoch": 5.612161051766639, "grad_norm": 0.010660712576528104, "learning_rate": 2.4788118345237216e-05, "loss": 0.0001, "step": 6830 }, { "epoch": 5.62037797863599, "grad_norm": 0.010784101069160784, "learning_rate": 2.46334662351633e-05, "loss": 0.0001, "step": 6840 }, { "epoch": 5.6285949055053415, "grad_norm": 0.02213237251284927, "learning_rate": 2.447914022742484e-05, "loss": 0.0001, "step": 6850 }, { "epoch": 5.636811832374692, "grad_norm": 0.008612140920933101, "learning_rate": 2.432514230597724e-05, "loss": 0.0001, "step": 6860 }, { "epoch": 5.645028759244043, "grad_norm": 0.02034265832772715, "learning_rate": 2.4171474450558135e-05, "loss": 0.0001, "step": 6870 }, { "epoch": 5.653245686113394, "grad_norm": 0.009204111443540561, "learning_rate": 2.401813863666198e-05, "loss": 0.0001, "step": 6880 }, { "epoch": 5.661462612982745, "grad_norm": 0.013511914811402799, "learning_rate": 2.386513683551462e-05, "loss": 0.0001, "step": 6890 }, { "epoch": 5.669679539852095, "grad_norm": 0.014607288531014027, "learning_rate": 2.3712471014047954e-05, "loss": 0.0001, "step": 6900 }, { "epoch": 5.677896466721446, "grad_norm": 0.018644940560811892, "learning_rate": 2.356014313487468e-05, "loss": 0.0001, "step": 6910 }, { "epoch": 5.686113393590797, "grad_norm": 0.00994720975542607, "learning_rate": 2.3408155156263017e-05, "loss": 0.0001, "step": 6920 }, { "epoch": 5.694330320460148, "grad_norm": 0.016054284348331855, "learning_rate": 2.3256509032111573e-05, "loss": 0.0001, "step": 6930 }, { "epoch": 5.702547247329499, "grad_norm": 0.00829429716545095, "learning_rate": 2.3105206711924192e-05, "loss": 0.0001, "step": 6940 }, { "epoch": 5.710764174198849, "grad_norm": 0.01735766208353996, "learning_rate": 2.2954250140784917e-05, "loss": 0.0001, "step": 6950 }, { "epoch": 5.7189811010682, "grad_norm": 0.01565638335164819, "learning_rate": 2.280364125933297e-05, "loss": 0.0001, "step": 6960 }, { "epoch": 5.727198027937551, "grad_norm": 0.012206027103081935, "learning_rate": 2.265338200373781e-05, "loss": 0.0001, "step": 6970 }, { "epoch": 5.735414954806902, "grad_norm": 0.010378249653357067, "learning_rate": 2.250347430567425e-05, "loss": 0.0001, "step": 6980 }, { "epoch": 5.743631881676253, "grad_norm": 0.006359873583217471, "learning_rate": 2.2353920092297597e-05, "loss": 0.0001, "step": 6990 }, { "epoch": 5.751848808545604, "grad_norm": 0.012002908695478095, "learning_rate": 2.220472128621891e-05, "loss": 0.0001, "step": 7000 }, { "epoch": 5.760065735414955, "grad_norm": 0.04637529779678336, "learning_rate": 2.205587980548027e-05, "loss": 0.0001, "step": 7010 }, { "epoch": 5.768282662284306, "grad_norm": 0.00459641901819906, "learning_rate": 2.1907397563530123e-05, "loss": 0.0001, "step": 7020 }, { "epoch": 5.776499589153657, "grad_norm": 0.011799335992636032, "learning_rate": 2.1759276469198663e-05, "loss": 0.0001, "step": 7030 }, { "epoch": 5.784716516023007, "grad_norm": 0.013842242357078047, "learning_rate": 2.161151842667333e-05, "loss": 0.0001, "step": 7040 }, { "epoch": 5.792933442892358, "grad_norm": 0.02223974163111052, "learning_rate": 2.1464125335474295e-05, "loss": 0.0001, "step": 7050 }, { "epoch": 5.801150369761709, "grad_norm": 0.01461479044325639, "learning_rate": 2.1317099090430064e-05, "loss": 0.0001, "step": 7060 }, { "epoch": 5.80936729663106, "grad_norm": 0.017247073463554623, "learning_rate": 2.1170441581653105e-05, "loss": 0.0001, "step": 7070 }, { "epoch": 5.817584223500411, "grad_norm": 0.010689827597451915, "learning_rate": 2.1024154694515607e-05, "loss": 0.0001, "step": 7080 }, { "epoch": 5.825801150369761, "grad_norm": 0.017038936691414295, "learning_rate": 2.087824030962509e-05, "loss": 0.0001, "step": 7090 }, { "epoch": 5.834018077239112, "grad_norm": 0.014829738225140464, "learning_rate": 2.07327003028004e-05, "loss": 0.0001, "step": 7100 }, { "epoch": 5.842235004108463, "grad_norm": 0.022955795752085145, "learning_rate": 2.0587536545047515e-05, "loss": 0.0001, "step": 7110 }, { "epoch": 5.8504519309778145, "grad_norm": 0.019818557639416843, "learning_rate": 2.044275090253549e-05, "loss": 0.0001, "step": 7120 }, { "epoch": 5.8586688578471655, "grad_norm": 0.013761719082181639, "learning_rate": 2.0298345236572475e-05, "loss": 0.0001, "step": 7130 }, { "epoch": 5.866885784716516, "grad_norm": 0.005655862914703428, "learning_rate": 2.015432140358178e-05, "loss": 0.0001, "step": 7140 }, { "epoch": 5.875102711585867, "grad_norm": 0.01585051429223647, "learning_rate": 2.0010681255078083e-05, "loss": 0.0001, "step": 7150 }, { "epoch": 5.883319638455218, "grad_norm": 0.023995255775545964, "learning_rate": 1.9867426637643465e-05, "loss": 0.0001, "step": 7160 }, { "epoch": 5.891536565324569, "grad_norm": 0.010990130346269344, "learning_rate": 1.972455939290383e-05, "loss": 0.0001, "step": 7170 }, { "epoch": 5.89975349219392, "grad_norm": 0.009793664805869804, "learning_rate": 1.958208135750515e-05, "loss": 0.0001, "step": 7180 }, { "epoch": 5.90797041906327, "grad_norm": 0.012032246379234486, "learning_rate": 1.9439994363089886e-05, "loss": 0.0001, "step": 7190 }, { "epoch": 5.916187345932621, "grad_norm": 0.007623975990021015, "learning_rate": 1.9298300236273404e-05, "loss": 0.0001, "step": 7200 }, { "epoch": 5.924404272801972, "grad_norm": 0.009607593038927106, "learning_rate": 1.9157000798620528e-05, "loss": 0.0001, "step": 7210 }, { "epoch": 5.932621199671323, "grad_norm": 0.01692880308322747, "learning_rate": 1.9016097866622157e-05, "loss": 0.0001, "step": 7220 }, { "epoch": 5.940838126540674, "grad_norm": 0.01741141289566327, "learning_rate": 1.8875593251671768e-05, "loss": 0.0001, "step": 7230 }, { "epoch": 5.949055053410024, "grad_norm": 0.02185383770745977, "learning_rate": 1.873548876004231e-05, "loss": 0.0001, "step": 7240 }, { "epoch": 5.957271980279375, "grad_norm": 0.008362862414980667, "learning_rate": 1.859578619286285e-05, "loss": 0.0001, "step": 7250 }, { "epoch": 5.9654889071487265, "grad_norm": 0.010982663113194705, "learning_rate": 1.8456487346095558e-05, "loss": 0.0001, "step": 7260 }, { "epoch": 5.9737058340180775, "grad_norm": 0.010797682137258725, "learning_rate": 1.83175940105124e-05, "loss": 0.0001, "step": 7270 }, { "epoch": 5.981922760887429, "grad_norm": 0.00875083688913143, "learning_rate": 1.8179107971672334e-05, "loss": 0.0001, "step": 7280 }, { "epoch": 5.990139687756779, "grad_norm": 0.012995937680644732, "learning_rate": 1.804103100989828e-05, "loss": 0.0001, "step": 7290 }, { "epoch": 5.99835661462613, "grad_norm": 0.004598613756243608, "learning_rate": 1.7903364900254155e-05, "loss": 0.0001, "step": 7300 }, { "epoch": 6.006573541495481, "grad_norm": 0.010315331393499121, "learning_rate": 1.7766111412522168e-05, "loss": 0.0001, "step": 7310 }, { "epoch": 6.014790468364832, "grad_norm": 0.002662479348045777, "learning_rate": 1.762927231117999e-05, "loss": 0.0001, "step": 7320 }, { "epoch": 6.023007395234182, "grad_norm": 0.0035447142433265733, "learning_rate": 1.7492849355378176e-05, "loss": 0.0001, "step": 7330 }, { "epoch": 6.031224322103533, "grad_norm": 0.012572897851314714, "learning_rate": 1.7356844298917368e-05, "loss": 0.0, "step": 7340 }, { "epoch": 6.039441248972884, "grad_norm": 0.0020264119715836773, "learning_rate": 1.7221258890225904e-05, "loss": 0.0001, "step": 7350 }, { "epoch": 6.047658175842235, "grad_norm": 0.0022216848164958183, "learning_rate": 1.7086094872337348e-05, "loss": 0.0001, "step": 7360 }, { "epoch": 6.055875102711586, "grad_norm": 0.00545679910896966, "learning_rate": 1.6951353982867924e-05, "loss": 0.0, "step": 7370 }, { "epoch": 6.064092029580936, "grad_norm": 0.002535609843230281, "learning_rate": 1.6817037953994346e-05, "loss": 0.0, "step": 7380 }, { "epoch": 6.0723089564502875, "grad_norm": 0.0024991492544328997, "learning_rate": 1.6683148512431458e-05, "loss": 0.0, "step": 7390 }, { "epoch": 6.0805258833196385, "grad_norm": 0.0026174380256037896, "learning_rate": 1.6549687379410116e-05, "loss": 0.0, "step": 7400 }, { "epoch": 6.0887428101889896, "grad_norm": 0.005531046520292157, "learning_rate": 1.6416656270654936e-05, "loss": 0.0, "step": 7410 }, { "epoch": 6.096959737058341, "grad_norm": 0.001996191226369207, "learning_rate": 1.6284056896362315e-05, "loss": 0.0, "step": 7420 }, { "epoch": 6.105176663927691, "grad_norm": 0.006346935940343153, "learning_rate": 1.6151890961178507e-05, "loss": 0.0, "step": 7430 }, { "epoch": 6.113393590797042, "grad_norm": 0.002676180284496888, "learning_rate": 1.60201601641776e-05, "loss": 0.0, "step": 7440 }, { "epoch": 6.121610517666393, "grad_norm": 0.014802165792258938, "learning_rate": 1.5888866198839673e-05, "loss": 0.0, "step": 7450 }, { "epoch": 6.129827444535744, "grad_norm": 0.007012198035783848, "learning_rate": 1.5758010753029107e-05, "loss": 0.0, "step": 7460 }, { "epoch": 6.138044371405094, "grad_norm": 0.0031271437577596937, "learning_rate": 1.5627595508972888e-05, "loss": 0.0, "step": 7470 }, { "epoch": 6.146261298274445, "grad_norm": 0.0027647829384675176, "learning_rate": 1.5497622143238837e-05, "loss": 0.0, "step": 7480 }, { "epoch": 6.154478225143796, "grad_norm": 0.003803065530812307, "learning_rate": 1.5368092326714232e-05, "loss": 0.0001, "step": 7490 }, { "epoch": 6.162695152013147, "grad_norm": 0.005264713571522259, "learning_rate": 1.5239007724584259e-05, "loss": 0.0, "step": 7500 }, { "epoch": 6.170912078882498, "grad_norm": 0.003345771722621515, "learning_rate": 1.5110369996310575e-05, "loss": 0.0, "step": 7510 }, { "epoch": 6.179129005751848, "grad_norm": 0.004309961466755241, "learning_rate": 1.4982180795609979e-05, "loss": 0.0, "step": 7520 }, { "epoch": 6.1873459326211995, "grad_norm": 0.0013193296697262904, "learning_rate": 1.4854441770433191e-05, "loss": 0.0, "step": 7530 }, { "epoch": 6.1955628594905505, "grad_norm": 0.001997910532498007, "learning_rate": 1.4727154562943702e-05, "loss": 0.0, "step": 7540 }, { "epoch": 6.203779786359902, "grad_norm": 0.0019134554143960024, "learning_rate": 1.4600320809496516e-05, "loss": 0.0, "step": 7550 }, { "epoch": 6.211996713229253, "grad_norm": 0.0029360750785403806, "learning_rate": 1.4473942140617253e-05, "loss": 0.0, "step": 7560 }, { "epoch": 6.220213640098603, "grad_norm": 0.012586896061178078, "learning_rate": 1.4348020180981181e-05, "loss": 0.0001, "step": 7570 }, { "epoch": 6.228430566967954, "grad_norm": 0.004457926890803278, "learning_rate": 1.422255654939224e-05, "loss": 0.0, "step": 7580 }, { "epoch": 6.236647493837305, "grad_norm": 0.004332221642359731, "learning_rate": 1.4097552858762276e-05, "loss": 0.0001, "step": 7590 }, { "epoch": 6.244864420706656, "grad_norm": 0.003876624079918508, "learning_rate": 1.3973010716090323e-05, "loss": 0.0, "step": 7600 }, { "epoch": 6.253081347576007, "grad_norm": 0.0053245520229774235, "learning_rate": 1.3848931722441977e-05, "loss": 0.0, "step": 7610 }, { "epoch": 6.261298274445357, "grad_norm": 0.0015663905732848297, "learning_rate": 1.3725317472928729e-05, "loss": 0.0, "step": 7620 }, { "epoch": 6.269515201314708, "grad_norm": 0.002041961818922734, "learning_rate": 1.3602169556687455e-05, "loss": 0.0, "step": 7630 }, { "epoch": 6.277732128184059, "grad_norm": 0.002124592433048766, "learning_rate": 1.3479489556860136e-05, "loss": 0.0, "step": 7640 }, { "epoch": 6.28594905505341, "grad_norm": 0.006483680702020652, "learning_rate": 1.335727905057334e-05, "loss": 0.0, "step": 7650 }, { "epoch": 6.294165981922761, "grad_norm": 0.008824383033340742, "learning_rate": 1.323553960891798e-05, "loss": 0.0, "step": 7660 }, { "epoch": 6.3023829087921115, "grad_norm": 0.005006078352606088, "learning_rate": 1.3114272796929177e-05, "loss": 0.0, "step": 7670 }, { "epoch": 6.3105998356614625, "grad_norm": 0.003911819277385541, "learning_rate": 1.2993480173566142e-05, "loss": 0.0, "step": 7680 }, { "epoch": 6.318816762530814, "grad_norm": 0.0015974523743080714, "learning_rate": 1.2873163291692065e-05, "loss": 0.0, "step": 7690 }, { "epoch": 6.327033689400165, "grad_norm": 0.003478313259129815, "learning_rate": 1.2753323698054138e-05, "loss": 0.0, "step": 7700 }, { "epoch": 6.335250616269515, "grad_norm": 0.008339588384415274, "learning_rate": 1.2633962933263826e-05, "loss": 0.0, "step": 7710 }, { "epoch": 6.343467543138866, "grad_norm": 0.0073636890827757, "learning_rate": 1.25150825317769e-05, "loss": 0.0, "step": 7720 }, { "epoch": 6.351684470008217, "grad_norm": 0.0011296785997836302, "learning_rate": 1.2396684021873734e-05, "loss": 0.0, "step": 7730 }, { "epoch": 6.359901396877568, "grad_norm": 0.003364679415997076, "learning_rate": 1.2278768925639722e-05, "loss": 0.0, "step": 7740 }, { "epoch": 6.368118323746919, "grad_norm": 0.0029050026983225624, "learning_rate": 1.2161338758945717e-05, "loss": 0.0, "step": 7750 }, { "epoch": 6.376335250616269, "grad_norm": 0.006790168668087443, "learning_rate": 1.2044395031428469e-05, "loss": 0.0, "step": 7760 }, { "epoch": 6.38455217748562, "grad_norm": 0.002118105174363574, "learning_rate": 1.1927939246471209e-05, "loss": 0.0, "step": 7770 }, { "epoch": 6.392769104354971, "grad_norm": 0.0022596599417157165, "learning_rate": 1.181197290118446e-05, "loss": 0.0, "step": 7780 }, { "epoch": 6.400986031224322, "grad_norm": 0.0020094366738312806, "learning_rate": 1.1696497486386631e-05, "loss": 0.0, "step": 7790 }, { "epoch": 6.409202958093673, "grad_norm": 0.0018075207570340274, "learning_rate": 1.158151448658496e-05, "loss": 0.0, "step": 7800 }, { "epoch": 6.4174198849630235, "grad_norm": 0.013845261767469415, "learning_rate": 1.146702537995632e-05, "loss": 0.0, "step": 7810 }, { "epoch": 6.425636811832375, "grad_norm": 0.0024476056166081827, "learning_rate": 1.1353031638328377e-05, "loss": 0.0, "step": 7820 }, { "epoch": 6.433853738701726, "grad_norm": 0.0033626822291673237, "learning_rate": 1.1239534727160534e-05, "loss": 0.0, "step": 7830 }, { "epoch": 6.442070665571077, "grad_norm": 0.001651014524301296, "learning_rate": 1.1126536105525092e-05, "loss": 0.0, "step": 7840 }, { "epoch": 6.450287592440427, "grad_norm": 0.0013975090416830352, "learning_rate": 1.1014037226088609e-05, "loss": 0.0, "step": 7850 }, { "epoch": 6.458504519309778, "grad_norm": 0.0019680528443716187, "learning_rate": 1.0902039535093107e-05, "loss": 0.0, "step": 7860 }, { "epoch": 6.466721446179129, "grad_norm": 0.0018501351630501506, "learning_rate": 1.0790544472337543e-05, "loss": 0.0, "step": 7870 }, { "epoch": 6.47493837304848, "grad_norm": 0.0009179913041330538, "learning_rate": 1.0679553471159238e-05, "loss": 0.0, "step": 7880 }, { "epoch": 6.483155299917831, "grad_norm": 0.003462689620436183, "learning_rate": 1.0569067958415552e-05, "loss": 0.0, "step": 7890 }, { "epoch": 6.491372226787181, "grad_norm": 0.0011821249977402727, "learning_rate": 1.0459089354465463e-05, "loss": 0.0, "step": 7900 }, { "epoch": 6.499589153656532, "grad_norm": 0.0013376525283599608, "learning_rate": 1.0349619073151273e-05, "loss": 0.0, "step": 7910 }, { "epoch": 6.507806080525883, "grad_norm": 0.0019887046498825553, "learning_rate": 1.0240658521780571e-05, "loss": 0.0, "step": 7920 }, { "epoch": 6.516023007395234, "grad_norm": 0.00127743474839085, "learning_rate": 1.0132209101107998e-05, "loss": 0.0, "step": 7930 }, { "epoch": 6.524239934264585, "grad_norm": 0.0021277443408164076, "learning_rate": 1.0024272205317325e-05, "loss": 0.0, "step": 7940 }, { "epoch": 6.5324568611339355, "grad_norm": 0.0016621441734981314, "learning_rate": 9.91684922200345e-06, "loss": 0.0, "step": 7950 }, { "epoch": 6.540673788003287, "grad_norm": 0.0018962326896739227, "learning_rate": 9.809941532154688e-06, "loss": 0.0, "step": 7960 }, { "epoch": 6.548890714872638, "grad_norm": 0.0025784439557522583, "learning_rate": 9.703550510134884e-06, "loss": 0.0, "step": 7970 }, { "epoch": 6.557107641741989, "grad_norm": 0.0012667199775145647, "learning_rate": 9.597677523665828e-06, "loss": 0.0, "step": 7980 }, { "epoch": 6.56532456861134, "grad_norm": 0.0013448286354715777, "learning_rate": 9.492323933809626e-06, "loss": 0.0, "step": 7990 }, { "epoch": 6.57354149548069, "grad_norm": 0.0032836140072708625, "learning_rate": 9.387491094951245e-06, "loss": 0.0, "step": 8000 }, { "epoch": 6.581758422350041, "grad_norm": 0.0009231606318727643, "learning_rate": 9.28318035478108e-06, "loss": 0.0, "step": 8010 }, { "epoch": 6.589975349219392, "grad_norm": 0.0034101077581938236, "learning_rate": 9.179393054277563e-06, "loss": 0.0, "step": 8020 }, { "epoch": 6.598192276088743, "grad_norm": 0.002291959822705487, "learning_rate": 9.076130527690091e-06, "loss": 0.0, "step": 8030 }, { "epoch": 6.606409202958094, "grad_norm": 0.0038596942348020848, "learning_rate": 8.973394102521703e-06, "loss": 0.0, "step": 8040 }, { "epoch": 6.614626129827444, "grad_norm": 0.001480194684233909, "learning_rate": 8.871185099512102e-06, "loss": 0.0, "step": 8050 }, { "epoch": 6.622843056696795, "grad_norm": 0.002805593189836145, "learning_rate": 8.769504832620645e-06, "loss": 0.0, "step": 8060 }, { "epoch": 6.631059983566146, "grad_norm": 0.006041014301825215, "learning_rate": 8.668354609009471e-06, "loss": 0.0, "step": 8070 }, { "epoch": 6.639276910435497, "grad_norm": 0.004297332286879946, "learning_rate": 8.567735729026704e-06, "loss": 0.0, "step": 8080 }, { "epoch": 6.647493837304848, "grad_norm": 0.0014920743333259015, "learning_rate": 8.467649486189665e-06, "loss": 0.0, "step": 8090 }, { "epoch": 6.655710764174199, "grad_norm": 0.0011500466201945887, "learning_rate": 8.36809716716837e-06, "loss": 0.0, "step": 8100 }, { "epoch": 6.66392769104355, "grad_norm": 0.005164111831234945, "learning_rate": 8.269080051768874e-06, "loss": 0.0, "step": 8110 }, { "epoch": 6.672144617912901, "grad_norm": 0.001262778142079656, "learning_rate": 8.170599412916863e-06, "loss": 0.0, "step": 8120 }, { "epoch": 6.680361544782252, "grad_norm": 0.0011679924772059953, "learning_rate": 8.072656516641303e-06, "loss": 0.0, "step": 8130 }, { "epoch": 6.688578471651603, "grad_norm": 0.016604120699383596, "learning_rate": 7.975252622058139e-06, "loss": 0.0, "step": 8140 }, { "epoch": 6.696795398520953, "grad_norm": 0.0040222538276890275, "learning_rate": 7.878388981354124e-06, "loss": 0.0, "step": 8150 }, { "epoch": 6.705012325390304, "grad_norm": 0.0020366695111365112, "learning_rate": 7.782066839770714e-06, "loss": 0.0, "step": 8160 }, { "epoch": 6.713229252259655, "grad_norm": 0.002683881336494846, "learning_rate": 7.686287435588063e-06, "loss": 0.0, "step": 8170 }, { "epoch": 6.721446179129006, "grad_norm": 0.0014642182397332326, "learning_rate": 7.5910520001091e-06, "loss": 0.0, "step": 8180 }, { "epoch": 6.729663105998356, "grad_norm": 0.0013438760932439806, "learning_rate": 7.49636175764371e-06, "loss": 0.0, "step": 8190 }, { "epoch": 6.737880032867707, "grad_norm": 0.0028709120496749874, "learning_rate": 7.402217925492982e-06, "loss": 0.0, "step": 8200 }, { "epoch": 6.746096959737058, "grad_norm": 0.0017762161725250717, "learning_rate": 7.308621713933567e-06, "loss": 0.0, "step": 8210 }, { "epoch": 6.754313886606409, "grad_norm": 0.0036501354123176385, "learning_rate": 7.215574326202118e-06, "loss": 0.0, "step": 8220 }, { "epoch": 6.76253081347576, "grad_norm": 0.0010842849261300245, "learning_rate": 7.123076958479819e-06, "loss": 0.0, "step": 8230 }, { "epoch": 6.770747740345111, "grad_norm": 0.0015077198839445545, "learning_rate": 7.031130799877012e-06, "loss": 0.0, "step": 8240 }, { "epoch": 6.778964667214462, "grad_norm": 0.0009969873311477047, "learning_rate": 6.939737032417909e-06, "loss": 0.0, "step": 8250 }, { "epoch": 6.787181594083813, "grad_norm": 0.03386953622819271, "learning_rate": 6.848896831025397e-06, "loss": 0.0, "step": 8260 }, { "epoch": 6.795398520953164, "grad_norm": 0.0008499452362792623, "learning_rate": 6.758611363505923e-06, "loss": 0.0, "step": 8270 }, { "epoch": 6.803615447822514, "grad_norm": 0.0027566772965418635, "learning_rate": 6.668881790534503e-06, "loss": 0.0, "step": 8280 }, { "epoch": 6.811832374691865, "grad_norm": 0.008194331655265475, "learning_rate": 6.5797092656397775e-06, "loss": 0.0, "step": 8290 }, { "epoch": 6.820049301561216, "grad_norm": 0.0012482399076770363, "learning_rate": 6.491094935189201e-06, "loss": 0.0, "step": 8300 }, { "epoch": 6.828266228430567, "grad_norm": 0.008029877560972723, "learning_rate": 6.403039938374284e-06, "loss": 0.0, "step": 8310 }, { "epoch": 6.836483155299918, "grad_norm": 0.008428519811452156, "learning_rate": 6.315545407195972e-06, "loss": 0.0, "step": 8320 }, { "epoch": 6.844700082169268, "grad_norm": 0.001316378738713296, "learning_rate": 6.228612466450085e-06, "loss": 0.0, "step": 8330 }, { "epoch": 6.852917009038619, "grad_norm": 0.0025395607649104646, "learning_rate": 6.14224223371283e-06, "loss": 0.0, "step": 8340 }, { "epoch": 6.86113393590797, "grad_norm": 0.0017105764774220768, "learning_rate": 6.0564358193264894e-06, "loss": 0.0, "step": 8350 }, { "epoch": 6.869350862777321, "grad_norm": 0.0014419035520029362, "learning_rate": 5.971194326385093e-06, "loss": 0.0, "step": 8360 }, { "epoch": 6.8775677896466725, "grad_norm": 0.0023864667796594757, "learning_rate": 5.886518850720268e-06, "loss": 0.0, "step": 8370 }, { "epoch": 6.885784716516023, "grad_norm": 0.0017185641948611085, "learning_rate": 5.802410480887138e-06, "loss": 0.0, "step": 8380 }, { "epoch": 6.894001643385374, "grad_norm": 0.002272620165118971, "learning_rate": 5.718870298150342e-06, "loss": 0.0, "step": 8390 }, { "epoch": 6.902218570254725, "grad_norm": 0.0009557354301123444, "learning_rate": 5.635899376470116e-06, "loss": 0.0, "step": 8400 }, { "epoch": 6.910435497124076, "grad_norm": 0.006317137120830284, "learning_rate": 5.553498782488498e-06, "loss": 0.0, "step": 8410 }, { "epoch": 6.918652423993427, "grad_norm": 0.0024636344491217868, "learning_rate": 5.471669575515625e-06, "loss": 0.0, "step": 8420 }, { "epoch": 6.926869350862777, "grad_norm": 0.0008714134451272828, "learning_rate": 5.3904128075160875e-06, "loss": 0.0, "step": 8430 }, { "epoch": 6.935086277732128, "grad_norm": 0.001329015572419464, "learning_rate": 5.30972952309543e-06, "loss": 0.0, "step": 8440 }, { "epoch": 6.943303204601479, "grad_norm": 0.0014464223279794595, "learning_rate": 5.22962075948672e-06, "loss": 0.0, "step": 8450 }, { "epoch": 6.95152013147083, "grad_norm": 0.001379359487093772, "learning_rate": 5.150087546537202e-06, "loss": 0.0, "step": 8460 }, { "epoch": 6.959737058340181, "grad_norm": 0.0009152371898863152, "learning_rate": 5.071130906695054e-06, "loss": 0.0, "step": 8470 }, { "epoch": 6.967953985209531, "grad_norm": 0.0014296051781477813, "learning_rate": 4.992751854996275e-06, "loss": 0.0, "step": 8480 }, { "epoch": 6.976170912078882, "grad_norm": 0.001981205199228891, "learning_rate": 4.914951399051604e-06, "loss": 0.0, "step": 8490 }, { "epoch": 6.9843878389482335, "grad_norm": 0.004504440784128484, "learning_rate": 4.837730539033564e-06, "loss": 0.0, "step": 8500 }, { "epoch": 6.9926047658175845, "grad_norm": 0.002182683987406521, "learning_rate": 4.761090267663626e-06, "loss": 0.0, "step": 8510 }, { "epoch": 7.000821692686935, "grad_norm": 0.0009001568544135854, "learning_rate": 4.685031570199477e-06, "loss": 0.0, "step": 8520 }, { "epoch": 7.009038619556286, "grad_norm": 0.0030032457674659726, "learning_rate": 4.609555424422241e-06, "loss": 0.0, "step": 8530 }, { "epoch": 7.017255546425637, "grad_norm": 0.0005553769128019869, "learning_rate": 4.534662800624034e-06, "loss": 0.0, "step": 8540 }, { "epoch": 7.025472473294988, "grad_norm": 0.0007834258457853978, "learning_rate": 4.460354661595434e-06, "loss": 0.0, "step": 8550 }, { "epoch": 7.033689400164339, "grad_norm": 0.000703826245852935, "learning_rate": 4.386631962613091e-06, "loss": 0.0, "step": 8560 }, { "epoch": 7.041906327033689, "grad_norm": 0.0012483085545840625, "learning_rate": 4.313495651427485e-06, "loss": 0.0, "step": 8570 }, { "epoch": 7.05012325390304, "grad_norm": 0.0007891283316159369, "learning_rate": 4.2409466682506944e-06, "loss": 0.0, "step": 8580 }, { "epoch": 7.058340180772391, "grad_norm": 0.0006788863859907665, "learning_rate": 4.168985945744392e-06, "loss": 0.0, "step": 8590 }, { "epoch": 7.066557107641742, "grad_norm": 0.0007917638510980589, "learning_rate": 4.097614409007722e-06, "loss": 0.0, "step": 8600 }, { "epoch": 7.074774034511093, "grad_norm": 0.0013736885502228914, "learning_rate": 4.026832975565542e-06, "loss": 0.0, "step": 8610 }, { "epoch": 7.082990961380443, "grad_norm": 0.0006387918794311795, "learning_rate": 3.956642555356549e-06, "loss": 0.0, "step": 8620 }, { "epoch": 7.091207888249794, "grad_norm": 0.0007192459021354059, "learning_rate": 3.887044050721606e-06, "loss": 0.0, "step": 8630 }, { "epoch": 7.0994248151191455, "grad_norm": 0.0010564868922662203, "learning_rate": 3.818038356392134e-06, "loss": 0.0, "step": 8640 }, { "epoch": 7.1076417419884965, "grad_norm": 0.0007279734624921566, "learning_rate": 3.749626359478614e-06, "loss": 0.0, "step": 8650 }, { "epoch": 7.115858668857848, "grad_norm": 0.0006607070160852934, "learning_rate": 3.681808939459214e-06, "loss": 0.0, "step": 8660 }, { "epoch": 7.124075595727198, "grad_norm": 0.000995870132333505, "learning_rate": 3.614586968168393e-06, "loss": 0.0, "step": 8670 }, { "epoch": 7.132292522596549, "grad_norm": 0.0008174432674370016, "learning_rate": 3.547961309785802e-06, "loss": 0.0, "step": 8680 }, { "epoch": 7.1405094494659, "grad_norm": 0.0006294058308118187, "learning_rate": 3.4819328208250903e-06, "loss": 0.0, "step": 8690 }, { "epoch": 7.148726376335251, "grad_norm": 0.0005353995290335276, "learning_rate": 3.4165023501229685e-06, "loss": 0.0, "step": 8700 }, { "epoch": 7.156943303204601, "grad_norm": 0.000613477673730241, "learning_rate": 3.351670738828211e-06, "loss": 0.0, "step": 8710 }, { "epoch": 7.165160230073952, "grad_norm": 0.0006072783598119823, "learning_rate": 3.2874388203909034e-06, "loss": 0.0, "step": 8720 }, { "epoch": 7.173377156943303, "grad_norm": 0.0009556385487319156, "learning_rate": 3.2238074205517443e-06, "loss": 0.0, "step": 8730 }, { "epoch": 7.181594083812654, "grad_norm": 0.0009023722561614857, "learning_rate": 3.160777357331346e-06, "loss": 0.0, "step": 8740 }, { "epoch": 7.189811010682005, "grad_norm": 0.0009126656094550076, "learning_rate": 3.098349441019804e-06, "loss": 0.0, "step": 8750 }, { "epoch": 7.198027937551355, "grad_norm": 0.0004952787276721692, "learning_rate": 3.036524474166225e-06, "loss": 0.0, "step": 8760 }, { "epoch": 7.206244864420706, "grad_norm": 0.0009256102431305894, "learning_rate": 2.975303251568462e-06, "loss": 0.0, "step": 8770 }, { "epoch": 7.2144617912900575, "grad_norm": 0.0005814294597868892, "learning_rate": 2.9146865602628358e-06, "loss": 0.0, "step": 8780 }, { "epoch": 7.2226787181594085, "grad_norm": 0.00115427274100865, "learning_rate": 2.8546751795140438e-06, "loss": 0.0, "step": 8790 }, { "epoch": 7.23089564502876, "grad_norm": 0.0007510569251756993, "learning_rate": 2.79526988080519e-06, "loss": 0.0, "step": 8800 }, { "epoch": 7.23911257189811, "grad_norm": 0.0006618215203906693, "learning_rate": 2.7364714278277636e-06, "loss": 0.0, "step": 8810 }, { "epoch": 7.247329498767461, "grad_norm": 0.0005890867382435551, "learning_rate": 2.6782805764719254e-06, "loss": 0.0, "step": 8820 }, { "epoch": 7.255546425636812, "grad_norm": 0.0014738614686725525, "learning_rate": 2.6206980748167076e-06, "loss": 0.0, "step": 8830 }, { "epoch": 7.263763352506163, "grad_norm": 0.0009830383674401353, "learning_rate": 2.563724663120476e-06, "loss": 0.0, "step": 8840 }, { "epoch": 7.271980279375514, "grad_norm": 0.0008129541015488385, "learning_rate": 2.5073610738113373e-06, "loss": 0.0, "step": 8850 }, { "epoch": 7.280197206244864, "grad_norm": 0.0011584128140593698, "learning_rate": 2.4516080314777647e-06, "loss": 0.0, "step": 8860 }, { "epoch": 7.288414133114215, "grad_norm": 0.000851812503026612, "learning_rate": 2.3964662528592974e-06, "loss": 0.0, "step": 8870 }, { "epoch": 7.296631059983566, "grad_norm": 0.0007901549249006642, "learning_rate": 2.3419364468372894e-06, "loss": 0.0, "step": 8880 }, { "epoch": 7.304847986852917, "grad_norm": 0.0008174256500227816, "learning_rate": 2.2880193144257932e-06, "loss": 0.0, "step": 8890 }, { "epoch": 7.313064913722268, "grad_norm": 0.0017107641934934197, "learning_rate": 2.2347155487625893e-06, "loss": 0.0, "step": 8900 }, { "epoch": 7.3212818405916185, "grad_norm": 0.0007189312677017951, "learning_rate": 2.1820258351002652e-06, "loss": 0.0, "step": 8910 }, { "epoch": 7.3294987674609695, "grad_norm": 0.0007600272692784257, "learning_rate": 2.1299508507973576e-06, "loss": 0.0, "step": 8920 }, { "epoch": 7.337715694330321, "grad_norm": 0.0006810628944823836, "learning_rate": 2.078491265309701e-06, "loss": 0.0, "step": 8930 }, { "epoch": 7.345932621199672, "grad_norm": 0.0011329805688820426, "learning_rate": 2.027647740181815e-06, "loss": 0.0, "step": 8940 }, { "epoch": 7.354149548069023, "grad_norm": 0.0011583381568143654, "learning_rate": 1.977420929038365e-06, "loss": 0.0, "step": 8950 }, { "epoch": 7.362366474938373, "grad_norm": 0.0006623008868649918, "learning_rate": 1.9278114775757804e-06, "loss": 0.0, "step": 8960 }, { "epoch": 7.370583401807724, "grad_norm": 0.0011443437517168107, "learning_rate": 1.878820023553951e-06, "loss": 0.0, "step": 8970 }, { "epoch": 7.378800328677075, "grad_norm": 0.0007241447557689012, "learning_rate": 1.8304471967880655e-06, "loss": 0.0, "step": 8980 }, { "epoch": 7.387017255546426, "grad_norm": 0.0008202533742528001, "learning_rate": 1.7826936191404297e-06, "loss": 0.0, "step": 8990 }, { "epoch": 7.395234182415776, "grad_norm": 0.0011035320754542752, "learning_rate": 1.735559904512546e-06, "loss": 0.0, "step": 9000 }, { "epoch": 7.403451109285127, "grad_norm": 0.000797888158920886, "learning_rate": 1.6890466588372178e-06, "loss": 0.0, "step": 9010 }, { "epoch": 7.411668036154478, "grad_norm": 0.000677049271616114, "learning_rate": 1.6431544800707078e-06, "loss": 0.0, "step": 9020 }, { "epoch": 7.419884963023829, "grad_norm": 0.0004004810900026675, "learning_rate": 1.5978839581850924e-06, "loss": 0.0, "step": 9030 }, { "epoch": 7.42810188989318, "grad_norm": 0.0026524632820758482, "learning_rate": 1.5532356751606692e-06, "loss": 0.0, "step": 9040 }, { "epoch": 7.4363188167625305, "grad_norm": 0.000647630622440846, "learning_rate": 1.509210204978495e-06, "loss": 0.0, "step": 9050 }, { "epoch": 7.4445357436318815, "grad_norm": 0.0006072480096154116, "learning_rate": 1.4658081136129432e-06, "loss": 0.0, "step": 9060 }, { "epoch": 7.452752670501233, "grad_norm": 0.0005258695810010135, "learning_rate": 1.4230299590245022e-06, "loss": 0.0, "step": 9070 }, { "epoch": 7.460969597370584, "grad_norm": 0.000785045611825937, "learning_rate": 1.3808762911525664e-06, "loss": 0.0, "step": 9080 }, { "epoch": 7.469186524239935, "grad_norm": 0.0010951684195373366, "learning_rate": 1.3393476519083736e-06, "loss": 0.0, "step": 9090 }, { "epoch": 7.477403451109285, "grad_norm": 0.002370463698692359, "learning_rate": 1.298444575168012e-06, "loss": 0.0, "step": 9100 }, { "epoch": 7.485620377978636, "grad_norm": 0.0018687243042547035, "learning_rate": 1.2581675867655963e-06, "loss": 0.0, "step": 9110 }, { "epoch": 7.493837304847987, "grad_norm": 0.0009009475327646263, "learning_rate": 1.2185172044865135e-06, "loss": 0.0, "step": 9120 }, { "epoch": 7.502054231717338, "grad_norm": 0.0007974077364470711, "learning_rate": 1.1794939380607106e-06, "loss": 0.0, "step": 9130 }, { "epoch": 7.510271158586688, "grad_norm": 0.0010000990075520598, "learning_rate": 1.1410982891561895e-06, "loss": 0.0, "step": 9140 }, { "epoch": 7.518488085456039, "grad_norm": 0.0010434464499455513, "learning_rate": 1.1033307513725443e-06, "loss": 0.0, "step": 9150 }, { "epoch": 7.52670501232539, "grad_norm": 0.0009734934140918075, "learning_rate": 1.0661918102346292e-06, "loss": 0.0, "step": 9160 }, { "epoch": 7.534921939194741, "grad_norm": 0.0021584258479075994, "learning_rate": 1.0296819431862781e-06, "loss": 0.0, "step": 9170 }, { "epoch": 7.543138866064092, "grad_norm": 0.0006342352616484835, "learning_rate": 9.938016195841948e-07, "loss": 0.0, "step": 9180 }, { "epoch": 7.5513557929334425, "grad_norm": 0.0006583685493983199, "learning_rate": 9.585513006919456e-07, "loss": 0.0, "step": 9190 }, { "epoch": 7.5595727198027936, "grad_norm": 0.0005905406976248578, "learning_rate": 9.239314396739751e-07, "loss": 0.0, "step": 9200 }, { "epoch": 7.567789646672145, "grad_norm": 0.0007994387794704483, "learning_rate": 8.899424815898006e-07, "loss": 0.0, "step": 9210 }, { "epoch": 7.576006573541496, "grad_norm": 0.0007100409206390407, "learning_rate": 8.565848633883156e-07, "loss": 0.0, "step": 9220 }, { "epoch": 7.584223500410847, "grad_norm": 0.0005577825747324518, "learning_rate": 8.238590139021396e-07, "loss": 0.0, "step": 9230 }, { "epoch": 7.592440427280197, "grad_norm": 0.0008326758307201644, "learning_rate": 7.917653538421221e-07, "loss": 0.0, "step": 9240 }, { "epoch": 7.600657354149548, "grad_norm": 0.0006265487662558193, "learning_rate": 7.603042957919248e-07, "loss": 0.0, "step": 9250 }, { "epoch": 7.608874281018899, "grad_norm": 0.0005892258784719833, "learning_rate": 7.294762442027314e-07, "loss": 0.0, "step": 9260 }, { "epoch": 7.61709120788825, "grad_norm": 0.0007622753204554103, "learning_rate": 6.992815953880405e-07, "loss": 0.0, "step": 9270 }, { "epoch": 7.625308134757601, "grad_norm": 0.0016100108456822975, "learning_rate": 6.697207375185532e-07, "loss": 0.0, "step": 9280 }, { "epoch": 7.633525061626951, "grad_norm": 0.0006474254009705203, "learning_rate": 6.407940506172217e-07, "loss": 0.0, "step": 9290 }, { "epoch": 7.641741988496302, "grad_norm": 0.0033795421320809618, "learning_rate": 6.125019065543302e-07, "loss": 0.0, "step": 9300 }, { "epoch": 7.649958915365653, "grad_norm": 0.0013752693884521677, "learning_rate": 5.848446690427223e-07, "loss": 0.0, "step": 9310 }, { "epoch": 7.658175842235004, "grad_norm": 0.0008427531420627247, "learning_rate": 5.578226936331365e-07, "loss": 0.0, "step": 9320 }, { "epoch": 7.666392769104355, "grad_norm": 0.000538981586154433, "learning_rate": 5.314363277096224e-07, "loss": 0.0, "step": 9330 }, { "epoch": 7.674609695973706, "grad_norm": 0.0007042065014478145, "learning_rate": 5.05685910485093e-07, "loss": 0.0, "step": 9340 }, { "epoch": 7.682826622843057, "grad_norm": 0.0007142517408986823, "learning_rate": 4.805717729969295e-07, "loss": 0.0, "step": 9350 }, { "epoch": 7.691043549712408, "grad_norm": 0.000907907850170342, "learning_rate": 4.560942381027611e-07, "loss": 0.0, "step": 9360 }, { "epoch": 7.699260476581759, "grad_norm": 0.0011361983534516529, "learning_rate": 4.322536204763028e-07, "loss": 0.0, "step": 9370 }, { "epoch": 7.70747740345111, "grad_norm": 0.0009770194473479631, "learning_rate": 4.0905022660329694e-07, "loss": 0.0, "step": 9380 }, { "epoch": 7.71569433032046, "grad_norm": 0.0007233697485938528, "learning_rate": 3.8648435477759427e-07, "loss": 0.0, "step": 9390 }, { "epoch": 7.723911257189811, "grad_norm": 0.0008146416373778586, "learning_rate": 3.6455629509730136e-07, "loss": 0.0, "step": 9400 }, { "epoch": 7.732128184059162, "grad_norm": 0.0017122674505446879, "learning_rate": 3.432663294610672e-07, "loss": 0.0, "step": 9410 }, { "epoch": 7.740345110928513, "grad_norm": 0.0009995921186196919, "learning_rate": 3.226147315644412e-07, "loss": 0.0, "step": 9420 }, { "epoch": 7.748562037797863, "grad_norm": 0.0006893374195408295, "learning_rate": 3.02601766896371e-07, "loss": 0.0, "step": 9430 }, { "epoch": 7.756778964667214, "grad_norm": 0.0007082755866281884, "learning_rate": 2.8322769273578244e-07, "loss": 0.0, "step": 9440 }, { "epoch": 7.764995891536565, "grad_norm": 0.0008724141540157085, "learning_rate": 2.644927581482659e-07, "loss": 0.0, "step": 9450 }, { "epoch": 7.773212818405916, "grad_norm": 0.0007403140247827292, "learning_rate": 2.463972039828899e-07, "loss": 0.0, "step": 9460 }, { "epoch": 7.781429745275267, "grad_norm": 0.0005724266236550641, "learning_rate": 2.2894126286909234e-07, "loss": 0.0, "step": 9470 }, { "epoch": 7.789646672144618, "grad_norm": 0.00066295643642293, "learning_rate": 2.121251592136886e-07, "loss": 0.0, "step": 9480 }, { "epoch": 7.797863599013969, "grad_norm": 0.0009883186398901085, "learning_rate": 1.9594910919800147e-07, "loss": 0.0, "step": 9490 }, { "epoch": 7.80608052588332, "grad_norm": 0.0005105245267251144, "learning_rate": 1.8041332077506356e-07, "loss": 0.0, "step": 9500 }, { "epoch": 7.814297452752671, "grad_norm": 0.0005523836279212665, "learning_rate": 1.6551799366695265e-07, "loss": 0.0, "step": 9510 }, { "epoch": 7.822514379622021, "grad_norm": 0.0007690638108040738, "learning_rate": 1.5126331936222704e-07, "loss": 0.0, "step": 9520 }, { "epoch": 7.830731306491372, "grad_norm": 0.0010022538716547183, "learning_rate": 1.3764948111346098e-07, "loss": 0.0, "step": 9530 }, { "epoch": 7.838948233360723, "grad_norm": 0.0006269855520444756, "learning_rate": 1.2467665393488536e-07, "loss": 0.0, "step": 9540 }, { "epoch": 7.847165160230074, "grad_norm": 0.0008751257110076748, "learning_rate": 1.1234500460014507e-07, "loss": 0.0, "step": 9550 }, { "epoch": 7.855382087099425, "grad_norm": 0.0006753350921018061, "learning_rate": 1.0065469164015074e-07, "loss": 0.0, "step": 9560 }, { "epoch": 7.863599013968775, "grad_norm": 0.0005753341606363669, "learning_rate": 8.960586534103587e-08, "loss": 0.0, "step": 9570 }, { "epoch": 7.871815940838126, "grad_norm": 0.0009432271066295922, "learning_rate": 7.919866774223073e-08, "loss": 0.0, "step": 9580 }, { "epoch": 7.880032867707477, "grad_norm": 0.0013549500961964926, "learning_rate": 6.943323263464141e-08, "loss": 0.0, "step": 9590 }, { "epoch": 7.888249794576828, "grad_norm": 0.0006072743257469168, "learning_rate": 6.030968555892358e-08, "loss": 0.0, "step": 9600 }, { "epoch": 7.8964667214461794, "grad_norm": 0.000792253023335194, "learning_rate": 5.182814380385592e-08, "loss": 0.0, "step": 9610 }, { "epoch": 7.90468364831553, "grad_norm": 0.000757134620534027, "learning_rate": 4.398871640486357e-08, "loss": 0.0, "step": 9620 }, { "epoch": 7.912900575184881, "grad_norm": 0.0006652616105216724, "learning_rate": 3.679150414258037e-08, "loss": 0.0, "step": 9630 }, { "epoch": 7.921117502054232, "grad_norm": 0.0007026535908356302, "learning_rate": 3.02365995415943e-08, "loss": 0.0, "step": 9640 }, { "epoch": 7.929334428923583, "grad_norm": 0.0010259009240952259, "learning_rate": 2.4324086869204065e-08, "loss": 0.0, "step": 9650 }, { "epoch": 7.937551355792934, "grad_norm": 0.0008119720328361312, "learning_rate": 1.9054042134397652e-08, "loss": 0.0, "step": 9660 }, { "epoch": 7.945768282662284, "grad_norm": 0.0007244333467401896, "learning_rate": 1.4426533086819849e-08, "loss": 0.0, "step": 9670 }, { "epoch": 7.953985209531635, "grad_norm": 0.0006484021131008125, "learning_rate": 1.0441619215939558e-08, "loss": 0.0, "step": 9680 }, { "epoch": 7.962202136400986, "grad_norm": 0.0005127588548068826, "learning_rate": 7.0993517502671024e-09, "loss": 0.0, "step": 9690 }, { "epoch": 7.970419063270337, "grad_norm": 0.0011553438763330672, "learning_rate": 4.39977365669364e-09, "loss": 0.0, "step": 9700 }, { "epoch": 7.978635990139688, "grad_norm": 0.0006641962071988268, "learning_rate": 2.342919639958252e-09, "loss": 0.0, "step": 9710 }, { "epoch": 7.986852917009038, "grad_norm": 0.0006978104616291894, "learning_rate": 9.288161421705521e-10, "loss": 0.0, "step": 9720 }, { "epoch": 7.995069843878389, "grad_norm": 0.00081968448531169, "learning_rate": 1.5748134250537227e-10, "loss": 0.0, "step": 9730 }, { "epoch": 8.0, "step": 9736, "total_flos": 935284648181760.0, "train_loss": 0.13075080679284903, "train_runtime": 28642.4295, "train_samples_per_second": 173.996, "train_steps_per_second": 0.34 } ], "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": true }, "attributes": {} } }, "total_flos": 935284648181760.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }